article
stringlengths 0
745k
| abstract
stringlengths 0
92.7k
|
|---|---|
atomic ensembles in optical resonators offer a promising platform for studying the physics of long - range interacting systems @xcite .
the long - range interaction here originates from multiple scattering of cavity photons , which carry the information about the positions of the scattering atoms and thus mediate an optomechanical interparticle potential @xcite . in a single - mode cavity
the photons are coherent over the cavity mode volume , which makes the interaction range as large as the system size .
therefore , the energy is non - additive like in gravitational and coulomb systems in two or more dimensions @xcite . in equilibrium
statistical mechanics , consequences of non - additivity are for instance the super - linear scaling of thermodynamic quantities with the system size and the inequivalence of the statistical ensembles @xcite , one manifestation of which are negative specific heats in the microcanonical ensemble @xcite .
differing from these systems , however , the dynamics of atomic gases in optical cavities is typically dissipative and non - trivial effects can only be observed if either the atoms or the cavity are pumped by light @xcite .
the steady state , when it exists , results from the dynamical interplay between drive and losses and its properties thus depend on the drive and on the cavity parameters .
it is therefore often not possible to draw a direct connection with equilibrium statistical mechanics of long - range interacting systems . in this context
it is remarkable that for some parameter regimes the dynamics of atoms spatial selforganization in an optical resonator can be mapped to long - range interacting systems at equilibrium @xcite .
selforganization of the atomic gas in ordered spatial patterns occurs in a single - mode standing - wave resonator when the atoms are driven by lasers whose intensity exceeds a threshold value , which depends also on the cavity decay rate @xcite . by suitably tuning the laser frequency , moreover , a stationary state exists which is characterised by a maxwell - boltzmann distribution of the atomic momentum . in ref .
@xcite it was shown that the stationary dynamics can be mapped to the one of the hamiltonian - mean - field model in a canonical ensemble @xcite and in particular that the transition to spatial order can be described in terms of a landau second - order phase transition .
the dynamics leading to equilibrium , moreover , exhibits a slow relaxation that is due to the interplay between the conservative and dissipative cavity - mediated long - range forces @xcite . in this paper
we extend the model of ref .
@xcite and consider a gas of cold atoms that interact with two cavity modes and are transversally driven by lasers .
a possible setup is illustrated in fig .
[ fig:1 ] .
we determine the parameter regimes where the dynamics asymptotically tends to a stationary state and show that its phase diagram as a function of the lasers and of the cavity parameters can be mapped to the one of the generalized hamiltonian mean field model ( ghmf ) in a canonical ensemble @xcite .
this model describes the dynamics of @xmath0 particles with canonically conjugated variables @xmath1 , @xmath2 constrained on a circle that interact via competing long - range forces . in the form studied in ref .
@xcite its hamiltonian reads @xmath3 where @xmath4 and @xmath5 is a dimensionless parameter that can vary continuously in the interval @xmath6 $ ] . the phase diagram as a function of the temperature and @xmath5 is characterised by ( i ) a paramagnetic , ( ii ) a nematic , and ( iii ) a ferromagnetic phase , with first and second - order transitions .
in our case the phases correspond to density modulations of the atoms at different periodicity and can be detected through the light emitted by the cavity .
our motivation draws from ongoing experimental investigations @xcite .
theoretical studies of this system focused on the dynamics leading to equilibrium and reported the existence of several metastable states @xcite .
these properties are at the basis of proposals for using these systems to simulate a quantum hopfield associative memory scheme @xcite .
the determination of the condition for a stationary state of the setup in fig .
[ fig:1 ] and the analysis of the corresponding phase diagram is the main result of the present manuscript .
the mapping to the ghmf model shows that cavity quantum electrodynamics ( cqed ) setups offer a versatile platform for studying the statistical mechanics of systems with long - range interactions .
this article is organised as follows . in sec .
[ sec:2 ] we introduce the physical model and sketch the derivation of a fokker - planck equation governing the dynamics of the atoms external variables in the semiclassical limit .
we then determine the parameters regime for which the fokker - planck equation allows for a stationary state which is a maxwell - boltzmann distribution . in sec .
[ sec:3 ] we define the free energy , which we can associate to the stationary state , and introduce an appropriate thermodynamic limit .
we then show that the free energy can be mapped to the one of the ghmf model .
we determine the phase diagram as a function of the system s parameters and identify the observables which allow one to measure the predicted phases . in sec .
[ sec:4 ] we discuss possible implementations of the setup which could correspond to the realization of the microcanonical ensemble of the ghmf model , for which ensemble inequivalence has been predicted @xcite , and identify the parameters regimes for which it could be measured .
-axis and interacts with the modes of two optical cavities , whose wave vectors form the angle @xmath7 .
the cavities emit photons at rate @xmath8 and @xmath9 , respectively , and are pumped via coherent scattering of laser photons by the atoms .
the atoms , in turn , experience the optical potential and the dissipative forces which result from the mechanical effects of the two cavity fields .
inset : linearly - polarized lasers propagate along the direction orthogonal to the plane defined by the two cavity wave vectors .
we assume that cavity 1(2 ) couples with the electronic transition @xmath10 ( @xmath11 ) , as illustrated in subplot * ( b)*. the electronic transitions are also driven by the laser fields at rabi frequency @xmath12 ( @xmath13 ) .
the cavity and laser fields are far - off resonance from the dipolar transition they couple to and quasi resonant with each other . in this limit
the scattering is prevailingly coherent . ]
the system we consider consists of a gas of @xmath0 cold atoms of mass @xmath14 , whose motion is confined to occur along one dimension parallel to the unit vector @xmath15 .
we denote by @xmath16 and @xmath17 the canonically conjugated position and momentum operators ( @xmath18 ) , such that @xmath19=i\hbar\delta_{ij}$ ] .
the atoms experience the optomechanical forces due to the interaction with the lasers and with the quantized fields of two high - finesse cavities in the setup of fig .
[ fig:1 ] . specifically , the axes of the two cavities are in the @xmath20 plane , the wave vector @xmath21 of cavity 1 forms an angle @xmath7 with the @xmath22-axis and thus the force the atoms experience is the projection of the mechanical force along @xmath22 , while the wave vector @xmath23 of cavity 2 is parallel to @xmath15 .
the laser fields are linearly polarised and propagate in the direction orthogonal to the plane , they pump the cavity fields by means of coherent scattering via the atoms .
the amplitude of coherent scattering , in turn , is maximal when the atoms form bragg gratings , whose stability depends on the mechanical forces of the cavity light .
as we will show , a nematic phase corresponds to a stable bragg grating which supports the build - up of the field of only one cavity mode . in the ferromagnetic phase , instead , the atoms form stable bragg gratings for both modes . below we describe the setup in detail and introduce the master equation for the density matrix @xmath24 of atoms and cavity fields which governs the system s dynamics .
we then sketch the derivation of a fokker - planck equation for the motion of the atoms external degrees of freedom , which is valid when the atomic variables can be treated as semiclassical variables and the cavity fields can be eliminated from the equations of motion in a coarse - grained time scale .
we finally determine the stationary state of the atoms and identify the regime in which it is a thermal state .
the state of the atoms external degrees of freedom and of the cavity modes is described by the density operator @xmath24 , whose dynamics is governed by a born - markov master equation of the form : @xmath25-\sum_{n=1,2}\kappa_n\left(\hat{a}^\dagger_n\hat{a}_n\hat{\rho}+\hat{\rho}\hat{a}^\dagger_n\hat{a}_n-2\hat{a}_n\hat{\rho}\hat{a}^\dagger_n\right)\ , , \label{eq : master_equation}\ ] ] where @xmath26 is the hamiltonian of the system , which we introduce below , and the second term of the right - hand side describes photon emission by the cavity modes at rate @xmath27 ( @xmath28 ) . here ,
@xmath29 and @xmath30 denote the creation and annihilation operators of a photon of the standing - wave cavity mode @xmath31 ( @xmath28 ) , with wave vector @xmath32 , frequency @xmath33 , and linear polarization in the @xmath20 plane .
the commutation relations are @xmath34=\delta_{ij}$ ] .
hamiltonian @xmath26 describes the optomechanical coupling between atoms degrees of freedom and cavity modes .
it is an effective hamiltonian derived in the limit where the atoms internal degrees of freedom can be adiabatically eliminated , such that the scattering processes are coherent and the relevant parameters of the atomic internal structure is the atoms polarizability @xcite . here ,
cavity mode @xmath31 couples with the electronic transition @xmath35 at frequency @xmath36 with vacuum rabi frequency @xmath37 .
we further assume that the coupling of mode 1 ( 2 ) with @xmath38 ( @xmath39 ) is off - resonance by orders of magnitude and can be discarded ( nonetheless , the wave numbers are assumed to be @xmath40 ) . in this limit cavity @xmath31
is pumped by coherent scattering of the laser , which couples to the transition @xmath35 with rabi frequency @xmath41 and frequency @xmath42 .
the condition for adiabatic elimination of the internal excited state is given by the inequality @xmath43 , where @xmath44 is the mean intracavity photon number in cavity @xmath31 and @xmath45 is the detuning of the laser from the cavity mode it pumps @xcite .
in this regime @xmath26 reads : @xmath46 and is here reported in the frame where each atomic transition and cavity mode rotates at the corresponding laser frequency . beside the kinetic energy of the atoms
, it contains the resonators energy , which is shifted by the dynamical stark shift with amplitude @xmath47 induced by the coupling between cavity mode and the atoms at the position @xmath48 within the cavity spatial mode function @xmath49 .
this term is also a periodic potential for the atoms whose depth is a dynamical variable .
the last term on the right - hand side , finally , describes coherent scattering by the atoms between laser and cavity mode with coupling strength @xmath50 .
it is an effective pump of the resonator whose amplitude is maximal when the atoms form bragg gratings maximizing the expectation value of the operator @xmath51 .
note that in eq .
( [ eq : effective_hamiltonian ] ) we introduced the notation @xmath52 and @xmath53 . in the following
we will set @xmath54 , thus @xmath55 .
we now discuss the assumptions at the basis of the derivation of a fokker - planck equation ( fpe ) for the dynamics of the atomic external variables .
a semiclassical description of the atoms center - of - mass motion is justified when the width @xmath56 of the single - atom momentum distribution is much larger than the linear momentum @xmath57 carried by a cavity photon @xmath58 @xcite . in this limit
it is convenient to use the wigner function @xmath59 for the atomic variables @xmath60 , @xmath61 : @xmath62 with @xmath63 .
we further assume that the cavity field relaxes very fast to a local steady state depending on the atomic distribution , which is verified when the inequality @xmath64 is fulfilled , namely , when the dimensionless parameter @xmath65 is small .
this allows us to identify a coarse - grained time scale @xmath66 that is infinitesimal for the external degrees of freedom but over which the cavity degrees of freedom can be eliminated from the equations of the atomic dynamics .
in particular , @xmath67 , where @xmath68 is the term in zero order in the retardation effect , corresponding to the cavity field following adiabatically the atomic motion , and @xmath69 represents the non - adiabatic corrections scaling with @xmath70 .
the latter can be expressed in terms of @xmath68 using perturbation theory @xcite .
the derivation is lengthy but is also a straightforward extension of the derivation for a single - mode cavity , which is extensively reported in ref .
we thus refer the interested reader to this work and present here the resulting fpe , which reads @xmath71 \nonumber\\ & + \sum_{i , j=1}^n\sum_{n=1,2}\frac{\partial}{\partial p_j}\left[\eta_n\sin(k_nx_i)\sin(k_nx_j)\frac{\partial}{\partial x_i}f({\bf x},{\bf p},t)\right]\ , .
\label{eq : fpe}\end{aligned}\ ] ] in detail , hamiltonian ( [ eq : hamiltonian ] ) results from the adiabatic component of the dynamics and describes coherent long - range , two - body interactions which are mediated by the cavity photons : @xmath72 where @xmath73 is a scalar and @xmath74 the quantities @xmath75 are order parameters for spatial selforganization .
in fact , they vanish for homogeneous spatial distributions , they are both different from zero when the atomic density forms spatial gratings with periodicity @xmath76 , while @xmath77 and @xmath78 when the spatial grating has periodicity @xmath79 , as illustrated in fig .
[ fig:2 ] .
the cavity mode field amplitudes @xmath80 , in turn , are proportional to @xmath75 and in leading order in the expansion in @xmath70 read @xcite @xmath81 the bragg gratings can be stable provided that @xmath82 .
the sign of @xmath73 is here controlled by the detuning @xmath83 , and hence by the frequency of the pumping laser . for later convenience
we write @xmath84 with @xmath85 friction and diffusion are instead due to retardation effects between atoms and cavity dynamics and describe cross - correlations between the atoms , which can play a relevant role in stabilizing the system in non - thermal metastable states @xcite .
their coefficients take the form @xmath86 and are here reported in the limit @xmath87 , where we neglected the contribution of the dynamical stark shift to the dynamics .
-axis and as a function of @xmath88 .
the spatial configurations leading to non - vanishing values of @xmath89 and/or @xmath90 are illustrated by the bullet points along each horizontal line .
as we show below , configurations where @xmath77 while @xmath78 correspond to the nematic phase of the ghmf .
configurations where both @xmath91 are ferromagnetic phases .
each configuration gives rise to different cavity field amplitudes and can thus be detected by monitoring the fields at the cavities outputs.,scaledwidth=100.0% ] the fpe ( [ eq : fpe ] ) allows for a stationary solution satisfying the condition @xmath92 .
we first consider two limiting cases , in which only one cavity mode is pumped .
these situations correspond to the dynamics of atoms in a single - mode standing - wave cavity investigated in refs .
@xcite .
let us first assume that @xmath93 but @xmath94 .
in this case a stationary state exists provided that @xmath95 and the stationary distribution reads @xmath96 , with normalization constant @xmath97 and @xmath98 given in eq .
( [ betan ] ) . ) that scale with @xmath99 .
this is exact if @xmath100 . in general , these terms give rise to corrections to the coherent dynamics that scale with @xmath101 in the thermodynamic limit we use in sec .
[ sec:3 ] , see ref .
@xcite for details . ]
vice versa , when @xmath102 , @xmath103 , and @xmath104 , the stationary state is @xmath105 with @xmath106 from eq .
( [ betan ] ) . when both resonators are pumped , a stationary solution can be found provided that @xmath107 are negative and @xmath108 , namely : @xmath109 this is the situation we consider in the following . in particular ,
@xmath110 , where @xmath111 the detunings and cavity loss rates thus determine an effective temperature @xmath112 characterizing the stationary state .
the stationary state is given by @xmath113
we now consider the stationary state of eq .
( [ eq : steady ] ) and draw a formal analogy to a canonical ensemble at equilibrium . for this purpose
we define the thermodynamic limit , according to which @xmath114 is constant .
this assumption warrants that the energy is extensive , it is thus equivalent to kac scaling @xcite and physically corresponds to scaling the cavity mode volume linearly with the number of particles @xcite . in this thermodynamic limit
we obtain an explicit expression of the free energy per particle which allows us to perform a mapping of the steady state in eq .
( [ eq : steady ] ) to the canonical ensemble realization of the ghmf .
we determine the free energy per particle @xmath115 using the relation @xmath116 , where @xmath117 is the canonical " partition function and reads : @xmath118 with @xmath119 .
we integrate over the momenta @xmath120 and apply the hubbard - stratonovich transformation to eliminate @xmath121 , obtaining @xmath122\nonumber \,,\end{aligned}\ ] ] with @xmath123 , @xmath124 , and @xmath125\ , .
\label{eq : canonical_integral}\ ] ] in the thermodynamic limit @xmath126 we perform a saddle - point approximation , which leads to the expression for the free energy @xmath127 this expression coincides , apart for irrelevant constants , with the mean free energy of the ghmf model in the canonical ensemble @xcite . in particular , the extrema @xmath128 of the free energy fulfill the relation @xmath129}{\int_{0}^{2\pi}dq \exp\left[2(\alpha_1y_1^*\cos(q)+\alpha_2y_2^*\cos(2q))\right]}\ , , \label{eq : fixpoint}\ ] ] their values lie in the interval @xmath130 $ ] and they can be identified with the variables @xmath75 : @xmath131 .
this shows that @xmath132 are analogous to the magnetization in the ghmf model . by means of this mapping ,
moreover , we connect the lasers and cavity parameters with the dimensionless parameter @xmath5 of the ghmf in eq .
( [ eq : ghmf ] ) : @xmath133 and @xmath134 . therefore , varying the lasers amplitudes would allow one to span over the values of @xmath5 in eq .
( [ ghmf ] ) , while the effective temperature can be tuned varying the detunings @xmath135 and @xmath136 , and thus the lasers frequencies ( provided condition ( [ condition ] ) is fulfilled ) .
the phase diagram is obtained following the same analysis as in ref . @xcite and it is illustrated in fig .
[ fig:3 ] .
the system exhibits second and first order phase transitions ( see [ app : a ] for details ) , which are found for the same corresponding values of the phase diagram as in ref .
the phases can be measured by monitoring the amplitude and the phase of the fields at the cavity output , since these are proportional to the order parameters @xmath89 and @xmath90 , as visible in eq .
( [ eq : steady_cavity_field ] ) . and @xmath137 .
the phases are identified by numerically determining the global minima of the free energy of eq .
( [ eq : free_energy ] ) .
the system shows second - order and first - order phase transitions ( see inset ) . in the inset : p denotes paramagnetic , f ferromagnetic , and n nematic phase .
the dark and light gray areas , labeled by a and b , respectively , indicate the parameter regions where ensemble inequivalence is expected . in a ( b )
the microcanonical ensemble exhibits three ( two ) phases @xcite.,scaledwidth=80.0% ]
the mapping of the stationary dynamics of eq .
( [ eq : fpe ] ) to the canonical ghmf shows that cavity qed can be a versatile platform for studying equilibrium statistical mechanics of long - range interacting systems .
in this perspective it is important to identify the parameter regimes for which this setup could simulate the microcanonical ghmf model .
this would allow one to experimentally investigate the ensemble inequivalence that has been predicted for the ghmf @xcite . within the validity of the semiclassical description ,
the microcanonical ghmf could be realised in the regime where the parameter @xmath70 , eq . ( [ eq : epsilon ] ) , becomes smaller than the small parameter @xmath138 of the semiclassical expansion .
this would require one to choose the detunings @xmath139 @xcite . in this limit
there is a well defined time scale over which the dynamics is coherent and solely dominated by hamiltonian ( [ eq : hamiltonian ] ) , while the right - hand side of the fpe ( [ eq : fpe ] ) , which scales with @xmath70 , can be discarded .
moreover , in order to prepare the system in a microcanonical ensemble , the atomic gas shall be in the asymptotic boltzmann - gibbs distribution of the corresponding hamiltonian dynamics @xcite . by identifying the mean energy per particle in the canonical ensemble @xmath140 with the constant energy @xmath141 in the microcanonical ensemble
, we obtain the relation @xmath142 which allows us to directly compare the results obtained in both ensembles .
ensemble inequivalence is predicted in the regions where the three phases meet and is indicated by the shaded area in fig .
[ fig:3 ] . in an experimental realization
it would become evident by detecting different phases in the canonical and microcanonical realizations for the same values of @xmath143 and @xmath137 @xcite
. it would be interesting to identify observables of this system which provide a measure of the specific heat , thus allowing one to determine whether it may become negative in the microcanonical ensemble .
this is an open question which we will address in future work .
the authors acknowledge discussions with shamik gupta , julian lonard , stefan schtz , valentin torggler , and tobias donner .
this work was supported by the german research foundation ( dfg dach `` quantum crystals of matter and light '' ) and by the european commission ( itn network `` colopt '' ) .
for the calculation of the phase diagram in fig .
[ fig:4 ] we numerically calculated the global minimum of the free energy , eq .
( [ eq : free_energy ] ) , for each pair @xmath144 .
the global minimum describes a paramagnetic phase when @xmath145 .
the phase is nematic when @xmath146 and @xmath147 and ferromagnetic when @xmath91 .
a phase transition occurs where the properties of the minimum change by varying @xmath148 .
the order of the transition is determined by calculating numerically the first derivatives of @xmath149 with respect to @xmath143 and @xmath137 : if they are discontinuous at the phase transition the transition is of first order , while if they are continuous the transition is of second order .
this determines the phase diagram in fig .
[ fig:4 ] .
+ we also checked our results by analytically calculating the hessian matrix of the free energy , eq .
( [ eq : free_energy ] ) , at the extrema where eq .
( [ eq : fixpoint ] ) is fulfilled .
we note that @xmath150 is always a solution of eq .
( [ eq : fixpoint ] ) .
it is a minimum when the following inequalities hold : @xmath151 the thin black dashed line in fig .
[ fig:4 ] delimitates the region @xmath152 , where the paramagnetic phase is a local minimum of the free energy .
we now consider the nematic phase , set @xmath77 in eq .
( [ eq : fixpoint ] ) and obtain the equation for @xmath90 : @xmath153 where @xmath154 is the modified bessel function of order @xmath31 .
the hessian matrix is positive definite when the following inequalities are fulfilled : @xmath155 one can show that imposing @xmath78 in eq .
( [ besselfunction ] ) is equivalent to eq .
( [ eq : canonical_nematic_conditions2 ] ) .
inequality ( [ eq : canonical_nematic_conditions1 ] ) determines an upper threshold @xmath156 on @xmath143 above which the nematic configuration is no longer a minimum of the free energy . the thin black solid line in fig
. [ fig:4 ] shows @xmath157 in the @xmath148-plane .
note that @xmath158 ( @xmath159 ) for @xmath160 ( @xmath161 ) . in the limit @xmath162 , eq .
( [ besselfunction ] ) delivers @xmath163 and @xmath164 , giving @xmath165 and @xmath166 , respectively .
+ we note that the conditions we determine analytically do not overlap for all values of @xmath167 with the numerically calculated phase transition lines ( see fig .
[ fig:4 ] ) .
for instance there is an area where the inequalities ( [ eq : canonical_para_conditions ] ) hold but the global minimum is in the ferromagnetic phase .
moreover the condition in eq .
( [ eq : canonical_nematic_conditions1 ] ) predicts a parameter region where the nematic phase with @xmath161 is a local minimum but the global minimum is a ferromagnetic phase . and the results of our analytical analysis .
the dashed black lines delimitate the region determined by the inequality eq.([eq : canonical_para_conditions ] ) , where a paramagnetic phase ( @xmath168 ) is a local minimum of the free energy .
the area below the solid black line ( determined by the inequality eq.([eq : canonical_nematic_conditions1 ] ) ) is the region where a nematic phase ( @xmath77 , @xmath78 ) is a local minimum of the free energy.,scaledwidth=80.0% ] | atoms in high - finesse optical resonators interact via the photons they multiply scatter into the cavity modes .
the dynamics is characterized by dispersive and dissipative optomechanical long - range forces , which are mediated by the cavity photons , and exhibits a steady state for certain parameter regimes . in standing - wave cavities
the atoms can form stable spatial gratings . moreover ,
their asymptotic distribution is a maxwell - boltzmann whose effective temperature is controlled by the laser parameters . in this work we show that in a two - mode standing - wave cavity the stationary state possesses the same properties and phases of the generalized hamiltonian mean field model in the canonical ensemble .
this model has three equilibrium phases : a paramagnetic , a nematic , and a ferromagnetic one , which here correspond to different spatial orders of the atomic gas and can be detected by means of the light emitted by the cavities .
we further discuss perspectives for investigating in this setup the ensemble inequivalence predicted for the generalized hamiltonian mean field model . |
in various economic contexts , institutions hold assets to cover future random liabilities .
banks and insurance companies are required by the authorities to hold regulatory capital to cover the risks they take .
pension funds face random future liabilities due to longevity risk and the structure of the pension plans which may involve variable annuity - type features .
the problem of managing a portfolio of assets under the condition of covering future liabilities or benchmarks , in particular in the context of pension plans , is commonly known as asset - liability management ( alm ) @xcite . the present study is mainly motivated by the alm problem of a nuclear power plant operator . in several countries ,
energy companies operating nuclear power plants are required by law to hold decomissioning funds , to cover the future costs of decontaminating and dismantling the plants , as well as the treatment and long - term storage of the radioactive waste . in france , this obligation was introduced by the law n@xmath0 2006 - 739 of june 28 , 2006 relative to sustainable management of radioactive substances and waste . according to this law , the three major players of the civil nuclear industry in france ( edf , areva and cea ) must hold portfolios of assets dedicated to future nuclear decomissioning charges @xcite
. the value of these dedicated assets must be sufficient to cover the discounted value of future liabilities .
the discount rate is determined by the operator , but the law stipulates that `` it may not exceed the rate of return , as anticipated with a high degree of confidence , of the hedging portfolio , managed with sufficient security and liquidity '' . computing the highest possible discount rate is therefore equivalent to finding the cheapest portfolio which covers the future liabilities with a high degree of confidence .
the law does not define the notions such as `` high degree of confidence '' in a precise manner , but it is clear that since the future liabilities are subject to considerable uncertainty , some kind of probabilistic risk measure criterion such as shortfall probability or value at risk should be used . in other contexts ,
the regulator stipulates that a specific probabilistic criterion should be used to measure the potential losses arising from not being able to cover the liabilities .
basel ii framework uses the value at risk to determine regulatory capital for banks . under the european solvency ii directive ,
insurance companies are required to evaluate the amount of capital necessary to cover their liabilities for a time horizon of one year with a probability of @xmath1 . under such frameworks
, companies must therefore hold enough assets to limit the losses with respect to a random future benchmark in the sense of a probabilistic loss measure .
motivated by these issues , in this paper , we consider the problem of an economic agent , who has random liabilities payable at a finite set of future dates , with expected shortfall constraints imposed at each payment date .
we consider two related questions : how to find the least expensive portfolio which allows to satisfy the constraints ( hedge the liabilities ) at each date , and how to determine the relationship between the probabilistic constraints at different dates .
for the latter issue , we propose three different formulations : with the european style constraints , the bound is imposed on the expected shortfall at each date , computed at @xmath2 ; in the time - consistent constraint , the bound is imposed at each payment date on the next period s shortfall ; finally , in the lookback - style constraint , the bound is imposed on the expectation of the maximal shortfall over all dates . in the literature ,
the problem of hedging a single random liability under a probabilistic constraint has been introduced and studied mainly in the complete market continuous - time setting in @xcite , where explicit solutions were obtained in the black - scholes model .
this was generalized to markovian contexts in @xcite using stochastic control and viscosity solutions and later extended to other classes of stochastic target problems in @xcite . a related strain of literature deals with portfolio management under additional constraint of outperforming a stochastic benchmark at a given date @xcite or a deterministic benchmark at all future dates @xcite .
the problem of hedging multiple stochastic benchmarks at a finite set of future dates , has , however , received little attention in the literature . in our paper , motivated by practical concerns that the liability constraints must be verified at pre - specified given dates , we place ourselves in the discrete - time setting .
to find the least expensive hedging portfolio , we adopt the classical complete market framework , which allows us to reformulate the problem as the one of finding the smallest discrete - time supermartingale satisfying a set of stochastic constraints .
this can be seen as an extension of the classical notion of snell envelope , which arises in the problem of superhedging an american contingent claim see ( * ? ? ?
* section 6.5 ) .
the market completeness assumption implies that in our context , the randomness of the future liabilities is mainly determined by `` hedgeable '' risk factors , such as the market risk , the interest rate risk , the inflation etc .
, rather than by unpredictable random events like changes in the regulatory framework . under the market completeness assumption , it is easy to find a strategy for hedging all future liabilities almost surely .
the main difficulty of our problem is due to the fact that the constraints are given by expected loss functions and therefore probabilistic in nature .
furthermore , we are concerned with general discrete - time processes where the markovian property holds no longer .
our main contribution is to investigate the interplay between the probabilistic constraints at different dates , and to characterize the solution via dynamic programming in a general context .
we consider three types of different risk constraints which concern respectively the expected loss at all liability payment dates , the time - consistent conditional loss in a dynamic manner and the maximum of all loss scenarios .
for each of the three constraint styles that we use , we obtain a recursive formula for the least expensive hedging portfolio .
explicit examples are then developed for specific loss functions .
our paper is structured as follows . in section [ problem.sec ]
we introduce three different formulations of the multi - objective hedging problem according to the three different constraint styles .
section [ solution.sec ] presents a non - markovian dynamic programming approach to the solution .
section [ examples2d.sec ] provides explicit examples in the case of two payment dates , and section [ examplesnd.sec ] contains some explicit results for the @xmath3-dimensional case .
finally , technical lemmas are proven in the appendix .
we start with a filtered probability space @xmath4 , where @xmath5 is a trivial @xmath6-field and @xmath7 . on this space
, we consider a financial market model with a risk - free asset @xmath8 and @xmath9 risky assets @xmath10 .
the risky assets are assumed to be adapted to the filtration @xmath11 . without loss of generality ,
we take the risk - free asset to be a constant process : @xmath12 .
we assume that there exists a class of admissible portfolio strategies which does not need to be made precise at this point . to guarantee absence of arbitrage and completeness
, we make the following assumption : there exists a probability @xmath13 equivalent to @xmath14 such that all admissible self - financing portfolios are @xmath13-supermartingales , and for any @xmath13-supermartingale @xmath15 , there exists an admissible portfolio @xmath16 , which satisfies @xmath17 for all @xmath18 $ ] . given a finite sequence of deterministic moments @xmath19
, we study the problem of an economic agent , who is liable to make a series of payments @xmath20 at dates @xmath21 , where for each @xmath22 , @xmath23 is @xmath24-measurable and satisfies @xmath25<\infty$ ] .
this may model for example the cashflows associated to a variable annuity insurance contract , or to a long - term investment project .
the portfolio of the agent , whose value is denoted by @xmath26 is effectively used to make the payments , and therefore satisfies @xmath27 .
we assume that the agent has some tolerance for loss , which means that the negative part of @xmath28 must be bounded in a certain probabilistic sense ( to be made precise later in this section ) . to work with self - financing portfolios we introduce the portfolio augmented with cumulated cash flows : @xmath29 as well as the benchmark process @xmath30 the agent is therefore interested in finding the cheapest portfolio process @xmath31 which outperforms , in a certain probabilistic sense , at dates @xmath21 , the benchmark process @xmath32 .
due to our market completeness assumption , this is equivalent to finding the @xmath13-supermartigale @xmath33 with the smallest initial value which dominates the benchmark in a probabilistic sense at dates @xmath21 .
introduce a discrete filtration @xmath34 defined by @xmath35 . for a @xmath11-supermartingale @xmath33 ,
the discrete - time process @xmath36 defined by @xmath37 is a @xmath38-supermartingale , and conversely , from a @xmath38-supermartingale @xmath39 one can easily construct a @xmath40-supermartingale @xmath41 which coincides with @xmath39 at dates @xmath42 .
therefore , our problem can be reformulated in the discrete time setting as the problem of finding a discrete - time @xmath13-supermartingale @xmath41 with respect to the filtration @xmath43 with the smallest initial value , such that for @xmath44 , @xmath45 dominates in a probabilistic sense the discrete - time benchmark @xmath46 ( we use the same letter for discrete and continuous - time benchmark ) .
there are many natural ways of introducing risk tolerance into our multiperiod hedging problem . in this paper ,
following @xcite , we consider that a bound is imposed on the expected shortfall , weighted by a loss function @xmath47 . throughout this paper
, we will always assume that the following condition on @xmath48 is satisfied .
[ assumption 1 ] the function @xmath47 is convex , decreasing and bounded from below .
the above assumptions are natural to describe a loss function .
a typical example will be a `` call '' function .
in particular , when @xmath49 , we take the positive part of the loss and the situations when the liability is hedged will not be penalized . in some cases , we need stronger assumptions to get explicit results . the following assumption
[ assumption 2 ] allows us to include widely - used utility functions .
for example , when @xmath50 with @xmath51 , then assumption [ assumption 2 ] is satisfied .
[ assumption 2 ] the function @xmath52 is strictly convex , strictly decreasing , bounded from below and of class @xmath53 .
in addition , the derivative @xmath54 satisfies the inada s condition @xmath55 and @xmath56 .
we next describe the three different constraints for our problem as follows .
[ [ european - style - constraint ] ] european - style constraint + + + + + + + + + + + + + + + + + + + + + + + + + * find the minimal value of @xmath57 such that there exits a @xmath13-supermartingale @xmath58 with @xmath59 \leq \alpha_k\ \text{for}\ k=1,\dots , n.\label{consteu}\end{aligned}\ ] ] we denote the set of all @xmath13-supermartingales satisfying by @xmath60 . [ [ time - consistent - constraint ] ] time - consistent constraint + + + + + + + + + + + + + + + + + + + + + + + + + + * find the minimal value of @xmath57 such that there exits a @xmath13-supermartingale @xmath58 with @xmath61 \leq \alpha_k,\,\ , a.s . \,\
k=1,\dots , n.\label{consttc}\end{aligned}\ ] ] we denote the set of all @xmath13-supermartingales satisfying by @xmath62 .
the time - consistent constraint has an interesting interpretation as an `` american - style '' guarantee .
[ prop american ] let @xmath41 be an @xmath43-adapted process .
then the condition @xmath63\leq \alpha_k , \ , a.s .
\quad k=1,\cdots , n\ ] ] is equivalent to : for all @xmath43-stopping times @xmath64 and @xmath6 taking values in @xmath65 such that @xmath66 , @xmath67\leq 0.\ ] ] let @xmath68 , @xmath69 .
the condition is equivalent to the fact that the process @xmath70 is a @xmath14-supermartingale .
equation then follows from doob s theorem .
conversely , assume that holds and let @xmath71 for @xmath72 .
taking @xmath73 and @xmath74 , we get from , @xmath75\leq 0.\ ] ] since @xmath76 and @xmath77 are arbitrary , this proves the supermartingale property of @xmath70 . [
[ lookback - style - constraint ] ] lookback - style constraint + + + + + + + + + + + + + + + + + + + + + + + + + * find the minimal value of @xmath57 such that there exits a @xmath13-supermartingale @xmath58 with @xmath78 \leq 0.\label{constlb}\end{aligned}\ ] ]
we denote the set of all @xmath13-supermartingales satisfying by @xmath79 .
[ [ relationship - between - different - types - of - constraints ] ] relationship between different types of constraints + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + for a given family of bounds @xmath80 , we denote by @xmath81 , @xmath82 and @xmath83 ( or @xmath84,@xmath85 and @xmath86 when there is no ambiguity ) the infimum value of @xmath57 where @xmath58 belongs to , respectively , @xmath60 , @xmath87 and @xmath79 .
the following proposition shows that @xmath88 [ pro : inclusion ] the following inclusion holds true : @xmath89 .
clearly , @xmath90 .
the inclusion @xmath91 follows from the alternative representation for the time - consistent constraint . when @xmath92 , the three constraint types coincide .
in addition , the value function has an explicit form in some particular cases .
[ n1.prop]let @xmath92 and assume that @xmath93 . 1 .
let assumption [ assumption 2 ] hold true and assume that there exists @xmath94 with @xmath95<\infty$ ] , where @xmath96 . then @xmath97 + \mathbb e^{\mathbb q}[i(\lambda^ * z)],\ ] ] where @xmath98 is the inverse function of @xmath99 and @xmath100 is the unique solution of @xmath101 = \alpha$ ] .
2 . assume that @xmath102 .
then , @xmath103 + l^{-1}(\alpha)\ ] ] where @xmath104 .
the first part is a particular case of proposition [ prop : value function time consistent ] which will be shown below .
this result is also very similar to the solution of the classical concave utility maximization problem for which we refer the reader for example to ( * ? ? ?
* theorem 2.0 ) .
the second part is obtained by using the jensen s inequality .
the following proposition is a natural generalization of the second case of proposition [ n1.prop ] and describes another situation when the three value functions coincide .
however , we need an extra hypothesis for @xmath105 .
assume that @xmath106 , @xmath107 for @xmath108 and @xmath109 \geq s_k + l^{-1}(\alpha_k)\label{assterm.eq}\end{aligned}\ ] ] for any @xmath110 , where @xmath104 .
then , @xmath111+l^{-1}(\alpha_n).\ ] ] assumption is implied for example by the following assumptions . *
the process @xmath112 is a submartingale . *
the process @xmath113 is a submartingale and @xmath114 for @xmath115 . from proposition [ n1.prop ] , by removing all constraints except the terminal one , we get @xmath116+l^{-1}(\alpha_n).\ ] ] to show the reverse inequality , let @xmath117.\ ] ] by assumption , @xmath118 , which means that @xmath119
. therefore , @xmath120+l^{-1}(\alpha_n).\ ] ] the proof is completed by applying proposition [ pro : inclusion ] .
in this section , we solve our optimization problem for the three types of constraints .
the main method consists of using the dynamic programming principle where the constraints are to be verified at each payment date .
we begin by the european - style constraint which consists of a family of expectation constraints .
denote by @xmath121 the infimum value of @xmath57 where @xmath122 .
the following result characterizes @xmath121 by using a family of @xmath14-supermartingales , each of which corresponds to one expectation constraint .
[ pro : first condition v0 ] @xmath121 equals the infimum value of @xmath57 such that there exist a @xmath13-supermatingale @xmath123 and a family of @xmath14-supermartingales @xmath124 , @xmath110 , satisfying @xmath125 and @xmath126 .
denote by @xmath127 the set of all @xmath128 which satisfy the conditions in the proposition and by @xmath129 the infimum value of the related @xmath57 . if @xmath130 , then @xmath131\geq\mathbb e^{\mathbb p}[l(m_k - s_k)].\ ] ] therefore @xmath122 , which implies @xmath132 .
conversely , given a @xmath133-supermartingale @xmath122 , we construct processes @xmath124 such that @xmath125 and @xmath134 $ ] for @xmath135 .
the condition @xmath136\leq\alpha_k$ ] implies that @xmath124 is a @xmath14-supermartingale .
it follows that the inverse inequality @xmath137 holds true as well . for the dynamic solution
, we will introduce the value - function process @xmath138 for @xmath110 .
note that , instead of a family of real - valued bounds @xmath80 for @xmath139 , the value function @xmath138 will take as parameters a family of random variables which describe the evolution of the successive bounds on the expected shortfall .
for @xmath140-measurable random variables @xmath141 , let @xmath142 be the essential infimum of all @xmath143 such that there exists a @xmath13-supermartingale @xmath144 with @xmath145\leq n^t$ ] for any @xmath146 .
we denote by @xmath147 the set of all such @xmath13-supermartingales which start from @xmath77 . by convention ,
let @xmath148 and @xmath149 coincide with the set of all @xmath150-measurable random variables .
we have in particular @xmath151 : \mathbb e^{\mathbb p}[l(m_n - s_n)|\mathcal f_{n-1}]\leq n^n\}\\ & = \mathbb e^{\mathbb q}[s_n|\mathcal f_{n-1 } ] + \text{essinf}_{m\in\mathcal f_n}\{\mathbb e^{\mathbb q}[m|\mathcal f_{n-1 } ] : \mathbb e^{\mathbb p}[l(m)|\mathcal f_{n-1}]\leq n^n\}. \ ] ] in fact , if @xmath152 is a @xmath13-supermartingale such that @xmath153\leq n^n$ ] , then @xmath154,m_n)$ ] is also a @xmath13-martingale satisfying this condition and one has @xmath155 $ ] .
similarly to proposition [ pro : first condition v0 ] , we can prove that @xmath142 equals the essential infimum of @xmath143 such that there exists a @xmath13-supermartingale @xmath144 together with a family of @xmath14-supermartingales @xmath156 such that @xmath157 and that @xmath158 for any @xmath146 .
[ lem : filtrant ] let @xmath159 , then there exists a supermartingale @xmath160 such that @xmath161 . by definition , @xmath145\leq n^t$ ] and @xmath162\leq n^t$ ] for any @xmath146 .
let @xmath76 be the set @xmath163 , which is @xmath140-measurable .
for any @xmath164 , let @xmath165 .
since @xmath71 , @xmath166 is a @xmath13-supermartingale .
moreover , @xmath167 and @xmath168= \indic_a\mathbb e^{\mathbb p}[l(m_t'-s_t)|\mathcal f_k]+\indic_{a^c}\mathbb e^{\mathbb p}[l(m_t - s_t)|\mathcal f_k]\leq n^t,\ ] ] therefore @xmath160 .
the following result characterizes the value - function process in a backward and recursive form .
[ eu main result ] @xmath142 equals the essential infimum of all @xmath169 such that there exist @xmath170 and a family of @xmath14-supermartingales @xmath171 , @xmath172 , @xmath173 which satisfy the following conditions : @xmath174=m_k,\\ n_k^i = n^i \text { for $ i\in\{k+1,\ldots , n\}$},\\ n_{k+1}^{k+1}\geq l(m_{k+1}-s_{k+1}),\\ m_{k+1}\geq v_{k+1}(n_{k+1}^{k+2},\ldots , n_{k+1}^n ) . \end{cases}\ ] ] by definition , the theorem is true for @xmath175 . in the following ,
we treat the case @xmath176 by induction .
denote by @xmath177 the essential infimum defined in the theorem .
let @xmath178 be an @xmath179-measurable random variable and @xmath171 , @xmath172 , @xmath173 be @xmath14-supermartingales which verify the conditions in .
note that the fourth condition , together with the essential infimum property , imply ( e.g. ( * ? ? ?
* v.18 ) ) the existence of a sequence of random variables @xmath180 in @xmath181 such that @xmath182 .
moreover , the previous lemma shows that the sequence @xmath180 can be supposed decreasing without loss of generality . by a similar argument as in proposition [ pro : first condition v0 ] , there exist @xmath13-supermartingales @xmath183 and @xmath14-supermartingales @xmath184 , @xmath172 , @xmath185 such that @xmath186 we extend each @xmath187 to a @xmath14-supermartingale on @xmath188 by taking @xmath189 , @xmath190 .
let @xmath191 $ ] for any @xmath192 .
note that the @xmath13-supermartingale @xmath193 and the @xmath14-supermartingales @xmath171 , @xmath194 verify the conditions which characterize @xmath142 .
therefore @xmath195 for any @xmath192 , which implies that @xmath196\leq \mathbb e^{\mathbb q}[m_{k+1}|\mathcal f_t],\ ] ] where the second inequality comes from the relation @xmath182 and the fact that the sequence @xmath180 is decreasing .
hence we obtain @xmath197 .
conversely , let @xmath144 be a @xmath13-supermartingale and @xmath198 be @xmath14-supermartingales such that @xmath199 and that @xmath200 for any @xmath201 . without loss of generality , we assume @xmath202 $ ] .
the conditions in are easily verified for @xmath178 , @xmath198 except the last one .
note that for any @xmath201 , the condition @xmath200 implies @xmath203\geq\mathbb e^{\mathbb p}[l(m_i - s_i)|\mathcal f_{k+1}],\ ] ] where the first inequality comes from the fact that @xmath204 is a @xmath14-supermartingale .
therefore , @xmath205 , which leads to the inverse inequality . the supermartingale condition in the previous theorem can be made simpler where only @xmath179-measurable random variables are concerned .
in particular , we consider the case of real - valued bounds .
[ cor : dynamic ] @xmath206 equals the essential infimum of all conditional expectations @xmath207 $ ] , @xmath208 such that there exist @xmath179-measurable random variables @xmath209 satisfying the following conditions : @xmath210=\alpha_i \text { for } i\in\{k+2,\ldots , n\},\\ \mathbb e^{\mathbb p}[l(m - s_{k+1})|\mathcal f_k]\leq \alpha_{k+1},\\ m\geq v_{k+1}(n^{k+2},\ldots , n^n).\\ \end{cases}\ ] ] denote by @xmath211 the essential infimum defined in the corollary .
we begin with @xmath179-measurable random variables @xmath41 , @xmath212 satisfying the conditions in .
let @xmath213 and @xmath214 $ ] .
we then construct @xmath14-supermartingales @xmath198 with @xmath215 for @xmath201 , @xmath216 and @xmath217 for @xmath218 .
therefore , @xmath219 .
note that the essential infimum @xmath211 remains unchanged if we replace the first condition of by `` @xmath220\leq \alpha_i$ ] _ for any _
@xmath218 '' since the function @xmath221 is descreasing with respect to each coordinate .
let @xmath170 and @xmath198 satisfy the conditions in .
we choose @xmath222 and @xmath223 for @xmath218 . by , @xmath224 and @xmath220\leq n_k^i=\alpha_i$ ] .
moreover , since @xmath225 , one has @xmath226\geq\mathbb
e^{\mathbb p}[l(m - s_{k+1})|\mathcal f_k].\ ] ] hence @xmath227 . for the time - consistent constraint ,
we consider the loss function at each payment date given the market information at the previous time step .
the constraint is written by using conditional expectations .
the dynamic programming structure in this setting is relatively simple since it only involves two successive dates .
recall that @xmath62 is the set of all @xmath13-supermartingales @xmath58 with @xmath61 \leq \alpha_k\ \text{for}\ k=1,\dots , n.\end{aligned}\ ] ] in a dynamic manner , denote by @xmath228 the set of all @xmath13-supermartingales @xmath144 verifying the condition @xmath229\leq\alpha_t\text { for } t = k+1,\ldots , n.\ ] ] note that @xmath230 is the set of all integrable @xmath150-measurable random variables and @xmath231 coincides with @xmath62 . define the value - function process @xmath232 in a backward manner as follows : @xmath148 , and for @xmath110 , let @xmath233 : m\geq v_k,\text { and } \mathbb e^\mathbb p[l(m - s_k)|\mathcal f_{k-1 } ] \leq \alpha_k\}\ ] ] [ pro : tcprogdy ] for any @xmath234 , @xmath235 the case where @xmath236 is trivial . in the following , we assume that the equality @xmath237 has been proved .
let @xmath238 be an element in @xmath239 .
there then exists a @xmath13-supermartingale @xmath240 such that @xmath229\leq\alpha_t\text { for } t = k,\ldots , n,\ ] ] which implies that @xmath241 .
therefore @xmath242 by the induction hypothesis .
combining this with the condition @xmath243 $ ] , we obtain by definition that @xmath244\leq m_{k-1}$ ] .
hence @xmath245 . for the converse inequality ,
let @xmath246 such that @xmath247 and @xmath248\leq\alpha_k$ ] . by the induction hypothesis and an argument similar to lemma [ lem : filtrant ]
, there exists a family @xmath193 of @xmath13-supermartingales indexed by @xmath192 such that @xmath249\leq\alpha_t\text { for } t = k+1,\ldots , n,\ ] ] and @xmath140-measurable sequence @xmath250 is decreasing and @xmath251 . for each @xmath192 , let @xmath252 , @xmath253 $ ] and @xmath254 for @xmath146 . then @xmath255 is a @xmath13-supermartingale .
moreover , @xmath256\leq\alpha_t\text { for } t = k,\ldots , n,\ ] ] which implies @xmath257 . therefore @xmath258\ ] ] for any @xmath192 . by taking the limit when @xmath259 goes to the infinity , one obtains @xmath260 $ ] by @xmath251 . since @xmath41 is arbitrary , @xmath261 . the result is thus proved . by the previous proposition , we can calculate the smallest initial capital @xmath139 by using the recursive formula for the value function . with extra regularity condition on the loss function @xmath48
, we can obtain a more explicit result as follows .
[ prop : value function time consistent ] let @xmath48 be a loss function satisfying assumption [ assumption 2 ] and let @xmath262 be the inverse function of @xmath99 .
suppose in addition that for any @xmath110 , @xmath263 and there exists a strictly negative random variable @xmath264 such that @xmath265<+\infty\,\text { where } z_k = \mathbb e^\mathbb p[\frac{d\mathbb q}{d\mathbb p}\,|\ , \mathcal f_k].\ ] ] then the value function @xmath266 satisfies @xmath267 a.s . for @xmath268 where @xmath269 is given by @xmath270\end{aligned}\ ] ] where @xmath271 is the solution of @xmath272 = \alpha_n,\label{lambdastep1}\end{aligned}\ ] ] and for @xmath273 , @xmath274 \\+ \mathbb e^\mathbb q\left[\left\{s_k - \widehat v_k+ i(\lambda_{k-1 } z_k / z_{k-1})\right\}^+\big| \mathcal f_{k-1}\right]\mathbf 1_{\mathbb e^\mathbb p[l(\widehat v_k - s_k)|\mathcal f_{k-1 } ] >
\alpha_k},\end{gathered}\ ] ] where @xmath275 is the solution of @xmath276 = \alpha_k.\end{aligned}\ ] ] @xmath277 + _ step 1 .
existence of @xmath278 .
_ for any @xmath110 ,
let @xmath279 be the set of strictly negative @xmath280-measurable random variables @xmath281 such that @xmath282 \leq \alpha_k.\end{aligned}\ ] ] by using the dominated convergence theorem , the condition @xmath283 and assumption imply that the family @xmath279 is not empty .
let @xmath284 be the essential infimum of this family .
note that @xmath279 is stable by taking the infimum of finitely many random variables .
therefore @xmath284 can be written as the limit of a decreasing sequence in @xmath279 .
the continuity of the functions @xmath98 and @xmath48 , together with the monotone convergence theorem , show that @xmath284 lies in the family @xmath279 .
it remains to show the equality .
if the equality does not hold , then for sufficiently small @xmath285 , the set @xmath286 of @xmath287 such that @xmath288(\omega ) \leq \alpha_k\ ] ] has a strictly postive measure .
this implies that @xmath289 also lies in @xmath279 , which leads to a contradiction .
_ step 2 . representation for @xmath290 . _ by proposition [ pro : tcprogdy ] , @xmath291+\essinf_{m\in \mathcal f_n}\{\mathbb e^{\mathbb q } [ m|\mathcal f_{n-1 } ] : \mathbb e^{\mathbb p}[l(m)|\mathcal f_{n-1}]\leq \alpha_n\}\ ] ] taking @xmath292 , we see that @xmath293 . on the other hand , for every strictly negative random variable @xmath294 , @xmath295 where @xmath296 is the legendre transformation of @xmath48 .
note that one has @xmath297 .
so @xmath298\\ & \geq \essinf_{m\in \mathcal f_n}\{\mathbb e^{\mathbb q } [ m|\mathcal f_{n-1 } ] : \mathbb e^{\mathbb p}[l^*(\lambda z_n / z_{n-1 } ) + \lambda m z_n / z_{n-1}|\mathcal f_{n-1}]\leq \alpha_n\}\\ & = \essinf_{m\in \mathcal f_n}\{\mathbb e^{\mathbb q } [ m|\mathcal f_{n-1 } ] : \mathbb e^{\mathbb p}[l^*(\lambda z_n / z_{n-1})|\mathcal f_{n-1 } ] + \lambda \mathbb e^{\mathbb q}[m |\mathcal f_{n-1}]\leq \alpha_n\}\\ & = \frac{\alpha_n -\mathbb e^{\mathbb p}[l^*(\lambda z_n / z_{n-1 } ) |\mathcal f_{n-1 } ] } { \lambda}\\ & = \frac{\alpha_n -\mathbb e^{\mathbb p}[l(i(\lambda z_n / z_{n-1 } ) ) |\mathcal f_{n-1 } ] } { \lambda } + \mathbb e^{\mathbb q } [ i(\lambda z_n / z_{n-1 } ) |\mathcal f_{n-1 } ] , \ ] ] where the last equality comes from the relation @xmath299 .
taking @xmath278 to be the solution of , we find that @xmath300 , which proves the desired representation of @xmath290 .
general case .
_ we now proceed to the proof of the general case by induction on @xmath77 .
assume that we have already established the equality @xmath301 .
by proposition [ pro : tcprogdy ] and this induction hypothesis , @xmath302\,:\,m\geq\widehat v_k,\ ; \mathbb e^{\mathbb p}[l(m - s_k)|\mathcal f_{k-1}]\leq\alpha_k\}.\ ] ] we choose @xmath303 > \alpha_k},\ ] ] which is bounded from below by @xmath304 and satisfies @xmath305\leq\alpha_k$ ] .
in fact , it is clear that this inequality holds on the set @xmath306\leq \alpha_k\}$ ] .
moreover , on the set @xmath306
> \alpha_k\}$ ] , one has @xmath307 , so @xmath308=\mathbb e^{\mathbb p}[l(i(\lambda_{k-1}z_k / z_{k-1})\vee ( \widehat{v}_k - s_k))|\mathcal f_{k-1}]\]]which implies @xmath305\leq\alpha_k$ ] on this set by the definition of @xmath284 .
hence @xmath309=\widehat{v}_{k-1}.\ ] ] the opposite inequality @xmath310 is straightforward on the set @xmath306\leq \alpha_k\}$ ] .
it remains to establish the inequality on the set @xmath306
> \alpha_k\}$ ] .
we have by a variable change @xmath311\\ & = \essinf_{}\{\mathbb e^{\mathbb q}[m|\mathcal f_{k-1}]:m\geq 0,\ ; \mathbb e^{\mathbb p}[l(m+\widehat{v}_k - s_k)|\mathcal f_{k-1}]\leq\alpha_k\},\\ & \geq\essinf_{m\in\mathcal f_k , m\geq 0}\{\mathbb e^{\mathbb q}[m|\mathcal f_{k-1}]:\mathbb e^{\mathbb p}[l^*\big(\lambda { z_{k}\over z_{k-1}}\big)+\lambda(m+\widehat{v}_k - s_k ) { z_{k}\over z_{k-1}}\big|\mathcal f_{k-1}]\leq\alpha_k\}\\ & = \essinf_{m\in\mathcal f_k , m\geq 0}\{\mathbb e^{\mathbb q}[m|\mathcal f_{k-1}]:\mathbb e^{\mathbb p}[l^*\big(\lambda { z_{k}\over z_{k-1}}\big)|\mathcal f_{k-1}]+\lambda\mathbb e^{\mathbb q}[m+\widehat{v}_k - s_k|\mathcal f_{k-1}]\leq\alpha_k\}\\ & = \essinf_{m\in\mathcal f_k , m\geq 0}\big\{\mathbb e^{\mathbb q}[m|\mathcal f_{k-1}]:\\&\qquad \mathbb e^{\mathbb q}[m|\mathcal f_{k-1}]\geq\frac{\alpha_k-\mathbb e^{\mathbb p}[l(i(\lambda { z_k\over z_{k-1}}))|\mathcal f_{k-1}]}{\lambda } + \mathbb e^{\mathbb q}[s_k-\widehat{v}_k+ i\big(\lambda { z_k\over z_{k-1}}\big)\big|\mathcal f_{k-1}]\big\ } \end{split}\ ] ] for any strictly negative @xmath312 .
on the set @xmath306 > \alpha_k\}$ ] , the last constraint above can be replaced by @xmath313\geq\frac{\alpha_k-\mathbb e^{\mathbb p}[l(i(\lambda z_k / z_{k-1})\vee(\widehat{v}_k - s_k))|\mathcal f_{k-1}]}{\lambda}\\+\mathbb e^{\mathbb q}[\big(s_k-\widehat{v}_k+ i(\lambda z_k / z_{k-1})\big)+|\mathcal f_{k-1}].\end{gathered}\ ] ] now we choose @xmath278 to be the solution of to obtain the following constraint @xmath314\geq \mathbb e^{\mathbb q}[\big(s_k-\widehat{v}_k+i(\lambda_{k-1}z_k / z_{k-1})\big)+|\mathcal f_{k-1}].\ ] ] therefore on the set
@xmath306 > \alpha_k\}$ ] , we have @xmath315\geq \mathbb e^{\mathbb q}[\big(s_k-\widehat{v}_k+i(\lambda_{k-1}z_k / z_{k-1})\big)^+|\mathcal f_{k-1}],\ ] ] which implies the inequality @xmath310 .
the theorem is thus proved . in the risk - neutral case
where @xmath106 , we can relax assumption [ assumption 2 ] and obtain a similar result under assumption [ assumption 1 ] only .
[ time_consistent_riskneutral : propo ] we assume that @xmath102 and that for any @xmath110 , @xmath263 . 1 .
the value function satisfies @xmath267 a.s .
for @xmath268 where @xmath316 + l^{-1}(\alpha_n)\\ \widehat v_{k-1 } & = \mathbb e [ \widehat v_k|\mathcal f_{k-1 } ] \\ & + \mathbb e\left[\left\{s_k - \widehat v_k+ \lambda_k\right\}^+\big| \mathcal f_{k-1}\right]\mathbf 1_{\mathbb e[l ( \widehat v_k - s_k)|\mathcal f_{k-1 } ] > \alpha_k},\ , k < n,\end{aligned}\ ] ] where @xmath317 is the solution of @xmath318 = \alpha_k.\label{eqlambda}\end{aligned}\ ] ] 2 .
if the loss function @xmath319 , with @xmath320 for all @xmath77 , then the value function satisfies @xmath267 a.s .
for @xmath268 where @xmath321 - \alpha_n\\ \widehat v_{k-1 } & = \max\{\mathbb e[\widehat v_k|\mathcal f_{k-1 } ] , \mathbb e[\widehat v_k \vee s_k - \alpha_k|\mathcal f_{k-1}]\}. \end{aligned}\ ] ] 1 .
the formula for @xmath290 follows directly from jensen s inequality .
we prove the general case by induction .
assume that we have already established the equality @xmath322 . to prove the formula for @xmath323
, we first follow step 1 in the proof of proposition [ prop : value function time consistent ] to establish the existence of @xmath324 . next , let @xmath325>\alpha_k}.\ ] ] clearly , @xmath247 .
in addition , @xmath326 & = \mathbb e[l(\widehat v_k -s_k + \left\{s_k - \widehat v_k + \lambda_k\right\}^+\mathbf 1_a ) |\mathcal f_{k-1}]\\ & = \mathbf 1_a \mathbb e[l(\lambda\vee(\widehat v_k -s_k ) ) |\mathcal f_{k-1 } ] + \mathbf 1_{a^c } \mathbb e[l(\widehat v_k -s_k ) |\mathcal f_{k-1}]\\ & \leq \alpha_k,\end{aligned}\ ] ] where we denote @xmath327>\alpha_k\}\in \mathcal f_{k-1}$ ] .
this proves that @xmath328 .
+ the opposite inequality only needs to be shown on the set @xmath76 . as in the proof of proposition [ prop : value function time consistent ] , we have , @xmath329 = \essinf_{m\in \mathcal f_k , m\geq 0}\{\mathbb e[m|\mathcal f_{k-1 } ] : \mathbb e[l(m+\widehat{v}_k - s_k)|\mathcal f_{k-1}]\leq\alpha_k\}.\ ] ] let @xmath41 be any random variable satisfying the constraints . then , by convexity , @xmath330 \\
& \leq \mathbb e[l'(\lambda_k \vee ( \widehat v_k - s_k ) ) ( \lambda_k \vee ( \widehat v_k - s_k)- m-\widehat v_k + s_k)|\mathcal f_{k-1 } ] \\ & = \mathbb e[\mathbf 1_{\lambda_k \geq \widehat v_k - s_k}l'(\lambda_k ) ( \lambda_k -m - \widehat v_k + s_k)-\mathbf
1_{\lambda_k < \widehat v_k - s_k}l'(\widehat v_k - s_k ) m|\mathcal f_{k-1}]\\ & = l'(\lambda_k)\mathbb e [ ( \lambda_k - \widehat v_k + s_k)^+-m|\mathcal f_{k-1 } ] -\mathbb e[(l'(\widehat v_k - s_k)-l'(\lambda_k))^+ m|\mathcal f_{k-1}],\end{aligned}\ ] ] where @xmath331 denotes any number belonging to the subdifferential of the convex function @xmath48 at the point @xmath332 , defined by @xmath333 since @xmath334 , this implies @xmath335 \geq 0.\ ] ] by the assumption of the proposition , on the set @xmath336 , @xmath337 and @xmath324 may not be the solution of equation .
therefore , @xmath338 almost surely , and @xmath339 \leq 0\ ] ] on @xmath76 , which finishes the proof .
2 . the formula for @xmath290 follows from part 1 . when @xmath340\leq \alpha_k$ ] , @xmath341\geq \mathbb e[v_k\vee s_k - \alpha_k|\mathcal f_{k-1}]$ ] , and the formulas hold true .
assume then that @xmath342>\alpha_k$ ] .
first we observe that @xmath343 a.s .
, otherwise on the set where @xmath344 , @xmath345 and the conditional expectation can not be equal to @xmath346 .
therefore , @xmath347 = \mathbb e[s_k|\mathcal f_{k-1 } ] + \mathbb e[\lambda_k \vee ( v_k - s_k)|\mathcal f_{k-1 } ] \\ & = e[s_k|\mathcal f_{k-1 } ] + \mathbb e[(\lambda_k \vee ( v_k - s_k))^+|\mathcal f_{k-1 } ] - \mathbb e[(\lambda_k \vee ( v_k - s_k))^-|\mathcal f_{k-1}]\\ & = e[s_k|\mathcal f_{k-1 } ] + \mathbb e [ ( v_k - s_k)^+|\mathcal f_{k-1 } ] - \mathbb e[l(\lambda_k \vee ( v_k - s_k))|\mathcal f_{k-1}]\\ & = \mathbb e [ v_k \vee s_k|\mathcal f_{k-1 } ] - \alpha_k.\end{aligned}\ ] ] the lookback constraint involves the maximum value of the loss function during the whole period .
so the dynamic programming structure takes into account this variable .
recall that @xmath79 denotes the set of all @xmath13-supermartingales @xmath58 with @xmath348 \leq 0.\ ] ] define @xmath349 , where @xmath350 is a pair of @xmath140-measurable random variables , to be the essential infimum of all @xmath143 such that there exist a @xmath13-supermartingale @xmath144 and a @xmath14-supermartingale @xmath351 verifying @xmath352 we denote the set of all such @xmath13-supermartingales @xmath144 by @xmath353 . by convention , let @xmath354 @xmath355 let @xmath58 and @xmath356 be respectively @xmath13 and @xmath14 supermartingales which verify with @xmath357
. then @xmath358 therefore @xmath359\geq\mathbb e^{\mathbb p}[\max_{k\in\{1,\ldots , n\}}\{l(m_k - s_k)-\alpha_k\}].\ ] ] hence @xmath360 . therefore @xmath361 conversely , given a @xmath13-supermartingale @xmath58 in @xmath79 , we can choose @xmath362 and @xmath363 $ ] , @xmath364 and @xmath365 to construct a @xmath14-supermartingale @xmath356 verifying the condition with @xmath357 .
thus @xmath366 .
the result is proved .
[ lookback : recursive ] @xmath349 equals the essential infimum of all @xmath207 $ ] , @xmath208 such that there exists a random variable @xmath367 satisfying@xmath368=n_k,\\ m\geq v_{k+1}\big(n,\max(z_k , l(m - s_{k+1})-\alpha_{k+1})\big ) .
\end{cases}\ ] ] denote by @xmath369 the essential infimum defined in the theorem .
let @xmath41 and @xmath39 be random variables which verify . by an argument similar to lemma [ lem :
filtrant ] , there exists a decreasing sequence of random variables @xmath180 in @xmath370 such that @xmath371 . moreover , there exist @xmath13-supermartingales @xmath183 and @xmath14-supermartingales @xmath372 verifying @xmath373 and @xmath374 . let @xmath191 $ ] and @xmath375 . the processes @xmath193 and @xmath376 are respectively @xmath13 and @xmath14 supermartingales which satisfy the condition .
therefore , @xmath377 for any @xmath192 , which implies @xmath207\geq v_k(n_k , z_k)$ ] .
so @xmath378 .
conversely , let @xmath144 and @xmath379 be respectively @xmath13-supermartingale and @xmath14-supermartingale which verify the condition .
without loss of generality , we may assume @xmath202 $ ] and that @xmath379 is a @xmath14-martingale .
note that @xmath380 can also be written as @xmath381 by the definition of @xmath221 , @xmath382 which shows that @xmath178 and @xmath383 verify the conditions in .
therefore @xmath202\geq\widetilde v_k(n_k , z_k)$ ] .
since @xmath144 is arbitrary , we obtain @xmath384 .
in this section , we apply the dynamic programming results obtained in the previous section under the three types of constraints and we give the explicit form of the smallest hedging portfolio value in the setting of two time steps . in the following , we assume that the function @xmath48 satisfies assumption [ assumption 2 ] .
we have by corollary [ cor : dynamic ] , @xmath386 \text { s.t . }
\mathbb e^{\mathbb p}[l(m - s_2)|\mathcal f_1]\leq \alpha_2\}\ ] ] by proposition [ prop : value function time consistent ] , @xmath387+\mathbb e^{\mathbb q}[i(c(\alpha_2)z_2/z_1)|\mathcal f_1],\ ] ] where @xmath388 is the @xmath389-measurable random variable such that @xmath390=\alpha_2.\ ] ] and @xmath391.\ ] ] we then use again to compute the value of @xmath392 , which identifies with the essential infimum of @xmath393 $ ] , @xmath394 such that @xmath395\leq \alpha_1 $ ] and that there exists @xmath396 verifying @xmath397=\alpha_2 $ ] and @xmath398 .
namely , @xmath399\,:\ , \mathbb e^{\mathbb p}[l(m - s_1)]\leq \alpha_1,\,\mathbb e^{\mathbb p}[v_1^{-1}(m)]\leq \alpha_2\}.\ ] ] consider the loss function where @xmath50 , with @xmath51 .
direct computations yield @xmath400 , @xmath401 and @xmath402
. therefore , @xmath403 from which we deduce @xmath404-\frac{\log(\alpha_2 + 1)}{p}-\frac 1p{\mathbb e^{\mathbb p}[z_2/z_1\log(z_2/z_1)|\mathcal f_1]}.\ ] ] denote by @xmath405}.\ ] ] from we have that @xmath406-d_1).\ ] ] hence @xmath392 is the infimum of the expectations
@xmath393=\mathbb e^{\mathbb p}[z_1m]$ ] , where @xmath394 such that @xmath407\leq \alpha_1\;\text { and } \ ; \mathbb e[l(m-\mathbb e^{\mathbb q}[s_2|\mathcal f_1]-d_1)]\leq \alpha_2.\ ] ] we will use lagrange multiplier method ( with kuhn - tucker conditions ) to study this problem . if @xmath394 realizes the infimum of @xmath393 $ ] subjected to the constraints , then @xmath41 verifies the first order condition @xmath408-d_1)}=0,\ ] ] where @xmath278 and @xmath409 are respectively positive solutions to the equations @xmath410-s_1+d_1)))}\bigg]=1+\alpha_1\ ] ] and @xmath411+d_1)+\mu)}\bigg]=1+\alpha_2.\ ] ] we introduce the notation @xmath412-s_1+d_1))$ ] and rewrite the above equations as @xmath413=p(1+\alpha_1)\;\quad\text{and}\quad\;\mathbb e^{\mathbb q}\big[\frac{x}{\lambda+\mu x}\big]=p(1+\alpha_2),\ ] ] from which we obtain @xmath414 by taking a linear combination .
moreover , the quotient of the two equations shows that @xmath415 verifies the following equation ( provided that @xmath416 ) @xmath417 where @xmath418 $ ] .
note that @xmath419,\qquad \lim_{t\rightarrow+\infty}\frac{f(t)}{1-tf(t)}=\mathbb e^{\mathbb q}[x]^{-1}.\ ] ] * suppose that @xmath420>\frac{1+\alpha_1}{1+\alpha_2}>\mathbb e^{\mathbb q}[x]^{-1}.\ ] ] the two constraints in are saturated .
we can use numerical methods to compute the value of @xmath415 and then use the relation to find explict value of @xmath278 and @xmath409 . *
if @xmath421^{-1},\ ] ] then only the first constraint in is saturated and one should take @xmath422 . in this case the first order condition becomes @xmath423 with @xmath424 .
therefore , one has @xmath425 thus we obtain @xmath426-\frac{1}{p}\log(1+\alpha_1)-\frac{1}{p } \mathbb e^{\mathbb p}[z_1\log z_1].\ ] ] * if @xmath427,\ ] ] then only the second constraint in is saturated , and one has @xmath428 . in this case the first order condition becomes @xmath429-d_1)}=0\ ] ] with @xmath430 .
therefore , one obtains @xmath431+d_1-\frac{1}{p}\log((1+\alpha_1)z_1)\\ & = \mathbb e^{\mathbb q}[s_2|\mathcal f_1]-\frac{1}{p}\mathbb e^{\mathbb p}[z_2/z_1\log ( z_2/z_1)|\mathcal f_1]-\frac{1}{p}\log((1+\alpha_1)z_1 ) .
\end{split}\ ] ] finally , we obtain @xmath432-\frac{1}{p}\log(1+\alpha_2)-\frac 1p\mathbb e^{\mathbb p}[z_2\log z_2].\ ] ] the computation of @xmath433 is identical to in the european case : @xmath434 . by proposition
[ pro : tcprogdy ] , @xmath435 : \mathbb e^{\mathbb p}[l(m - s_1)]\leq \alpha_1 , v_1^{-1}(m ) \leq \alpha_2\}.\ ] ] from as in the european case , @xmath436-\frac{\log(\alpha_2 + 1)}{p } + d_1.\ ] ] by proposition [ prop : value function time consistent ] , @xmath139 may be found using the following algorithm : * compute @xmath437 $ ] . *
if @xmath438 \leq \alpha_1 $ ] then the constraint at date @xmath439 is not binding , and the solution is given by @xmath440 $ ] . *
if @xmath438 > \alpha_1 $ ] then the constraint at date @xmath439 is binding and we proceed as follows : * * compute @xmath441 by solving the equation @xmath442 = \alpha_1.\ ] ] * * the solution is given by @xmath443 + \mathbb e^{\mathbb q}[\{s_1 - v_1 + i(\lambda z_1)\}^+].\ ] ] by proposition [ lookback : recursive ] , @xmath444 : \mathbb e^{\mathbb p}[\max(z , l(m - s_2 ) - \alpha_2)|\mathcal f_1 ] < y\right\}\ ] ] so the value function @xmath445 coincides with as in the european and the american cases .
we then have the minimal capital at the initial time as @xmath446 : \exists
n \in \mathcal f_1 : \mathbb e^{\mathbb p}[n]=0 , m\geq v_1(n , l(m - s_1)-\alpha_1)\}.\ ] ] however , it is more difficult to obtain explicit results as ou for the lookback - style constraint . as we have already mentioned , in the three cases , the value functions @xmath433 at time @xmath447 coincide . in the risk - neutral case
where @xmath106 , we have @xmath448 , so @xmath449+l^{-1}(\alpha_2).\ ] ] for the initial value at time @xmath2 , we have the following results : * _ european - style constraint : _ @xmath450 : \mathbb e[l(m - s_1)]\leq \alpha_1 , \mathbb e[l(m-\mathbb e[s_2|\mathcal f_1])]\leq \alpha_2\}\ ] ] * _ american - style constraint : _ @xmath451 : \mathbb e[l(m - s_1)]\leq \alpha_1 , l(m-\mathbb e[s_2|\mathcal f_1])\leq \alpha_2\}\ ] ] * _ lookback - style constraint : _ @xmath452 : \mathbb e[\max\ { l(m - s_1)-\alpha_1 , l(m-\mathbb e[s_2|\mathcal f_1])-\alpha_2\}]\leq 0\}.\ ] ] in the risk - neutral case , the results hold under assumption [ assumption 1 ] instead of assumption [ assumption 2 ] .
we are particularly interested in the loss function @xmath453 where explicit results are obtained , thanks to lemma [ 2dim ] in the appendix ( for the time - consistent constraint the result follows by taking @xmath454 and @xmath455 - \alpha_2 $ ] ) : @xmath456-\alpha_1,\mathbb e[s_2]-\alpha_2,\mathbb e[(s_1\vee\mathbb e[s_2|\mathcal f_1])]-\alpha_1-\alpha_2\big),\\ v_0^{tc } & = \max\big(\mathbb e[s_2 ] - \alpha_2 , \mathbb e[s_1 \vee ( \mathbb e[s_2|\mathcal f_1 ] - \alpha_2)]-\alpha_1\big),\\ v_0^{lb } & = \mathbb e[\max\big(s_1 - \alpha_1,(\mathbb e[s_2|\mathcal f_1 ] - \alpha_2)\big ) ] .
\end{aligned}\ ] ] we compare the cost for hedging two objectives under the three probabilistic constraints in a numerical example . in figure [ exploss.fig ] , we consider the risk - neutral case where @xmath102 and the loss function is given by @xmath457 .
the model is as follows : @xmath458 and @xmath459 where @xmath460 , @xmath461 and @xmath462 and @xmath463 are standard normal random variables with correlation @xmath464 .
we fix the loss tolerance at the first date to be @xmath465 and plot the hedging value @xmath139 for three different constraint styles as a function of the loss tolerance at the second date @xmath466 .
not surprisingly , all the three curves are decreasing w.r.t . the constraint level @xmath466
moreover , among the three constraints we consider , the european constraint corresponds to the lowest hedging cost and the lookback constraint to the highest one , which is coherent with proposition [ pro : inclusion ] .
finally , the cost of almost sure hedging is higher than the three probabilistic constraints . in this case.,scaledwidth=80.0% ]
for the multi - objective hedging problem , it is in general difficult to get explicit solutions by using the recursive formulas . in this section ,
we first present a situation where the explicit solution may be obtained for the expected loss and then extend our framework to discuss the conditional value at risk . our first result deals with the expected loss constraint given as a call function under a risk - neutral probability .
[ incloss.cor]let @xmath49 .
assume that @xmath106 , @xmath467 and that the process @xmath468 is non - decreasing . then , @xmath469 - \alpha_k\}\label{exploss}\end{aligned}\ ] ] for any @xmath470
, we have @xmath471\leq\alpha_k,\,k=1,\dots , n\ } \\ & \geq \inf_{m\in \mathcal f_k}\ { \mathbb e[m ] : \mathbb e[(s_k - m)^+]\leq \alpha_k\ } = \mathbb e[s_k ] - \alpha_k,\end{aligned}\ ] ] and therefore , @xmath472 - \alpha_k\}\ ] ] on the other hand , jensen s inequality yields @xmath473 \leq \mathbb e[(s_k - m_n)^+]$ ] , so @xmath474\leq \alpha_k,\,k=1,\dots , n\ } \\ & = \inf_{m\in \mathcal f_n}\ { \mathbb e[m ] : \mathbb e[(s_k - m)^+]\leq \alpha_k,\,k=1,\dots , n\}.\end{aligned}\ ] ] since @xmath475 is non - decreasing , then follows by lemma [ exploss.lm ] in appendix .
we have up to now considered risk constraints given by ( conditional ) expected loss . in the literature and in practice ,
the risk constraint is often described by using risk measures .
let us now extend our framework to cover loss constraints expressed in terms of the conditional value at risk .
we once again assume that @xmath476 and denote by @xmath477 the set of all supermartingales @xmath58 such that @xmath478\leq\alpha_k$ ] for any @xmath110 , where @xmath479 is the risk measure defined by @xmath480 : \frac{d \tilde{\mathbb p}}{d\mathbb p } \leq \frac{1}{1-\lambda}\},\label{cvarproba}\end{aligned}\ ] ] where @xmath481 represents the loss .
it is also well known to satisfy @xmath482\ ] ] and @xmath483 + z\right\}.\end{aligned}\ ] ] the minimum above is attained , so that one can write @xmath484 + z\right\}.\label{varexp}\end{aligned}\ ] ] moreover , the simplified formula @xmath485\ ] ] holds whenever @xmath70 has continuous distribution .
we define @xmath486 let us start with some simple observations : * from one deduces immediately that @xmath487 $ ] , which means ( see the proof of proposition [ incloss.cor ] ) that @xmath488 - \alpha_k\}.\label{simpleobs}\end{aligned}\ ] ] * for @xmath92 , the choice @xmath489 attains the optimal value in . but in this case @xmath490 = \alpha_1 $ ] , which implies that the lower bound above is also attained : @xmath491 - \alpha_1.\end{aligned}\ ] ] * since in our case the loss is always positive , the equality may be simplified as @xmath492 + z\right\}.\end{aligned}\ ] ] the following representation reduces the problem to that of hedging under expected loss constraints . [ glemma ] @xmath493 where @xmath494 is a shorthand for @xmath495 $ ] , @xmath496 and @xmath497 + z_k \leq \alpha_k , \,k=1\ldots n\}\end{gathered}\ ] ] for every @xmath498 fixed , the set @xmath477 contains the set of supermartingales @xmath58 satisfying @xmath499 + z_k \leq \alpha_k,\ , k\in\{1\ldots n\}$ ] , which shows that @xmath500 on the other hand , for any @xmath501 , there exists an @xmath43-supermartingale @xmath58 satisfying @xmath502\leq \alpha_k,\ , k\in \{1,\dots , n\}$ ] and such that @xmath503 since the infimum in is attained , we can also find @xmath498 such that @xmath499 + z_k \leq \alpha_k,\ , k\in\{1\ldots n\}$ ] .
but this means that @xmath504 which completes the proof since @xmath505 is arbitrary .
we close this section by presenting a particular case when the solution of problem is completely explicit .
if @xmath506 is an increasing process , then @xmath507 - \alpha_n.\ ] ] if the process @xmath508 is increasing , by proposition [ incloss.cor ] , @xmath509 - ( 1-\lambda ) \alpha_k - \lambda z_k\ } = \mathbb e[s_n ] - ( 1-\lambda)\alpha_n - \lambda z_n.\ ] ] observe now that @xmath510 - \alpha_n.\ ] ] combining this with , the proof is complete .
[ 2dim ] let @xmath511 be a probability space , @xmath512 and @xmath513 .
then , @xmath514 : \mathbb e[(x - m)^+]\leq \alpha,\mathbb e[(y - m)^+]\leq \beta \ } \\= \max\{\mathbb e[x-\alpha ] , \mathbb e[y-\beta ] , \mathbb e[x\vee y - \alpha - \beta]\}. \end{gathered}\ ] ] and @xmath515 : \mathbb e[\max((x - m)^+- \alpha,(y - m)^+- \beta)]\leq 0 \ } \\= \mathbb e[(x-\alpha)\vee ( y - \beta ) ] . \end{gathered}\ ] ] _ first part . _ by considering only one of the two constraints , we have clearly @xmath516 , \mathbb e[y-\beta]\}.\ ] ] in addition , @xmath517 and so @xmath518.\ ] ] to finish the proof , we need to find @xmath519 which satisfies the constraint and realizes the equality .
assume first that @xmath520 \geq \mathbb e[x-\alpha]\quad \text{et}\quad \mathbb e[x\vee y - \alpha - \beta ] \geq \mathbb e[y-\beta].\ ] ] this is equivalent to @xmath521\geq \alpha\quad \text{and}\quad \mathbb e[(y - x)^+]\geq \beta.\ ] ] in this case we can take @xmath522 } - \beta \frac{(y - x)^+}{\mathbb e[(y - x)^+]}\ ] ]
if @xmath523\geq \mathbb e[x\vee y - \alpha - \beta ] \quad \text{and}\quad \mathbb \mathbb e[x-\alpha ] \geq \mathbb e[y-\beta],\ ] ] we take @xmath524.\ ] ] if finally @xmath525\geq \mathbb e[x\vee y - \alpha - \beta ] \quad \text{and}\quad \mathbb \mathbb e[y-\beta ]
\geq \mathbb e[x-\alpha],\ ] ] then we can take @xmath526.\ ] ] _ _ second part.__since @xmath527 satisfies the constraint , @xmath528 $ ] .
on the other hand , @xmath529 : \mathbb e[\max((x - m)^+- \alpha,(y - m)^+- \beta)]\leq 0 \ } \\ & \geq \inf_{m\in \mathcal f}\ { \mathbb e[m ] : \mathbb e[\max(x - m- \alpha , y - m- \beta)]\leq 0 \ } \\ & = \inf_{m\in \mathcal f}\ { \mathbb e[m ] : \mathbb e[\max(x- \alpha , y- \beta)]\leq m \}\\ & = \mathbb e[(x-\alpha)\vee ( y-\beta)].\end{aligned}\ ] ] [ exploss.lm ] let @xmath511 be a probability space , @xmath530 with @xmath531 a.s . and @xmath532
. then , @xmath533 : \mathbb e[(z_k - m)^+]\leq \alpha_k , k=1,\dots , n\ } = \max_{k=1,\dots , n } \{\mathbb e[z_k ] - \alpha_k\}\ ] ] 1 .
assume that there exists @xmath470 such that @xmath534 - \alpha_k \geq \mathbb e[z_{k+1 } ] - \alpha_{k+1}\label{wrong}\end{aligned}\ ] ] then the constraint @xmath535
\leq \alpha_{k}$ ] implies @xmath536 \leq \alpha_{k+1}$ ] since @xmath537 \leq \mathbb e[(z_{k } - m)^+ ] + \mathbb e[z_{k+1}-z_k ] \leq \alpha_{k } + \alpha_{k+1 } - \alpha_{k } = \alpha_{k+1}.\ ] ] we can then remove the constraint @xmath536 \leq \alpha_{k+1}$ ] without modifying the value function . repeating the same argument for all other indices @xmath77
satisfying , we can assume with no loss of generality that , @xmath538 - \alpha_1 < \dots
< \mathbb e[z_n ] - \alpha_n.\label{right}\end{aligned}\ ] ] we then need to prove that @xmath533 : \mathbb e[(z_k - m)^+]\leq \alpha_k,\,k=1,\dots , n\ } = \mathbb e[z_n ] - \alpha_n.\ ] ] 2 . removing all the constraints except the last one , it is easy to see that @xmath533 : \mathbb e[(z_k - m)^+]\leq \alpha_k,\,k=1,\dots , n\ } \geq \mathbb e[z_n ] - \alpha_n.\ ] ] to finish the proof , it is then enough to find @xmath41 which satisfies the constraints and is such that @xmath539 = \mathbb e[z_n ] - \alpha_n$ ] .
3 . if @xmath540\geq \alpha_n$ ] , we let @xmath541 \geq \alpha_n\}\ ] ] and @xmath542 - \mathbb e[z_n ] + \alpha_n}{\mathbb e[z_{k^*+1 } ] - \mathbb e[z_{k^ * } ] } \in [ 0,1].\ ] ] otherwise , let @xmath543 and @xmath544.\ ] ] by construction , @xmath545 = \mathbb e[z_n ] - \alpha_n $ ] and since @xmath546 , also @xmath547 = \alpha_n$ ] . to check the other constraints , observe that if @xmath548 then @xmath549 a.s . and so @xmath550 = 0 $ ] . on the other hand , if @xmath551 then @xmath552 a.s .
and so by @xmath553 = \mathbb e[z_k ] - \mathbb e[m ] = \mathbb e[z_k ] - \mathbb e[z_n ] + \alpha_n < \alpha_k.\ ] ] | motivated by the asset - liability management of a nuclear power plant operator , we consider the problem of finding the least expensive portfolio , which outperforms a given set of stochastic benchmarks . for
a specified loss function , the expected shortfall with respect to each of the benchmarks weighted by this loss function must remain bounded by a given threshold .
we consider different alternative formulations of this problem in a complete market setting , establish the relationship between these formulations , present a general resolution methodology via dynamic programming in a non - markovian context and give explicit solutions in special cases . _
key words : _ multiple risk constraints , expected loss , asset liability management , snell envelope , dynamic programming . |
spin- heisenberg chains adopt a non - magnetic ground state , which qualitatively can be seen as a linear superposition of states representing all the possible ways of forming singlets in the system.@xcite with quantum fluctuations suppressing any long - range order , the translational symmetry is preserved . at large length scales ( low energies ) _ any _ amount of disorder in the exchange parameters is predicted to dominate over the quantum or thermal fluctuations@xcite and the resulting system is known as `` random heisenberg chain '' ( rhc ) . in practice , the disorder leads to an inhomogeneous ground state by associating to every random configuration of exchange paths a unique way of forming singlets , regardless of the distance between the involved spins and of their interactions .
this new type of ground state , specific to random heisenberg chains , is called _ random - singlet state _ ( rss).@xcite it is widely accepted@xcite that the interest in rhcs originally arose from a novel approach to deal with random exchanges in the isotropic spin- heisenberg model , introduced in 1979 by dasgupta and ma.@xcite their new physical insight into the effects of disorder was later developed into a formal theory by fisher@xcite and , subsequently , applied to a large variety of problems involving magnets with quenched disorder.@xcite however , it is only the recent combined availability of materials representing physical realizations of disordered quasi-1d quantum magnets@xcite and of novel stochastic numerical methods@xcite that made possible the first quantitative studies concerning the impact of disorder on materials whose magnetic properties are well described by an antiferromagnetic heisenberg - chain hamiltonian [ see eq . ] .
the aim of the present work was to identify a model material featuring the properties of an rhc and , through experimental and numerical methods , to demonstrate that the chosen rhc hamiltonian describes its physical behavior . here
we present a set of field- and temperature - dependent magnetization and @xmath3si nuclear magnetic resonance ( nmr ) data of the electronic insulator bacu@xmath0sigeo@xmath1.@xcite its relevant structural unit is shown in fig .
[ fig : struct_basge ] and its physical properties can be discussed in terms of the one - dimensional hamiltonian : @xmath4 \nonumber \\ & - g \mu_{\mathrm{b } } \sum_{i}h s^z_i + \mu_{\mathrm{b } } h_{\perp } \sum_{i } ( -1)^i s^x_i . \label{rhc1}\end{aligned}\ ] ] here @xmath5 represent the random exchange couplings along the spin - chain sites @xmath6 , @xmath7 is the bohr magneton , @xmath8 the uniform applied magnetic field , @xmath9 a locally induced transverse staggered field,@xcite and the spin operators refer to a spin @xmath10 .
( color online ) sketch of the relevant structural unit in bacu@xmath0sigeo@xmath1 highlighting the different @xmath11 and @xmath12 exchange couplings .
spin chains run along the crystallographic @xmath13-axis.,scaledwidth=42.0% ] in order to achieve a description of the large length - scale / low - energy physics contained in eq . with @xmath14 , dasgupta and
ma developed a _
real - space _ renormalization procedure .
they started from the local bonds rather than from the large spatial blocks , as is usually done for homogeneous systems.@xcite starting from the distribution of unrenormalized exchange couplings @xmath15 , where @xmath16 is the energy cutoff , an effective low - energy theory is constructed by tracking , via flow equations , the behavior of @xmath15 as the cutoff @xmath17 is progressively reduced . in the rss theory ,
the flow of the distribution of unrenormalized bond energies @xmath18 depends on the considered energy scale and temperature.@xcite as a result , the static magnetic properties are dominated by a temperature - dependent population of effectively paramagnetic moments @xmath19 .
their characteristic energy- ( @xmath20 ) and temperature - dependent average spatial separation @xmath21 diverges as @xmath22 , leading to a quasi long - range ordered ground state.@xcite in the expression for @xmath23 , @xmath24 is the largest exchange coupling in the chain prior to the renormalization.@xcite due to weak residual interchain interactions , a magnetic order at a nel temperature @xmath25 k is established also in the random heisenberg chains , albeit with a lower ordering temperature in comparison with the disorder - free case.@xcite in the present work we focus our attention on the local and bulk static magnetic properties of bacu@xmath0sigeo@xmath1 for @xmath26 and @xmath27 .
the experimental work on bacu@xmath0sigeo@xmath1 began with studies on tunable super - exchange interactions in spin - chain systems by yamada _
et al_. [ ] . across the series from bacu@xmath0si@xmath0o@xmath1
to bacu@xmath0ge@xmath0o@xmath1 the bacu@xmath0(si@xmath28ge@xmath29)@xmath0o@xmath1 compounds ( @xmath30 ) crystallize with a _ pnma _ space group . for @xmath31 ,
corresponding to maximum disorder in the exchange couplings , the values of the lattice parameters are @xmath32 , @xmath33 , and @xmath34 ,@xcite in between the parameters of the two end members of the series .
the spin chains run along the crystallographic @xmath13 axis , as shown in fig .
[ fig : struct_basge ] . early magnetization measurements of polycrystalline samples revealed the often observed broad maximum of @xmath35 , also known as the bonner - fisher peak.@xcite this maximum shifts linearly with @xmath36 towards higher temperatures as the ge concentration is enhanced . from this feature
it was initially deduced that the high - temperature properties of bacu@xmath0sigeo@xmath1 effectively reflect those of a standard spin- heisenberg chain with @xmath37 mev , where @xmath38 mev and @xmath39 mev represent the intrachain exchange interactions for no and for complete ge substitution , respectively .
the enhancement of the afm exchange , resulting from a higher ge content , was explained in terms of a change of the cu - o - cu bonding angle , from @xmath40 = 124@xmath41 in bacu@xmath0si@xmath0o@xmath1 to 135@xmath41 in the ge parent compound.@xcite earlier bulk measurements already revealed two distinct features:@xcite ( _ i _ ) a low - temperature divergence of the spin susceptibility in bacu@xmath0sigeo@xmath1 , interpreted as a simple curie - type behavior due to uncompensated magnetic moments,@xcite and ( _ ii _ ) a low - temperature magnetic order , expected even in the case of random heisenberg chains ( rhcs),@xcite with an onset at @xmath42 k in bacu@xmath0sigeo@xmath1.@xcite in zero magnetic field , bacu@xmath0si@xmath0o@xmath1 is known to order antiferromagnetically at @xmath43 k.@xcite the more than tenfold reduction of @xmath44 in bacu@xmath0sigeo@xmath1 , may be related to the fact that the introduction of ge modifies not only the intrachain , but also the interchain coupling which , along the @xmath45-axis , changes from fm- to afm - type.@xcite the latter implies that for bacu@xmath0sigeo@xmath1 the interchain coupling alternates in sign and , therefore , it averages out to a mean - field value.@xcite the interest in bacu@xmath0sigeo@xmath1 grew significantly as soon as results of magnetic - susceptibility measurements on single crystals appeared in the literature.@xcite they suggested a logarithmic dependence of the low - temperature spin susceptibility,@xcite consistent with predictions of the rhc model.@xcite subsequent inelastic neutron - scattering ( ins ) studies focused on the measurement of the energy dependence of the correlation length @xmath46 , which at zero - temperature is inversely proportional to the free - spin concentration @xmath47 , with @xmath48 the energy of the probing neutrons . in ins , @xmath49 is extracted from the inverse width of the energy - integrated scattering intensity @xmath50 which , in case of a disordered phase , is typically of lorentzian shape.@xcite following some earlier misinterpretation of the data due to problems with background subtraction,@xcite the ins results were found to be quantitatively consistent with a luttinger - liquid behavior reflecting a disorder - free spin chain with a single effective coupling @xmath51 mev.@xcite it was concluded that the expected rhc - related physics must manifest itself at energies much lower than those accessible by neutron experiments . in a previous nmr study@xcite we already pointed out how the spin dynamics of bacu@xmath0sigeo@xmath1 , probed at the characteristic larmor energy @xmath52 @xmath53ev ( with @xmath54 the nmr resonance frequency ) , turns out to be notably different with respect to that of the regular heisenberg chain bacu@xmath0si@xmath0o@xmath1 . on the other hand , considering existing data on bacu@xmath0si@xmath0o@xmath1 and bacu@xmath0ge@xmath0o@xmath1 , a low - temperature divergence of the magnetic susceptibility also in bacu@xmath0sigeo@xmath1 is not really surprising.@xcite in addition , since ins failed to detect any sort of disorder - induced spin dynamics in this compound , the existence of a random - singlet state ( rss the ground state of the isotropic rhc hamiltonian ) was considered questionable . in the present work
we show how the effect of disorder on the _ static _ magnetic properties can be quantitatively modeled and compared with experiments , thereby supporting the rhc scenario .
( color online ) field and temperature dependences of the magnetic susceptibility @xmath55 per cu ion in bacu@xmath0sigeo@xmath1 ( and in bacu@xmath0si@xmath0o@xmath1 , at 1 t ) with @xmath56 in normalized units , assuming @xmath57.@xcite the fit of @xmath58 with an rhc model was adopted from ref . , while the @xmath35 curves for chains with a single isotropic exchange value were calculated as in ref . .
qmc simulations of the @xmath55 susceptibility at @xmath59 and 14 t for a simple @xmath11@xmath12 rhc model is also shown ( see text for details).,scaledwidth=45.0% ] as a first step , we extended the measurements of the magnetic susceptibility @xmath55 to wider temperature and field ranges than those covered in previous studies .
single crystals of bacu@xmath0sigeo@xmath1 were grown by the floating - zone technique .
one of them was cut and the smaller 6.5-mg part was used in magnetometry measurements ; the remaining 20-mg piece was used for the nmr investigations .
the magnetization of the chosen sample was measured by applying a magnetic field @xmath56 .
magnetic susceptibility data @xmath60 are displayed as bold solid lines in fig .
[ fig : rhc_chi_meas ] . for @xmath61 t and at temperatures
@xmath62 k , the measurements were performed using a standard dc squid magnetometer , whereas to cover temperatures down to 0.5 k we employed a squid device with a @xmath63he insert .
for all the other fields and for temperatures @xmath62 k we used a vibrating - sample magnetometer ( vsm ) .
since in the latter case data are collected `` on the fly '' , i.e. , while the temperature is being swept at 0.25 k / min , this causes a small ( known ) temperature difference between sample and thermometer position .
in addition , a certain dispersion in the magnetization values was reduced by applying a moving - average filter with a 200-mk width to the raw experimental data .
magnetic hysteresis , defined as the difference between the zero - field - cooled and field - cooled magnetization data , was observed for @xmath64 mk .
the hysteretic behavior is probably related to ferromagnetic domains with a rather small net magnetic moment present in the ordered low - temperature phase . since only the zero - field - cooled scans are relevant for our discussion , the field - cooled data are not shown . as shown in fig .
[ fig : rhc_chi_meas ] , for temperatures above 30 k the @xmath55 curves are field - independent and , for @xmath65 k , they approach the theoretical prediction for an _ isotropic _ heisenberg chain @xcite with an exchange parameter @xmath37 mev ( see the green dashed line in fig . [
fig : rhc_chi_meas ] ) . for comparison
, the red dashed line indicates the calculated @xmath35 for an isotropic chain with @xmath66 mev , corresponding to the @xmath67 case .
we first discuss two major differences between the magnetic susceptibilities of bacu@xmath0si@xmath0o@xmath1 ( @xmath67 ) and bacu@xmath0sigeo@xmath1 ( @xmath31 ) : ( _ i _ ) our previous successful attempt to interpret the low - temperature increase / divergence of @xmath35 in bacu@xmath0si@xmath0o@xmath1 and bacu@xmath0ge@xmath0o@xmath1 as simply being caused by a local transverse staggered field ( ltsf,@xcite see sec .
[ ssec : nmr_analysis ] ) is inadequate for treating the more complex case of bacu@xmath0sigeo@xmath1 .
first of all , the magnetic susceptibility @xmath68 is strongly _ reduced _ even by moderate fields . as a result ,
the magnetization deviates from the linear response with respect to @xmath8 ( see fig . [
fig : m_h_qmc ] and ref . ) , expected either for an isotropic chain in the ll regime , or for the sine - gordon ( sg ) model in the high - temperature limit , the latter describing the physics when an ltsf is present.@xcite the sg model , even at intermediate temperatures , is inadequate to describe the physics of bacu@xmath0sigeo@xmath1 . indeed , in this case @xmath35 is known to exhibit a symmetric peak centered at a temperature @xmath69 , roughly corresponding to half of the field - induced gap @xmath70.@xcite to exemplify this fact , we consider the hamiltonian in eq . with @xmath71 and @xmath27 .
in the reduced units @xmath72 , with @xmath73 , this model features a gap @xmath70 in the spin - excitation spectrum which , for @xmath74 , scales as:@xcite ( color online ) first derivative of the sample magnetization vs. temperature , @xmath75 , in bacu@xmath0sigeo@xmath1 for @xmath76 .
the graphic compares several cross sections of the color map shown in fig .
[ fig : phase_bound ] .
the assigned errors reflect the line - slope uncertainty for fits within the binning range ( see text for details ) .
the shown qmc simulation results were obtained _ without _ using any free parameters.,scaledwidth=48.0% ] @xmath77 equation implies that for @xmath78 ( the value estimated for bacu@xmath0si@xmath0o@xmath1@xcite if @xmath56 ) , the ltsf yields @xmath79 mev at @xmath80 t , corresponding to @xmath81 k. on the other hand , as can be seen in fig .
[ fig : rhc_chi_meas ] , the susceptibility recorded at 14 t tends to saturate at low temperatures , rather than display a peak at @xmath69 .
( _ ii _ ) the inadequacy of the sg model in interpreting the @xmath55 data of bacu@xmath0sigeo@xmath1 is particularly evident from the @xmath82-dependence of @xmath83 , the inflection point which marks the start of saturation ( corresponding to the maximum of @xmath84 in a fixed field @xmath8 see fig .
[ fig : phase_bound_cuts ] ) . to overcome noise problems in the calculation of @xmath84 from raw magnetization data
, the experimental points were linearly interpolated using a 200-mk binning and the resulting line slope was taken as the derivative with respect to temperature .
the results of this procedure are shown as colored circles in fig .
[ fig : phase_bound_cuts ] for four selected fields .
a @xmath82 color map of the @xmath84 values is shown in fig .
[ fig : phase_bound]a . here
the inflection points ( assigned with an error corresponding to the peak width at 95% of the peak height in fig . [ fig : phase_bound_cuts ] ) are shown as open circles .
the resulting @xmath85 values follow a linear field dependence , such that @xmath86 . besides requiring a nonuniversal dimensionless prefactor,@xcite the field dependence of @xmath83 does not follow the @xmath87 power law that the small @xmath88 value would imply for @xmath89.@xcite the calculated dependences of @xmath90 for @xmath91 with @xmath92 or 0.052 are both also shown in fig .
[ fig : phase_bound]a .
the first @xmath13 value represents the ltsf proportionality constant in bacu@xmath0si@xmath0o@xmath1.@xcite because of the similar local environments , it seems reasonable to use the same value also for bacu@xmath0sigeo@xmath1 .
the reason for choosing @xmath93 instead , will be clarified in sec .
[ sec : nmr1 ] .
( color online ) ( a ) first derivative of the sample magnetization vs. temperature , @xmath75 , in bacu@xmath0sigeo@xmath1 for @xmath76 ( data scaled using @xmath94).@xcite the open circles represent cross - over temperatures . for details on data treatment and for a description of the plotted lines see text .
( b ) qmc simulation of the @xmath11@xmath12 rhc model for a system of 6000 spins , using 20 realizations of disorder for each @xmath82-point .
the color map was generated using the same data treatment as in panel ( a).,scaledwidth=48.0% ] in view of these two inadequacies , we modified our previous approach to the experimental data analysis@xcite and instead chose to compare the experimental results with the predictions of the rhc model hamiltonian in eq .
.@xcite a first comparison is carried out by fixing @xmath95 and by assuming the @xmath96 values to be uniformly distributed between the two limits , @xmath97 and @xmath98 . in case of nonzero temperatures and magnetic fields , the renormalization flow of the decimation procedure
is interrupted either at the thermal energy @xmath99 , or at the magnetic energy @xmath100.@xcite in the first case , the resulting susceptibility is:@xcite @xmath101 in the second case , the system is driven away from the zero - field fixed point into a regime with field - aligned undecimated spins , a saturated magnetization and zero entropy.@xcite in spin-@xmath102 afm - rhcs the decimation procedure involving the total set of spins leaves the renormalized magnetization magnitude at @xmath103 , unlike , for example , in an rhc with fm - afm coupling.@xcite for this reason , the onset of saturation of the low - temperature magnetization occurs at @xmath104.@xcite a fit to the susceptibility data at the smallest applied field , using the theoretical random - singlet expression for @xmath105 , is shown in fig .
[ fig : rhc_chi_meas ] as a black dashed line .
a rather good match is found above 4 k with a cutoff @xmath106 mev . to achieve a more quantitative comparison
, we carried out a series of quantum monte carlo ( qmc ) simulations based on eq . , where @xmath96 couplings alternate randomly from @xmath107 mev to @xmath108 mev ( we call this a @xmath11@xmath12 model ) , mimicking the situation expected for bacu@xmath0sigeo@xmath1 .
we chose 6000 spin sites with randomly distributed but equally probable @xmath11 and @xmath12 couplings .
the simulations for @xmath55 were averaged over at least 25 random realizations of disorder , as obtained from a directed - loop algorithm @xcite within the alps 2.0 package.@xcite technical details related to the qmc simulations are discussed in app .
[ app : a ] .
disorder in the exchange couplings was imposed by assuming an equal number ( @xmath109 , with @xmath110 as the chain length ) of @xmath111 and @xmath112 values , randomly permuted to construct the rhc chain that was to be simulated .
selected results of these simulations are shown as red and yellow circles in fig .
[ fig : rhc_chi_meas ] , as interconnected diamonds in fig .
[ fig : phase_bound_cuts ] , and as a color map in fig .
[ fig : phase_bound]b . in view of the simplicity of the model ,
the qualitative and quantitative agreement of our parameter - free qmc simulations with the data is remarkable .
incidentally , the best fit of the rs prediction to the low-@xmath113 qmc results in an applied field @xmath61
t requires @xmath114 .
this corresponds to the largest energy scale in the system before the decimation starts , thus reinforcing the pertinence of the theory used to capture the magnetic properties of the system . as may be seen in figs .
[ fig : phase_bound_cuts ] and [ fig : phase_bound]b , by applying the moving - average procedure described above also to the qmc results , numerical calculations predict the same free - spin - like linear field - dependence for @xmath83 as observed experimentally .
this is an important result , because the existence of paramagnetic entities at @xmath115 k in a strongly - correlated system , with a characteristic energy scale corresponding to a temperature exceeding 200 k , is highly counterintuitive .
although the measurements of @xmath55 and the comparison with qmc calculations already provide a strong indication for the formation of an rs phase in bacu@xmath0sigeo@xmath1 at low temperatures , it is tempting to directly verify the presence of a random - singlet state on a local scale . with decreasing temperature an increasing number of spins form pairs by adopting a nonmagnetic singlet ground state .
the divergent low - temperature tail of @xmath55 is due to the residual unpaired spins .
the tendency to form a random - singlet ground state should be directly manifest in nmr data , reflecting a growing number of @xmath2si sites which experience a zero net transferred local field .
height - normalized @xmath2si nmr lines , as recorded in a 7-t external field applied along the crystallographic @xmath116 axis , are plotted in fig .
[ fig : nmr_lines ] .
the spectra were obtained by superposing several acquisitions in the frequency - sweep mode , as described in ref . .
the effects of disorder are evident already by comparing the high - temperature line shapes of bacu@xmath0sigeo@xmath1 and bacu@xmath0si@xmath0o@xmath1 ( see inset of fig . [
fig : nmr_lines ] ) .
we note that the @xmath2si resonance of bacu@xmath0sigeo@xmath1 , apart from being broader than that of its pristine counterpart , is also shifted to higher frequencies .
we recall that nmr data of bacu@xmath0si@xmath0o@xmath1 ( taken at @xmath117 t , with @xmath56 ) exhibit a positive orbital shift of @xmath118 mhz , with a _ negative _ hyperfine - coupling to the longitudinal magnetization.@xcite ( color online ) temperature dependence of the @xmath2si nmr lines of the disordered heisenberg chain compound bacu@xmath0sigeo@xmath1 in a magnetic field of 7.05 t applied along the @xmath116-axis .
the zero frequency marks the resonance of @xmath2si nuclear spins in a standard reference sample such as si(ch@xmath119)@xmath120.@xcite the white dashed line marks the position of the resonance at room temperature , while the yellow points represent the position of the maxima , displaying a shift towards higher frequencies at low temperatures .
the inset compares the cases of maximum and no disorder at high temperature .
the green dashed line marks the orbital shift as obtained from ref .
( see text for details).,scaledwidth=45.0% ] as shown in fig .
[ fig : rhc_chi_meas ] , the high - temperature magnetization of bacu@xmath0sigeo@xmath1 is equivalent to that of an isotropic chain with @xmath121 mev and is distinctly smaller than that of bacu@xmath0si@xmath0o@xmath1 , well modelled by choosing @xmath122 mev . consequently
, the negative hyperfine shift in case of a sample with disorder is smaller and , hence , the resonance is located closer to @xmath123 ( see inset of fig .
[ fig : nmr_lines ] ) .
this result is ultimately an independent confirmation of the validity of the model which allowed us to extract the parameter @xmath123 in ref . .
we emphasize that , due to a random variation of the hyperfine interactions , a broadening of the resonance in the @xmath31 case may well be of structural origin , rather than due to magnetic disorder .
realistically , both structural and magnetic disorder have to be taken into account and one of the aims of the analysis outlined below is to disentangle them .
another striking feature of the nmr signals is the temperature - dependent shift of the line maxima . in fig .
[ fig : nmr_lines ] this shift is evidenced by comparing the sequence of yellow points , marking the maxima , with the vertical dashed line , representing the peak position of the 300-k line . it indicates a growing nonzero uniform magnetization as the temperature is lowered , in obvious contradiction to the rs - phase hypothesis , where the formation of singlets would imply the absence of a local magnetization .
as expected , unlike to what is observed in bacu@xmath0si@xmath0o@xmath1,@xcite the nmr data show no indications of a phase transition around @xmath115 k for the disordered compound . in order to interpret the recorded nmr line shapes and positions , we compare the experimental data with the results of qmc simulations of the @xmath11@xmath12 model , addressed in sec .
[ sec : magnqmc ] . in fig .
[ fig : lines_nmr_simulation]a we show the temperature - dependent , height - normalized histograms , which reflect the occurrences of longitudinal local magnetization values @xmath124 in @xmath125 @xmath7 intervals for @xmath117 t ( @xmath126 is the spin quantization axis along which the external field was applied ) .
the histograms were obtained by using a single configuration of disorder in a system of @xmath127 spins .
we note that at @xmath128 k the @xmath124 values are distributed around a nonzero , albeit small , uniform magnetization . in this regime
the details of disorder are not essential for describing the physics . at low temperatures , however
, a broadened magnetization peak appears , centered at zero local moment , as expected in the case of a continuous formation of singlets . at the same time we observe weak but extremely broad tails ( see top histogram in fig . [
fig : lines_nmr_simulation]a ) , indicating the presence of incompletely compensated spins , which are progressively polarized as the temperature is reduced . as shown in fig .
[ fig : lines_nmr_simulation]b , the local - moment distribution obtained from these simulations ( which do not include an ltsf ) does not match the experimental nmr data .
the formation of random singlets implies that at a significant number of sites the moments of the cu atoms are quenched at low temperatures , corresponding to a zero shift of the simulated nmr line , in disagreement with observations .
it turns out that it is essential to consider the combined effect of random exchange interactions and an ltsf , namely to set @xmath129 in the hamiltonian . to include an ltsf we use the same approximation made in the case of no disorder,@xcite i.e.
, we assume the total local magnetization @xmath130 to be the sum of a non - uniform longitudinal and a transverse contribution , in the form : @xmath131 where for the magnitude of the local transverse magnetization we write @xmath132 , and analogously for the longitudinal component , with @xmath133.@xcite ( color online ) field dependence of the magnetization of bacu@xmath0sigeo@xmath1 at 2 and 10 k ( dots ) and comparison with qmc simulations of the @xmath134 rhc model without ( @xmath95 ) and with ( @xmath135 ) ltsf .
the @xmath13 coefficient refers to the staggered field value @xmath136 , with @xmath8 the uniform field .
also shown is the field dependence of the ground state magnetization ( @xmath137 k ) of an isotropic heisenberg spin- chain , as from ref . .
the magnetization is expressed in normalized units ( @xmath138 is the total number of copper sites and @xmath7 is the bohr magneton ) , assuming @xmath94.,scaledwidth=48.0% ] because of this particular approximation , two qmc simulations , one for @xmath139 and another for @xmath140 , can be carried out separately on a diagonal basis .
+ with @xmath27 , an ltsf acts on paired spins by mixing the nonmagnetic singlet states @xmath141 with the magnetic triplet states @xmath142 and @xmath143 , thereby inducing a nonzero magnetization also in the ground state of the spin pair.@xcite if both the uniform field and the ltsf are considered , in a first - order perturbation approximation , the new ground state can be written as:@xcite @xmath144 the uniform field appearing in the denominator can be neglected if @xmath145 , which is the case for the strongest coupled singlets in the rs phase .
therefore , at temperatures @xmath146 the longitudinal local magnetization varies linearly with the applied field . here
min(@xmath16 ) refers to the minimum exchange coupling value of the already formed singlets . the distribution of local - magnetization values in a @xmath11@xmath12 rhc model which includes an ltsf is shown in fig .
[ fig : lines_nmr_simulation]c .
clearly , the presence of an ltsf substantially modifies the magnetization profiles shown in fig .
[ fig : lines_nmr_simulation]a . by assuming that the local symmetry of the oxygen atoms surrounding a cu@xmath147 ion is largely unaffected by the si - to - ge substitution , the same value of @xmath92 obtained for bacu@xmath0si@xmath0o@xmath1 ought to be valid also in our case .
as seen in fig .
[ fig : lines_nmr_simulation]c , at @xmath128 k and @xmath117 t , the chosen ltsf is too weak to induce a nonzero magnetization .
however , as the temperature is progressively lowered , singlets with a _ distribution _ of effective exchange couplings are formed and a staggered field @xmath148 acting on them creates a distributed local magnetization of the form @xmath149 .
therefore , the histograms in fig . [
fig : lines_nmr_simulation]c split and broaden with decreasing temperature .
the symmetrical splitting reflects the alternating sign of the ltsf from site to site , with a mean absolute value of the local moment which progressively shifts away from @xmath150 . in order to compare the simulated local magnetization profiles with the nmr line shapes , structural details have to be considered .
eight configurations of disorder in systems with @xmath127 sites were simulated . in order to reproduce 6000 unit cells , the local moments
were then arranged into four different chains , with two inequivalent copper atoms each . at this first stage ,
the dipolar coupling was neglected , but we considered the hyperfine coupling to both the longitudinal and the transverse magnetization . following the notation of ref .
, for @xmath56 , the local hyperfine field @xmath151 at a general @xmath2si nuclear site can be written as : @xmath152 where the index @xmath6 runs over the four nearest - neighbor copper sites . with this notation
, the relative nmr resonance frequency @xmath153 can be written as @xmath154 , with @xmath155 as the @xmath2si nuclear gyromagnetic ratio . from our previous study@xcite of bacu@xmath0si@xmath0o@xmath1 we can fix the parameters @xmath156 t/@xmath7 , @xmath157 t/@xmath7 , whereas @xmath118 mhz at 7 t. each of the above hyperfine parameters , which couple the silicon nuclear magnetism to the longitudinal and transverse electronic magnetization of copper , represents the sum of four individual copper - to - silicon couplings . given the broken translational symmetry due to disorder , the individual cu - si couplings are essential for a quantitative comparison between data and theory .
unfortunately , these can not be evaluated by experiment , leaving six unknown free parameters in the model . for a comparison of the simulated nmr lines with data recorded at 5 k
, we used the following parameters ( in t/@xmath7 units ) : @xmath158 , @xmath159 for the coupling to the transverse magnetization , and @xmath160 , @xmath161 , @xmath162 for the coupling to the longitudinal one .
the sign of the parameters @xmath163 changes according to the considered si nucleus . by making use of them and with no other free parameters , we succeeded in simulating the complete temperature dependence of the nmr lines , which account for the simultaneous presence of both an ltsf and a uniform magnetic field . the result is shown in fig . [ fig : lines_nmr_simulation]d , where the agreement between data and the theoretical model is again remarkable . to be more specific , for a comparison with the experiment , we calculated the line shapes by convoluting the distribution of local fields at the @xmath2si sites with a 2-khz wide gaussian , which is equivalent to the width of the line profile of pure bacu@xmath0si@xmath0o@xmath1 , shown in the inset of fig . [
fig : nmr_lines ] .
the choice of the optimal binning range , which strongly affects the shape of the tails of the @xmath124 histograms in fig .
[ fig : lines_nmr_simulation ] , was done by employing the algorithms proposed in refs . .
note that the lines displayed in fig .
[ fig : lines_nmr_simulation]d show that the averaging process over four cu sites removes the splitting in the original histograms in fig .
[ fig : lines_nmr_simulation]c .
in addition , the simulated lines shift with temperature , as indeed observed in the experiment .
the minimally - shifted line shape at @xmath128 k reflects magnetic disorder . by comparing figs .
[ fig : lines_nmr_simulation]b and [ fig : lines_nmr_simulation]d , it may be seen that the inclusion of an ltsf in the simulations significantly improves the agreement between theory and experiment .
therefore , from both a macroscopic and a local point of view , nmr data confirm that the static magnetic properties of bacu@xmath0sigeo@xmath1 are those of an rhc , with the addition of a residual ltsf .
to cross check the consistency of the proposed model , it seems natural to consider the ltsf - induced effects also in the magnetometry data .
since the influence of the ltsf grows with the applied field @xmath8 , we consider the calculated @xmath164 curves ( shown by dotted lines in fig . [
fig : m_h_qmc ] ) at two representative temperatures , @xmath165 k and 10 k. the calculation of @xmath164 at @xmath165 k , using the average exchange coupling @xmath166 mev , reflects the _ linear _ behavior expected for an isotropic heisenberg chain in its ll regime@xcite and differs substantially from the experimental data .
in fact , this oversimplified model completely neglects the field - induced alignment of the still uncompensated spins that , in a random heisenberg chain , survive even at the lowest temperatures . on the other hand , qmc simulations using the @xmath11@xmath12 model discussed in sec . [
sec : magnqmc ] and in ref .
show significant improvements over the simplified `` average-@xmath16 '' model ( solid blue and red lines in fig .
[ fig : m_h_qmc ] for @xmath165 k and 10 k , respectively ) .
the next refinement was to include both a uniform and a staggered magnetization in the @xmath11@xmath12 model and combine them@xcite to obtain : @xmath167,\ ] ] with @xmath168 the rhc magnetization in a uniform field .
as can be seen in fig .
[ fig : m_h_qmc ] , the qmc calculation for @xmath169 with @xmath78 unexpectedly fails to reproduce the data .
a good agreement with the magnetization scans at both temperatures was , however , obtained using @xmath170 .
the discrepancy in the prefactor is most probably due to the fact that the current model neglects the effects of structural disorder in bacu@xmath0sigeo@xmath1 .
a more refined model taking into account such subtleties , by including a site - dependent @xmath13 value , was not implemented in the present analysis .
nonetheless , the essence of rhc physics , which dominates the magnetic properties of the probed system , is born out by the combined investigation of bulk magnetometry , nmr measurements and qmc simulations . taken together
they strongly suggest that bacu@xmath0sigeo@xmath1 is an excellent realization of a random heisenberg chain model , as originally claimed in ref .
, thus providing the ideal testing ground for analytic theories modeling a random - singlet ground state .
in conclusion , we considered the bacu@xmath0sigeo@xmath1 compound , where spin-@xmath102 cu@xmath147 ions interact via random exchange couplings , as a representative of rhc systems with bond disorder . we demonstrate that this spin - chain material has a ground state which is very well described on the basis of the rhc hamiltonian in eq . .
by using temperature- and magnetic field - dependent microscopic @xmath3si nmr measurements , combined with macroscopic magnetic - susceptibility data and results of detailed numerical quantum monte carlo simulations , we find a coherent description of the physical properties of the system , compatible with that of an rs phase .
in particular , the considerable broadening of the nmr lines and the divergent magnetic response at low temperatures , combined with a nonlinear increase of the magnetization as a function of field at low @xmath113 , all reflect the presence of unpaired spins with arbitrary small couplings even at temperatures close to zero .
specific refinements in the simulations of @xmath11-@xmath12 model have shown that the addition of a local transverse staggered field ltsf ( similar to that present in the parent compound bacu@xmath0si@xmath0o@xmath1)@xcite is essential also for reproducing the observed nmr line shifts .
since the magnetization of a singlet in an ltsf is proportional to the local field , the detection of the field distribution by future @xmath171cu nmr measurements may provide a direct way to explore the @xmath16 distribution in an rhc .
in addition , besides the existing stretched - exponential nmr relaxation data,@xcite the spectral properties of random heisenberg chains would be more directly accessible if neutron scattering experiments in a lower energy range would be possible .
the authors thank a. feiguin ( northwestern university , boston ) , k. pra ( epfl , lausanne ) , and o. zaharko ( psi , villigen ) for useful discussions .
the bacu@xmath0sigeo@xmath1 samples used in this work were prepared in the early 2000 s in the group of prof .
k. uchinokura at the university of tokyo .
this work was financially supported in part by the schweizerische nationalfonds zur frderung der wissenschaftlichen forschung ( snf ) and the nccr research pool manep of snf .
in order to correctly interpret the numerical results , it is necessary to mention the theoretical expectations for the error bars in qmc simulations ( i.e. , specific to this method ) , in relation to the self - averaging induced errors ( i.e. , specific to disorder).@xcite for a spin chain of length @xmath110 , one can define the ratio @xmath172 , where @xmath173 is any macroscopic variable subject to self - averaging and @xmath174 denotes the same quantity averaged over the realizations of disorder.@xcite by using @xmath175 monte carlo updates " ( the number of random updates of the system performed by the algorithm before a new configuration is generated ) and @xmath176 independent monte carlo measurements in a system of 6000 spins , the relative error for @xmath177 in low - temperature qmc simulations was found to be 10 times lower than @xmath178 , the error due to the average on all the realizations of disorder . therefore , in our case @xmath178 is the dominant simulation error .
whenever the correlation length is @xmath179 , self - averaging can be justified based on the central - limit theorem.@xcite in case of critical systems , where the correlation length tends to diverge ( @xmath180 ) , the only way to obtain meaningful results in the presence of disorder consists in increasing the system size @xmath110 . by simulating systems with different spin - chain lengths @xmath110 , we find that @xmath181 , with @xmath182 the standard deviation of the local spin susceptibility .
this implies the correct evaluation of the thermodynamic limit even in the problematic @xmath183 case .
incidentally , since @xmath182 is smaller at higher temperatures and magnetic fields , this justifies the smaller error bars in these regions .
a. zheludev and t. roscilde , arxiv:1305.1194 [ cond-mat.str-el ] ( 2013 ) . s. ward , p. bouillot , h. ryll , k. kiefer , k. w. krmer , ch .
regg , c. kollath and t. giamarchi , j. phys . : condens .
matter * 25 * , 014004 ( 2013 ) .
g. simutis , s. gvasaliya , m. mnsson , a. l. chernyshev , a. mohan , s. singh , c. hess , a. t. savici , a. i. kolesnikov , a. piovano , t. perring , i. zaliznyak , b. bchner , a. zheludev , accepted in phys .
* 111 * ( 2013 ) .
b. bauer et al ( alps collaboration ) , j. stat .
p05001 ( 2011 ) . | the disordered quasi-1d magnet bacu@xmath0sigeo@xmath1 is considered as one of the best physical realizations of the random heisenberg chain model , which features an irregular distribution of the exchange parameters and whose ground state is predicted to be the scarcely investigated _ random - singlet _ state ( rss ) . based on extensive @xmath2si nmr and magnetization studies of bacu@xmath0sigeo@xmath1 , combined with numerical quantum monte carlo simulations , we obtain remarkable quantitative agreement with theoretical predictions of the random heisenberg chain model and strong indications for the formation of a random - singlet state at low temperatures in this compound . as a local probe , nmr is a well - adapted technique for studying the magnetism of disordered systems . in this case
it also reveals an additional local transverse staggered field ( ltsf ) , which affects the low - temperature properties of the rss .
the proposed model hamiltonian satisfactorily accounts for the temperature dependence of the nmr line shapes . |
supernovae ( sn ) are among the most spectacular events because they reach the same brightness as an entire galaxy .
this makes them good candidates to determine extragalactic distances and to measure the basic cosmological parameters .
moreover , they are thought to be the major contributors to the chemical enrichment of the interstellar matter with heavy elements .
energy injection by sn into the interstellar medium , triggers star formation and feedback in galaxy formation , and is regarded as a key for our understanding of the formation and evolution of galaxies .
core collapse supernovae are thought to be the final stages of the evolution of massive stars which live only @xmath0 to @xmath1 years .
such supernovae could be the brightest objects in the distant past when stars first began to form .
a detailed understanding of core collapse is essential to probe the very early phases of the universe right after the initial star forming period which occurs at redshifts @xmath2 .
understanding the mechanism of core collapse supernovae explosions is a problem that has challenged researchers for decades ( hoyle & fowler 1960 ) . in the general scenario for the explosion , the central region of a massive star collapses and forms a neutron star .
eventually , parts of the potential energy will cause the ejection of the envelope .
this general scenario has been confirmed by a wealth of observations including the direct detection of neutrinos in sn1987a and neutron stars in young supernovae remnants . in recent years , there has been a mounting evidence that the explosions of massive stars ( core collapse supernovae ) are highly aspherical .
( 1 ) the spectra of core - collapse supernovae ( e.g. , sn87a , sn1993j , sn1994i , sn1999em ) are significantly polarized indicating asymmetric envelopes ( mndez et al . 1988 ; hflich 1991 ; jeffrey 1991 ; wang et al .
1996 ; wang et al . 2000 ) .
the degree of polarization tends to vary inversely with the mass of the hydrogen envelope , being maximum for type ib / c events with no hydrogen ( wang et .
1999 , wang et al . 2000 ) .
for supernovae with a good time and wavelength coverage the orientation of the polarization vector tends to stay constant both in time and in the wavelength .
this suggests that there is a global symmetry axis in the ejecta ( wang et al .
( 2 ) observations of sn 1987a showed that radioactive material was brought to the hydrogen rich layers of the ejecta very quickly during the explosion ( lucy 1988 ; tueller et al .
( 3 ) the remnant of the cas a supernova shows rapidly moving oxygen - rich matter outside the nominal boundary of the remnant and evidence for two oppositely directed jets of high - velocity material ( fesen & gunderson 1997 ) . ( 4 ) .
recent x - ray observations with the chandra satellite have shown an unusual distribution of iron and silicon group elements with large scale asymmetries in cas a ( huges et al .
( 5 ) after the explosion , neutron stars are observed with high velocities , up to 1000 km / s ( strom et al .
1995 ) .
there is a general agreement that the explosion of a massive star is caused by the collapse of its central parts into a neutron star or , for massive progenitors , into a black hole .
the mechanism of the energy deposition into the envelope is still debated .
the process likely involves the bounce and the formation of the prompt shock ( e.g. van riper 1978 , hillebrandt 1982 ) , radiation of the energy in the form of neutrino ( e.g. bowers & wilson 1982 ) and the interaction of the neutrino with the material of the envelope and various types of convective motions ( e.g. herant et al .
1994 , burrows et al .
1995 , mller & janka 1997 , janka & mller 1996 ) , rotation ( e.g. leblanc & wilson 1970 , saenz & shapiro s.l . 1981 , mnchmeyer et al .
1991 ) and magnetic fields ( e.g. leblanc & wilson 1970 , bisnovati - kogan 1971 ) .
spherically symmetric explosion models rely on the neutrino deposition mechanism .
the results depend critically on the progenitor structure , equation of state , neutrino physics , and implementation of the neutrino transport .
currently , results are inconclusive even when using sophisticated boltzman solvers for the neutrino transport .
for example , mezzacappa et al . ( 2000 ) find an explosion whereas yamada et al .
( 1999 ) and rammp & janka ( 2000 ) do not .
even if successful , these models can not explain the observed asymmetries .
within the spherical core - collapse picture , additional mechanisms must be invoked which operate within the envelope itself .
two such mechanisms have been studied in some detail .
one is the rayleigh - taylor instability which causes mixing of the layers of different composition when the outgoing shock front passes through ( mller et al .
1989 , benz & thielemann 1990 , fryxell et al .
this effect can explain mixing of the carbon , oxygen and helium - rich layers required for the sn1987a , but none of the simulations were able to account for the high velocity of ni observed in sn1987a ( kifonidis et al .
rayleigh - taylor mixing provides a rather small - scale structures and can hardly account for the observed polarization which requires a global asymmetry of the expanding envelope ( hflich 1991 ) .
another mechanism involves an explosion inside a rapidly and differentially rotating supernova progenitor ( steinmetz & hflich 1992 ) . with this mechanism it was possible to account for the polarization in sn1987a which originated from a blue supergiant .
this mechanism may have difficulty accounting for the early polarization in some type ii supernovae ( wang et al .
2000 ) whose light curves indicate a red - giant progenitors . a strong differential rotation in red supergiants
can hardly be expected due to their convective envelopes ( steinmetz & hflich , 1991 ) .
attempts have also been made to include multi - dimensional effects into a model of collapse itself .
the collapsing core becomes unstable due to the gradients of both electron mole fraction @xmath3 and entropy .
the developing convection then affects both the neutrino flux and the energy deposition behind the stalled shock .
numerous studies have demonstrated the presence of this effect .
it is still debated whether convection combined with the neutrino transport provides the solution to the supernova problem ( rammp et al .
1998 and references therein ) . in the current calculations ,
the size and scale of the convective motions seem to be too small to explain the observed asymmetries .
the angular variability of the neutrino flux caused by the convection has been invoked to explain the neutron star kicks ( burrows et al .
1995 , janka & mller 1994 ) .
calculations give kick velocity up to @xmath4 km / s whereas ns with velocities of several @xmath5 km / s are common .
rotation of the collapsing core may also be important .
it tends to facilitate the explosion because the centrifugal barrier reduces the effective potential for the material moving in the equatorial plane and introduces an axial symmetry in the fluid motions .
simulations made so far indicate that the rotation alone has no or has only a weak effect on the explosion ( e.g. mnchmeyer et al .
1991 , zwerger & mller 1997 ) . in the latter case
it induces a rather weak asymmetry of the explosion with more energy going along the rotational axis .
it has long been suggested that the magnetic field can play an important role in the explosion ( leblanc & wilson 1970 ; ostriker & gunn 1971 , bisnovati - kogan 1971 , symbalisty 1984 ) . leblanc and wilson simulations showed the amplification of the magnetic field due to rotation and the formation of two oppositely directed , high - density , supersonic jets of material emanating from the collapsed core .
their simulations assumed a rather high initial magnetic field @xmath6 gauss and produced a very strong final fields of the order of @xmath7 gauss which seemed to be unreasonable at the time .
the recent discovery of pulsars with very high magnetic fields ( kouveliotou et al .
1998 , duncan & thomson 1992 ) revives the interest in the role of rotating magnetized neutron stars in the explosion mechanism .
it is not clear whether a high initial magnetic field required for the leblanc & wilson mechanism is realistic . on the other hand
it may not be needed .
recently mayer and wilson ( 2000 , private communication ) have suggested that the field amplification and the generation of the jet may continue during the first few seconds of the cooling of the neutron star when the neutron star shrinks rapidly .
the current picture of the core collapse process is unsettled .
a quantitative model of the core collapse must eventually include all the elements mentioned above . due to the difficulty of modeling core collapse from first principles ,
a very different line of attack on the explosion problem has been used extensively and proved to be successful in understanding of the supernova problem , sn1987a in particular ( arnett et al .
1990 , hillebrand & hflich 1991 ) .
the difference of characteristic time scales of the core ( a second or less ) and the envelope ( hours to days ) allows one to divide the explosion problem into two largely independent parts - the core collapse and the ejection of the envelope . by assuming the characteristics of the energy deposition into the envelope during the core collapse , the response of the envelope can be calculated .
thus , one can study the observational consequences of the explosion and deduce characteristics of the core collapse and the progenitor structure .
this approach has been extensively applied in the framework of the 1d spherically symmetric formulation .
the major factors influencing the outcome have been found to be the explosion energy and the progenitor structure .
the same approach can be applied in multi - dimensions to investigate the effects of asymmetric explosions . in this paper
we study the effects and observational consequences of an asymmetric , jet - like deposition of energy inside the envelope of a core - collapse supernova .
* 3-d hydrodynamics : * the explosion and jet propagation are calculated by a full 3-d code within a cubic domain of size d. the stellar material is described by the time - dependent , compressible , euler equations for inviscid flow with an ideal gas equation with @xmath8 plus a component due to radiation pressure with @xmath9 .
the euler equations are integrated using an explicit , second - order accurate , godunov type , adaptive - mesh - refinement , massively parallel , fully - threaded tree ( ftt ) program , alla ( khokhlov 1998 ) .
euler fluxes are evaluated by solving a riemann problem at cell interfaces .
ftt discretization of the computational domain allowes the mesh to be dynamically refined or coarsened at the level of individual cells . for more details , see khokhlov ( 1998 ) and khokhlov et al .
( 1999ab ) .
* 1-d radiation - hydrodynamics : * about 1000 seconds after the core collapse and in case of the explosion of red supergiants , the propagation of the shock front becomes almost spherical ( see below ) . to be able to follow the developement up to the phase of homologous expansion ( @xmath10 days ) , the 3-d structure is remapped on a 1-d grid , and the further evolution is calculated using a one - dimensional radiation - hydro code ( e.g. hflich et al .
1998 ) that solves the hydrodynamical equations explicitly in the comoving frame by the piecewise parabolic method ( colella and woodward 1984 ) .
* radiation transport : * detailed polarization and flux spectra for asymmetric explosions are calculated using our monte carlo code including detailed equations of state . for details , see hflich ( 1995 ) , hflich et al .
( 1995 ) & wang et al .
( 1998 ) . *
the setup : * the computational domain is a cube of size @xmath11 with a spherical star of radius @xmath12 and mass @xmath13 placed in the center . the innermost part with mass @xmath14 and radius @xmath15 cm , consisting of fe and si ,
is assumed to have collapsed on a timescale much faster than the outer , lower - density material .
it is removed and replaced by a point gravitational source with mass @xmath16 representing the newly formed neutron star .
the remaining mass of the envelope @xmath17 is mapped onto the computational domain . at two polar locations where the jets are initiated at @xmath18
, we impose an inflow with velocity @xmath19 @xmath20 . at @xmath21 ,
the jet density and pressure are the same as those of the background material .
for the first 0.5 s , the jet velocity at @xmath21 is kept constant at @xmath19 .
after 0.5 s , the velocity of the jets at @xmath21 was gradually decreased to zero at approximately 2 s. the total energy of the jets is @xmath22 .
these parameters are consistent within , but somewhat less than , those of the leblanc - wilson model .
as a baseline case , we consider a jet - induced explosion in a helium star .
jet propagation inside the star is shown in fig .
as the jets move outwards , they remain collimated and do not develop much internal structure . a bow shock forms at the head of the jet and spreads in all directions , roughly cylindrically around each jet .
the jet - engine has been switched off after about 2.5 seconds the material of the bow shock continues to propagate through the star .
the stellar material is shocked by the bow shock .
mach shocks travels two wards the equator resulting in a redistribution of the energy .
the opening angle of the jet depends on the ratio between the velocity of the bow shock to the speed of sound . for a given star ,
this angle determines the efficiency of the deposition of the jet energy into the stellar envelope . here , the efficiency of the energy deposition is about 40 % , and the final asymmetry of the envelope is about two .
2 shows two examples of an explosion with with a low and a very high jet velocity compared to the baseline case ( fig .
( fig.1 ) .
fig.2 demonstrates the influence of the jet velocity on the opening angle of the jet and , consequently , on the efficiency of the energy deposition .
for the low velocity jet , the jet engine is switched off long before the jet penetrates the stellar envelope .
almost all of the energy of the jet goes into the stellar explosion . on a contrary , the fast jet ( 61,000 km / sec ) triggers only a weak explosion of 0.9 foe although its total energy was @xmath23 . for a very extended star , as in case of normal type ii supernovae ,
the bow shock of a low velocity jet stalls within the envelope , and the entire jet energy is used to trigger the ejection of the stellar envelope . in our example ( fig .
3 ) , the jet material penetrates the helium core at about 100 seconds .
after about 250 seconds the material of the jet stalls within the hydrogen rich envelope and after passing about 5 solar masses in the radial mass scale of the spherical progenitor . at this time
, the isobars are almost spherical , and an almost spherical shock front travels outwards .
consequently , strong asphericities are limited to the inner regions .
after about 385 seconds , we stopped the 3-d run and remaped the outer layers into 1-d structure , and fullowed the further evolution in 1-d .
after about 1.8e4 seconds , the shock front reaches the surface .
after about 3 days , the envelope expands homologously .
the region where the jet material stalled , expands at velocities of about 4500 km / sec .
* fallback : * jet - induced supernovae have very different characteristics with respect to fallback of material and the innermost structure . in 1-d calculations and for stars with main sequence masses of less than 20 @xmath24 and explosion energies in excess of 1 foe ,
the fallback of material remains less than 1.e-2 to 1.e-3 @xmath24 and an inner , low density cavity is formed with an outer edge of @xmath25 . for explosion energies between 1 and 2 foe ,
the outer edge of the cavity expands typically with velocities of about 700 to 1500 km / sec ( e.g. woosley 1997 , hflich et al .
in contrast , we find strong , continuous fallback of @xmath26 in the the 3-d hydro models , and no lower limit for the velocity of the expanding material ( fig .
4 ) . this significant amount of fallback must have important consequences for the secondary formation of a black hole .
the exact amount and time scales for the final accretion on the neutron star will depend sensitively on the rotation and momentum transport .
* chemical structure : * the final chemical profiles of elements formed during the stellar evolution such as he , c , o and si are butterfly- shaped whereas the jet material fills an inner , conic structure ( figs . 3 and 5 ) .
the composition of the jets must reflect the composition of the innermost parts of the star , and should contain heavy and intermediate - mass elements , freshly synthesized material such as @xmath25 and , maybe , r - process elements because , in our examples , the entropy at the bow shock region of the jet was as high as a few hundred . in any case , during the explosion , the jets bring heavy and intermediate mass elements into the outer h - rich layers .
we have numerically studied the explosion of core collapse supernovae caused by supersonic jets generated in the center of the supernova as a result of the core collapse into a neutron star .
we simulated the process of the jet propagation through the star , and the redistribution of elements . a strong explosion and a high efficiency for the conversion of the jet energy requires low jet velocities or a low , initial collimation of the jet .
with increasing extension of the envelope , the conversion factor increases .
typically , we would expect higher kinetic energies in sne ii compared to sne ib / c if a significant amount of explosion energy is carried away by jets . within the framework of jet - induced sn
, the lack of this evidence suggests that the jets have low velocities . for the compact progenitors of sne
ib / c , the final departures of the iso - density contours from sphericity are typical a factor of two .
this will produce a linear polarization of about 2 to 3 % ( fig .
6 ) consistent with the values observed for type ib / c supernovae . in case of a red supergiant , i.e. sne ii
, the asphericity is restricted to the inner few solar masses . in the latter case ,
the iso - densities show an axis ratios of up to @xmath27 1.4 at the innermost , hydrogen - rich layers .
the outer layers remain spherical .
this has strong consequences for the observations , in particular , for polarization measurements . in general
, the polarization should be larger in sne ib / c compared to classical sne ii which is consistent with the observations by wang et al .
early on , we expect no or little polarization in supernovae with a massive , hydrogen rich envelope which will increase with time to about 1 % ( hflich 1991 ) , depending on the inclination the supernovae is observed .
this is also consistent both with the long - term time evolution of sn1987a ( e.g. jefferies 1991 ) and , in particular , the plateau supernova 1999em which has been observed recently with vlt and keck ( wang et al .
2000 ; leonard et al .
2000 ) .
the he , c , o and si rich layers of the progenitor show characteristic , butterfly - shape structures .
this overall morphology and pattern should be observable in supernovae remnants , e.g. with the chanda observatory despite some modifications and instabilities when the expanding medium interacts with the interstellar material . during the explosion ,
the jets bring heavy and intermediate mass elements into the outer layers including @xmath25 . due to the high entropies of the jet material close to the center
, this may be a possible site for r - process elements .
spatial distribution of the jet material will influence the properties of a supernova . in our model for a sn ii , the jet material stalled within the expanding envelope corresponding to a velocity of @xmath28 during the phase of homologous expansion . in sn1987a ,
a bump in spectral lines of various elements has been interpreted by material excited by a clump of radioactive @xmath25 ( lucy 1988 ) . within our framework
, this bump may be a measure of region where the jet stalled .
this could also explain the early appearance of x - rays in sn1987a which requires strong mixing of radioactive material into the hydrogen - rich layers ( see above ) .
we note that , if this interpretation is correct , the mystery spot ( nisenson et al . 1988 ) would be unrelated .
in contrast to 1-d simulations , we find in our models strong , continuous fallback over an extended period of time , and a lack of an inner , almost empty cavity .
this significant amount of fallback and the consequences for the secondary formation of a black hole shall be noted .
moreover , fallback and the low velocity material may alter the escape probablity for @xmath29-rays produced by radioactive decay of @xmath25 . in general ,
the lower escape probability is unimportant for the determination of the total @xmath25 production by the late lcs because full thermalization can be assumed in core collapse sn during the first few years .
however , in extreme cases such as sn98bw ( e.g. schaefer et al . 1999 ) , only a small fraction of gamma s are trapped .
effects of multi - dimensionallity will strongly alter the energy input by radioactive material and disallow a reliable estimate for the total @xmath25 mass .
finally , we want to emphasize the limits of this study and some of the open questions which will be addressed in future .
we have assumed that jets are formed in the course of the formation of a neutron star , and have addressed observational consequences and constrains .
however , we have not calculated the jet formation , we do not know if they really form , and , if they form , whether they form in all core - collapse supernovae
. we can not claim that the jets are the only mechanism that can explain asphericity in supernovae although we are not aware of the others .
qualitatively , the observational properties of core collapse supernovae are consistent with jet - induced supernovae and support strongly that the explosion mechanism is highly aspherical but no detailed comparison with an individual object has been performed .
_ acknowledgments : _ we want to thank our collegues for helpful discussions , in particular , e.s .
oran , l. wang , j.c .
wheeler , inzu yi a. , c. mayers , j.c .
wilson , a. chieffi , m. limongi , and o. straniero .
this work is supported in part by nasa grant lsta-98 - 022 .
arnett w.d . , bahcall j.n . ,
kirshner , r.p . ,
woosley , s.e .
1990 , araa 27 , 62 benz w. , thielemann f .- k .
1990 , apj 348 , 17 bisnovatyi - kogan 1971 , soviet astronomy aj , 14 , 652 bowers r.l . ,
wilson j.r .
1982 apjs 50 , 115 colella , p. ; woodward , p.r .
1984 , j.comp.phys .
54 , 174 duncan , r. c. & thompson , c. 1992 , apj , 392 , l9 fesen , r. a. & gunderson , k. s. 1996 , apj , 470 , 967 fryxell b. , arnett d. , mller e. 1991 , apj 367 , 619 herant m. , benz w. , hix w.r .
, fryer c.l .
, colgate s.a .
1994 , apj 435 , 339 hillebrandt w. , hflich 1991 , nuclear physics b 19 , 113 hillebrandt w. 1982 , apj 103 , 147 hflich p. , straniero o. , limongi m. dominguez i. chieffi a. 2000 , 7th texmex- conference , eds
. w. lee & s. torres - peimbert , unam - publ .
, in press & astro - ph/005037 hflich , p. , wheeler , j. c. , and thielemann , f.k .
1998 , apj 495 , 617 hflich , p. 1995
, apj 443 , 89 hflich p. , wheeler , j.c . , hines , d. , trammell s. 1995b , apj 459 , 307 hflich , p. 1991
a&a 246 , 481 hoyle f. , fowler w. a. 1960 , apj 132 , 565 huges j.p . , rakowski c.e .
, burrows d.n .
, slane p.o .
2000 , aj , in press & astro - ph/9910474 janka h.t . & mller e. , 1994 a&a 290 , 496 jeffrey d.j . , 1991 , apj , 375 , 264 kifonidis k. , plewa t. , janka h.t .
, mller e. 2000 , apj 531 , 123 khokhlov , a.m. , oran , e.s .
2000 , combustion & flame , in press .
khokhlov a. , hflich p. , oran e.s . , wheeler j.c . , p. wang l. , 1999 , apj 524 , l107 khokhlov , a.m. , oran , e.s . , chtchelkanova , a.yu . , wheeler , j.c . 1999a , combustion & flame , 117 , 99 khokhlov , a.m. , oran , e.s . , thomas , g.o . , 1999b , combustion & flame , 117 , 323 khokhlov , a.m. 1998 , j.comput.phys .
, 143 , 519 kouveliotou , c. , strohmayer , t. , hurley , k. , van paradijs , j. , finger , m. h. , dieters , s. , woods , p. , thompson , c. & duncan , r. c. 1998 , apj , 510 , 115 lucy l.b .
1988 , proc . of the 4th george mason conference , ed . by m. kafatos , cambridge university press , p. 323
leblanc , j. m. & wilson , j. r. 1970 , apj , 161 , 541 leonard d.c .
, filippenko , a.v . , barth a.j . , matheson t. 2000 , apj 536 , 239 mezzacappa a. , liebendoerfer m. , bronson messer o.e .
hix r. , thielemann f - k , bruenn s.w .
2000 , phy.rev.let . , accepted mnchmeyer r. , schaefer g. , mueller e. , kates r.e .
1991 a&a 246 , 417 mller e. , janka h.t .
1997 , a&a 317 , 140 mller e. , hillebrandt w. , orio m. , hflich p. , mnchmeyer r. , fryxell b.a .
1989 , a&a 220 , 167 nisenson p. , papaliolios c. , karovska m. , noyes r. 1988 , apj 324 , 35 ostriker , j. p. & gunn , j. e. 1971 , apj , 164 , l95 rammp m. , mller e. , ruffert m. 1998 , a&a 332 , 969 rammp m. and janka , h .-
2000 , apj 593 , l33 schaefer b. 2000 , apj 533 , 21 saenz r.a . ,
shapiro s.l .
1981 , apj 244 , 1033 steinmetz m. , hflich p. 1992 , a&a 257 , 641 symbalisty e.m.d .
1984 , apj 285 , 729 straniero , o. chieffi , a. and limongi m. 1999 apj , in press strom r. , johnston h.m .
, verbunt f. , aschenbach b. 1995 , nature , 373 , 587 symbalisty e.m.d .
1984 , apj 285 , 729 trammell s. , hines d. , wheeler j.c .
1993 , apj 414 , 21 tueller j. , barthelmy s. , gehrels n. , leventhal m. , maccallum c.j . , teegarden b.j . 1991 , in : supernovae , ed .
woosley , springer press , p. 278
van riper k.a .
1978 , apj 221 , 304 wang l. , howell a. , hflich p. , wheeler j.c .
2000 , apj , in press wang , l. , wheeler , j. c. , li , z. w. , & clocchiatti , a. 1996 , apj , 467 , 435 wang , l. , wheeler , j.c . ,
hflich , p. 1997 , apj , 476 , 27 yamada s. , janka h.t . , suzuki h. 1999 , a&a 344 , 533 zwerger t. , mller e. 1997 , a&a 320 , 209 | core collapse supernovae ( sn ) are the final stages of stellar evolution in massive stars during which the central region collapses , forms a neutron star ( ns ) , and the outer layers are ejected .
recent explosion scenarios assumed that the ejection is due to energy deposition by neutrinos into the envelope but detailed models do not produce powerful explosions .
there is mounting evidence for an asphericity in the sn which is difficult to explain within this picture .
this evidence includes the observed high polarization , pulsar kicks , high velocity iron - group and intermediate - mass elements material observed in remnants , etc .
the discovery of highly magnetars revived the idea that the basic mechanism for the ejection of the envelope is related to a highly focused mhd - jet formed at the ns .
our 3-d hydro simulations of the jet propagation through the star confirmed that the mechanism can explain the asphericities . in this paper ,
detailed 3-d models for jet induced explosions of `` classical '' core collapse supernovae are presented .
we demonstrate the influence of the jet properties and of the underlaying progenitor structure on the final density and chemical structure .
finally , we discuss the observational consequences , predictions and tests of this scenario . |
patterns in effective two - dimensional ( 2d ) fluids flows are found in nature in various contexts @xcite : atmospheric @xcite and oceanic flows @xcite are examples . to understand such large scale patterns
theoretically , the 2d euler equation describing perfect fluids flows is a simplified starting point . in this context
, the study of nonlinear structures , such as kelvin s cat s eyes , over a shear flow has a long history @xcite , which includes more recent mathematical developments @xcite .
statistical physics has often been invoked to explain the formation of large vortices since onsager @xcite .
the miller - robert - sommeria ( mrs ) theory @xcite , which constructs the microcanonical measure for 2d euler flows by taking all the invariants into account , is a particularly successful achievement .
a difficulty in applying this theory in practice is the fact that the 2d euler equation has infinitely many invariants .
more importantly , the theory assumes the vorticity field on the large scales to be stationary , and therefore can not describe non - stationary asymptotic behaviors .
for these cases , the statistical physics approach should be supplemented by a dynamical understanding of pattern formations . in this paper , we consider a shear base flow to which a perturbation is added , and propose a phenomenological approach to the pattern formations by analyzing the landau pole of the linearized equation with its bifurcation , and by taking into account nonlinear effects .
detailed numerical simulations of this situation are presented in ref .
@xcite : starting from a kolmogorov flow on the doubly periodic domain @xmath0 , a perturbation of size @xmath1 is added ; depending on the parameters @xmath2 , the perturbation may fully damp , evolve into a stationary dipole , or create four long lived traveling vortices .
the regions on the @xmath2 plane where the dipole or traveling vortices appear are numerically investigated in ref .
@xcite , but a theoretical understanding is lacking . in this article , we aim at providing such a theoretical explanation , by exploiting the analogy between the 2d euler equation for the vorticity field and the vlasov equation for plasmas . indeed , a similar phenomenon has been investigated in one - dimensional plasmas , described by the vlasov equation .
a naive linear theory predicts that perturbations added to stable stationary states damp exponentially @xcite .
it is well known , however , that , if the perturbation is large enough , nonlinear effects come into play , prevent complete damping , and may create traveling clusters @xcite ; see also ref .
@xcite for the simpler cases of ferromagnetic and anti - ferromagnetic hamiltonian mean - field model . to be more precise , according to refs .
@xcite , such a phenomenon occurs when the two following criteria are satisfied : 1 . the landau damping time scale is longer than a nonlinear timescale , called `` trapping timescale '' .
if several clusters are formed , they should be so small that they do not overlap and that a nonlinear superposition approximation @xcite may work .
although these criteria are in view of nonlinear dynamics , they can be expressed using the dominant landau pole computed from the linearized euler equation .
our goal is to draw the phase diagram in the @xmath2 plane by using the above two criteria in the context of the euler equation , that is , by combining the phenomenological nonlinear estimate and the linear landau damping theory .
this article is organized as follows . in sec .
[ sec : conditions ] , we qualitatively analyze criteria 1 and 2 , and show that they are related to the imaginary and the real parts of the dominant landau pole of the initial kolmogorov flow .
we therefore briefly review the linear theory for 2d incompressible and inviscid fluids , and derive the dispersion relation in sec .
[ sec : linear ] ; the computation of the landau pole requires an analytic continuation procedure similar to the one used in ref .
@xcite . using these computations
, we draw a phase diagram in the @xmath2 plane in sec . [
sec : phased ] , and compare this phenomenological estimate with numerical simulations in sec .
[ sec : num ] .
section [ sec : sum ] is devoted to the summary and discussions .
we start from the 2d euler equation in the domain @xmath3 @xmath4 where the vorticity field @xmath5 and the velocity field @xmath6 are related to a stream function @xmath7 through @xmath8 the periods for @xmath9 and @xmath10 axes are set as @xmath11 and @xmath12 respectively .
we consider the stationary kolmogorov flow , called hereafter the `` base flow '' , whose stream function is @xmath13 and vorticity and velocity fields are respectively @xmath14 with @xmath15 .
we add a large scale perturbation to the base flow , and expand the stream function as @xmath16 with @xmath17 in our analytical computations , @xmath1 is assumed to be small ; this restriction obviously does not hold for numerical simulations .
nevertheless , the theory will be qualitatively in good agreement with the simulations .
when the base flow is stable ( this corresponds to an aspect ratio @xmath18 ) , the linear theory typically predicts that the perturbation damps and possibly oscillates at complex frequency @xmath19 , where @xmath20 is the root of the dispersion relation yielding the slowest damping , that is , the root closest to the real axis , or , in other words , with the smallest @xmath21 .
the idea is that while it damps , the perturbation will tend to create structures traveling at velocity @xmath22 ; if nonlinear effects are strong enough , these structures may persist for long times .
if @xmath23 , one may expect traveling vortices , while the formation of two stationary vortices ( a dipole ) should be favored if @xmath24 .
the damping time scale @xmath25 is easily estimated as the inverse of the landau damping rate @xmath26 .
actually , in addition to the exponential landau damping described by @xmath27 , there is an algebraic damping coming from the branch points of the dispersion function ( see sec . [
sec : linear ] ) ; however , the exponential landau damping is expected to be dominant on the time scale considered in this paper .
next , we estimate the `` trapping time scale '' , @xmath28 , which is the characteristic time scale concerning nonlinearities @xcite , as the period of a test point vortex trapped around the edge of a small vortex .
the temporal evolution of the position of a test point vortex is governed by the velocity field @xmath29 where the dot denotes @xmath30 .
we approximate the perturbation @xmath31 phenomenologically .
we are interested in the macroscopic behavior corresponding to the @xmath32 mode , where @xmath33 is the wave number with respect to @xmath9 , and we assume that the damping perturbation has created small vortices whose velocity is @xmath22 in the @xmath9 direction ; note that @xmath22 can be @xmath34 .
the first order @xmath31 can be , therefore , approximated by @xmath35 where @xmath36 is of @xmath37 . substituting eq . into eq .
, we have the approximate equations of motion @xmath38 where the prime represents @xmath39 .
thanks to the phenomenological approximation of @xmath40 , we can compute the motion of the test point vortex by perturbation techniques . expanding @xmath9 and @xmath10 into series of @xmath1 as @xmath41
we obtain the solutions @xmath42 and @xmath43 with initial conditions @xmath44 and @xmath45 : @xmath46 equations for the first order in @xmath1 are : @xmath47 from the frequency of @xmath48 in eq . , we estimate the trapping time scale @xmath28 as @xmath49 where @xmath50 satisfies @xmath51 , and @xmath52 represents the width of a vortex . the unknown quantity , the width of a vortex @xmath53 , is determined self - consistently . the solution @xmath54 is , from eq . ,
@xmath55,\ ] ] where the initial condition is @xmath56 .
the amplitude in the @xmath10 direction , @xmath57 , must be the same as the width @xmath53 at the edge of the vortex , @xmath58 .
thus , the self - consistent equation for @xmath59 is @xmath60 where we have introduced another phenomenological approximation by replacing @xmath61 with @xmath62 , since it is of @xmath37 and we are looking for an order - of - magnitude estimate .
the width @xmath59 is , therefore , estimated as @xmath63 for the kolmogorov base flow , @xmath15 , we have @xmath64 and the trapping time scale is @xmath65 criterion 1 reads @xmath66 , that is : @xmath67
if @xmath24 , this condition simplifies into @xmath68 . according to eq . , we consider the modes @xmath32 .
we assume that the frequency @xmath22 of the perturbation reflects the existence of periodically moving vortices , thanks to the periodic boundary condition for the @xmath9 axis . the period @xmath11 permits to identify the frequency with the velocity of the moving vortices , and therefore the @xmath10 positions of the vortices are estimated as the solutions of the equation @xmath69 .
we recall the base flow is @xmath15 .
if @xmath24 , one expects that vortices are formed at @xmath70 and @xmath71 .
the estimate of the vortices width shows that the vortices will never overlap for any reasonably small @xmath1 ( say for instance @xmath72 ) .
hence criterion 2 brings no restriction in this case .
if @xmath23 , in contrast , the vortices can be close one to another , and criterion 2 leads to a restriction .
we name the four vortices a , b , c , and d , whose @xmath10 positions are , respectively , @xmath73 , @xmath74 , @xmath75 and @xmath76 in the interval @xmath77 , where we take the branch of solutions @xmath78 ; see fig .
[ fig : schematic ] . ,
suitably rescaled .
the two vertical dotted lines correspond to @xmath79 ; their intersections with the velocity curve , @xmath69 , yield the estimated @xmath10 positions of the four vortices , which travel along the horizontal dotted lines .
, width=264 ] we assume that each vortex has the same width @xmath59 .
then the distance between two nearby vortices must be larger than @xmath80 in order to avoid an overlap .
using @xmath81 , this condition is expressed as @xmath82 the first inequality comes from the distance in @xmath10 between vortices a and b ( identically c and d ) , and the second from the distance between b and c ( identically a and d ) . the pair ( b , c ) moves toward the left while ( a , d ) toward the right according to the base flow @xmath83 . moreover
, the numerical observations @xcite say that the difference in @xmath9 among each pair stays around @xmath84 .
thus the distance between b and c ( a and d ) is large enough to neglect the effect of overlapping ; see fig .
[ fig : schematic ] .
hence we can omit the second inequality of eq .
( 19 ) , and criterion 2 is finally expressed as : @xmath85 the two conditions and involve the parameters @xmath86 and @xmath1 , as well as the dominant landau pole @xmath19 . in the following section , we turn to the computation of this landau pole .
this section contains a classical computation for the linearized 2d euler equation , as well as an analytic continuation in the spirit of ref .
@xcite , for the self - consistency of the paper .
we perform it for a general base flow and perturbation , and then specialize it to eqs . and .
we linearize eqs . and around the base flow .
we add a small perturbation @xmath40 to @xmath88 , whose associated vorticity and velocity fields are denoted respectively as @xmath89 and @xmath90 .
substituting @xmath91 and @xmath92 into the euler equation gives the linearized 2d euler equation @xmath93 using the perturbation @xmath94 and eq .
, we rewrite the linearized euler equation as @xmath95 our goal is now to derive the dispersion relation for the linearized euler equation ; we use a fourier - laplace transformation , and follow the route of ref .
thanks to the periodicity with respect to @xmath9 , we expand @xmath96 and @xmath94 into fourier series as @xmath97 substituting the fourier expansions into the linearized euler equation , we obtain the equation for @xmath33-th fourier mode as @xmath98 the laplace transform of a function @xmath99 with respect to @xmath100 is defined as @xmath101 where the condition @xmath102 is introduced to ensure the convergence of the integral . performing the laplace transform of the linearized euler equation , we obtain an algebraic equation for @xmath103 and @xmath104 : @xmath105 where @xmath106 ; note that @xmath107 .
the fourier transform of eq . with respect to @xmath9 @xmath108
gives a closed equation for @xmath109 , which is the rayleigh equation @xmath110 the stream function @xmath111 is computed through an inverse laplace transform .
its asymptotic behavior is determined by the singularities of @xmath109 with respect to the complex variable @xmath20 .
for simplicity , we introduce the functions @xmath112 and @xmath113 as @xmath114 these functions have no singularity for real @xmath10 , since @xmath20 is defined in the region @xmath115 . fixing the complex variable @xmath20
, we analyze the rayleigh equation of the form @xmath116 and the corresponding homogeneous rayleigh equation @xmath117 let @xmath118 and @xmath119 be independent solutions to the homogeneous equation with boundary conditions @xmath120 the particular solution @xmath121 to the inhomogeneous equation is then given by @xmath122 indeed , the double derivative of @xmath123 is @xmath124 where @xmath125 is the wronskian @xmath126 and is constant @xmath127 hence , the general solution @xmath128 to eq . is : @xmath129 where @xmath130 and @xmath131 are constants determined from the periodic boundary condition @xmath132 from the boundary conditions and @xmath133 , eqs . and lead to : @xmath134 remembering that @xmath135 and hence @xmath136 and @xmath137 depend on @xmath20 , we define the function @xmath87 as @xmath138 to show the last equality in eq .
, we have used the fact that the wronskian @xmath125 is unity .
the values of @xmath139 and @xmath140 are computed by integrating the homogeneous rayleigh equation from @xmath34 to @xmath12 .
the general solution has singularities for @xmath20 satisfying @xmath141 , and these singularities yield non - trivial modes proportional to @xmath142 by the inverse laplace transform .
this justifies our terminology `` dispersion relation '' for @xmath141 , and `` dispersion function '' for the function @xmath87 .
recall that the dispersion function @xmath87 is a priori defined only in the region @xmath143 to ensure the convergence of the laplace transform .
now , to find roots giving stable modes , we analytically continue @xmath87 to the whole complex @xmath20 plane , in a similar manner to what has been performed for the euler equation on the 2d disc @xcite .
hereafter we set @xmath144 without loss of generality , and take the particular form of the base flow : @xmath15 .
corresponding to a given @xmath20 , the functions @xmath135 and @xmath145 have singularities at which the equation @xmath146 is satisfied .
such singular points are , together with the integration paths of the homogeneous rayleigh equation , schematically illustrated in fig .
[ fig : continuation ] for both the cases @xmath147 and @xmath148 .
the integration paths are determined as follows .
for @xmath20 on the upper half plane , @xmath87 is simply defined by integrating eq . along the real @xmath10 axis , namely along the integration path : @xmath149 continuously moving @xmath20 to the lower half plane allows to define @xmath87 for @xmath150 .
if the path taking @xmath20 from the upper to the lower half plane crosses the singular interval @xmath151 $ ] , we have to avoid the singularity that would be created . for this purpose
, we deform the integration path @xmath152 to @xmath153 using a @xmath154-class , real - valued , and 1-periodic function @xmath155 satisfying @xmath156 ; correspondingly , this amounts to deform the singular line @xmath157 which doubly covers the interval @xmath158 $ ] , to the curve @xmath159 we choose the function @xmath155 so that the deformed curve @xmath160 is below the @xmath20 for which we want to define @xmath87 .
this procedure is similar to the spectral deformation for the vlasov - poisson equation in refs .
@xcite .
if the path taking @xmath20 from the upper to the lower plane crosses the real axis through either one of the half lines @xmath161 or @xmath162 , no singularity is crossed , and the analytic continuation does not require any deformation of the integral path ; in the lower half plane , @xmath87 is multi valued unless a riemann surface associated with the branch points @xmath163 is introduced .
we choose the particular deformation function @xmath155 in eq . as @xmath164
the extended domains of @xmath87 are shown in fig .
[ fig : sp - deform ] for two different values of @xmath165 .
we use @xmath165 large enough so that @xmath166 is below @xmath20 ; the computational cost increases exponentially with increasing @xmath165 . and
@xmath10 planes for @xmath144 .
the left two panels are for @xmath167 , and the right two panels for @xmath168 .
the integration path @xmath152 on the @xmath10 plane , consisting of the solid and the dotted lines , is mapped to the double covering of the interval @xmath158 $ ] on the @xmath20 plane by the mapping @xmath169 . corresponding to a given cross point on the @xmath20 plane ,
there are two singular points ( crosses ) of @xmath135 and @xmath145 on the @xmath10 plane in general .
as the point on the @xmath20 plane goes down , the singular points on the @xmath10 plane pass the real axis and the integration path @xmath152 must be deformed to @xmath170 to avoid the singularities .
the mapping of @xmath170 , denoted by @xmath171 , determines the boundary of the continued domain of the dispersion function @xmath87 .
[ fig : sp - deform ] . a branch cut on the @xmath20 plane is set as the boundary.,width=264 ] .
the gray regions represent the domains of @xmath87 for ( a ) @xmath172 and ( b ) @xmath173 .
the curve in the @xmath20 plane represents @xmath174 , where @xmath175 is the deformed integration path for the homogeneous rayleigh equation . with increasing @xmath165 , the domain of @xmath87
is extended to a larger area in the lower half @xmath20 plane , but the computational cost also increases.,width=264 ] by introducing the following functions of @xmath176 : @xmath177 the homogeneous rayleigh equation is rewritten as @xmath178 where the subscripts `` r '' and `` i '' represent the real and imaginary parts respectively , and the prime denotes the derivative with respect to @xmath176 .
the continued solutions @xmath179 and @xmath180 are computed by solving eq . with the boundary conditions @xmath181 respectively .
solving eq .
, we obtain @xmath182 and @xmath183 , and the dispersion function is expressed as @xmath184 whose real and imaginary parts are respectively @xmath185 we now specialize the above computations to the initial perturbation @xmath186 as in the numerics . using the analytical continuation of @xmath87 into the lower half plane , we numerically compute landau poles , i.e. the roots of the equation @xmath141 with @xmath168 .
we are interested in the traveling vortices , whose velocity is @xmath22 , in resonance to the base flow , whose velocity @xmath83 is in the range @xmath158 $ ] .
it is therefore natural to choose the analytical continuation of @xmath87 defined by deforming the integral path as in eq . .
we look for the dominant landau pole , i.e. , the root of @xmath87 with the largest imaginary part ; in a stable situation , it corresponds to the slowest damping .
the variation of this pole as a function of the aspect ratio @xmath86 is shown on fig .
[ fig : bifurcation ] . for @xmath187 ,
the flow is unstable . with increasing @xmath86 above @xmath62 ,
the flow becomes stable , with the main landau pole on the imaginary axis . with further increasing @xmath86 ,
the dominant landau pole undergoes a bifurcation at an aspect ratio @xmath188 and acquires a non - zero real part .
we will see that this bifurcation is crucial to understand the appearance of a stationary dipole or of the four traveling vortices . plane . when @xmath187 , one dominant root is obtained on the upper half plane , where the base flow is unstable , and the root is at the origin for @xmath189 . with increasing @xmath86
, the main root goes down to the lower half plane along the imaginary @xmath20 axis , accompanied by another root going up the axis .
when @xmath190 , these two roots collide at @xmath191 and they bifurcate to the right and left directions .
the gray region is given by the inequality , @xmath192 coming from eq .
, and @xmath193 required by the stability of the flow . for later convenience , we marked the bifurcation point corresponding @xmath194 by the cross , and the point at which the main pole enters the gray region by thestars .
, width=302 ]
using the theoretical analyses above , we construct the phase diagram on the @xmath2 plane , which consists of zonal , dipole and oscillation phases @xcite .
the last oscillation phase corresponds to the four traveling vortices . to obtain the region on the @xmath2 plane where the appearance of four traveling vortices is expected ,
we combine criteria 1 and 2 , that is eqs . and respectively , with the result on the dominant landau pole . satisfying both criteria then reads : @xmath195 the right inequality requires a landau pole with non - zero @xmath22 , which implies @xmath196 .
the region which satisfies eq . on the @xmath197 plane
is reported in fig .
[ fig : phasediagram ] , which is qualitatively in good agreement with the numerically obtained oscillatory region @xcite .
we remark that @xmath198 increases when @xmath86 is increased above @xmath199 , staying smaller than @xmath200 in the reported parameter region .
this fact validates that we have omitted the lower condition in eq . to derive the second necessary condition .
we expect to find a stationary dipole when criterion 1 is satisfied and @xmath24 : @xmath201 the region satisfying these requirements is highlighted in fig .
[ fig : phasediagram ] .
this theoretically estimated region is again in qualitative agreement with the region where a stationary dipole emerges in numerical simulations @xcite .
we expect a zonal flow when criterion 1 is not satisfied , that is @xmath202 physically , it means that nonlinearity is not dominant and hence simple damping of the perturbation is expected . in fig .
[ fig : phasediagram ] the zonal region appears for small @xmath1 . on fig .
[ fig : phasediagram ] , there is a region where the theory makes no prediction ; indeed , the vortices that would be created by the phenomenological mechanism considered above would be so close one to another that they would strongly interact ; the final fate of the system is then out of scope of the present theory .
we note that in this region , zonal flows , dipoles and traveling vortices are observed numerically @xcite . plane . the gray , polka - dotted and checkered regions correspond to the oscillatory , zonal and dipole phases , respectively .
the upper panel is drawn by the present theory and the lower is a schematic picture of the phase diagram obtained with the numerical simulations by morita @xcite . in the upper panel , the gray region
is obtained from the inequality .
the left green boundary of the gray region is given by @xmath203 ( see eq . ) , and the lower red by @xmath204 ( see eq . ) .
the left blue boundary of the checkered region is given by @xmath205 , and the upper purple by @xmath194 ( see eq . ) , whose value is determined at the bifurcation point marked by the cross in fig . [
fig : bifurcation ] .
the star point corresponds to the star point of fig .
[ fig : bifurcation ] .
the broken line represents @xmath206 on which the theoretical and numerical frequencies are compared in fig .
[ fig : comparison ] .
the points marked by the cross and the circle correspond to those in fig .
[ fig : comparison ] , respectively.,title="fig:",width=302 ] plane .
the gray , polka - dotted and checkered regions correspond to the oscillatory , zonal and dipole phases , respectively .
the upper panel is drawn by the present theory and the lower is a schematic picture of the phase diagram obtained with the numerical simulations by morita @xcite . in the upper panel , the gray region
is obtained from the inequality .
the left green boundary of the gray region is given by @xmath203 ( see eq . ) , and the lower red by @xmath204 ( see eq . ) .
the left blue boundary of the checkered region is given by @xmath205 , and the upper purple by @xmath194 ( see eq . ) , whose value is determined at the bifurcation point marked by the cross in fig . [
fig : bifurcation ] .
the star point corresponds to the star point of fig .
[ fig : bifurcation ] .
the broken line represents @xmath206 on which the theoretical and numerical frequencies are compared in fig .
[ fig : comparison ] .
the points marked by the cross and the circle correspond to those in fig .
[ fig : comparison ] , respectively.,title="fig:",width=302 ]
according to the phenomenological prediction of the preceding sections , the four traveling vortices run in the @xmath9 direction along the four lines @xmath207 in @xmath208 , where @xmath209 , and their velocities are @xmath210 .
the period in the @xmath9 direction is @xmath11 ; hence the phenomenological theory predicts the frequency @xmath211 for the vorticity field @xmath5 .
we examine our phenomenological prediction by numerically observing the @xmath10-positions of the four vortices and the oscillation frequency .
the initial condition of the vorticity field is : @xmath212 fixing the small parameter @xmath213 . in numerical simulations ,
we add a hyper - viscous term @xmath214 to the right hand side of the 2d euler equation , with @xmath215 and @xmath216 , for stabilizing the numerical scheme , and use the pseudo - spectral method with the resolution @xmath217 .
first , we observe the @xmath10-positions by computing the averaged vorticity field @xmath218 defined by @xmath219 at the initial time @xmath220 , the averaged vorticity field is @xmath221 , and it evolves in time .
asymptotically , we observe four bumps in @xmath218 as shown in fig . [ fig : baromega ] for @xmath222 and @xmath213 .
the theoretical prediction is in good agreement with the numerically computed bump positions . .
four bumps are observed , corresponding to four vortices in the long time regime .
the vertical four dotted lines indicate the theoretical predictions for the @xmath10-positions of the vortices : @xmath223 . @xmath222 and @xmath213.,width=302 ] next , we obtain the frequency by computing the power spectra of the @xmath224 fourier mode of @xmath225 defined by @xmath226 as shown in fig . [
fig : comparison ] , the dependence on @xmath86 of @xmath227 is in qualitative agreement with the prediction , in the sense that the frequency increases as @xmath86 increases .
however , it is not in good agreement quantitatively .
one possible explanation for the quantitative discrepancy in frequency is that @xmath213 is too large to be considered a small perturbation to the base flow .
the present theory is based on the linear analysis of the euler equation .
it gives good predictions qualitatively , but quantitatively , nonlinear effects may kick in for rather large @xmath1 .
dependence of the frequencies obtained from the theoretical prediction ( bold curve ) and the numerical simulations ( red points ) .
clear oscillation is not found by numerical simulations in the small @xmath86 region , @xmath228 .
the points marked by the cross and the circle correspond to those in fig .
[ fig : phasediagram ] , respectively.,width=302 ]
inspired by previous works on the vlasov equation @xcite , we have considered i ) the balance between landau damping and nonlinear effects and ii ) the non - overlapping criterion between nonlinear structures , in order to analyze the formation of a stationary dipole and traveling vortices in the 2d euler equation starting from the perturbed stable kolmogorov flow .
the detailed linear analysis provides information on this nonlinear dynamical phenomenon .
we emphasize that the bifurcation of the dominant landau pole as the aspect ratio is varied plays a crucial role in the theory .
since this kind of bifurcation is probably not a special feature of the kolmogorov flow , we expect that a similar analysis , predicting the appearance , or not , of a dipole or traveling vortices , would be fruitful for other base flows .
we note that our prediction is in qualitative , though not quantitative , agreement with the numerical results of ref .
@xcite , not only for the shape of the phase diagram , but also for the vortex frequency .
the quantitative discrepancy for the phase diagram is rather natural since we have performed phenomenological order - of - magnitude estimates and not taken the prefactors into consideration .
the reason for the quantitative disagreement in the frequency is probably that the size of the perturbation @xmath1 used in the numerical simulation is large enough to trigger nonlinear effects that we do not take into account , while such a large @xmath1 is necessary to realize the oscillating phase numerically .
the authors are grateful to f. bouchet for fruitful discussions and comments on this study .
so thanks for the hospitality of universit de nice sophia - antipolis and acknowledges the support of the jsps research fellowships for young scientists ( grant no .
this work was supported by the anr-09-jcjc-009401 interlop project .
yyy acknowledges the support of a grant - in - aid for scientific research ( c ) 23560069 . | a phenomenological theory is proposed to analyze the asymptotic dynamics of perturbed inviscid kolmogorov shear flows in two dimensions .
the phase diagram provided by the theory is in qualitative agreement with numerical observations , which include three phases depending on the aspect ratio of the domain and the size of the perturbation : a steady shear flow , a stationary dipole , and four traveling vortices .
the theory is based on a precise study of the inviscid damping of the linearized equation and on an analysis of nonlinear effects . in particular , we show that the dominant landau pole controlling the inviscid damping undergoes a bifurcation , which has important consequences on the asymptotic fate of the perturbation . |
after discovery of the conspicuous radial velocity ( rv ) pulsational variations in a sample of rapidly oscillating magnetic peculiar ( roap ) stars ( kanaan & hatzes , savanov , malanushenko & ryabchikova , kochukhov & ryabchikova for @xmath0 equ ; baldry et al . , baldry & bedding , kochukhov & ryabchikova for @xmath1 cirand hd 83368 ) , attempts to search for magnetic field variations over the pulsational period have been made .
first , hubrig et al .
@xcite tried to measure pulsational variability of the longitudinal magnetic field @xmath2 in six roap stars .
their sample included @xmath0 equ probably one of the most favourable stars for this kind of investigation .
equ is a bright northern roap star with a strong magnetic field , and with one of the largest pulsational rv amplitudes which exceeds 1000 ms@xmath3 in individual spectral lines .
the extremely slow rotation of @xmath0 equ , leading to very sharp spectral lines , makes this star the best candidate for any study of the pulsational variability in spectroscopy .
hubrig et al .
@xcite used low - resolution zeeman time - series observations and measured @xmath2 using hydrogen lines and unresolved blends of metal lines .
they failed to detect any variability beyond the formal errors of their measurements which were 40100 g. according to a coarse theoretical estimate made by hubrig et al .
@xcite there should exist a linear relation between magnetic field variability over the pulsational cycle and the rv amplitudes . in roap
stars the largest rv amplitudes are observed in the lines of first and second ions of rare - earth elements ( ree ) , while they are usually below measurement errors for the lines of iron group elements .
thus , high - resolution spectroscopy and spectropolarimetry of selected ree spectral lines is a more promising tool for an investigation of possible rapid magnetic oscillations in roap stars . taking this into account leone &
kurtz @xcite obtained a high - resolution ( r=115000 ) , high s / n time - series of observations of @xmath0 equ with a circular polarization analyzer , and measured @xmath2 using four ndiii lines .
they reported the discovery of pulsational variations with amplitudes between 112 to 240 g and , more surprisingly , with discrepant phases of magnetic maximum for different ndiii lines .
leone & kurtz s result was based on only 18 time - resolved spectra .
a year later kochukhov , ryabchikova & piskunov @xcite obtained a time - series of polarimetric observations of @xmath0 equ with a smaller resolving power ( r=38000 ) , but acquired more than 200 spectra over 3 nights , more than compensating for lesser quality of individual spectra .
kochukhov et al .
@xcite used simultaneously 13 ndiii lines for magnetic measurements which allowed them to achieve a substantial reduction of the error of the @xmath2 determinations .
they did not confirm longitudinal field variability over the pulsational period in @xmath0 equand gave a conservative upper limit of @xmath440 g for the amplitude of pulsational modulation of @xmath2 determined from ndiii lines .
another attempt to search for possible rapid magnetic variability in @xmath0 equ was made by savanov , musaev & bondar @xcite .
they measured the surface magnetic field @xmath5 variations over the pulsational period using the feii6149.25 line observed in unpolarized light .
due to a very simple zeeman pattern ( two equally separated @xmath6- and @xmath7-components ) this line is ideal for @xmath5 measurements ( see mathys et al . ) .
savanov et al .
reported a 1.8@xmath7 detection of @xmath5 variability with an amplitude of 99@xmath853 g. at the same time they did not find periodic variations of rv measured for the individual zeeman resolved components of the feii6149.25 doublet exceeding their error limit ( 100120 ms@xmath3 ) .
the authors used high - resolution r=120000 time - series observations , but the s / n of a single spectrum did not exceed 4060 .
clearly , the result of savanov et al .
@xcite is marginal and needs to be confirmed or rejected with data of better quality . in this paper
we present the results of a new search for pulsational variations of @xmath5 in @xmath0 equ using high - resolution and high s / n time - resolved observations of this star .
these observational data allowed us to obtain precise measurements of the magnetic field in @xmath0 equ , and strongly constrain possible changes of @xmath5 over the pulsation cycle of the star .
the time - resolved observations of @xmath0 equ were obtained using the single - order @xmath9 gecko coud spectrograph with the eev1 ccd at the 3.6-m canada - france - hawaii telescope .
table [ tbl1 ] gives details of this spectroscopic time - series dataset .
the spectra cover approximately the spectral window 61046194 .
this wavelength interval contains two zeeman resolved lines , feii 6149.25 and fei 6173.34 .
it also has strong ndiii and priiilines , optimal for investigation of pulsational variability of roap stars , as well as strong and weak lines of other elements such as si , ca , cr , and ba .
the spectra were reduced using standard iraf tasks .
each stellar , flat and calibration frame had a mean bias subtracted and was then cleaned of cosmic ray hits and extracted to one dimension .
extracted stellar spectra were divided by an extracted mean flat field , and the continuum was fit with a third - order legendre polynomial , using the same rejection parameters for all spectra so that the continuum fit is as uniform as possible .
the wavelength scale was established using about 40 lines of a thar emission lamp , resulting in an rms scatter about the adopted pixel - wavelength polynomial ( a sixth - order legendre polynomial ) of about @xmath10 .
the wavelength scale was linearly interpolated between thar lamp spectra taken before and after the stellar series , but the spectra were not resampled to a linear wavelength spacing . in this paper
we also used 31 very high resolution time - resolved observations of @xmath0 equ analysed by kochukhov & ryabchikova @xcite .
these time - resolved data were obtained with the coud echelle spectrograph ( ces ) , fiber - linked to the cassegrain focus of the eso 3.6-m telescope .
the highest resolution ces image slicer and the eso ccd#38 were used , allowing us to reach a resolving power of @xmath11 and record spectra in the 61406165 wavelength interval .
we refer the reader to kochukhov & ryabchikova @xcite for other details of the acquisition and reduction of the ces spectra of @xmath0 equ .
llcc line & @xmath12 ( min ) & @xmath13 ( ms@xmath3 ) & @xmath14 + + ndiii 6145.07 & [email protected] & [email protected] & [email protected] + feii 6149.25@xmath15 & 12.290 & [email protected] & [email protected] + feii 6149.25@xmath16 & 12.290 & [email protected] & [email protected] + + ndiii 6145.07 & [email protected] & [email protected] & [email protected] + feii 6149.25@xmath15 & 12.281 & [email protected] & [email protected] + feii 6149.25@xmath16 & 12.281 & [email protected] & [email protected] + fei 6173.34@xmath17 & 12.281 & [email protected] & [email protected] + fei 6173.34@xmath16 & 12.281 & [email protected] & [email protected] +
as a first step in our analysis of @xmath0 equ we computed synthetic spectra in the region of the zeeman resolved iron lines feii 6149.25 and fei6173.34 ( hereafter we refer to these spectral features simply as the feii and fei lines ) using abundances , fe stratification and model atmosphere from the study of ryabchikova et al .
synthetic spectra of @xmath0 equ were calculated with the code ( piskunov ) , modified to take into account vertical stratification of chemical abundances .
the simplest model , with a constant magnetic field over the stellar surface , was adopted .
the splitting pattern of the fei line is a pure zeeman triplet with unshifted central @xmath6-component and two @xmath7-components . to fit the relative intensity of the @xmath6- and @xmath7-components of this line the magnetic field vectors in our model have to be inclined by @xmath18 to the stellar surface . to look more carefully at the blending effects we took into account the hyperfine structure of nearby laii 6172.72 and euii 6173.03 .
an unidentified line at @xmath19 6148.85 , which shows a pulsational behaviour typical for ree lines , was synthesized with arbitrary atomic parameters to take into account its possible blending of the blue - shifted feii component . the comparison of these calculations with the average cfht spectrum is presented in fig . [ fig1 ] .
this figure shows that the feii line is free from significant blends , which enables accurate pulsational analysis of both the red and blue components of this zeeman doublet .
in particular , we found that pulsational variability of the 6148.85 feature has negligible influence on the measurements of the blue component of the feii line . on the other hand ,
the blue @xmath7-component of the fei line is blended by the euii 6173.03 and hence is not suitable for time - resolved measurements of @xmath5 .
consequently , rv analysis and field modulus measurements using the fei line were based on the @xmath6- and red @xmath7-component .
we determined the time - dependent position of the centres of zeeman components using centre - of - gravity measurements within the spectral regions indicated in fig .
we note that an alternative technique of fitting a superposition of three gaussian profiles to the 6148.85 and feii line represents a better way to derive @xmath5 from _ time - averaged _ spectra because it allows to model partially resolved feii components and remove the blending contribution of the 6148.85 feature .
however , this method may encounter difficulties in describing _ time - resolved _ profiles of ree lines , including the 6148.85 line , which show substantial rv shifts and exhibit profile asymmetries associated with non - radial pulsation velocity field . consequently
, we prefer to use the centre - of - gravity technique throughout this paper and show the gaussian fitting results in figs .
[ fig2][fig4 ] for comparison purpose only .
the rv obtained from the iron lines were compared with the outstanding pulsational variation seen in the ndiii 6145.07 line .
this feature allows us to verify the presence of rapid spectroscopic variability during our observations of @xmath0 equ and to determine the oscillation period and amplitude with high accuracy . for each available dataset we established the period of the spectroscopic pulsational variation using this ndiii line and
then fitted the phase ( counted from the start hjd , see table [ tbl1 ] ) and the amplitude of the fe rv measurements using a non - linear least - squares technique .
results of this analysis are summarized in table [ tbl2 ] .
we do not find any pulsational variation of the rv determined for the zeeman resolved components of the fe lines exceeding about 510 ms@xmath3for the feii and @xmath20 ms@xmath3 for the fei line .
the measurements of the feii line reported in table [ tbl2 ] supersede tentative rv amplitude of 64 ms@xmath3 given by kochukhov & ryabchikova @xcite , who used the same ces dataset as studied here .
this difference mainly comes from the fact that in the present paper we adopt a fixed pulsation period ( @xmath21 min ) for the analysis of the individual feii zeeman components , while a set of four photometric periods was tested in our previous paper and the corresponding maximum rv amplitude for @xmath22 min and for the whole line was reported .
the measurements of the line centre positions were converted to surface field using the expression : @xmath23 where @xmath24 is the half of the wavelength difference between the red and blue zeeman components of the feii line or the distance between the red @xmath7- and the @xmath6-component of the fei line , @xmath25 is the mean land factor and @xmath26 is the laboratory wavelength of a line .
based on the information available from the vald database ( kupka et al . )
we adopted @xmath27 and 1.35 for the fei and feii lines respectively .
the internal precision of our magnetic measurements is estimated to be @xmath2810 g for the feiiline and @xmath470 g for the weaker fei line . at the same time
the average field strengths derived from the two lines are substantially different in the cfht spectra : centre - of - gravity measurements of the feii line give @xmath29 g ( @xmath30 g is obtained with multiple fit of three gaussians ) , while @xmath31 g ( @xmath32
g ) is derived using the weaker fei line .
this @xmath33 g difference is confirmed by the spectrum synthesis .
the discrepant @xmath5 might be related to inaccuracy of the tabulated land factors , or it could be real and reflect different horizontal and/or vertical formation regions of the two diagnostic lines , or the effects of saturation in the feii line .
an average field strength of @xmath34 g ( @xmath35 g ) was obtained from the feii line in the eso spectra of @xmath0 equ .
llccc line & @xmath36@xmath5 ( g ) & @xmath14 & @xmath7@xmath5 ( g ) & @xmath37@xmath5 ( g ) + + feii 6149.25 & [email protected] & [email protected] & 13.0 & 13 + + feii 6149.25 & [email protected] & [email protected] & 9.2 & 5 + fei 6173.34 & [email protected] & [email protected] & 66.4 & 28 + table [ tbl3 ] and figs .
[ fig2][fig4 ] present results of our time - series analysis of the surface field measurements .
we see absolutely no evidence for any magnetic variations during the pulsation cycle in @xmath0 equ .
formal results of the least - squares fits with a fixed pulsation period as derived from the ndiiiline indicate insignificant amplitudes , all below 10 g. our most precise time - resolved magnetic measurements are derived from the cfht dataset .
analysis of the 6149.25 line in these spectra results in formal @xmath5 amplitude of just @xmath41 g , with the highest noise peaks in the amplitude spectrum ( fig .
[ fig3]d ) not exceeding 5 g over the whole period domain typical for roap pulsations .
a 3@xmath7 upper limit of the amplitude of magnetic variability was estimated from monte carlo simulations by sampling a sinusoidal signal at the phases of our observations and adding a random noise , characteristic of the scatter of the @xmath5 measurements .
this rigorous statistical estimate is reported in table [ tbl3 ] and indicates that , at the 3@xmath7 confidence level , no magnetic variability with the amplitude @xmath385 g is seen in @xmath0 equ during the night of our cfht observation .
our time - resolved magnetic measurements of @xmath0 equ have achieved the highest precision for a roap star , but reveal no evidence of pulsational modulation of the field strength .
we constrain possible magnetic variability to be below @xmath45 g in fe lines .
these results suggest that the marginal detection of @xmath4100 g @xmath5 variability during the pulsation cycle reported in @xmath0 equ by savanov et al .
@xcite is spurious , and probably stems from insufficient precision of the @xmath5 measurements in that study .
the null result reported in the present paper complements non - detection of the pulsational variability of @xmath2 determined from ndiii lines ( kochukhov et al . ) .
it should be recalled that the fe lines studied in our paper and the ree lines showing strong pulsational rv modulation are formed at substantially different atmospheric depths due to the extreme stratification of chemical abundances in cool ap stars and in @xmath0 equ in particular .
stratification analysis presented by ryabchikova et al .
@xcite allows us to conclude that ndiii lines sample very high atmospheric layers with optical depths @xmath39 , while the fe lines are formed below @xmath40 . the striking difference in the rv amplitudes of fe and ndiii lines
is then attributed to an outward increase of pulsational amplitude by roughly a factor of 50100 .
this increase is none the less not accompanied by an increase or even the presence of any observable oscillations of the magnetic field structure . combining the results of this paper with the study of kochukhov et al .
@xcite , and taking into account that for @xmath0 equ@xmath41 , we find that at all observed atmospheric depths @xmath5changes by less than about @xmath42 g per ms@xmath3 of corresponding velocity oscillations .
hence , possible magnetic variability is constrained to be below 1% of the field strength .
we note that @xmath0 equ is distinguished among the roap stars by its strong magnetic field , sharp lines and the exceptionally high amplitude of pulsational line profile changes .
yet this star defies any attempts to detect magnetic variability with pulsation phase .
this suggests that rapid magnetic modulation may be even more difficult to detect in other roap pulsators .
the outcome of our monitoring of @xmath2 and @xmath5 in @xmath0 equ demonstrates that very accurate measurements of the zeeman resolved lines can yield more precise magnetic time - series compared with the polarimetric @xmath2 observations .
therefore , the most promising direction for future attempts to detect magnetic oscillations in @xmath0 equ ( which are in any case unlikely to exceed a few tens of gauss ) would be to observe those ndii , ndiii and priii lines which are characterized by large pulsational rv shifts and at the same time show zeeman resolved profiles .
unfortunately , the extra broadening of the pulsating lines ( see kochukhov & ryabchikova ) strongly smears observed zeeman structure and makes proposed study of @xmath5oscillations extremely difficult .
this paper is based on observations obtained at the european southern observatory ( la silla , chile ) and at the canada - france - hawaii telescope .
we acknowledge support by the lise meitner fellowship to ok ( fwf project m757-n02 ) , by the fwf project _
p 14984 _ , by the natural sciences and engineering research council of canada , by the russian federal program ` astronomy ' ( part 1102 ) and by rfbr ( grant 04 - 02 - 16788 ) .
baldry i. k. , bedding t. r. , viskum m. , kjeldsen h. , frandsen s. , 1998 , mnras , 295 , 33 baldry i. k. , bedding t. r. , 2000 , mnras , 318 , 341 hubrig s. , kurtz d. w. , bagnulo s. , szeifert t. , schoeller m. , mathys g. , dziembowski w. a. , 2004 , a&a , 415 , 661 kanaan a. , hatzes a. p. , 1998 ,
apj , 503 , 848 kochukhov o. , ryabchikova t. , 2001a , a&a , 374 , 615 kochukhov o. , ryabchikova t. , 2001b , a&a , 377 , l22 kochukhov o. , ryabchikova t. , piskunov n. , 2004 , a&a , 415 , l13 kupka f. , piskunov n. , ryabchikova t. a. , stempels h. c. , weiss w. w. , 1999 , a&as , 138 , 119 leone f. , kurtz d. w. , 2003 , a&a , 407 , l67 martinez p. et al .
, 1996 , mnras , 282 , 243 mathys g. , hubrig s. , landstreet j. d. , lanz t. , manfroid j. , 1997 , a&as 123 , 353 piskunov n. e. , 1999 , in : 2nd workshop on solar polarization , j. stenflo and k.n .
nagendra ( eds . ) , kluwer academic publishers , dodrecht , 515 ryabchikova t. , piskunov n. , kochukhov o. , tsymbal v. , mittermayer p. , weiss w. w. , 2002 , a&a , 384 , 545 savanov i. , musaev f. a. , bondar a. v. , 2003 , ibvs , 5468 savanov i. s. , malanushenko v. p. , ryabchikova t. a. , 1999 , astron .
lett . , 25 , 802 | we describe an analysis of the time - resolved measurements of the surface magnetic field in the roap star @xmath0 equ .
we have obtained a high - resolution and high s / n spectroscopic time - series , and the magnetic field was determined using zeeman resolved profiles of the feii 6149.25 and fei 6173.34 lines .
contrary to recent reports we do not find any evidence of magnetic variability with pulsation phase , and derive an upper limit of 510 g for pulsational modulation of the surface magnetic field in @xmath0 equ .
stars : chemically peculiar stars : oscillations stars : magnetic field stars : individual : @xmath0 equ [ cols="^,^,^,^,^,^,^,^,^ " , ] |
in the ponzano regge model of 3-dimensional riemannian quantum gravity @xcite , an amplitude is associated with each tetrahedron in a triangulation of spacetime .
the amplitude depends on the tetrahedron s six edge lengths , which are assumed to be quantized , taking values proportional to @xmath4 where @xmath5 is a half - integer spin .
one can compute this amplitude either by evaluating an @xmath6 spin network shaped like a tetrahedron , or by doing an integral . approximating this integral by the stationary phase method , ponzano and
regge argued that when all six spins are rescaled by the same factor @xmath7 , the @xmath8 asymptotics of the amplitude are given by a simple function of the volume of the tetrahedron and the regge calculus version of its einstein action .
nobody has yet succeeded in making their argument rigorous , but roberts @xcite recently proved their asymptotic formula by a different method . this result lays the foundation for a careful study of the relation between the ponzano regge model and classical general relativity in 3 dimensions .
our concern here is whether a similar result holds for the barrett
crane model of _
4-dimensional _ riemannian quantum gravity @xcite . in this model
an amplitude is associated with each 4-simplex in a triangulation of spacetime .
this amplitude , known as a @xmath0 symbol , is a function of the areas of the 10 triangular faces of the 4-simplex .
each triangle area is proportional to @xmath4 , where @xmath5 is a spin labelling the triangle .
the amplitude can be computed by evaluating an @xmath9 spin network whose edges correspond to the triangles of the 4-simplex : @xmath10 there is also an integral formula for the @xmath0 symbol @xcite .
the problem is to understand the asymptotics of the @xmath0 symbol as all 10 spins are rescaled by a factor @xmath7 and @xmath8 .
barrett and williams @xcite applied a stationary phase approximation to the integral for the @xmath0 symbol , focusing attention on stationary phase points corresponding to nondegenerate 4-simplices with the specified face areas .
they showed that each such 4-simplex contributes to the @xmath0 symbol in a manner that depends on its regge action .
they pointed out the existence of contributions from degenerate 4-simplices , but did not analyse them . in section 2 of this paper
we begin by applying barrett and williams estimate of the @xmath0 symbols to the case where all 10 spins are equal .
we find that the contribution of their stationary phase points to the @xmath0 symbols is of order @xmath11 .
however , our numerical calculations show that the @xmath0 symbols are much larger , of order @xmath12 .
this means we must look elsewhere to explain the asymptotics of the @xmath0 symbol . in section 3
we analyse the contribution of ` degenerate 4-simplices ' to the integral for the @xmath0 symbols .
they do not correspond to stationary phase points in the integral for the @xmath0 symbols ; instead , the integrand has a strong _ maximum _ at these points .
we argue that the contribution of a small neighborhood of these points is asymptotically proportional to @xmath12 .
we give a formula expressing the constant of proportionality as an integral over the space of degenerate 4-simplices .
we also reduce this to an explicit integral in 5 variables . in section 4 , we numerically compare these results to the @xmath0 symbols as calculated using the algorithm developed by christensen and egan @xcite .
our formula for the contribution of degenerate 4-simplices closely matches the actual asymptotics of the @xmath0 symbols .
thus , even though our argument that these asymptotics are dominated by degenerate 4-simplices is not rigorous , we feel confident that the resulting formula is correct . in section 5 we discuss a new sort of spin network , associated to the representation theory of the euclidean group , which arises naturally in our analysis of the contribution of degenerate 4-simplices . generalizing our results on the riemannian @xmath0 symbols , we conjecture formulas for the asymptotics of a large class of riemannian spin networks in terms of these new ` degenerate spin networks ' .
we verify this conjecture in a number of simple cases . in section 6
we formulate a similar conjecture for lorentzian spin networks . taken with the previous one this conjecture implies that the @xmath8 asymptotics of a lorentzian spin network in this class are the same , up to a constant , as those of the corresponding riemannian spin network .
for example , as @xmath13 , the lorentzian @xmath0 symbol should be asymptotic to @xmath3 times the corresponding riemannian @xmath0 symbol !
we conclude by presenting some numerical evidence that this is the case .
the @xmath0 symbols can be defined using a riemannian spin network also known as a ` balanced ' spin network @xcite whose underlying graph is the complete graph on five vertices .
the ten edges of the graph are labelled with half - integer spins @xmath14 , where @xmath15 and @xmath16 refer to the vertices connected by each edge . in this approach , an edge labelled by
the spin @xmath5 corresponds to the representation @xmath17 of @xmath18 , and the @xmath0 symbols are computed using the representation theory of this group .
however , to analyse the asymptotics of the @xmath0 symbol , it is easier to use the integral formula due to barrett @xcite .
this is : @xmath19 where @xmath20 is the unit 3-sphere in @xmath21 equipped with its usual lebesgue measure , @xmath22 is the angle between the unit vectors @xmath23 and @xmath24 , and the kernel @xmath25 is given by : @xmath26 the normalizing factors in the integral come from the fact that the volume of the unit 3-sphere is @xmath27 .
the @xmath0 symbol gives the amplitude for a 4-simplex with specified triangle areas .
each vertex of the above graph corresponds to a tetrahedron in this 4-simplex , and each edge of the graph corresponds to the unique triangle shared by two of these tetrahedra . in this picture , the spin @xmath28 determines the area of the triangle shared by the @xmath15th and @xmath16th tetrahedra .
the precise formula for this area is somewhat controversial @xcite , but given the integral formula for the @xmath0 symbols , we find it convenient to assume the area is proportional to @xmath29 .
ignoring the constant factor , we thus define triangle areas by : @xmath30 in what follows , we study the behaviour of the @xmath0 symbol as all these triangle areas @xmath31 are multiplied by a large integer @xmath7 .
this is not the same as multiplying the spins @xmath28 by @xmath7 .
however , note that as @xmath28 ranges over all spins , @xmath31 ranges over all positive integers .
this means that if we multiply the areas @xmath31 by any positive integer @xmath7 , we can find new spins @xmath32 corresponding to the new areas by solving @xmath33 .
it is shown in @xcite that the integral in is nonnegative .
accordingly , we will concentrate on analysing the asymptotics of the absolute value of the @xmath0 symbol , given by the integral alone : @xmath34 where for simplicity we have left out the spins labelling the spin network edges , which are now @xmath32 .
barrett and williams @xcite express the numerator in the kernel @xmath25 as a difference of exponentials , which allows them to rewrite the integrand in equation as a sum of @xmath35 terms , each consisting of an exponential times the function : @xmath36 this function is unbounded as any of the @xmath22 tend to zero or @xmath37 , but if the regions of the domain where that occurs are set aside for a separate analysis , the integral over the remaining region becomes amenable to a stationary phase approximation @xcite .
the relevant phase is simply : @xmath38 the phases of the @xmath35 exponentials arising from the product of the kernels include variants of @xmath39 with all possible signs for the ten terms .
however , barrett and williams use the invariance of the integrand under the transformations @xmath40 to break the integral into @xmath41 identical parts , according to whether @xmath42 , where @xmath43 are the outward normals of a 4-simplex whose five tetrahedra lie in the hyperplanes normal to the @xmath23 .
they then focus on the region where all the @xmath44 , and show that of the @xmath35 phases , only @xmath39 and its opposite have stationary points in this region , and they occur when the @xmath22 are the angles between the outward normals to the tetrahedra of a 4-simplex whose ten faces have areas given by @xmath45 . in this case
the @xmath23 can be interpreted as the outward normals to the five tetrahedra , the @xmath22 are the angles between these normals , and @xmath45 is the area of the face shared by the tetrahedra numbered @xmath15 and @xmath16 .
interpreted this way , @xmath39 is precisely the regge action for the 4-simplex .
this means that the relevant stationary phase points are those of a much simpler integral : @xmath46 to carry out a stationary phase approximation of this integral we must note that the stationary phase ` point ' determined by the geometry of a 4-simplex is not actually a single point in the 15-dimensional manifold @xmath47 , but rather a whole 6-dimensional submanifold , since the geometry of the 4-simplex is invariant under the 6-parameter rotation group @xmath1 .
however , the same invariance can be used to remove this complication .
since the functions @xmath48 and @xmath39 are invariant under the action of @xmath1 , they pass to the quotient space @xmath49 , and we obtain : @xmath50 where @xmath51 is the result of pushing forward lebesgue measure on @xmath47 to this quotient space .
a factor of @xmath52 here accounts for the complication that simultaneously negating all the @xmath23 has the same action as @xmath53 , an element of @xmath1 , and so the discrete symmetries used to obtain already entailed taking the quotient by a 2-element subgroup of @xmath1 .
now , the stationary phase approximation @xcite of an @xmath54-dimensional integral @xmath55 is a sum over stationary points @xmath56 of the function @xmath39 : @xmath57 , \ ] ] where @xmath58 is the matrix of second partial derivatives of @xmath39 at the point @xmath56 , and @xmath59 is the signature of this matrix , i.e. , the number of positive eigenvalues minus the number of negative eigenvalues .
this approximation assumes there are finitely many stationary points , all with @xmath60 . applying this to the case at hand , and neglecting points where some of the angles @xmath22 are @xmath61 or @xmath37 , we obtain barrett and williams stationary phase approximation of the @xmath0 symbols : @xmath62,\ ] ] where the sum is over the 4-simplices @xmath63 with the specified face areas .
we can easily say something about the @xmath8 behavior of this quantity without actually evaluating it .
each stationary point @xmath56 is independent of @xmath7 , so @xmath64 will be constant as @xmath8 , while @xmath65 will grow linearly , as will its matrix of second derivatives , @xmath58 .
this is a @xmath66 matrix , since @xmath49 is 9-dimensional , so its determinant will grow as @xmath67 .
it follows that : @xmath68 however , cancellation between different stationary points could in principle make this a misleading overestimate .
thus it seems worthwhile to explicitly evaluate the sum in equation , at least in a simple special case . in
what follows we evaluate this sum for the case of a @xmath0 symbol with all edges labelled by the same spins .
in other words , we calculate the stationary phase approximation of the amplitude for a regular 4-simplex .
the first step , which would be useful more generally , is to find explicit coordinates on the quotient space @xmath49 and describe the measure @xmath51 in these terms .
to do this , we exploit the fact that any configuration of the five unit vectors @xmath23 can be rotated into one that belongs to a 9-dimensional subspace of the original domain .
specifically , if we express the vectors @xmath23 in polar coordinates @xmath69 where @xmath70 and @xmath71 , any set of @xmath23 can be rotated in such a way that the following restrictions are met : [ e : gaugefix ] @xmath72 by means of this ` gauge - fixing ' we can use the remaining 9 variables @xmath73 @xmath74 @xmath75 @xmath76 @xmath77 @xmath78 @xmath79 @xmath80 @xmath81 as coordinates on the quotient space . to describe the measure
@xmath51 in these coordinates recall that in polar coordinates , lebesgue measure on the unit 3-sphere is given by [ e : measures ] @xmath82 since @xmath83 is completely fixed we omit @xmath84 from the formula for @xmath51 , only inserting a factor @xmath27 due to the volume of the 3-sphere .
since @xmath85 and @xmath86 are partially fixed we replace @xmath87 and @xmath88 by @xmath89 we thus obtain @xmath90 next , we determine the stationary points of the function @xmath39 in the case where all ten spins are equal , neglecting points where some of the angles @xmath22 are @xmath61 or @xmath37 that is , where some of the vectors @xmath23 are parallel or anti - parallel .
recall that these stationary points correspond to nondegenerate 4-simplices having all 10 face areas equal .
one obvious candidate is the regular 4-simplex .
however , we must rule out the possibility that there are other , _ non - regular _ 4-simplices for which all the faces have identical areas .
since we are excluding degenerate 4-simplices from the current analysis , we can appeal to a theorem of bang @xcite which states that if all the faces of a non - degenerate tetrahedron have the same area , they are all congruent .
it follows that if all ten triangles in a non - degenerate 4-simplex have the same area , they too are all congruent .
now , each of the ten edges of a 4-simplex is shared by three triangles , so we can treat each edge as a triple of congruent line segments that happen to be superimposed , giving a total of 30 in all . if all the triangles were congruent isoceles triangles , each with one side of length @xmath91 , then ten of these 30 line segments would be of length @xmath91 .
however , there is no way to partition ten line segments into congruent triples .
the same argument rules out scalene triangles .
so the faces must be equilateral , and the 4-simplex must be regular .
we must therefore find all sets of unit vectors @xmath23 which satisfy the gauge - fixing conditions in equation and form the outward normals of a regular 4-simplex .
one choice is : @xmath92 these vectors have mutual dot products of @xmath93 .
they represent a single point in our 9-dimensional domain : [ e : statphasepoint ] @xmath94 the only other choice comes from interchanging @xmath95 and @xmath81 .
this yields another regular 4-simplex .
now we take the phase and specialise to the case where all the spins @xmath28 are equal , say to @xmath5 . setting @xmath96 , and working in our chosen coordinates ,
we obtain @xmath97 the partial derivatives of @xmath39 are all zero at the point described by .
symbolic computer calculations show that at this point the matrix of second derivatives of @xmath39 has 5 positive eigenvalues and 4 negative ones , and a determinant of : @xmath98 applying this result to at the point described by , working in our chosen coordinates , redefining the function @xmath99 to include the measure as well as the kernel denominators , and multiplying by a factor of two to account for the second point where @xmath95 and @xmath81 are interchanged , we obtain : @xmath100\ ] ] the above expression consists of an oscillating term times a function proportional to @xmath11 . in short ,
the estimate @xmath101 is sharp , at least for the case of 10 equal spins .
however , our numerical calculations of the @xmath0 symbols exhibit very different behaviour . rather than being of order @xmath11
, they appear to be much larger , of order @xmath12 . also , they exhibit no discernable oscillations . in the next section
we argue that these results are explained by the contribution of ` degenerate 4-simplices ' .
the absolute value of the kernel @xmath102 has maxima when the angle @xmath103 equals @xmath61 or @xmath37 , and these maxima become ever more sharply peaked as @xmath104 .
this means that as @xmath8 , the integrand in equation becomes very large at ` degenerate points ' , where some of the vectors @xmath23 are either parallel or anti - parallel .
we have made a detailed study of the ` fully degenerate ' points , where _ all _ the vectors @xmath23 are either parallel or anti - parallel .
it is plausible to expect the contribution from a neighborhood of these points to dominate the integral for the @xmath0 symbol , at least asymptotically as @xmath13 , since in this region we are integrating a product of kernels , all of which are near their greatest possible absolute value . in the next section we shall present numerical evidence that this is in fact the case .
the value of the kernel is positive at @xmath105 , but at @xmath106 its sign is positive when @xmath107 is even and negative otherwise , where @xmath108 . at first glance , this appears to allow the possibility that parallel and anti - parallel degenerate points might cancel each other for certain values of the spins .
in fact , cancellation can occur only when the @xmath0 symbol vanishes .
the integrand in is the product of ten kernels , so it will be positive when all the @xmath23 are parallel .
if @xmath83 is replaced by its opposite , the four @xmath109 will change from zero to @xmath37 , leading to an overall sign change of @xmath110 .
however , for the @xmath0 symbol to be non - zero , the spins at each vertex must sum to an integer @xcite .
therefore , if the @xmath0 symbol is non - zero , the sign of the integrand will remain positive . the same argument applies if any subset of the @xmath23 are reversed . in the integral in equation ,
fully degenerate points occur in 3-dimensional submanifolds of the domain , but we can apply the same gauge - fixing principles as we used to analyse the stationary points , in this case simply fixing @xmath111 .
there are then 16 discrete fully degenerate points in the restricted domain : those where @xmath112 for @xmath113 . restricting the integral to the vicinity of these points we obtain @xmath114 where @xmath115 is a small open ball around the point @xmath116 .
we approximate this quantity by noting that when @xmath103 is small , the kernel in is close to : @xmath117 we call this quantity the ` degenerate kernel ' . as we shall see in section [ s : euclidean ] , this is the analog of the original kernel in a spin - network formalism where the space of constant curvature , @xmath20 , is replaced by three - dimensional euclidean space , and the group @xmath1 , the isometry group of @xmath20 , is replaced by the euclidean group of isometries of @xmath2 .
similarly , the integral over a small subset of @xmath118 can be approximated by an integral over a subset of @xmath119 , and the angle @xmath22 between unit vectors @xmath23 and @xmath24 in @xmath20 replaced by the euclidean distance @xmath120 between vectors @xmath121 and @xmath122 in @xmath2 . in terms of these new euclidean variables , the restriction @xmath111 is replaced by @xmath123 .
thus , when @xmath7 is large , we have : @xmath124 where now @xmath115 is a small open ball around the origin of @xmath119 .
the approximation exhibits very simple scaling behaviour .
first , note that the degenerate kernel obeys the identity : @xmath125 this allows the scaling of to be deduced from a linear change of variables , @xmath126 : @xmath127 where @xmath128 is the result of rescaling the neighborhood @xmath115 by a factor of @xmath7 .
if the integral on the right hand side of this equation converges in the limit @xmath129 , we obtain the asymptotic formula : @xmath130 we call this integral the ` degenerate @xmath0 symbol ' .
apart from the factor of @xmath131 , the integral here can be interpreted as the evaluation of a spin network with edges labelled by unitary irreducible representations of the euclidean group : the group of isometries of euclidean 3-space .
we discuss this interpretation in more detail in section [ s : euclidean ] . while elegant , the integral in is difficult to compute numerically , very much like the integral for the lorentzian @xmath0 symbol . to obtain a more convenient form for evaluation ,
we make use of the kirillov trace formula : @xmath132 where @xmath133 is a vector in @xmath2 , and the integral is over the unit 2-sphere with its standard measure .
this formula is easily confirmed by choosing spherical coordinates such that the @xmath134-axis is parallel to the vector @xmath133 . for our purposes
we shall rewrite it as follows : @xmath135 where @xmath136 is the 2-sphere of radius @xmath137 embedded in @xmath2 , but where @xmath138 is the induced lebesgue measure divided by @xmath139 , to hide some annoying constants that would otherwise appear in this formula . using this
we can rewrite as : @xmath140 where @xmath141 is a cartesian product of 2-spheres with the measures described above , and @xmath142 is defined by : @xmath143 if @xmath144 then the final integral in will be well - behaved so long as at each point @xmath145 the differential of @xmath146 has maximal rank , namely 12 .
when this is the case @xmath147 is an 8-dimensional submanifold of @xmath148 , and the integral reduces to : @xmath149 here @xmath138 is the lebesgue measure on @xmath147 induced by the riemannian metric on @xmath148 , but divided by a factor of @xmath150 , since we have divided the lebesgue measure on each sphere by a factor of @xmath151 .
if we choose local coordinates @xmath152 on @xmath148 near @xmath145 such that @xmath153 are zero on @xmath147 , then @xmath154 is the jacobian determinant of @xmath146 as a function of the remaining 12 variables @xmath155 .
one way to get solutions of @xmath156 is to start with 4-simplices in @xmath2 : that is , 5-tuples of points in @xmath2 , together with the ten triangles and five tetrahedra determined by these points .
given such a 4-simplex , let @xmath157 be the vector that is normal to the triangle shared by the @xmath15th and @xmath16th tetrahedra , and has length equal to the area of this triangle . as shown in @xcite , the four vectors @xmath157 normal to the triangles in any one tetrahedron must sum to zero , with appropriate signs : @xmath158 each vector appears twice in these formulas , with opposite signs , since the outwards - pointing normal to one tetrahedron is the inwards - pointing normal to another .
the first four equations say that @xmath156 ; the last is an algebraic consequence of the rest . if @xmath159 , the vectors @xmath157 thus determine a point in @xmath147 .
this suggests that we think of points of @xmath147 as ` degenerate 4-simplices ' .
however , not every point of @xmath147 comes from a 5-tuple of points in @xmath2 this way . to see this , note that two 5-tuples in @xmath2 determine the same point in @xmath147 if they are translates of each other .
the space of 5-tuples modulo translation has dimension @xmath160 , but the space of solutions of equation has dimension @xmath161 .
thus there are simply not enough 5-tuples of points in @xmath2 to account for all solutions of equation .
we shall still call points of @xmath147 ` degenerate 4-simplices ' , because it is a useful heuristic .
however , it is important to take this phrase with a grain of salt . of course
, for the space @xmath147 to be nonempty , there must _ exist _ a solution of @xmath156 .
this imposes certain restrictions on the numbers @xmath31 .
for example , there will be a solution of @xmath162 if and only if these ` tetrahedron inequalities ' hold : @xmath163 in general @xmath147 will be empty if the four numbers @xmath31 corresponding to the faces of any one tetrahedron violate the tetrahedron inequalities . in this case
the degenerate @xmath0 symbols vanish .
similarly , the riemannian @xmath0 symbols vanish if the spins @xmath28 violate the tetrahedron inequalities @xcite , and for the same sort of geometrical reason @xcite . for computational purposes
it is helpful to rewrite the integral for the degenerate @xmath0 symbols in yet another way . to do this
, first we note that we can exploit @xmath164 invariance to convert to an integral over the 5-dimensional manifold @xmath165 , by rotating the @xmath157 in @xmath148 so that they lie in the 17-dimensional subspace where @xmath166 is fixed at @xmath167 , and @xmath168 lies in the @xmath169 half of the @xmath170-plane .
this corresponds to integrating over all possible values for @xmath166 , inserting a factor equal to the volume of @xmath171 with our chosen measure , which is @xmath172 , and also integrating over all possible azimuthal coordinates for @xmath168 and inserting a factor of @xmath173 . in what follows , we will choose coordinates so that the @xmath174 matrix @xmath175 whose determinant we require is block diagonal , with two @xmath176 blocks and one @xmath177 block .
it turns out to be convenient to parameterise @xmath168 , not by its angle from the @xmath134-axis , but by the length @xmath178 since @xmath179 , these four vectors can be positioned to form the sides of a ( possibly non - planar ) quadrilateral .
then @xmath180 is the length of one of the diagonals . with the vectors @xmath166 and @xmath168 fixed and the lengths of @xmath181 and @xmath182 specified , the only remaining freedom this quadrilateral has , if it is to remain closed , is the ` hinge angle ' , @xmath183 , between the two triangles that meet along the diagonal .
the vectors @xmath181 and @xmath182 have 4 degrees of freedom in all , and specifying @xmath183 removes one of them , leaving 3 which break the quadrilateral . with our chosen measure on @xmath147 , the product of the measure for the coordinates we are integrating over , and
the jacobian determinant for the 3 that break the quadrilateral , is : @xmath184 we can treat a second 4-tuple of vectors more or less identically .
if the vector @xmath185 has an azimuthal angle of @xmath103 , and we define @xmath186 then the quadrilateral formed by @xmath187 and @xmath188 can be assigned a ` hinge angle ' of @xmath189 .
this specifies all the degrees of freedom that allow this quadrilateral to remain closed .
once again , the product of the measure for the coordinates we are integrating over , and the jacobian determinant for the 3 that break the quadrilateral , is : @xmath184 so far , we have specified 5 degrees of freedom for @xmath165 : @xmath180 , @xmath190 , @xmath183 , @xmath189 and @xmath103 .
no further continuous degrees of freedom remain .
the three vectors we have yet to specify must form triangles that complete two quadrilaterals with vectors that have already been parameterised , and because these two triangles have a vector in common , there is no ` hinge ' freedom left . specifically , the three vectors @xmath191 and @xmath192 must form a tetrahedron by fitting over a triangular base that has , as two of its sides , the vectors : @xmath193 given their fixed lengths , this determines @xmath191 and @xmath192 completely , apart from the freedom to locate the apex of the tetrahedron on either side of the plane spanned by @xmath194 and @xmath195 .
this freedom can be accounted for with a factor of 2 in the integral .
the final contribution to the jacobian comes from a @xmath177 block involving all the coordinates of @xmath191 and @xmath192 , and with our chosen measure this is : @xmath196 where @xmath197 is the volume of the tetrahedron as a function of its edge lengths . combining these results
, we can rewrite as : @xmath198 the integrals over the @xmath199 are taken over intervals determined by the four sides of the quadrilaterals for which they are the diagonal lengths , and the angular variables range from @xmath61 to @xmath173 , with the proviso that any part of the domain where the tetrahedron is not geometrically possible must be excluded . in numerical calculations ,
this can be dealt with by setting the integrand to zero wherever cayley s determinant formula for the squared volume of the tetrahedron yields a negative value .
we note that the integrand here is unbounded , and we have not proved that converges , but our numerical calculations suggest that it does .
to test our hypothesis that the asymptotics of @xmath0 symbols are dominated by the contribution of degenerate 4-simplices , we used the algorithm described in @xcite to calculate values for several sets of @xmath0 symbols .
the figure below shows log - log plots for the absolute values of the riemannian @xmath0 symbols as a function of @xmath7 , where @xmath7 is the parameter by which the areas @xmath31 were multiplied . the legend shows the base spins @xmath28 ; multiplying @xmath31 by @xmath7
was achieved by replacing the @xmath28 with : @xmath200 the lines on the plot show the @xmath12 asymptotic behaviour described by equation ; the integrals were evaluated numerically with lepage s vegas algorithm @xcite , and found to have values of 0.680 , 0.341 , 0.209 , 0.110 , 0.0841 and 0.0446 respectively . in summary ,
our numerical data supports the following conjecture .
let us say that the spins @xmath28 are ` admissible ' if for each vertex in the @xmath0 symbol , the spins labelling the four incident edges satisfy the tetrahedron inequalities and sum to an integer . then : if the ten spins @xmath28 are admissible , the @xmath8 asymptotics of the riemannian @xmath0 symbols are given by : @xmath201 in the next two sections we generalize this conjecture to a large class of riemannian and lorentzian spin networks , including the lorentzian @xmath0 symbols .
the reader may have wondered why we consider asymptotics of the @xmath0 symbols as _ areas _ are rescaled , instead of _ spins_. the reason is that they are much simpler .
we also calculated values for sets of @xmath0 symbols where the spins @xmath28 were multiplied by @xmath7 .
the figure below shows log - log plots for the absolute values of the riemannian @xmath0 symbols with spins @xmath202 , where the @xmath28 match those shown in the legend for the previous figure : 1em since the areas @xmath203 are asymptotic to a series of values @xmath204 that are proportional to @xmath205 , the data might be expected to exhibit @xmath12 scaling . and
in fact , in cases a - d , that is exactly what is seen .
furthermore , numerically evaluating the integral with @xmath206 provided the correct coefficients for lines on the plot which are asymptotic to the data ; these coefficients were 2.73 , 0.987 , 0.472 and 0.165 respectively . for cases
e and f , the @xmath0 symbols have one or more vertices at the ` border of admissibility ' : that is , vertices where three of the spins labelling incident edges sum to equal the fourth spin , making one of the tetrahedron inequalities a strict equality . these vertices are marked with heavy dots on the legend .
setting @xmath207 in in these cases yields a coefficient of zero , because the domain for at least one variable of integration , the quadrilateral diagonal length @xmath180 , was reduced to a single point .
empirically , the data here appears to scale as @xmath208 .
while @xmath0 symbols with one or two vertices on the border of admissibility have @xmath208 asymptotics with spin rescaling , @xmath0 symbols with three or more vertices on the border of admissibility appear to decay exponentially as their spins are multiplied by @xmath7 , as illustrated in the log - linear plot below : 8em 1em to prove convergence of the partition function in our newly formulated version of the barrett crane model @xcite , it will probably be necessary to understand the @xmath208 and exponential decay of borderline - admissible @xmath0 symbols under spin rescaling
. however , these are more delicate phenomena than we are prepared to tackle here .
though the riemannian @xmath0 symbols are given by an integral over a product of copies of @xmath20 , we have seen that when we rescale the areas by a large constant @xmath7 , this integral is dominated by the contribution of a very small patch of this space , which can be approximated by a product of copies of @xmath2 .
indeed , by a change of variables we can think of the @xmath8 limit as one in which the radius of @xmath20 approaches infinity , so that it degenerates to euclidean 3-space . the riemannian @xmath0 symbols can also be described using the representation theory of @xmath1 , the isometry group of @xmath20 .
this suggests that the @xmath8 asymptotics of the @xmath0 symbols can be described in terms of the representation theory of the euclidean group @xmath209 , the isometry group of @xmath2 . to do this ,
we introduce certain spin networks associated to @xmath209 which we call ` degenerate spin networks ' .
these are more closely analogous to the lorentzian spin networks defined in @xcite than to the riemannian spin networks we have been discussing so far , because the edge labels are not restricted to discrete values , and the group representations are infinite - dimensional .
however , all three sorts of spin network form part of a unified theory , as outlined below : 1em 1em the rotation group @xmath1 and the lorentz group @xmath210 consist of linear transformations of @xmath21 with determinant 1 that preserve metrics of signature ( + + + + ) and ( @xmath211+++ ) , respectively . similarly , the euclidean group @xmath209 is isomorphic to the group of linear transformations of @xmath21 with determinant 1 that preserve the singly degenerate metric with signature ( @xmath61+++ ) .
the groups @xmath1 and @xmath210 both ` contract ' to @xmath209 , meaning that they have it as a limiting case if one forms the isometry group of the metric @xmath212 and lets @xmath213 and @xmath214 , respectively .
this makes it plausible that the asymptotics of not only riemannian but also _ lorentzian _ spin networks can be calculated using degenerate spin networks . in this section we state a precise conjecture along these lines for a large class of riemannian spin networks , and present some supporting evidence . in section [ s : lorentzian ]
we do the same for lorentzian spin networks .
we begin by describing degenerate spin networks and how to evaluate them .
the representations @xmath17 labelling edges of a riemannian spin network are representations not just of @xmath215 , but actually of @xmath1 . as emphasized by freidel and krasnov @xcite , these representations are precisely the eigenspaces of the laplacian on @xmath20 , which is the homogeneous space @xmath216 .
similarly , the representations labelling edges of a lorentzian spin network are the eigenspaces of the laplacian on hyperbolic 3-space , @xmath217 here we must take the connected component of the lorentz group to get just one sheet of the hyperboloid . following this pattern , the representations labelling edges of a degenerate spin network should be the eigenspaces of the laplacian on @xmath218 .
in fact , there is one representation of this sort for each positive real number @xmath137 .
any complex function on @xmath2 with @xmath219 can be written as : @xmath220 where @xmath136 is the 2-sphere of radius @xmath137 centered at the origin of @xmath2 , and @xmath138 is the induced lebesgue measure divided by @xmath139 .
defining an inner product on these solutions by @xmath221 we form a hilbert space @xmath222 consisting of all solutions @xmath48 with @xmath223 .
this hilbert space becomes a representation of the euclidean group where each group element @xmath224 acts via @xmath225 .
in fact , this representation is unitary and irreducible .
the functions @xmath226 form a ` basis ' , in the sense that any element of @xmath222 can be expressed as in equation for a unique square - integrable function @xmath227 on the sphere . we define
a ` degenerate spin network ' to be a directed graph with each edge @xmath228 labelled by a representation of this form , or equivalently , a positive number @xmath229 .
an intertwiner between tensor products of these representations can be defined at each vertex by taking the product of functions from the representations labelling the _ incoming _ edges , multiplying it by the product of complex conjugates of functions from representations labelling the _ outgoing _ edges , and integrating the result over @xmath2 . given this , the standard way to evaluate such a spin network
@xcite would be to take a ` trace ' : that is , integrate over a basis label @xmath230 for each edge of the graph .
the result would be : @xmath231 \ ; \prod_{v \in v } \frac{dx_v}{2\pi^2 } .\ ] ] here @xmath232 denotes the set of edges of the graph , @xmath233 denotes its set of vertices , and the vertices @xmath234 and @xmath235 are the source and target of the edge @xmath228 .
however , just as in the lorentzian case @xcite , this gives a divergent integral , because the integrand is invariant when we simultaneously translate all the vectors @xmath236 by the same amount .
more generally , if the underlying graph of our spin network consists of several connected components , and we translate the vectors @xmath237 where @xmath194 lies in any one component , the integrand does not change . to keep things simple ,
let us consider only spin networks whose underlying graph is _
connected_. in this case we can sometimes obtain a well - defined integral by ` gauge - fixing ' one of the vectors rather than integrating over it : that is , setting @xmath238 for some vertex @xmath239 .
if we let @xmath240 , this gives the following formula for evaluating a degenerate spin network with edges labelled by the numbers @xmath229 : @xmath241 \ ; \prod_{v \in v ' } \frac{dx_v}{2\pi^2 } .\ ] ] as in the lorentzian case @xcite , one can show that if this integral converges , the result does not depend on our choice of the special vertex @xmath242 or the point @xmath243 . assuming the integral
does converge , we can use the kirillov trace formula to reexpress it as : @xmath244 \ ; \prod_{v\in v ' } \frac{dx_v}{2\pi^2 } .\ ] ] when the evaluation of a degenerate spin network converges , it always obeys a very simple scaling law as we multiply all the edge labels @xmath229 by the same constant @xmath7 . using the scaling property of the degenerate kernel noted in equation ,
we find that @xmath245 where @xmath246 is the number of edges in the underlying graph and @xmath247 is the number of vertices .
we have already argued that the asymptotics of the riemannian @xmath0 symbols are governed by the corresponding degenerate spin network . we can generalize this argument as follows .
fix a connected graph .
if we label each edge @xmath228 by a positive integer @xmath229 or equivalently a spin @xmath248 with @xmath249 we obtain a riemannian spin network , whose evaluation we define by : @xmath250 here @xmath251 is the distance between points @xmath252 in the unit 3-sphere in @xmath21 as measured by the induced riemannian metric .
this formula is equivalent to the standard integral formula @xcite , except that we have omitted the usual signs in order to simplify the relationship to degenerate spin networks .
fixing a small open ball @xmath115 around some point @xmath253 we define the ` degenerate contribution ' to this integral to be : @xmath254
just as we included a factor of @xmath255 in equation , here we include a factor of @xmath256 to take into account the contribution of anti - parallel degenerate points ; as before there is no cancellation between these if the spins labelling edges incident to each vertex sum to an integer , as they must for the spin network to have a nonzero value . using the same nonrigorous argument as in section [ s : degenerate ] ,
we see that as @xmath8 , @xmath257 where @xmath115 now denotes an open ball around the origin of @xmath258 , and we made the change of variables @xmath259 . in short
, this argument suggests that the asymptotics of the degenerate contribution to the value of a riemannian spin network are proportional to those of the corresponding degenerate spin network : @xmath260 and we know the latter are very simple : @xmath261 this is particularly interesting when we also have @xmath262 because then we can compute the asymptotics of a riemannian spin network by evaluating a degenerate spin network : @xmath263 when can we expect this to occur ?
clearly we should at least demand that the degenerate contribution outweigh the contribution of stationary phase points
. a simple power - counting argument as in section [ s : stationary ] suggests that the contribution of stationary phase points is of order : @xmath264 where the graphs with one or two vertices are different because there is less need for ` gauge - fixing ' . comparing these asymptotics to those of the degenerate contribution
, we can formulate the following : given a connected graph with more than two vertices and @xmath265 , or two vertices and @xmath266 , or one vertex and @xmath267 , as @xmath8 we have @xmath268 as long as the integral defining @xmath269 converges and the spins @xmath248 labelling the edges incident to each vertex are admissible .
here we say the spins labelling the edges incident to some vertex are ` admissible ' if they sum to an integer and each is less than or equal to the sum of the rest .
we do not yet have general criteria for when the integrals associated to euclidean spin networks converge , and as we shall see , the relevant theorems are bound to be a bit different than in the lorentzian case @xcite .
the simplest test of this conjecture is the ` theta network ' , with two vertices joined by three edges , labelled by positive integers @xmath137 , @xmath270 , and @xmath271 .
when the corresponding spins are admissible , the riemannian theta network evaluates to : @xmath272 ^{a } \ar@{- } @/_1.5pc/ [ r ] _ { c } \ar@{- } [ r]^{b } & * { \bullet } \\ } } \right)^r = 1 .\ ] ] the degenerate theta network can also be explicitly evaluated ; assuming without loss of generality that @xmath273 : @xmath274 ^{a } \ar@{- } @/_1.5pc/ [ r ] _ { c } \ar@{- } [ r]^{b } & * { \bullet } \\ } } \right)^d = \begin{cases } 0 & c > a+b \\ \frac{1}{4 } & c = a+b \\ \frac{1}{2 } & c < a+b .
\end{cases}\ ] ] since the riemannian network s spins are admissible , the third inequality must hold for the corresponding areas @xmath275 . thus in this case the conjecture gives an _ exact _ formula for the riemannian spin network .
the next simplest case is the ` 4@xmath5 symbol ' : the spin network with two vertices joined by four edges , labelled by positive integers @xmath137 , @xmath270 , @xmath271 and @xmath276 . without loss of generality
let us assume @xmath277 .
as noted in a previous paper in this series @xcite , the riemannian 4@xmath5 symbol counts the dimension of a space of @xmath6 intertwiners . using this
it follows that : @xmath278 ^{a } \ar@{- } @/_1.5pc/ [ r ] _ { d } \ar@{- } @/^/ [ r]^{b } \ar@{- } @/_/ [ r]_{c } & * { \bullet } \\ } } } } \right)^r = \begin{cases } 0 & b+c \leq d - a \\
\frac{1}{2}(a+b+c - d ) & d - a \leq b+c < d+a \\ a & d+a \leq b+c .
\end{cases}\ ] ] the corresponding degenerate spin network evaluates to : @xmath279 ^{a } \ar@{- } @/_1.5pc/ [ r ] _ { d } \ar@{- } @/^/ [ r]^{b } \ar@{- } @/_/ [ r]_{c } & * { \bullet } \\ } } } } \right)^d = \begin{cases } 0 & b+c \leq d - a \\
\frac{1}{4}(a+b+c - d ) & d - a \leq b+c
< d+a \\ \frac{1}{2 } a & d+a \leq b+c , \end{cases}\ ] ] so the conjecture is again exact .
an interesting check on our hypotheses is the tetrahedral spin network .
this has four vertices and @xmath280 , so the hypotheses of conjecture 2 do not apply : we expect the stationary phase contribution to the riemannian tetrahedral network to be comparable to the degenerate contribution .
the degenerate tetrahedral network evaluates to : @xmath281 where @xmath282 is defined as the volume of the tetrahedron dual to the tetrahedral network .
each triangle in this dual tetrahedron corresponds to a vertex of the tetrahedral network , and the three sides of the triangle have lengths equal to the labels on the three edges incident to the network vertex : @xmath283 on the other hand , the riemannian tetrahedral network evaluates to the square of the @xmath6 tetrahedral network , the basic building - block of the ponzano regge model .
thanks to the calculation of ponzano and regge @xcite , later made rigorous by roberts @xcite , this means that : @xmath284 here we are dealing with a dual tetrahedron whose edge lengths are @xmath285 , where @xmath31 ranges over @xmath286 as @xmath287 .
this tetrahedron has regge action @xmath288 where @xmath289 are the corresponding dihedral angles .
thus it appears that the asymptotics of the riemannian tetrahedral network are a sum of two parts : a part equal to 8 times the degenerate tetrahedral network , and an oscillatory part coming from the stationary phase points .
when the edge lengths of the dual tetrahedron are such that it can not exist in euclidean space , the degenerate tetrahedral network evaluates to zero , and the riemannian network obeys different asymptotics in which its evaluation exponentially decays with increasing spin .
now let us turn to lorentzian spin networks @xcite .
each positive real number determines a unitary irreducible representation of the connnected lorentz group @xmath290 corresponding to an eigenspace of the laplacian on @xmath291 ; however , we prefer to describe the evaluation using an integral formula . fixing a connected graph with vertex set @xmath233 and edge set @xmath232 , and labelling each edge @xmath228 by a positive real number @xmath229
, we obtain a so - called ` lorentzian spin network ' , whose evaluation is given by : @xmath292 as with degenerate spin networks , we have chosen a vertex @xmath242 , set @xmath293 , and let @xmath294 be any fixed point in @xmath295 . here
@xmath291 is hyperbolic 3-space , i.e. , the submanifold of minkowski spacetime given by : @xmath296 with its induced riemannian metric .
we define the lorentzian kernel @xmath297 by : @xmath298 we warn the reader that this convention differs from that of most previous papers @xcite , which include a factor of @xmath137 in the denominator . including that factor would divide any lorentzian spin network by the product of its edge labels , so for example , it would divide the asymptotics of the lorentzian @xmath0 symbols as defined here by a factor of @xmath299 .
the same line of argument by which we arrived at our conjecture concerning asymptotics of riemannian spin networks applies to lorentzian ones .
the most important difference is that no factor of @xmath256 appears , since there are no ` antipodal points ' in hyperbolic space .
we have not investigated criteria for the existence of stationary phase points , but where they are present their exponents will be the same as in the riemannian case , leading us to make : given a connected graph with more than two vertices and @xmath265 , or two vertices and @xmath266 , or one vertex and @xmath267 , as @xmath8 we have @xmath300 as long as the integral defining @xmath269 converges and the positive numbers @xmath229 labelling edges incident to each vertex are admissible .
here we say the positive numbers labelling the edges incident to some vertex are ` admissible ' if each is strictly less than the sum of the rest .
again the simplest test of this conjecture is the lorentzian theta network . translating their result into our notation
, a calculation of barrett and crane @xcite shows that for any @xmath301 , @xmath302 ^{a } \ar@{- } @/_1.5pc/ [ r ] _ { c } \ar@{- } [ r]^{b } & * { \bullet } \\ } } \right)^l = \frac{1}{4 } \left[f(-a + b + c ) + f(a - b +
c ) + f(a + b - c ) - f(a + b + c ) \right ] , \ ] ] where @xmath303 as the conjecture predicts , the asymptotics of this match those of the degenerate theta network , which are given by : @xmath302 ^{a } \ar@{- } @/_1.5pc/ [ r ] _ { c } \ar@{- } [ r]^{b } & * { \bullet } \\ } } \right)^d = \frac{1}{4 } \left[\operatorname{sign}(-a+b+c ) + \operatorname{sign}(a - b+c ) + \operatorname{sign}(a+b - c ) - \operatorname{sign}(a+b+c)\right ] .\ ] ] it is worth noting that while the integral for the lorentzian theta network converges even when we take the absolute value of the integrand , this fails for the degenerate theta network .
this makes it more challenging to find criteria for convergence of degenerate spin networks , since we can not simply mimic the theory that applies in the lorentzian case @xcite .
barrett and crane also worked out the lorentzian @xmath304 symbols , obtaining : @xmath305 ^{a } \ar@{- } @/_1.5pc/ [ r ] _ { d } \ar@{- } @/^/ [ r]^{b } \ar@{- } @/_/ [ r]_{c } & * { \bullet } \\ } } } } \!\right)^{\!l } \!= \frac{1}{4 } [ g(-&a+b+c+d)+g(a - b+c+d)+g(a+b - c+d)+g(a+b+c - d)\\ -g(&a+b - c - d)-g(a - b+c - d)-g(a - b - c+d)-g(a+b+c+d ) ] , \end{split}\ ] ] where @xmath306 from equation one can show : @xmath307 ^{a } \ar@{- } @/_1.5pc/ [ r ] _ { d } \ar@{- } @/^/ [ r]^{b } \ar@{- } @/_/ [ r]_{c } & * { \bullet } \\ } } } } \right)^d = \ ] ] @xmath308 , \end{split}\ ] ] where @xmath309 thus the conjecture is also confirmed in this case .
next , consider the tetrahedral spin network . as in the riemannian case ,
conjecture 3 does not apply , but we can still predict the asymptotic behaviour of the contribution of the fully degenerate point : @xmath310 where ` @xmath311 ' represents the contribution from stationary phase points , if any . below is a log - log plot comparing this prediction to numerical data . the horizontal axis in this graph
represents @xmath7 , while the vertical axis represents the value of the tetrahedral spin network .
the most interesting test of conjecture 3 is the @xmath0 symbol .
if the conjecture is true , the lorentzian @xmath0 symbol should be asymptotic to the degenerate @xmath0 symbol , and therefore asymptotic to @xmath3 times the riemannian @xmath0 symbol .
since the riemannian @xmath0 symbol is positive @xcite , this in turn would imply that the lorentzian and degenerate @xmath0 symbols are positive in the @xmath8 limit .
it is difficult to compute the lorentzian @xmath0 symbol , but we have numerically checked the conjecture in the special case where all the edges are labelled by the same number @xmath7 . in this case
the conjecture says that : @xmath312 where of course the constant is not exact .
below we show a log - log plot comparing this prediction to data points obtained by computing the lorentzian @xmath0 symbols numerically .
the horizontal axis represents @xmath7 , while the vertical axis represents the value of the @xmath0 symbol .
here we computed the lorentzian @xmath0 symbols by applying the vegas algorithm to evaluate the integral above using polar coordinates on @xmath291 .
the integral was reduced from 12 to 9 dimensions by exploiting the @xmath290 invariance ; the infinite domain was made compact by replacing the radial coordinate for each point , @xmath313 , with a new variable @xmath314 ; and the domain was further reduced by exploiting a 24-fold symmetry present in the regular case .
the large dimension and oscillatory nature of this integral make these calculations extremely computationally intensive .
the graph below shows the same lorentzian @xmath0 symbols multiplied by @xmath315 , plotted on a linear scale to give a clearer picture of the rate of convergence towards the asymptote .
the error bars are three times the standard deviation computed by the vegas algorithm .
the coordinate system we used yielded the lowest standard deviations of several we tried , but there is no reason to believe that the estimates of the integral are drawn from a normal distribution , so this data should be treated with caution .
it appears that the asymptotics of both riemannian and lorentzian @xmath0 symbols can be computed in terms of degenerate spin networks . for the mathematician
, this claim still requires proof .
indeed , there is even an issue of rigor concerning our argument that degenerate spin networks asymptotically describe the contribution of fully degenerate points , since we have not proved that the limit @xmath316 exists . the ambitious mathematician could also try to prove more general versions of our conjectures , in which all the areas @xmath229 approach infinity , but not in fixed proportion to one another . for the physicist ,
however , a more pressing question is : _
what do these results imply for the physics of the barrett
crane model ?
_ on the one hand , it is unsettling that the simple asymptotic behavior of the ponzano regge model is not found here .
the way degenerate geometries govern the asymptotics of the @xmath0 symbols raises the possibility that in the limit of large spins , the barrett crane model reduces to a theory of degenerate metrics .
however , it is important to bear in mind that the physics of the barrett crane model may not be controlled by the large - spin behavior of the @xmath0 symbols though certainly this affects the convergence of the partition function @xcite .
indeed , there are still major open questions about the right way to extract physics from spin foam models , and we hope that our work spurs on research on this subject . after we submitted this article for publication , preprints of two forthcoming papers appeared , one by john barrett and chris steele @xcite and another by laurent freidel and david louapre @xcite , both of which deal with the asymptotics of @xmath0 symbols by somewhat different methods than our own , and both of which obtain results in agreement with our conjectures .
we thank john barrett and chris steele for helpful conversations , and note that we corrected two errors in an earlier draft of this paper ( an excess factor of 2 in equation and a false conjecture about stationary phase points in lorentzian spin network evaluations ) after reading a preprint of their forthcoming paper on the asymptotics of @xmath0 symbols .
j. c. baez , an introduction to spin foam models of @xmath318 theory and quantum gravity , in _ geometry and quantum physics _ , eds .
helmut gausterer and harald grosse , springer lecture notes in physics * 543 * , berlin , 2000 . also available as gr - qc/9905087 . | the riemannian @xmath0 symbols are spin networks that assign an amplitude to each 4-simplex in the barrett - crane model of riemannian quantum gravity .
this amplitude is a function of the areas of the 10 faces of the 4-simplex , and barrett and williams have shown that one contribution to its asymptotics comes from the regge action for all non - degenerate 4-simplices with the specified face areas .
however , we show numerically that the dominant contribution comes from degenerate 4-simplices . as a consequence
, one can compute the asymptotics of the riemannian @xmath0 symbols by evaluating a ` degenerate spin network ' , where the rotation group @xmath1 is replaced by the euclidean group of isometries of @xmath2 .
we conjecture formulas for the asymptotics of a large class of riemannian and lorentzian spin networks in terms of these degenerate spin networks , and check these formulas in some special cases . among other things
, this conjecture implies that the lorentzian @xmath0 symbols are asymptotic to @xmath3 times the riemannian ones . |
the effect of atmospheric electric fields on the development of extensive air showers from high energy cosmic rays has not received much attention in the past . because of the large energies of shower particles , the electric fields present in the atmosphere are generally much too small to alter the particle energies significantly .
the largest fields are of the order of 1 kv / cm and only occur in thunderstorms @xcite .
in such fields the hadronic and muonic part of the shower are hardly affected , although a muon deficit due to increased decay rate has been reported by alexeenko et al .
the effects on the electromagnetic shower are much larger , but they are , as we will show in this work , local in the sense that the amount and energy distribution of electromagnetic particles quickly adapts to the local background field . below thunderstorms the electric field decreases , so particle detector arrays will in general not be strongly sensitive to the influence of electric fields .
mountain top experiments , however , can be very close to , or even inside thunderstorms .
an increase in the air shower detection rate and the single particle count has been reported by high altitude experiment such as eas - top @xcite and the carpet air shower array @xcite .
effects of atmospheric electric fields are generally not included in air shower simulation codes like corsika @xcite .
there are , however , two topics for which we demonstrate in this paper that it is important to take electric fields into consideration .
the first is the consequence of different shower developments in strong electric field regions on the radio emission of the shower . with experiments such as lopes @xcite and codalema @xcite , the technique of radio detection of air showers
has become mature in the last few years .
radio antennas , with their low costs and high duty cycles , have been demonstrated to be an attractive addition to large air shower arrays like the pierre auger observatory @xcite . already in the 1970s , experimental results by madolesi et al .
@xcite showed excessively large radio pulses during thunderstorms . together with a large range in measured radio intensity and the inability to filter out radio interference ,
the unknown effect of electrical conditions in the atmosphere was a reason to abandon the radio experiments @xcite until the development of digital radio arrays like lofar @xcite revived the interest .
a study of the effect of weather conditions on the radio pulse strength with lopes data confirmed the amplification of radio pulses during thunderstorm conditions , but also showed that no effect is observed under other weather conditions @xcite . in another study
@xcite it was shown that during thunderstorms the air shower arrival direction reconstructed with the lopes radio antennas can be a few degrees off with respect to the direction reconstructed by the kascade particle detector array .
it is now known that the radio emission of air showers can be described in terms of coherent synchrotron radiation of shower electrons and positrons that follow curved trajectories in the earth s magnetic field as first described in falcke et al .
@xcite and in more detail in huege et al .
alternatively , a macroscopic picture can be constructed in which the shower charges , that are separated in the magnetic field , support a transverse current producing radiation as proposed by kahn & lerche @xcite and more recently by scholten et al . @xcite .
although there are subtle differences between these descriptions , it is clear that an alteration in the electron and positron distribution of the shower at some altitude , can influence the power of the radio pulse in both models . in this paper
we limit ourselves to a study of electric field effects on these distributions , not on the radio pulse itself , which will be the subject of a forthcoming paper .
the second issue related to the electric field is the suggestion that air showers of sufficient energy can start an avalanche of runaway electrons in thunderstorm electric fields .
ionization electrons that are produced in collisions of shower particles with air molecules are accelerated in the thunderstorm electric field and can , under the right conditions , gain enough energy to ionize further molecules , an effect described by gurevich et al . @xcite . in thunderstorm research the field strength that can support such avalanches
is known as the threshold field , described in marshall et al .
@xcite . in their work , the authors present thunderstorm measurements which show that lightning often occurs when the thunderstorm field exceeds the breakeven field , suggesting that runaway electron breakdown plays a role in lightning initiation . by providing seed electrons for avalanches
, air showers from cosmic rays may play an important role in thunderstorm dynamics .
simulations by dwyer @xcite have shown that the threshold field strength for an avalanche to develop is slightly higher than the breakeven field , taking into account the effects of elastic scattering .
inside thunderstorms there have been measurements of x - ray bursts @xcite and gamma bursts , both from earth @xcite and space @xcite
. this emission can be explained in terms of bremsstrahlung emitted by the runaway breakdown electrons @xcite . in this work ,
we simulate the effect of electric fields on the development of air showers with corsika . in sec .
[ sec : sim ] we describe the setup of our simulation and the modification that has been made for corsika . in sec .
[ sec : effects ] we derive some simple analytic estimates and limits to compare with the results .
simulation results are presented in sec .
[ sec : res ] and we conclude with a discussion of the simulation limitations and the consequences for realistic field configurations .
corsika @xcite is a monte carlo code that simulates the development of air showers by tracing the individual particles and their interactions based on several sophisticated interaction models . in our simulations we use the high - energy hadronic interaction model qgsjet - ii @xcite and for low - energy hadronic interactions we use urqmd 1.3cr @xcite .
electromagnetic interactions are simulated by the standard corsika ( version 6.720 ) routines to treat electromagnetic particles .
these routines are taylor - made versions of the egs4-code @xcite adapted to the barometric atmosphere with a density decreasing exponentially with increasing altitude .
all possible interactions are considered and a proper treatment of ionization energy loss and multiple scattering is performed . by including some suitable extra statements into the transport routine electr for e@xmath0 particles the effects of an external electrical field
are taken into account which causes an acceleration ( energy gain rsp .
loss ) for particles moving parallel to the field and a deflection for those moving perpendicular to the field .
a suitable limitation of the transport step length guarantees small changes of the particle movements to neglect higher order effects on the particle traces . by these means
the energy gain / loss in the electrical field and the ionization energy loss can be treated independently for each transport step . in our simulations
we use the `` thinning '' option with thinning at @xmath1 level and optimized weight limitation @xcite to keep the computing times in a tolerable level .
the energy cutoff below which electrons and positrons are discarded from the simulation is 0.5 mev .
because in an electric field , particles below this threshold may be accelerated to higher energies , this introduces a limitation to the simulations , which is discussed in sec .
[ sec : dis ] .
simulations were done on five proton showers : vertical showers of @xmath2 ev , @xmath3 ev and respectively , @xmath4 ev and two inclined showers of @xmath2 ev , having zenith angles of 30 and 60 degrees .
all showers had a proton as primary particle .
we use homogeneous ambient electric fields of strengths up to 1 kv / cm . in reality
, fields of such strengths will only occur in small parts of a thunderstorm , but keeping the field constant in the simulations allows us to understand the electric field effect better . in sec .
[ sec : dis ] we will discuss the implications for realistic field configurations . when simulating showers with the same primary particle but different random seeds fluctuations will occur from shower to shower .
most importantly the altitude of the first interaction varies , but also the location of the shower maximum , for example , is dependent on number of particles that are produced in the first interaction , and the energy distribution of these particles .
we use conex @xcite to make 100 shower simulations for each of the five above - mentioned configurations . from these 100 simulations we select a shower with a large number of secondary particles in the first interaction and a fairly typical longitudinal shower profile .
conex produces a file that lists all secondary particles after the first interaction and their momenta , which can be used as an input stack for corsika using the stackin option . with a conex stack of particles created at the first interaction instead of one primary particle as input
, different random seeds will produce much smaller variations .
for each shower configuration we have selected a conex input stack and used this to produce ten showers with different random seeds . in the following plots of shower evolution
we plot the mean value of these ten showers and one sigma error bars . because the fluctuations between simulations are very small with this approach , we are more senstitive to changes that are introduced by the background electric field .
we use the coast interface code for corsika @xcite to get information on electron and positron distributions at 50 layers evenly distributed in atmospheric depth along the shower axis . at each layer
two three - dimensional histograms are written out .
one shows the distribution as a function of , respectively : * particle energy , * distance of the particle to the shower axis and , * time lag of the particle with respect to a plane that travels along the shower axis with the speed of light in vacuum .
the second histogram shows the distirbution as a function of : * particle energy , * angle between the particle momentum and shower axis and , * angle between the component of the particle momentum that is perpendicular to the shower axis and a vector pointing radially outwards from the shower axis to the particle .
both histograms are created separately for electrons and positrons . in fig .
[ noe_vs_smalle ] the number of electrons and positrons is plotted as a function of atmospheric depth for a @xmath4 ev shower .
simulation results with the original code , not including an electric field , are represented by the dashed blue line and the red solid line shows the results of the modified code , with a very small electric field added of 1 mv / cm . for small electric fields the original and modified corsika codes converge to within the limit of random fluctuations .
this is also observed for showers of different energy and zenith angle .
high energy electrons lose their energy primarily through the process of bremsstrahlung .
the radiation length for electrons in air is @xmath5 g/@xmath6 @xcite .
when an electric field @xmath7 is present the change in energy @xmath8 per unit of atmospheric depth @xmath9 is given by : @xmath10 where @xmath8 is the particle energy and @xmath11 the particle charge .
the vertical distance @xmath12 is related to the vertical atmospheric depth as @xmath13 .
the air density @xmath14 is approximately given by @xcite : @xmath15 where @xmath16 km is the scale height , so @xmath17 , which is used in the last step of eqn .
[ bremsloss ] .
the particle reaches an equilibrium energy when @xmath18 particles below this energy are accelerated , while for particles above this energy radiation losses dominate .
[ fig : equilibrium ] shows this equilibrium energy as well as trajectories ( in energy ) of particles that are created at a certain atmospheric depth with an energy of 10 mev and only lose energy through smooth bremsstrahlung losses as in eqn .
[ bremsloss ] .
it can be seen that particles are accelerated until the point where they cross the equilibrium energy .
although this picture is a strong simplification of the realistic case where the particles do not lose their energy smoothly over time but in discrete processes , we can use it to set two limits on the effect of an electric field on the energy distribution of shower particles : the distribution will be largely unaffected beyond the maximum energy that accelerated particles can reach ( e.g. @xmath19 mev for 1 kv / cm at x=100 g/@xmath20 , see fig . [
fig : equilibrium ] ) and the effect on the distribution will be most prominent below the equilibrium energy given by eqn .
[ equilibrium ] .
below @xmath21 mev , the energy loss due to collisions increases with decreasing energy . for a given electric field strength , only particles above a certain energy can be accelerated .
the breakeven field is defined as the field strength at which the energy loss due to collisions is equal to the energy gain in the electric field for an electron of 1 mev and is given by @xcite : @xmath22 electrons with energies higher than 1 mev will be accelerated in a breakeven field , while electrons with energy below 1 mev will decelerate . for fields exceeding the breakeven field ,
the mimimum electron energy required for acceleration will be lower . in collisions with air molecules air shower particles
produce many electrons with energies below 1 mev .
when the background electric field exceeds the breakeven field a part of these electrons will be accelerated to energies at which they can create new ionization electrons , resulting in a breakdown process .
it has been found by dwyer @xcite that the threshold field strength for the development of a runaway breakdown process is : @xmath23 a field of 1000 v / cm is equal to the threshold field at @xmath24 km or @xmath25 g/@xmath20 . above this altitude
we can encounter breakdown effects .
when the breakdown is efficient it gives an exponential increase of the number of electrons .
the breakdown electrons can be distinguished from the electrons from pair creation in three ways : ( a ) their energies are generally lower , ( b ) as they are created at lower energies they will have a larger time delay w.r.t .
the shower front and ( c ) they are deflected strongly ( or entirely ) into the electric field direction . at @xmath26 km or @xmath27
g/@xmath20 , a field of 1000 v / cm is equal to the breakeven field . below this altitude
no breakdown effects are expected .
the complications of electron runaway breakdown are described in much detail in various publications , see for example gurevich et al .
the simple features presented here will serve as a handle to interpret the simulation results .
as follows from eqns .
[ bremsloss ] and [ equilibrium ] a positive field points downwards and accelerates the positrons .
the errors bars are 1 sigma variations of 10 simulations with the same conex input stack of particles after the first interaction . from these 10 showers
we choose the one of which the longitudinal profile is closest to the mean profile , when producing energy distribution plots .
[ const_field_100 ] shows simulation results for a vertical @xmath2 ev air shower .
the number of electrons and positrons is plotted as a function of atmospheric depth .
the red data points correspond to the absence of an electric field and the green and blue points to fields of 100 v / cm and -100 v / cm respectively .
the variations are within the 1 sigma error bars .
when the field strength is increased by an order of magnitude the effect on the shower development becomes significant .
[ const_field_1000 ] shows the number of electrons and positrons for a @xmath2 ev vertical shower in an electric field of 1000 v / cm ( accelerating the positrons ) .
the lines without error bars represent the development in absence of an electric field .
when the field is switched on , the number of positrons outgrows the number of electrons , causing a positive charge excess . fig .
[ enerdist1000 ] shows the energy distribution of electrons and positrons at the shower maximum .
the distributions for the same shower in absence of an electric field is plotted for reference .
the sum of electron and positrons increases slightly but within the 1 @xmath28 error region .
[ const_field_-1000 ] shows the shower development in an electric field of -1000 v / cm ( accelerating the electrons ) for several shower energies . for each energy the lower line represents the same shower in absence of a field . above the altitude in which the electric field equals the threshold field ( eqn .
[ threshold ] ) an explosive increase in the number of electrons can be seen ( note we switched to logaritmic scale ) .
high up in the atmosphere the increase of electrons is nearly exponential .
interestingly , the largest electron content is reached by the shower that had its primary interaction highest up in the atmosphere , not the shower with the highest primary energy .
the latter does have the most electrons at lower altitudes , below the altitude at which the electric field equals the breakeven field , where the breakdown process has stopped and the electrons are injected by pion decay .
high up in the atmosphere , the number of electrons increases exponentially reaching a maximum at @xmath29 g/@xmath20 , where the electric field is slightly stronger than the threshold field .
the energy cutoff at 0.5 mev in our simulation may be of influence to the location of the maximum .
note that the point of first interaction for the showers of different energies is random due to the way we selected our showers , and does not follow the dependence of mean first interaction height on primary energy .
[ enerx152 ] through [ enerx794 ] show the energy distribution of electrons and positrons for the @xmath3 ev shower in fig .
[ const_field_-1000 ] . the same shower in absence of a field is plotted for reference .
two vertical lines represent the limits derived in sec .
[ sec : effects ] : the main effect of the particle acceleration is expected to occur below the equilibrium energy ( eqn .
[ equilibrium ] ) and no significant change is expected above the equilibrium energy at the altitude of first interaction ( dotted and solid line ) .
the plots represent the early breakdown phase ( fig .
[ enerx152 ] ) , the altitude at which the number of electrons reaches its maximum ( fig .
[ enerx332 ] ) and the altitude at which the shower reaches its maximum in absence of a field ( fig . [ enerx794 ] ) . at the point where the number of electrons reaches its maximum ,
the amount of positrons is also increased by a factor 100 .
apparently , the number of electrons in the breakdown process is so large that the pair creation from gamma emission of the breakdown electrons exceeds the pair creation from gamma emission of pion decay . to study the effect of a local electric field
, two field configurations are simulated in which the electric field is switched on or off , respectively , at the altitude of 3.2 km .
[ localeffect ] shows the number of positrons as a function of atmospheric depth for a 10@xmath30 ev shower in absence of a field , inside a field of 1000 v / cm that is homogeneous over the whole atmosphere and inside two fields that have a discontinuous jump at 3.2 km altitude .
one of the latter has field strength 1000 v / cm above 3.2 km and no electric field below that altitude , the other vice versa .
it can be seen that the number of particles quickly adapts itself to the expected number for the local electric field .
the reason is that the electromagnetic component is continually refreshed by pair production due to pion decay , while the electric particles lose their energy quickly through radiation and ionization losses . for inclined showers
the longitudinal development in slant depth is very similar to that of vertical showers in the absence of an electric field .
[ nofield3i ] shows the number of electrons as a function of atmospheric depth for a vertical shower of @xmath2 ev and two inclined showers of the same energy , with zenith angles of 30 and 60 degrees . for inclined showers the slanted atmospheric depth
is used , so their evolution should coincide with a vertical shower but the same depth corresponds to a higher altitude .
the development is the same within error bars .
the inclined showers continue to larger atmospheric depth , as the simulation stops when the particles hit ground level .
as is the case with vertical showers , the shower development is only slightly affected by a field of 100 v / cm .
[ smallfield30 ] shows the number of electrons and positrons for negatively and positively aligned fields of 100 v / cm for a shower with a 30 degrees zenith angle .
the variations are of the order of the 1 sigma error bars .
[ smallfield60 ] contains the same information for a shower with 60 degrees zenith angle .
again , the electric field effect is small .
[ incl1000 ] shows the shower development for vertical and inclined showers in a strong , positively directed , vertical field of 1000 v / cm ( accelerating the positrons ) .
the positive charge excess becomes even more pronounced for the 60 degrees shower .
the reason for this change is that for an inclined shower the atmosphere is less dense at the the same slant depth . in a region of low density
the energy loss due to collisions is smaller and acceleration in the electric field more efficient . the same effect but
even stronger can be seen in the development of inclined showers in a negatively aligned field ( see fig .
[ incl-1000 ] ) .
the slant depth at which the electric field equals the threshold field is larger for inclined showers and the number of breakdown electrons is increased by more than three orders of magnitude . also , the breakdown is sustained longer in terms of slant depth .
the breakdown electrons will be deflected in the electric field and will move away from the shower axis . the area that is affected by the breakdown effect can have a large horizontal extent . figs .
[ s257 ] through [ s700505 ] show electron distributions for different showers .
for each figure the left column corresponds to the breakdown region and the right column to the shower maximum . from top to bottom , plotted along the vertical axis are respectively , angle of the particle momentum with respect to the shower axis , distance from the shower axis , and delay time w.r.t . a hypothetical shower front travelling at the speed of light .
each plot has the particle energy on the horizontal axis .
[ s257 ] shows a vertical shower in the absence of an electric field .
of course , no breakdown occurs and the shape of the distributions at high altitude is roughly the same as the shape of the distributions at shower maximum . fig .
[ s503 ] shows the distributions for a vertical shower in a negative electric field of 1000 v / cm .
the characteristics of the shower maximum are not much different from those of the shower maximum in fig .
[ s257 ] but in the breakdown region clear differences can be seen . in the first place
the number of electrons is about a factor of 100 higher , but the characteristics of the distribution have also changed .
the bulk of the particles has a lower energy , peaking around 10 mev and the delay time is larger .
these are the characteristics that would be expected from an electron avalanche made up of electrons that start with low energies and are accelerated in the electric field .
[ s500500 ] corresponds to a shower in the same electric field with a zenith angle of 30 degrees .
it can now be seen that in the breakdown region the bulk of the electrons has an angle between 20 and 40 degrees with the shower axis .
the reason for this is that the low energy electrons are deflected into the direction of the electric field ( i.e. vertical in the case studied here ) .
for the same reason the distance to the shower axis and the delay time increases . at the shower maximum the distributions return roughly to the shapes of the vertical shower distributions . for the shower with 60 degrees zenith angle the distributions
are plotted in fig .
[ s700505 ] .
now , the shower maximum is located at such an altitude that the breakdown has not yet ended . in the left column
an electron avalanche is seen of which the particles have an even larger angle with respect to the shower axis and also larger distances and delay times . in the right column
, two contributions can be distinguished : breakdown electrons having the same characteristics as in the left column and shower electrons that are produced via pion decay and feature a more traditional distribution .
the electric fields we have used in our simulation are completely homogeneous and not realistic .
we used these simple fields to clarify the effects the electric field can have on the shower development and make it easier to compare the results to rough analytic limits . inside a thunderstorm , the electric field will strongly change in strength and direction with altitude .
the effects described in this paper will only occur locally at the region where the field is strongest .
electric field measurements inside thunderstorms are difficult to perform because of their violent nature .
marshall et al .
@xcite present soundings of the vertical component of the electric field measured with balloon - borne field meters .
the strength of the field generally fluctuates , but rarely exceeds the breakeven field strength .
the polarity can change multiple times .
experiments typically measure fields integrated over a few tens of meters and a few seconds .
it is uncertain if stronger field fluctuation exist on smaller spatial or temporal scales . for air shower experiments
this means that the particle count at detector level does not change strongly , except when the experiment is very close to or inside the thundercloud .
the observed @xmath31 increase in electron count at the high altitude experiment eas - top @xcite during thunderstorms can be explained by a nearby electric field with an order of magnitude of 1000 v / cm .
the additional particles are either electrons produced in a breakdown process , or low energy shower electrons and positrons that have gained sufficient energy in the background electric field to be above the detection threshold .
the same study also reports an increase in air shower detection rate .
this can be the result of an increase in the number of electrons and positrons in the front of showers which are , in fair weather conditions , below the detection threshold alternatively , it can be due to the deflection of breakdown electrons in the direction of the electric field coming from showers that would have otherwise missed the detector .
indeed , during the period of increased detection rate , the air showers cluster around a specific ( near - vertical ) arrival direction . the radio pulse of an air shower can be severely affected by strong electric fields .
all shower electrons and positrons contribute to the emission and the largest contribution to the pulse comes from the particles at and just before the shower maximum @xcite . when the shower maximum is inside a strong field region the radio pulse will not only be altered because the energy distribution of the electrons and positrons has changed , but also because the acceleration itself produces additional radiation .
we will study the combination of these effects in a future paper using the radio simulation code reas2 @xcite . the growth and decay of the current that is produced by runaway electrons , results in strong radio emission ( see e.g. @xcite ) .
the intensity of emitted fluorescence light is dependent on the energy of the electromagnetic shower and can in principle be influenced by atmospheric electric fields .
it is , however , unlikely that strong fields exist outside clouds , where such fluorescence effects can be observed .
we used corsika without the upward option , so the upward going particles are not tracked , which have a zenith angle of more than 90 degrees .
the electron avalanches that are observed in negative field simulations may also occur in the upward direction for positive fields , but are discarded by corsika .
upward going particles may produce new pairs via bremsstrahlung and pair production , of which one particle is accelerated downwards again .
it is shown by dwyer @xcite that such feedback mechanisms are important in breakdown processes .
these particles , however , are missed by our simulation . the lower energy cutoff in our corsika simulations is set to 0.5 mev for the electromagnetic shower .
this means that electrons below this energy are discarded before they may be accelerated to higher energies by the electric field . to compensate for collisional energy loss a field of a few times the breakeven field
is needed to accelerate these particles . for our simulation
this means the number of electrons above the altitude at which the field equals the breakeven field may be underestimated . by discarding particles that have low energies or propagate in the upward direction ,
the simulations can not be expected to give a complete description of the runaway breakdown process .
the inclusion of lower energy particles will lead to a much larger total number of runaway electrons because of exponential growth .
these low energy electrons will be accelerated to relativistic energies by the electric field , but due to elastic scattering the average position of the runaway electrons moves at 0.89c @xcite , considerably slower than the shower particles . for new generations of electrons produced by upward traveling particles
, the delay will be even larger .
moreover , the particles that start out with low energies will not in general be accelerated into the direction of shower propagation .
this means our simulations are reliable for all particles that are considered to be part of the shower ( i.e. located within the shower front , which has a typical thickness of @xmath21 m near the shower axis ) .
furthermore , the simulations show that a shower induces a runaway breakdown process if the background electric field is strong enough .
the size of the breakdown is , however , not reliably simulated and probably underestimated . because of its structure , corsika can not be easily used for simulation of the breakdown itself . inside an area where electron breakdown occurs the electric field will be shielded by the electrons and ions . in effect , the field may be amplified outside this region .
a simulation of the breakdown should have an electric field that varies over time as a function of the charge distribution ( see e.g. @xcite ) .
corsika traces one particle at a time , making it impossible to include this without major alterations to the code .
electrical processes , such as the formation of streamers and lightning initiation , are outside the reach of our simulations , but our results provide a strong hint that air showers can trigger such events when they pass through a strong field region of a thunderstorm .
we have conducted corsika simulations for vertical and inclined showers of several energies inside electric fields with varying strengths .
the evolution of the electromagnetic part of the shower does not change significantly below field strengths of 100 v / cm . for fields of the order of a 1000 v
/ cm the effect becomes important .
such fields only occur naturally inside thunderclouds .
the energy distribution can be altered up to energies of @xmath21 gev . for positive fields
this means positrons can outnumber the electrons over a large energy range , resulting in a positive charge excess . for negative fields electron breakdown
is observed for altitudes at which the field is larger than the breakeven field and is most efficient when the field is larger than the threshold field .
this electron avalanche is directed antiparallel to the electric field and can geometrically be detached from the shower .
the electric field effect on the shower evolution is local in the sense that in a low field region a shower that has traversed a high field region is not much different from a shower that has not . for air showers that traverse thunderstorms
this means they are generally affected only in the strong field regions of the cloud .
ground based particle detectors will not be very sensitive to these effects .
the radio signal from an air shower that travels through a strong field region can significantly change .
it has been established experimentally that under thunderstorm conditions the power of the radio pulse may be much larger than anticipated .
the order of magnitude of maximum electric fields that occur in thunderstorms coincide with the field strengths at which our simulations show a considerable change in particle distribution .
it should be noted , however , that for the strength of the radio pulse to increase , the shower evolution does not necessarily have to change .
the direct acceleration of the particles by the electric field produces radiation just like the magnetic deflection does .
the effect of electric fields on the radio emission of air showers will be explored in more detail in a forthcoming paper .
air showers can trigger electron avalanches in regions where the electric field exceeds the threshold field .
these avalanches may play a role in lightning initiation .
although electron avalanches can be initiated by any seed electron of sufficient energy in a field exceeding the threshold field , a passing air shower offers the unique scenario in which a huge number of high energy electrons is injected in a very small time . the possible interaction between air showers and thunderstorms
can be investigated with a hybrid array of particle detectors and radio receivers . in principle radio antennas
can pick up the signal of both air showers and electrical discharges , but it is probably not possible to unambiguously detect an air shower signal behind the violent radiative background of a thunderstorm . with a combination of particle detectors and an array of radio antennas electrical activity after an air shower passage could be imaged , allowing for a study of temporal and spatial coincidences .
macgorman , d. , & rust , w. , the electrical nature of storms ( new york : oxford univ . press ) ( 1998 ) vernetto , s. et al .
( eas - top coll . )
the eas counting rate during thunderstorms , proc . of 27th icrc ( 2001 )
alexeenko et al . , phys .
a , 301 ( 2002 ) 299 heck , d. et al .
report fzka 6019 ( 1998 ) falcke , h. et al .
( lopes coll . )
nature , 435 ( 2005 ) 313 ardouin , d. et al .
, astropart .
26 ( 2006 ) 341 van den berg , a. et al ( auger coll . ) proc . of 30th icrc ( 2007 ) 176 mandolesi , n. , morigi , g. & palumbo , g. atmos .
phys . , 36 ( 1974 ) 1431 baggio , r. , mandolesi , n. , morigi , g. & palumbo , g. , nuovo cimento b ser . , 40 ( 1977 ) 289 ` http://www.lofar.org ` buitink , s. et al .
( lopes coll . )
a&a , 467 ( 2007 ) 385 nigl , a. et al .
( lopes coll . )
a&a , 487 ( 2008 ) 781 falcke , h. & gorham , p. astropart .
phys . , 19 ( 2003 ) 477 huege , t. & falcke , h. a&a , 412 ( 2003 ) 19 kahn , f. , & lerche , i. roy .
a , 289 ( 1966 ) 206 scholten , o. , werner , k. , & rusydi , f. astropart .
phys . , 29 ( 2008 ) 94 gurevich , a. , milikh , g. , & roussel - dupre , r. , phys .
a , 165 ( 1992 ) 463 marshall , t. , mccarthy , m. , & rust , w. jgr , 100 ( 1995 ) 7097 dwyer , j. , geophys .
32 ( 2005 ) l20808 mccarthy , m. & parks , g. geophys .
, 12 ( 1985 ) 393 gurevich , a. et al .
lett . a 325 ( 2004 )
389 - 402 fishman , g. et al .
, science , 264 ( 1994 ) 1313 - 1316 gurevich , a. & milikh , g. , phys .
lett . a 262 ( 1999 )
457 dwyer , j. , j. geophys .
res . 113 ( 2008 )
ostapchenko , s. nucl .
b ( proc . suppl . )
151 , 143 and 147 ( 2006 ) bass , s. et al .
( 1998 ) 225 nelson , w. , hirayama , h. , & rogers , d. report slac 265 ( 1985 ) kobal , m. et al .
( auger coll . ) , astropart .
15 ( 2001 ) 259 bergman , t. et al .
, astropart .
( 2007 ) 420 huege , t. , ulrich , r. & engel , r. , astropart .
phys . , 27 ( 2007 ) 392 and ` http://www-ik.fzk.de/~rulrich/coast ` amsler , c. et al .
( particle data group ) , phys .
b , 667 ( 2008 ) 1 gurevich , a. , zybin , k. , & roussel - dupre , r. , phys .
a , 254 ( 1999 ) 79 gurevich , a. , zybin , k. & medvedev , yu .
lett . a 349 ( 2006 )
331 - 339 huege , t. , ulrich , r. & engel , r. , astropart .
phys . , 27 ( 2007 )
392 gurevich , a. , duncan , l. , medvedev , yu . & zybin , k. , phys . lett . a 301 ( 2002 )
320 - 326 tierney , h , roussel - dupr , e. , symbalisty , m. & beasley , w. , j. geophys .
110 d12 ( 2005 ) dwyer , j. , uman , n. and rassoul , h. , j. geophys .
res . 114 ( 2009 )
d09208 coleman , l. and dwyer , j. geophys .
33 ( 2006 ) l11810 dwyer , j. , geophys .
32 ( 2005 ) l20808 | the development of cosmic ray air showers can be influenced by atmospheric electric fields . under fair weather conditions these fields are small , but
the strong fields inside thunderstorms can have a significant effect on the electromagnetic component of a shower .
understanding this effect is particularly important for radio detection of air showers , since the radio emission is produced by the shower electrons and positrons .
we perform monte carlo simulations to calculate the effects of different electric field configurations on the shower development .
we find that the electric field becomes important for values of the order of 1 kv / cm .
not only can the energy distribution of electrons and positrons change significantly for such field strengths , it is also possible that runaway electron breakdown occurs at high altitudes , which is an important effect in lightning initiation . , , , , cosmic rays , extensive air showers , atmospheric electricity , radiation by moving charges , computer modeling and simulation 96.50.s- , 96.50.sd , 92.60.pw , 41.60.-m , 07.05.tp |
with the evidence for neutrino masses and mixings from solar and atmospheric neutrino data getting more and more firm , the nature of new physics that could explain the observations is under a great deal of scrutiny @xcite .
there are several issues that need to be understood , notably ( a ) the smallness of neutrino masses , and ( b ) the origin of the large atmospheric neutrino as well as the solar neutrino mixing angles , the latter being favored@xcite by the combined solar neutrino results@xcite including the recent sno neutral current data @xcite .
seesaw mechanism@xcite provides one of the simplest ways to understand the small neutrino masses .
it assumes the existence of a set of three right handed neutrinos ( @xmath0 ) that have masses at the scale @xmath6 corresponding to a new ( local ) @xmath7 symmetry of weak interactions .
atmospheric neutrino observations suggest that the scale @xmath6 is much lower than the planck scale , leading to a new threshold inferred solely from experiments . at low energies , these heavy right handed neutrinos induce operators of the form @xmath8 , where @xmath9 is the dirac yukawa coupling matrix connecting the @xmath0 with the left - handed neutrinos ( @xmath10 ) and the standard model ( sm ) higgs doublet @xmath11 .
@xmath12 denotes the lepton doublet which contains @xmath10 .
after electroweak symmetry breaking , these operators induce masses for the light @xmath10 of order @xmath13 .
since @xmath6 is much larger than the weak scale @xmath14 , the induced @xmath10 masses will be extremely small compared to all other masses of the sm fermions .
present neutrino oscillation data suggests a scale @xmath15 gev .
the precise seesaw formula for the light neutrino mass matrix with three generations is given by @xmath16 where @xmath17 is the dirac neutrino mass matrix and @xmath18 is the @xmath0 majorana matrix .
the neutrino mixing angles in such schemes would arise as a joint effect from two sources : ( i ) mixings among the right handed neutrinos present in @xmath19 and ( ii ) mixings among different generations present in the dirac mass matrix @xmath17 .
the ( physical ) neutrino oscillation angles will also receive contributions from mixings among the charged leptons .
the neutrino flavor mixings induced by the seesaw mechanism ( as needed by oscillation data ) can lead to lepton flavor violating ( lfv ) effects .
however , within the sm extended minimally to accommodate the seesaw mechanism , such effects are extremely small in any process other than neutrino oscillation itself , owing to a power suppression factor @xmath20 , as required by the decoupling theorem .
this situation is drastically different if there is low energy supersymmetry .
the power suppression then becomes @xmath21 which is very much weaker and can lead to observable lfv effects at low energies , as noted in a number of papers@xcite .
the main difference is that with low energy susy , lepton flavor violation can be induced in the slepton sector , which can then be transferred to the leptons suppressed only by a factor @xmath21 .
the way this comes about is as follows .
consider the minimal @xmath22 supergravity ( msugra ) models @xcite . at the fundamental scale @xmath23 where
susy breaking is communicated to the sm sector , all the soft susy breaking scalar masses are taken to be universal , and the trilinear @xmath24matrices are taken to be proportional to the respective yukawa matrices .
thus , at the planck scale , there is no flavor violation anywhere except in the yukawa couplings and susy preserving mass terms in the superpotential .
now , the right handed neutrinos have masses of order @xmath6 which is much lower than @xmath23 , as inferred from atmospheric neutrino data . in the momentum regime @xmath25 where the @xmath0 fields are active
, the soft masses of the sleptons will feel the effects of lfv in the neutrino yukawa sector through the renormalization group evolution . at the scale @xmath6 , the slepton mass matrix is no longer universal , and this non - universality will remain down to the weak scale .
this lfv in the slepton sector is subsequently transferred to the leptons through one loop diagrams involving the exchange of gauginos .
it should be noted that in the absence of neutrino masses , there is only one leptonic yukawa matrix , @xmath26 for the charged leptons , which can be diagonalized at @xmath23 .
@xmath26 will remain diagonal to the weak scale , and would not induce any flavor violation in the slepton sector in msugra models .
the experimental evidence for neutrino oscillation is thus a strong indicator that there might very well be lepton flavor violation , assuming the validity of low energy susy .
searches for lfv processes such as @xmath27 and/or @xmath28 can therefore be an important source of information on the @xmath0 mixings in @xmath18 and/or family mixings in @xmath17 .
( there are some exceptional circumstances , which we point out later in this section . )
it is possible to classify the seesaw models into two classes depending on the way the @xmath0 gets its majorana mass @xmath18 . in one case
, @xmath18 arises from renormalizable yukawa couplings involving a @xmath1 higgs field @xmath29 through the superpotential terms @xmath30 here @xmath31 is the majorana yukawa coupling matrix , @xmath32 is the dirac yukawa matrix and we use the standard notation appropriate for supersymmetry with @xmath33 being the conjugate of @xmath0 . in the second class of models ,
the matrix @xmath18 is either put in `` by hand '' as a bare mass term in the lagrangian ( such terms are super - renormalizable ) , or arises from non - renormalizable operators involving a @xmath34 higgs boson @xmath35 via the couplings @xmath36 . in models of the second class , the flavor violating slepton mixings can arise ( assuming msugra ) only from the dirac yukawa couplings of the neutrinos , @xmath32 . on the other hand , in the first class , which is the main focus of this paper , lfv can arise from both the dirac yukawa coupling @xmath32 as well as the majorana yukawa coupling @xmath31 . for the purpose of the present paper
, we will call the two lepton flavor violation alternatives as lfv in the majorana case and lfv in the dirac case .
the simplest example of models with renormalizable majorana yukawa couplings is provided by the minimal ( susy ) extension of the sm with a gauged @xmath7 symmetry .
bare masses for the @xmath0 are then forbidden by @xmath7 gauge invariance . if the higgs sector contains a @xmath1 field @xmath29 , the yukawa coupling @xmath37 as in eq .
( 2 ) will be allowed which will induce @xmath18 of order @xmath38 .
( if instead of @xmath29 , a @xmath34 higgs field @xmath35 is used , @xmath18 can arise from a non - renormalizable coupling @xmath39 .
this belongs to the second class .
) there are interesting extensions of this class of models with renormalizable majorana couplings , such as so(10 ) with a * 126 * higgs field@xcite or left - right symmetric model with a triplet higgs field . in these models the majorana yukawa couplings may play a dominant role in generating the required neutrino mixings .
the role of the dirac yukawa coupling @xmath32 for lfv may be subleading and may even be negligible . to be concrete , we shall assume in our analysis that the matrix @xmath32 is diagonal in a basis where the charged lepton yukawa matrix @xmath26 is also diagonal . for numerical purposes we shall further assume that @xmath32 and @xmath26 are proportional ,
a situation realized naturally if the gauge symmetry is left - right symmetric .
the neutrino mixings will all arise from the majorana coupling matrix @xmath31 , which will be determined ( upto an overall scale factor ) from neutrino oscillation data .
the proportionality assumption @xmath40 is not crucial to our main conclusions on lfv effects related to @xmath18 , but it greatly simplifies the presentation of our results .
this assumption may be viewed as analogous to the assumption often made in the case of dirac lfv models that the super - renormalizable majorana mass matrix @xmath18 is proportional to an identity matrix . in an earlier paper @xcite
we have presented results of our preliminary investigations on lfv involving the majorana yukawa matrix @xmath31 , assuming proportionality of @xmath32 and @xmath26 .
( we called this relation up down unification . ) there we studied a general class of susy breaking models and did not adhere to msugra .
the analysis of ref .
@xcite was carried out in the context of left
right symmetric gauge theories with an enhanced gauge structure .
in contrast to ref .
@xcite , in this paper we analyze lfv effects induced by the @xmath31 matrix within msugra models .
this means that the only source of lepton flavor violation is in the neutrino yukawa sector .
significant lfv effects are found even in this minimal scenario . as explained in sec .
ii.c , a novel way of fitting the low energy neutrino oscillation data has the effect of enhancing the lfv in this minimal scheme .
we also use a simplified gauge structure involving only an extra @xmath7 symmetry .
even this symmetry is not essential , the crucial ingredient is the yukawa coupling matrix @xmath31 of eq .
thus the models studied here are more minimal compared to the left - right models and are likely to be the low energy limits of a wider class of unified models .
our main result is that lfv effects associated with the majorana yukawa coupling @xmath31 are in the range accessible to forthcoming experiments for a large range of mssm parameters consistent with constraints from dark matter , the higgs boson mass and direct experimental search for supersymmetry .
we also point out that the detailed dependence of the branching ratios in processes such as @xmath41 and @xmath3 on the mssm parameters is different in this class of models compared to the case of flavor violation associated with the dirac yukawa coupling @xmath32 . with
some information of the mssm parameters , measurements of these radiative decays will enable one to pin down the nature of interactions that give rise to the right - handed neutrino masses .
obviously , this will have a profound impact on our understanding of the neutrino sector and can shed light on the origin of the seesaw mechanism .
it may be worth pointing out that there are other experimental predictions that may help to distinguish between these two possibilities .
for instance , two of us have recently pointed out that the dirac lfv models can lead to an observable signal in the baryon number non - conserving process neutron - anti - neutron oscillation@xcite whereas in the majorana lfv case , the n-@xmath42 oscillation is unobservable .
there is one caveat in the foregoing discussions that should be mentioned .
it applies equally well to the dirac and majorana lfv alternatives . in a special class of seesaw schemes it is possible that no lepton flavor violation shows up in the next round of experiments without contradicting the current neutrino oscillation data even in the presence of low energy susy .
while we feel that such situations would be non generic , it should be born in mind , nevertheless . in the dirac lfv alternative
it occurs as follows .
it is conceivable that all the flavor mixings arise from the mass matrix @xmath18 and that @xmath32 is diagonal .
flavor violation in the bare mass parameter ( or non - renormalizable term ) @xmath18 is not transmitted to the slepton sector via the rge , implying highly suppressed lfv effects .
one can easily fit all the neutrino oscillation data via the mixings present in @xmath18 .
it is also possible that the scale @xmath18 is much below @xmath43 gev , in which case the elements of @xmath32 will be much smaller than unity so as to fit the neutrino data , again suppressing lfv effects .
similarly , in the majorana lfv alternative , it might so happen that the elements of @xmath31 are very small , compensated by a higher value of @xmath6 so that @xmath18 is unchanged , or that the elements of @xmath32 are small .
such a scenario will suppress lepton flavor violation .
a null result in the rare lepton decay searches would point to a situation in which either of these two situations could be operative and we will have no way to tell which operators are inducing the @xmath0 masses . on the other hand ,
if there is a positive signal in the rare decays , our results will help in probing the origin of @xmath18 .
it appears to us natural to have at least the third family yukawa coupling in @xmath32 of order one ( analogous to the top quark yukawa coupling being of order one ) and similarly at least one element of @xmath31 ( @xmath44 ) to be of order one .
observable lfv effects follow in this case , which is what we analyze .
we have organized this paper as follows . in section
ii , we present the basic model and the relevant renormalization group equations ( rge ) that will be used in getting the slepton mixings and thus the lepton flavor violating rates for the case of majorana lfv . here
we also present analytic approximations to the lfv effects by integrating the rge .
these expressions , while only approximate , are useful in gauging the significance of lfv in the majorana case as well as in comparing it with the case of dirac lfv . in section
ii.a we present the relevant expressions for obtaining the @xmath31 matrix by inverting the seesaw formula . in ii.b
this information is used to compare analytically the predictions for @xmath2 and @xmath3 branching ratios in the two cases . in section
ii.c we present a convenient way to parametrize the majorana yukawa matrix @xmath31 that can be compared directly against low energy neutrino oscillation data on the one hand and lfv effects on the other .
section iii is devoted to our exact numerical results for lfv effects in the case of majorana as well as dirac yukawa couplings .
in iii.a we present our numerical fits to the neutrino parameters , iii.b has our results for the branching ratios @xmath41 and @xmath45 as well as the ratio of the two branching ratios .
section iv has our conclusions .
in order to illustrate in detail the source of lfv for the case of majorana yukawa couplings generating the @xmath0 mass matrix ( the majorana lfv alternative ) , we work with a simple extension of the minimal supersymmetric standard model that accommodates the seesaw mechanism . it is based on the weak gauge group @xmath46 with fermions assigned as follows : @xmath47 .
this is the simplest extension of the weak interaction sector that predicts the existence of the @xmath0 from gauge anomaly cancellations conditions and thus to small neutrino masses via the seesaw mechanism .
the higgs superfields are @xmath48 .
the superpotential for this theory is : @xmath49 here all the yukawa couplings are @xmath50 matrices ; @xmath51 is a singlet field introduced to help facilitate the breaking of @xmath52 down to @xmath53 in the susy limit .
one identifies the sm hypercharge as @xmath54 .
this superpotential gives equal vacuum expectation values ( vevs ) to @xmath55 .
the yukawa coupling matrix @xmath31 will induce large mass for the @xmath33 field , leading to the seesaw mechanism .
this model should be contrasted with the commonly discussed case in the literature where the heavy @xmath0 masses are either put in by hand or arise from a non - renormalizable term in the superpotential . in order to study flavor violation
, we will make the simplifying assumption that both @xmath9 and @xmath26 are diagonal at the planck scale ; therefore all neutrino mixings arise from the general flavor structure of the majorana yukawa couplings @xmath56 .
then from the approximate bi - maximal pattern for neutrino mixings deduced from oscillation data , we can obtain @xmath56 for different patterns of neutrino masses .
since the model is supersymmetric , the flavor violation present in @xmath56 will induce processes such as @xmath57 and @xmath28 decays .
we investigate these predictions using simple analytic approximations in this section .
we need to write down the rge for the yukawa couplings and slepton masses for the class of models under consideration .
denote the left
handed slepton mass squared matrix as @xmath58 and the trilinear @xmath24 terms in the slepton sector as @xmath59 ( in an obvious notation ) .
the relevant rge for @xmath32 and @xmath58 are : @xmath60 \\ \frac{dm^2_l}{dt}~=~\frac{1}{16\pi^2 } [ ( m^2_l + 2m^2_{h_d } ) y^{\dagger}_\ell y_\ell + ( m^2_l+2m_{h_u}^2)y^{\dagger}_{\nu}y_{\nu } + 2y^{\dagger}_\ell m^2_{e^c}y_\ell + y^{\dagger}_\ell y_\ell m^2_l\\ \nonumber+2y^{\dagger}_\nu m^2_{\nu^c}y_\nu + y^{\dagger}_\nu y_\nu m^2_l + 2a^{\dagger}_\ell a_\ell + 2a^{\dagger}_\nu a_\nu -6g^2_2m_2 ^ 2 - 3g^2_{b - l}m_{b - l}^2]\end{aligned}\ ] ] the rge for the relevant @xmath24 parameters are : @xmath61 + y_\ell[tr(6a_dy^{\dagger}_d+2 a_\ell y^{\dagger}_\ell)\\ \nonumber & + & 4 y^{\dagger}_\ell a_\ell+ 2y^{\dagger}_\nu a_\nu+ 6g^2_2m_2 + 2g^2_{r}m_{r}+3g^2_{b - l}m_{b - l}]]\end{aligned}\ ] ] @xmath62 + y_\nu[tr(6a_uy^{\dagger}_u+2 a_\nu
y^{\dagger}_\nu ) \\
\nonumber&+ & 4 y^{\dagger}_\nu a_\nu+8f^{\dagger}a_f+2 y^{\dagger}_\ell a_\ell+6g^2_2m_2 + 2g^2_{r}m_{r}+3g^2_{b - l}m_{b - l}]]\end{aligned}\ ] ] we are interested in lfv arising primarily from the matrix @xmath31 . adopting msugra , we have at the planck scale , @xmath63 .
furthermore , at @xmath23 , we have universality of all scalar masses ( denoted by @xmath64 ) and a common gaugino mass ( denoted as @xmath65 ) . under the set of assumption specified above ,
it is clear from eqs .
( 4)-(7 ) that the off
diagonal elements of @xmath31 will not induce slepton flavor violation to the lowest order in a perturbative expansion , since @xmath31 does not enter directly in the rge for @xmath58 or @xmath66 .
however , the one loop solution for @xmath67 involves @xmath68 , which receives a contribution from @xmath69 ( see eq .
( 4 ) ) . therefore in second order in the parameter @xmath70
, we do have flavor violating effects in @xmath58 and @xmath66 .
it is important to note that this is not a two loop rge effect , rather that it is a one loop rge improved effect , as indicated by the presence of the [ @xmath71 ^ 2 $ ] term .
we can estimate the strength of the slepton flavor violation by integrating the rge analytically .
we find from eq .
( 4)-(7 ) that @xmath72_{ij}\left(\ell n\frac{m_{p\ell}}{m_{b - l}}\right)^2\end{aligned}\ ] ] @xmath73_{ij}\left(\ell n\frac{m_{p\ell}}{m_{b - l}}\right)^2~.\end{aligned}\ ] ] the branching ratio for @xmath74 can be written approximately as @xmath75 here @xmath76 is the average slepton mass at the weak scale , which may be quite different from @xmath77 .
( 10 ) is obtained as an approximation to the chargio and neutralino exchange diagram proportional to lfv in eq .
( 8) with the enhancement factor of @xmath78 as indicated .
@xmath79 in eq .
( 10 ) is an @xmath80 coefficient .
lfv arising from eq .
( 9 ) can be comparable , but it will not change our approximate results by much . we have found that eq .
( 10 ) is a reasonable approximation to the exact numerical results to within a factor of few , provided that @xmath78 is not too large and that there is no large hierarchy between the susy breaking mass parameters ( @xmath81 ) . within msugra , @xmath65 must be larger than about 250 gev , to be consistent with the higgs boson mass constraint ( viz . , @xmath82 gev ) and the @xmath83 constraint .
since we also want to identify the lightest neutralino ( lsp ) as the dark matter in the universe , the relic abundance puts another constraint that the second lightest susy particle ( nlsp ) , the right handed stau in our case , should be nearly degenerate with the lsp to within about 20 - 30 gev .
this condition can be satisfied if @xmath84 is obeyed , which we shall adopt .
take for example a relatively light susy spectrum with @xmath85 gev and @xmath86 .
the charged wino mass is then approximately 200 gev , and the bino mass is about 100 gev .
choosing @xmath87 gev and @xmath88 leads to a mass of about 190 gev for the left handed sleptons and a mass of about 130 gev for the right handed sleptons , with the right handed stau being somewhat lighter around 115 gev .
co - annihilation of lsp will be efficient with such a spectrum .
next we need to choose the neutrino parameters . with @xmath40 , a good fit to the current neutrino oscillation data
can be obtained with the choice @xmath89 and @xmath90 ( see detailed numerical fits later ) .
furthermore , @xmath91 is quite plausible . with this choice
, we estimate @xmath92 for @xmath86 from eq .
( 10 ) . as @xmath93
is increased from 0 to 300 gev , this branching ratio increases to about @xmath94 .
thus we see that for interesting ranges of parameters , the process @xmath95 can be within experimental reach .
more careful examination of the decay is thus warranted .
similarly , the decay @xmath3 can have a branching ratio of order @xmath96 for reasonable ranges of model parameters .
some of this parameter space is already excluded by data , there is a good chance of being able to measure this decay in the planned experiments .
it is instructive to compare the branching ratios obtained here in the case of majorana lfv to that expected in the case of dirac lfv . in the case of dirac lfv , assuming that the @xmath0 majorana mass matrix is proportional to the identity matrix , we can write down the analog of eq .
( 8) as @xmath97 comparing eq .
( 11 ) with eq .
( 8) , we find that numerically the two effects can be comparable .
which effect is more dominant will depend on the overall strengths of the coupling matrices in the two cases .
this will be discussed further in sec .
we can invert the seesaw formula ( eq .
( 1 ) ) and obtain the matrix elements of the majorana yukawa coupling matrix @xmath31 : @xmath98 we also have the relation for the inverse of the light neutrino mass matrix @xmath99 in terms of the light neutrino mass eigenvalues @xmath100 and the leptonic mixing matrix @xmath101 as @xmath102 we ignore cp violation and use the near bi - maximal approximation for @xmath101 given by @xmath103 here @xmath104 from the chooz and palo - verde reactor experiments@xcite .
@xmath105 is the solar neutrino mixing angle for which the allowed range after the sno data is @xcite @xmath106 at three @xmath107 confidence level . multiplying all these together , we have @xmath108 where we have assumed @xmath17 to be diagonal with its entries denoted by @xmath109
. the definition @xmath110 has been used .
we have then @xmath111 using these expressions , for different scenarios of neutrino masses we can determine the values of @xmath56 which enter into the flavor changing slepton masses .
we will focus primarily on the hierarchical neutrino mass schemes where @xmath112 .
let us first discuss the case where the right handed neutrino masses arise from a renormalizable majorana yukawa coupling @xmath31 . as in the previous section , let us assume that the dirac neutrino mass matrix @xmath17 is diagonal in the same basis where @xmath113 is diagonal .
one interesting pattern is provided by the normal hierarchy @xmath112 . in this case
, it is plausible , but not mandatory , that in eq .
( 16 ) , the terms with @xmath114 in the denominator dominate @xmath115 with all other terms being negligible . in this approximation
, we obtain for the process @xmath2 process an amplitude proportional to @xmath116 similarly , the @xmath3 process has an amplitude given by @xmath117 we have set @xmath118 for simplicity here . from eq .
( 17)-(18 ) it might appear that the branching ratios can increase without limit as @xmath114 is decreased .
however , this is not true .
the coupling @xmath119 should remain perturbative ( @xmath120 ) .
this means that for a given @xmath114 , there is a lower limit on @xmath6 .
for example , if @xmath121 ev , and for @xmath122 gev , @xmath123 , it must be that @xmath124 gev so that @xmath125 .
it turns out that with this constraint , the branching ratios ( br ) for the decays @xmath95 and @xmath45 are suppressed within msugra , mainly because the logarithm upon which the br has a fourth power dependence , is small . we conclude that in this scenarios with @xmath126 dominance in eq .
( 16 ) , in the case of majorana lfv there is no significant lfv in lepton sector within msugra .
we have also examined lfv effects if the msugra assumption is relaxed somewhat for the case of @xmath126 dominance .
if for example , @xmath127 at @xmath23 with @xmath128 different from @xmath93 , we found that the br for @xmath95 can be in the range @xmath129 for relative large @xmath130 tev .
the br for @xmath45 is extremely small ( @xmath131 with this set of boundary conditions .
we shall not present details of this fit , and will focus solely on the msugra case .
for comparison , let us also consider the case of dirac lfv . for @xmath132 there is an unambiguous connection between the lfv effects and the neutrino mass matrix . to see this more explicitly ,
note that in this case , the seesaw formula reduces to @xmath133 thus the product @xmath134 can be completely expressed in terms of the neutrino masses and mixings .
since for the cp conserving case , the flavor violating slepton mixings effects always involve @xmath135 , the lfv amplitudes can be expressed directly in terms of the neutrino mass matrix .
assuming again that @xmath112 , we can write down the amplitude for the decay @xmath2 to be proportional to @xmath136 , while that for the decay @xmath45 is proportional to @xmath137 . these amplitudes increase with the light neutrino masses , unlike the case of majorana lfv , which scale inversely with @xmath114 .
both decays have branching ratios accessible in the planned experiments in the dirac lfv alternative , as will be further explained in sec .
the dominance of @xmath126 in @xmath138 and therefore in @xmath56 is not the only possibility .
we have found a large range of parameters where the lfv effects are significant while deviating from the @xmath126 dominance . to see this
explicitly it is more convenient to define the light neutrino mass matrix @xmath139 in the gauge basis as a perturbative expansion in a small parameter @xmath140 @xmath141 here @xmath142 are order one coefficients , @xmath143 and the exponents @xmath144 in the ( 1,1 ) and ( 1,2 ) entries have to be at least 1 , but can be larger .
this matrix provides a good fit to all neutrino data .
if @xmath145 , and @xmath146 it will reproduce the results of the previous section with @xmath126 dominance . but for other values of @xmath144 , which are also consistent with all neutrino oscillation data , we have found much larger lfv effects . consider the case @xmath147 for example . in this case
the ( 1,1 ) and the ( 1,2 ) entries of @xmath139 are not completely specified by neutrino data alone .
inverting eq .
( 20 ) we obtain for the matrix @xmath31 , @xmath148 here we have used the parametrization @xmath149 . all parameters in @xmath31 are fixed , except for @xmath150 .
if @xmath151 , we have a situation where the off diagonal entries as well as the diagonal entries of @xmath31 are of the same order leading to large lfv effects .
contrast this with the case where @xmath145 is adopted in eq .
( 20 ) , in which case we would have the ( 3,3 ) entry of @xmath31 to be order 1 , the ( 2,3 ) to be order @xmath152 , the ( 1,3 ) of order @xmath153 , and the ( 1,2 ) entry of order @xmath154 .
the choice ( which is equivalent to @xmath126 dominance in @xmath31 ) will lead to a suppression of the @xmath95 branching ratio by a factor @xmath155 which would put it in an unobservable range .
the decay @xmath45 will be suppressed even further , by a factor @xmath156 , compared to the case of @xmath157 .
it is clear then that the most interesting case for lepton flavor violation is when @xmath158 in eq .
this is the choice we shall make for our numerical fits in sec .
we start with a basis where the charged lepton mass matrix is diagonal . for the dirac neutrino mass matrix
we assume that @xmath159 where @xmath160 .
this assumption is well justified if the model is embedded into a susy left
right framework as we have shown earlier papers@xcite .
it is also possible that this assumption holds approximately in a wider class of models where the off diagonal entries in @xmath26 and @xmath32 are small , although not strictly proportional .
the proportionality constant @xmath161 is a parameter of the model .
if we assume that the @xmath162 dirac yukawa coupling is the same as the top
quark yukawa couplings , as happens in some versions of minimal @xmath163 models , then @xmath164 for large values of @xmath165 while @xmath166 for @xmath167 .
the light neutrino majorana mass matrix is obtained as @xmath168 in our analysis , we choose values of @xmath169 , since smaller @xmath78 values are less preferred by the recent higgs mass bounds from lep @xcite and the muon @xmath170 anomaly@xcite , if interpreted as a susy effect , would prefer a moderately large @xmath78 . following the analytic expression given in sec .
ii.c , we obtain a fit for the light neutrino spectrum as follows .
we choose tan@xmath171 = 10 and find a good fit to the solar and atmospheric neutrino data with hierarchical neutrino masses if the matrix @xmath31 defined at at @xmath172 gev has entries given by @xmath173 the resulting neutrino masses at the weak scale ( using the rge evolution @xcite ) are : @xmath174 the leptonic mixing matrix is given by : @xmath175 note that the mixing angles @xmath176 and @xmath177 are in very good agreement with the solar and atmospheric oscillation data .
note also the prediction for the parameter @xmath178 , which is in the range that can be tested in planned long baseline experiments .
( muegmlr1new.eps width 8 cm ) for comparison purposes we will also present lfv effects associated with dirac yukawa couplings , assuming no flavor violation in the majorana sector
. it will be desirable to have a neutrino mass fit for this case as well , where @xmath18 is proportional to an identity matrix and all flavor mixings arise from @xmath32 .
we have found such fits for the same set of light neutrino spectrum as in eq .
( 25 ) -(26 ) .
actually , there are discrete possibilities for @xmath32 since the equations are not linear in @xmath32 .
we have found that the predictions for lfv decays are essentially unchanged within these discrete possibilities .
so we focus on one of them . at @xmath179 gev , we find that @xmath32 is given by @xmath180 although @xmath32 in eq .
( 27 ) have complex elements , @xmath181 is real , as can be easily verified .
thus the light neutrino mass matrix @xmath139 is real ( see eq .
( 19 ) ) , and there is no cp violation in neutrino oscillations in this limit , as in the case of majorana lfv .
we shall use this fit for the case of dirac lfv case in our numerical analysis . in our calculation
we make the following assumptions about the supersymmetry breaking sector , adopting the msugra framework . we assume universal scalar masses @xmath64 , a common gaugino mass @xmath65 and trilinear soft susy breaking @xmath24 terms that are proportional to the respective yukawa coupling matrices , with a common proportinality constant @xmath93 .
we take the fundamental scale at which these relations hold to be the gut scale @xmath182 gev .
we impose various experimental constraints on the susy spectrum .
in particular , the higgs boson mass limit ( @xmath183 gev ) along with the @xmath184 constraint requires that @xmath185 gev .
the direct limits on susy searches are also taken into account .
furthermore , radiative electroweak symmetry breaking is demanded .
we choose the sign of @xmath186 to be positive , which is preferred by the @xmath184 constraint .
we do not explicitly impose constraints from muon @xmath170 , but we note that for the range of parameters we have chosen ( positive @xmath186 and moderate to large @xmath78 ) , this anomaly can be explained via susy exchange , provided that @xmath65 is limited to less than about 800 gev ( @xmath187 ) .
( muegmlr2new.eps width 8 cm ) additionally , we require that the lightest neutralino be the lightest supersymmetric particle ( lsp ) in the model .
furthermore , this lsp should constitute the dark matter of the universe .
however , for generic susy parameters the neutralino annihilation cross section is not strong enough to bring its relic abundance to the required level from dark matter constraints ( density in dark matter should be @xmath18830% @xcite of the critical density in order for the lsp to form a good dark matter candidate ) .
this remark applies for the range of parameters that we are interested in , viz .
, @xmath189 gev and @xmath167 .
the dark matter constraint can be satisfied by invoking the co - annihilation mechanism involving the lsp and the second lightest susy particle ( nlsp ) , which is the right handed stau in these models@xcite . in order for this co - annihilation to be efficient , we need the lighter stau mass to be within 25 - 30 gev of the lsp mass .
( taumugmlr3final1.eps width 8 cm ) we choose the universal scalar mass @xmath64 such that this constraint on the lighter stau mass is satisfied .
we allow the range @xmath190 .
for example , for the case of @xmath191 , @xmath84 will be required to satisfy this constraint .
we choose @xmath192 at the gut scale . for the common @xmath93 ,
we choose three different values , @xmath193 gev .
we find that the branching ratio increases with @xmath93 , as can be inferred from the analytic approximation , eq .
( 8) . in fig .
1 we plot the branching ratio for @xmath3 in the majorana lfv case as a function of @xmath65 , varying it from its lower limit of 250 gev to 1 tev . here
we chose @xmath191 .
we find that the br varies between @xmath194 and @xmath195 .
part of the parameter space is already ruled out by the non observation of @xmath196 .
this experimental lower limit on the br is indicated by a thick solid line in fig .
we conclude that the entire parameter space can be explored with the planned round of @xmath45 experiment @xcite . in fig .
2 , we show the @xmath78 dependence of the b@xmath197 for the same set of input parameters as in fig . 1 . here
we present results for two values of @xmath198 and 30 .
the branching ratio increases with @xmath78 , as anticipated from eq .
( 10 ) . as in fig . 1
, we demand the dark matter constraint be satisfied in fig .
2 as well .
( taumugmlr3final2.eps width 8 cm ) in fig .
3 we plot the branching ratio for @xmath199 for the case of @xmath191 .
the solid line corresponds to the majorana lfv case . for comparison
, we also present in the same graph the @xmath2 br in the case of the dirac alternative ( fit of eq .
( 27 ) ) . in fig .
4 we plot the same quantities for a different value of @xmath200 gev and @xmath86 .
we find from figs . 3 and 4 that the @xmath2 is typically smaller in the majorana case compared to the dirac case , although they can be comparable for some values of @xmath65 where there is an accidental cancellation between the chargino and the neutralino diagrams in the case of dirac lfv ( and not for the majorana lfv for this input choice ) .
the br for the case of majorana lfv is @xmath201 for @xmath88 , but can be as large as @xmath202 for @xmath203 gev . in fig . 5
, we show the b(@xmath204 ) as a function of @xmath65 for the majorana alternative ( solid lines ) and compare it with the dirac case ( dashed line and dot sashed line ) .
we have chosen @xmath191 and @xmath205 for illustration in this case .
the dashed line is drawn for the @xmath9 shown in eq .
we see that the branching ratio is large in the dirac case . however if we scale down the dirac neutrino coupling by a factor of 2.5 , which does not alter the light neutrino spectrum ( since the overall scale factor is not determined from low energy data ) , we find that the branching ratio ( denoted by the dot dashed line ) decreases .
thus , depending on the overall scale factor , @xmath206 can be larger in the majorana case compared to the dirac case ( or vice versa ) . as fig .
5 shows , there is some uncertainty in the br arising from the overall scale factor .
if we consider ratios of branching ratios , this uncertainty disappears .
we define the ratio @xmath207 and study its dependence on the susy spectrum as well as on the dirac / majorana cases . in figures 6 and 7 ,
we plot @xmath5 for the cases of dirac and majorana alternatives for @xmath191 ( fig .
6 ) and @xmath208 gev ( fig .
7 ) for @xmath205 .
we see from these figures that the ratio depends on the supersymmetry breaking parameters @xmath93 , @xmath64 , @xmath65 etc , in a way that is different for these two cases .
this difference shows up in almost the entire parameter space .
we also notice that the ratio @xmath5 is large for the majorana case for large @xmath93 .
we conclude that with some information on the susy breaking parameters , measurement of @xmath5 can provide crucial insight into the origin of the seesaw mechanism .
= 8.0 cm = 8.0 cm = 8.0 cm
in this paper we have studied the consequences of implementing the seesaw mechanism by renormalizable yukawa couplings involving a @xmath209 higgs field on lepton flavor violating processes . the @xmath0
majorana masses arise as renormalizable yukawa couplings , rather than bare mass terms , and therefore influence the renormalization group evolution of the soft susy masses of sleptons .
this slepton flavor violation is transmitted to the lepton sector and induces the rare leptonic decays @xmath45 and @xmath210 .
we have emphasized that this case is qualitatively as well as quantitatively different from the case where the @xmath0 majorana masses are directly put in by hand or are induced by non - renormalizable terms . we have found interesting structure for the majorana yukawa coupling matrix @xmath31 which is consistent with low energy neutrino oscillation data and that leads to observable lepton flavor violations in radiative lepton decays .
we have adopted a susy breaking scenario with minimal flavor violation , the popular msugra scheme , and have found even within this scheme that the radiative lepton decays are within reach of planned experiments .
the parameter space that we have explored is consistent with acceptable abundance of neutralino dark matter , as well as higgs boson mass limit , @xmath211 constraint , muon @xmath170 constraint as well as the direct experimental limits on susy particles .
we have compared the predictions from the majorana yukawa coupling related lfv with that related to the dirac yukawa coupling .
the ratio @xmath212 , being insensitive to the overall scale of @xmath7 symmetry breaking , along with some information on the susy spectrum , can provide deep insight into the origin of the seesaw mechanism .
the work of r. n. m. is supported by the national science foundation grant no .
ksb is supported by u.s .
department of energy grants de - fg03 - 98er41076 and de - fg02 - 01er45684 , and by a grant from the research corporation . v. barger , d. marfatia , k. whisnant , and b. p. wood , phys
b537 * , 179 ( 2002 ) ; j. bahcall , c. gonzalez - garcia and c. pena - garay , jhep * 0207 * , 054 ( 2002 ) ; a. bandopadhyay , s. choubey , s. goswami and d. p. roy , phys . lett . *
b540 * , 14 ( 2002 ) ; p. de hollanda and a. smirnov , hep - ph/0205241 .
cleveland et .
al . , astrophys .
j. * 496 * , 505 ( 1998 ) ; r. davis , prog .
. phys . * 32 * , 13 ( 1994 ) ; superkamiokande collaboration , y. fukuda et .
al . , phys .
lett . * 81 * , 1158 ( 1998 ) ; erratum @xmath213 . , * 81 * , 4279 ( 1998 ) and @xmath213 . , * 82 * , 1810 ( 1999 ) ; y. suzuki , _ nucl . phys . proc . suppl . _
* b 77 * , 35 ( 1999 ) ; y. fukuda et .
al . , phys .
* 86 * , 5651 ( 2001 ) ; sage collaboration , j.n .
abdurashitov et .
al . , phys . rev .
* c60 * , 055801 ( 1991 ) ; gallex collaboration , w. hampel et .
b447 * , 127 ( 1999 ) ; sno collaboration , q.r . ahmed et .
al . , phys .
lett . * 87 * , 071301 ( 2001 ) .
m. gell - mann , p. rammond and r. slansky , in _
supergravity _ ,
d. freedman et .
al . , ( north - holland , amsterdam , 1980 ) ; t. yanagida , in proc . kek workshop , 1979 ( unpublished ) ; r.n .
mohapatra and g. senjanovi , phys .
* 44 * , 912 ( 1980 ) .
f. borzumati and a. masiero , phys .
lett . * 57 * , 961 ( 1986 ) ; l. hall , v. kostelecky and s. raby , nucl . phys . *
b267 * , 415 , ( 1986 ) .
j. hisano , t. moroi , k. tobe , m. yamaguchi and t. yanagida , phys . lett . *
b357 * , 579 ( 1995 ) ; j. hisano , t. moroi , k. tobe and m. yamaguchi , phys . rev . * d53 * , 2442 ( 1996 ) . s. f. king and m. oliveira , phys . rev . *
d60 * , 035003 ( 1999 ) ; j. hisano and d. nomura , phys . rev . *
d59 * , 116005 ( 1999 ) ; k. s. babu , b. dutta and r. n. mohapatra , phys
b458 * , 93 ( 1999 ) ; w. buchmuller , d. delepine and f. vissani phys . lett . * b459 * , 171 ( 1999 ) ; j. ellis , m. gomez , g. leontaris , s. lola and d. nanopoulos , eur .
j. * c14 * , 319 ( 2000 ) ; s. lavgnac , i. masina and c. savoy , phys . lett . * b520 * , 269 ( 2001 ) ; nucl . phys .
* b633 * , 139 ( 2002 ) ; j. casas and a. ibarra , nucl . phys . * b618 * , 171 ( 2001 ) ; r. gonzalez felipe and f. joaquim , jhep * 0109 * , 015 ( 2001 ) ; j. sato , k. tobe and t. yanagida , phys . lett . *
b498 * , 189 ( 2001 ) ; x. j. bi , y. b. dai and x. y. qi , phys . rev . * d63 * , 096008 ( 2001 ) , phys . rev . *
d66 * , 076006 ( 2002 ) ; g. cvetic , c. dib , c.s .
kim and j.d .
kim , phys . rev .
* d66 * , 034008 ( 2002 ) ; j. ellis , j. hisano , s. lola and m. raidal , nucl . phys . * b621 * , 208 ( 2002 ) ; a. kageyama , s. kaneko , n .
shimoyama and m. tanimoto , phys . lett . *
b527 * , 206 ( 2002 ) ; d. chang , a. masiero and h. murayama , hep - ph/0205111 ; j. ellis , j. hisano , m. raidal and y. shimizu , hep - ph/0206110 ; f. deppisch , h. pas , a. redelbach , r. ruckl and y. shimizu , hep - ph/0206122 ; k. babu and c. kolda , hep - ph/0206310 ; a. rossi , hep - ph/0207006 ; a. masiero , s. vempati and o. vives , hep - ph/0209303 .
a. chamsheddine , r. arnowitt and p. nath , _
n=1 supergravity _ , world scintific , singapore ( 1984 ) ; r. barbieri , s. farrara and c. savoy , phys . lett . *
b119 * , 343 ( 1982 ) ; l. hall , j. lykken and s. weinberg , phys .
rev.*d27 * , 2359 ( 1983 ) .
k.s . babu , c.n . leung and j. pantaleone ,
phys . lett .
* b319 * , 191 ( 1993 ) ; p.h .
chankowski and z. pluciennik , phys .
b316 * , 312 ( 1993 ) ; s. antusch , m. drees , j. kersten , m. lindner and m. ratz , phys . lett . * b525 * , 130 ( 2002 ) .
m. turner 2002 , astro - ph/0202007 .
j. ellis , t. falk and k. olive , phys .
b444 * , 367 ( 1998 ) ; j. ellis , t. falk , k. olive and m. srednicki , astropart .
* 13 * , 181 ( 2000 ) ; r. arnowitt , b. dutta and y. santoso , hep - ph/0010244 ; nucl . phys . *
b606 * , 59 ( 2001 ) ; j ellis , t. falk , g. ganis , k. olive and m. srednicki , phys .
* b570 * 236 ( 2001 ) ; erratum - ibid . * 15 * , 413 ( 2001 ) ; m. gomez and j. vergados , hep - ph/0012020 ; m. gomez , g. lazarides and c. pallis , phys . rev . * d61 * , 123512 ( 2000 ) ; phys . lett .
* b487 * , 313 ( 2000 ) ; l. roszkowski , r. austri and t. nihei , jhep * 0108 * , 024 ( 2001 ) ; a. lahanas , d. nanopoulos and v. spanos , phys . lett . * b518 * , 94 ( 2001 ) . | the right handed neutrino mass matrix that is central to the understanding of small neutrino masses via the seesaw mechanism can arise either ( i ) from renormalizable operators or ( ii ) from nonrenormalizable or super - renormalizable operators , depending on the symmetries and the higgs content of the theory beyond the standard model . in this paper , we study lepton flavor violating ( lfv ) effects in the first class of seesaw models wherein the @xmath0 majorana masses arise from renormalizable yukawa couplings involving a @xmath1 higgs field .
we present detailed predictions for @xmath2 and @xmath3 branching ratios in these models taking the current neutrino oscillation data into account . focusing on minimal supergravity models , we find that for a large range of mssm parameters suggested by the relic abundance of neutralino dark matter and that is consistent with higgs boson mass and other constraints , these radiative decays are in the range accessible to planned experiments .
we compare these predictions with lepton flavor violation in the second class of models arising entirely from the dirac yukawa couplings .
we study the dependence of the ratio @xmath4 on the mssm parameters and show that measurement of @xmath5 can provide crucial insight into the origin of the seesaw mechanism .
epsf.tex ( # 1 width # 2)=#2 1.0 in |
a possible existence of a tribaryon state with strangeness @xmath8 and isospin 0 has been predicted theoretically by akaishi and yamazaki as a deeply - bound @xmath9 nuclear state in @xmath10he nuclei @xcite .
recently , the kek - ps e471 group reported a distinct peak in a missing mass spectrum of the @xmath0he(@xmath11 reaction @xcite , which was interpreted to be a signal of a tribaryon state with strangeness @xmath21 and isospin 1 , called s@xmath12 ( m = 3117.0@xmath13 mev/@xmath3 and @xmath14 21 mev/@xmath3 ) .
the same group obtained another candidate of the tribaryon state , s@xmath15(3140 ) , in the @xmath0he(@xmath16 missing mass spectrum , but its statistical significance was not sufficient to claim a definitive evidence for its existence @xcite .
these results have triggered many theoretical interpretations @xcite . meanwhile the finuda group observed a peak in the proton momentum spectrum of @xmath17li(@xmath18 ) @xcite .
they attributed the peak to the proton emission from the @xmath19 capture on a `` quasi '' deuteron in @xmath17li as , @xmath20 ( @xmath21 mev/@xmath22 ) , which was introduced as an interpretation of s@xmath23(3115 ) by ref .
@xcite , and they claimed that the peak structures observed in e471 and finuda should be of the same nature .
counter arguments to this interpretation followed @xcite .
therefore , the confirmation and further study of these states were strongly awaited .
this situation has lead us to carry out a new experimental search with improved resolution and higher statistics based on the e471 experimental setup .
the reaction we adopted is @xmath0he(@xmath24 ) , the same as in e471 ; @xmath25 where the state with isospin 1 and charge 0 ( s@xmath23 ) is populated in the proton emission reaction and the state with isospin 1 and 0 and charge @xmath261 ( s@xmath27 ) state in the neutron emission reaction .
the state originally predicted by akaishi and yamazaki has be searched for using neutron spectroscopy .
the objectives of the proton spectroscopy are , first , to confirm s@xmath12 and determine its width and formation ratio with improved resolution by an inclusive measurement . since the experimental setup of e471
was originally designed to be optimized for the neutron tof measurement , the experimental resolution for protons was not satisfactory .
moreover , the limited momentum acceptance for protons made the search for excited states of s@xmath28 difficult .
thus , the second objective of the present experiment is to search for other candidates for tribaryon states , including excited states of s@xmath12 . in this letter
, we give the first report , namely , on experimental results on the missing mass analysis of the @xmath0he(@xmath29 ) reaction , and results on the neutron measurement will be given in a paper to follow .
the experiment ( called `` e549 '' ) was performed at the k5 beamline of the kek 12 gev proton synchrotron .
the experimental setup is schematically illustrated in fig .
[ setup ] together with a detailed view around the target in the inset . a @xmath19 beam with momentum 650 mev/@xmath22 was stopped inside a super - fluid helium-4 target ( 20 cm in diameter , 15 cm long at a density of 0.145 g / cm@xmath10 ) .
we determined the energy of protons emitted from the @xmath19 absorption reaction by means of the time - of - flight ( tof ) method .
the experimental apparatus was based on that of the former experiment e471 , and its description can be found in ref @xcite .
since the setup of e471 was optimized for neutron spectroscopy , considerable improvements have been made to achieve better energy resolution for the proton tof measurement . in e549 ,
tof detectors dedicated for the proton measurement were installed .
they consist of a tracking drift chamber ( pdc ) and a set of tof start / stop counters ( @xmath31 and @xmath32 ) , which are plastic scintillator walls with 8 and 17 segments , respectively .
the missing mass resolution was expected to be twice that of e471 thanks to the precise determination of the tof and the flight path . beside this , the @xmath0he(@xmath29 ) data were taken inclusively , whereas the proton data in e471 were measured in coincidence with secondary charged particles ( `` semi - inclusive '' ) .
the present inclusive measurement enables us to determine the formation ratio of the state precisely .
together with the beam timing counter ( t0 ) , @xmath31 also gives us information on the reaction timing which is defined by the time difference between the stopped kaon and outgoing charged particles .
furthermore , we extended the momentum acceptance for the proton to perform the experimental search with a wider missing mass region than in e471 .
this became possible by minimizing the amount of material between the target and @xmath32 .
data were taken for a month from the end of may , 2005 , and the accumulated data correspond to ( 1.03 @xmath33 0.15 ) @xmath34 stopped @xmath19 inside the target .
there is a well - known fact that a few percent of stopped @xmath19 s in liquid helium form long - lived atomic states with large angular momenta .
these `` _ metastable _ '' kaonic he atoms were directly observed in a time spectrum by a past kek experiment with @xmath19 absorption on @xmath0he @xcite . since the partial lifetime of this state ( @xmath35=59@xmath334 ns@xcite ) is much longer than the kaon lifetime , most kaons in this state decay at - rest without nuclear absorption .
a monochromatic muon associated with a two body decay of stopped @xmath19 ( @xmath36 , called @xmath37 ) was used for determining the timing resolution of the present tof system and its long - term stability throughout the experiment .
the yield of @xmath38 was also used for the normalization of the missing mass spectrum as described later .
these decay components , especially @xmath37 , can be clearly identified by applying the timing analysis between the stopped kaons and the outgoing muons .
this is because the lifetime of a kaon in the metastable atomic state together with the kaon decay lifetime ( @xmath39 ns @xcite ) is much longer than the typical hyperon lifetimes which are of the order of 100 ps .
thus , by selecting the delayed timing events , we can drastically suppress the events from the prompt nuclear absorption and successive hyperon decays .
figure [ meta ] shows the inverse velocity ( 1/@xmath40 ) spectrum of secondary charged particles from the @xmath0he(@xmath41 ) reaction with delayed timing selection ( @xmath42 ns ) .
we can clearly see three peaks at @xmath43 .
they correspond to electrons ( mainly @xmath44 ) , muons ( @xmath38 ) and pions ( @xmath45 ) , respectively .
we estimated the tof resolution from the width of the @xmath37 muons , and found it to be 0.020 ( @xmath46 ) after fitting the spectrum with three gaussians and a second - order polynomial function .
this value is equivalent to that obtained from @xmath47 events in the stopped @xmath48 calibration data .
spectrum from the @xmath0he(@xmath49 , @xmath50 ) reaction with the delayed event selection .
the spectrum was fitted by a three gaussian and a second - order polynomial function .
fitting results are overlaid by dotted lines . ] protons from the @xmath0he(@xmath41 ) reaction were clearly identified by the correlation between 1/@xmath40 and the total light output in p@xmath51 and nc .
figure [ pid ] shows the criteria for proton selection .
this selection method is essentially the same as that of e471 , but we can detect slower protons , i.e. a higher missing mass region , thanks to the extension of the momentum acceptance ( see fig . 2 in ref .
the contamination due to reacted pions was estimated to be about 5 % of protons in the present proton selection . and the total light output in @xmath32 and nc in units of mev@xmath52 ( mev electron equivalent ) .
the two lines in the figure define the proton region . ]
figure [ mom ] shows the obtained momentum distribution under the inclusive condition . in the figure , a correction for the energy loss between the reaction vertex and the tof counters was applied .
the overall spectrum shape is understood well by the two nucleon absorption reaction , @xmath53 , for the higher momentum side and quasi - free hyperon production and its decay , @xmath54 and @xmath55 , for the lower one .
the upper figure shows the geometrical acceptance of protons , which was obtained by a monte carlo simulation using the geant 4 package taking into account the realistic stopped @xmath19 distribution and the e549 setup .
the vertical line in the figure shows the lower acceptance limit in e471 .
the present sensitive region was extended below 415 mev/@xmath22 .
as shown in the figure , there is no distinct peak . in the following sections
, we will determine the upper limit of the formation branching ratio for a tribaryon state .
he@xmath56 reaction .
the upper figure represents the geometrical acceptance for protons as a function of the proton momentum .
the vertical line in the figure shows the acceptance limit of e471 . ]
the proton counts in the missing mass spectrum were normalized by the number of stopped @xmath19 .
for the normalization , we used the metastable state described above , because the fraction of free decays , @xmath57 , of stopped @xmath19 on @xmath0he was measured to be 3.5 @xmath33 0.5 % /(@xmath7 ) in ref .
the proton yield , @xmath58 , is given as @xmath59 where @xmath60 and @xmath61 are the number of counts of the @xmath37 muons and protons respectively , @xmath62 is the branching ratio of the @xmath37 decay , @xmath63 is the reduction factor resulting from the delayed timing gate , i.e. @xmath64 , and @xmath65 and @xmath66 are detection efficiencies of muons and protons respectively .
figure [ mass ] shows the normalized missing mass spectrum obtained from the inclusive @xmath0he@xmath56 reaction together with the proton momentum scale and energy thresholds of the @xmath67 and @xmath68 . the ordinate unit shows the proton yield per stopped @xmath19 , and the systematic error of the ordinate corresponds to 14 % in the relative error , which is dominated by the error of @xmath57 .
an inset in the figure is the close - up view of the lower mass region .
the upper figure in fig [ mass ] shows the missing mass resolution as a function of the missing mass .
it was derived from a monte carlo simulation with an assumption that the tof resolution of protons is equivalent to that of @xmath37 muons .
this resolution is about twice better than that of ref .
this was used in a spectral fitting for the determination of upper limits described in the following section .
he@xmath56 reaction by the inclusive measurement .
the abscissa is the missing mass and the ordinate is the proton yield normalized by the number of stopped kaons .
the systematic error of the ordinate contains a 14 percent relative error .
the upper figure shows the overall missing mass resolution in the present experiment . ] for quantifying the search results , we derived upper limits for the formation branching ratio of a tribaryon state .
they were obtained by fitting the spectrum shown in fig .
[ mass ] with an assumed peak and a smooth background function .
the fitting procedure is as follows ; * the fitting function was set to be a voigtian ( convolution of gaussian and lorentzian ) as a peak and a polynomial function as a background .
the yield of the state was determined by the area of the lorentzian .
the natural width and center of the peak were fixed for each fitting .
the standard deviation of the gaussian was set to be the experimental resolution taking the mass dependence into account as shown in fig .
* since the background shape can not be fully reproduced by a monte carlo simulation , we assume that the background can be expressed as a smooth function within the fit region .
* we tested three fit regions determined by the width of the peak function to be ( @xmath69 hwhm ( @xmath70 ) ) .
conservative upper limits were determined using the highest value of the results from each fit region .
we used this fit method with widths of 0 , 20 and 40 mev/@xmath3 by changing the mass from 3000 to 3200 mev/@xmath3 .
systematic errors , mainly dominated by an uncertainty in @xmath57 , were taken into account in the total error of the peak intensity by summing quadratically the systematic error and fitting error ( @xmath71 ) . in the fitting
, the yield of the state sometimes results in an unphysical value , namely , a negative lorentzian area . in that case
, we followed a prescription of _ `` unified approach '' _ assuming gaussian statistics @xcite .
the upper limit on the formation branching ratio at 95 % confidence level was obtained using the table x in ref .
@xcite with the resultant peak yield and the total error .
figure [ ul ] shows a plot of the upper limits versus the missing mass for widths of the assumed states of 0 , 20 and 40 mev/@xmath3 .
the present procedure gives a higher upper limit especially for the state with a larger width .
thus , it must be emphasized that the present procedure using the inclusive proton spectrum only is insensitive for a broad peak structure , which will be investigated with exclusive measurements .
the present upper limit of the formation branching ratio at the mass of 3115 mev/@xmath3 , where the e471 group observed s@xmath12 , is @xmath720.2 % /(@xmath73 ) with the width of @xmath74 = 20 mev/@xmath3 .
although the trigger condition was different , the e471 group reported a formation branching ratio of about 1 % /(@xmath73 ) in ref .
@xcite , which can be excluded by more than a 95 % confidence level .
we studied the reason for this discrepancy and found as a most likely cause for a fake peak formation , an erroneous slewing correction of the proton time of flight spectrum for the large proton signals compared to those of well calibrated minimum ionizing particles @xcite .
it is worth mentioning that the inclusive proton momentum spectrum from @xmath0he(@xmath49 , @xmath75 ) has no structure whereas the finuda group found a peak in the proton momentum spectrum for @xmath17li(@xmath49 , @xmath75 ) .
there is a controversial discussion concerning the production of monochromatic proton by antikaon absorption on low momentum deuteron clusters in nuclei @xcite . in the case of @xmath17li ,
the peak observed in the finuda spectrum may be attributed to such a process , however , it is obvious that the monochromatic process does not occur in the @xmath76he absorption case at the level of @xmath77 .
= 0 , 20 and 40 mev/@xmath3 states , respectively . ]
in conclusion , we have searched for neutral tribaryon states with strangeness @xmath21 and isospin 1 by missing - mass analysis of the inclusive @xmath0he@xmath56 reaction .
quite high statistics for protons ( more than one million ) was accumulated in the momentum range of 300 700 mev/@xmath22 .
no significant narrow structure was observed .
the upper limits for the formation branching ratio in the mass range 3000 @xmath78 m @xmath78 3200 mev/@xmath3 with an assumed width of 0 , 20 and 40 mev/@xmath3 were determined to be approximately @xmath79 , @xmath80 and @xmath81/(@xmath7 ) , respectively , with a 95 % confidence level .
this result clearly indicates that a large formation ratio ( order of 1 % /(@xmath73 ) ) of a narrow tribaryon state , such as s@xmath12 is excluded .
we would like to thank all of the staff in kek - ps for their continuous support .
this research was partially supported by riken , kek and the ministry of education , science , sports and culture , grand - in - aid for scientific research ( s ) , 14102005 .
00 y. akaishi and t. yamazaki , phys .
c 65 ( 2002 ) 044005 .
t. suzuki et al . , phys .
b 597 ( 2004 ) 26 . m. iwasaki et al . , arxiv : nucl - ex/0310018 .
y. akaishi , a. dote and t. yamazaki , phys .
b 613 ( 2005 ) 140 .
y. maezawa et al . , prog .
phys . 114 (
2005 ) 317 .
e. oset and h. toki , phys .
c 74:015207 , 2006 , nucl - th/0509048 .
m. agnello et al .
( finuda collaboration ) , nucl .
phys . a 775 ( 2006 ) 35 .
t. yamazaki and y. akaishi , nucl - ex/0609041 , nucl .
a 792 ( 2007 ) 229 .
m. iwasaki et al . , nucl .
inst . meth .
a 473 ( 2001 ) 286 .
t. yamazaki et al . , phys .
63 ( 1989 ) 1590 .
h. outa , doctoral dissertation , univ . of tokyo ( 2003 ) .
g. j. feldman and r. d. cousins , phys .
d 57 ( 1998 ) 3873 .
m. iwasaki et al . , arxiv:0706.0297 [ nucl - ex ] , to be published in epj . | a search for tribaryon states was performed at kek - ps .
we adopted the @xmath0he(@xmath1 reaction to populate the states with strangeness @xmath21 , charge 0 and isospin 1 .
no significant narrow structure was observed in the mass region from 3000 to 3200 mev/@xmath3 in an inclusive missing mass spectrum . the upper limit of the formation branching ratio was determined to be ( @xmath4 , ( @xmath5 and ( @xmath6/(@xmath7 ) with 95 % confidence level for narrow states with an assumed width of 0 , 20 and 40 mev/@xmath3 , respectively .
, , , , , , , , , , , , , , , , , , , , , , , , , 13.75.jz , 25.80.nv |
grs 1915 + 105 is a galactic black - hole binary with a number of highly interesting and unique properties .
it is a long - lived transient , which outburst began in 1992 @xcite and is still lasting .
it was the first galactic source discovered to show superluminal motion ( @xcite , hereafter mr94 ) .
it is unusually highly variable in x - rays ( e.g. , @xcite ) , and it is one of the intrinsically brightest galactic x - ray binaries , occasionally radiating above the eddington limit @xcite . the minimum jet power in the superluminal ejection in 1994 was claimed to be highly super - eddington @xcite . here , we reconsider the minimum kinetic jet power , the mass flow rate , and the total energy and mass of the major mass ejections of grs 1915 + 105 observed in 1994 ( mr94 ) and in 1997 ( @xcite , hereafter f99 ) .
we use the method developed in @xcite ( hereafter z14 ) .
it minimizes the kinetic jet power associated with both the jet internal energy content and the ion rest mass , correcting some earlier estimates .
we note , in particular , that finding the minimum comoving energy , adding the associated ions and transforming to the observer frame does not give the minimum of the power ( z14 ) .
the jet power becomes extremely high close to the maximum kinematically allowed distance for a given ejection , but it also decreases very fast with the decreasing distance .
recently , the black - hole mass and distance to grs 1915 + 105 were measured by @xcite ( hereafter s13 ) .
we use their results to derive a distance - dependent estimate of the black - hole mass .
we also use their measurement of the donor radius to derive a constraint on the distance .
the jet inclination , @xmath1 , and the velocity , @xmath2 , can be measured using the angular velocities of the approaching and receding condensations , @xmath3 and @xmath4 , respectively , under the assumption that the ejections were symmetric ( mr94 ) .
we have the relationships , @xmath5^{1/2 } } , \label{incl_gamma}\ ] ] where @xmath6 is the source distance and @xmath7 is the jet lorentz factor .
[ note that @xmath8 , which can be used in equations ( [ q_r2][mbh ] ) below for a given measurement of @xmath9 . ] from @xmath10 , we see there is the maximum possible distance , @xmath11 , corresponding to @xmath12 .
the doppler factors , @xmath13 , of the approaching and receding ejecta are , @xmath14^{1/2}\!,\quad \delta_{\rm r}=\delta_{\rm a}{\mu_{\rm r}\over \mu_{\rm a}}= \left[{\mu_{\rm r}\over \mu_{\rm a}}\left(1-{d^2\over c^2}\mu_{\rm a}\mu_{\rm r}\right)\right]^{1/2}\ ! , \label{delta}\ ] ] respectively .
we then calculate the source - frame spectral luminosity from the approaching blob at @xmath15 as @xmath16 where @xmath17 is the measured flux of the approaching blob at @xmath18 .
we note that mr94 and f99 found that the observed ratios between the fluxes of the approaching and receding ejecta are somewhat lower than the ratio , @xmath19 , where @xmath20 is the energy spectral index ( @xmath21 ) , implied by equation ( [ lum ] ) .
then , based on the results of mr94 only , @xcite proposed that the ejections were not intrinsically symmetric , which could explain the lower observed flux ratio .
this seems not confirmed by the results of f99 , who also found a lower flux ratio .
the ratio could be lower if the emission were from continuous jet and counterjet @xcite .
however , the radio images , showing individual blobs , of mr94 and f99 appear not to be compatible with that
. we will use equation ( [ lum ] ) , bearing in mind this issue as a caveat .
since the obtained values of @xmath22 are close to unity for the values of @xmath6 we determine , a possible error due to this assumption is insignificant .
hereafter , we use the standard propagation of errors to calculate the uncertainties on the obtained quantities . allowed by this criterion is @xmath23 kpc .
( c ) the black - hole mass obtained using the mass function of s13 .
the meaning of the curves is the same as in the panel ( a ) . , width=234 ] allowed by this criterion is @xmath23 kpc .
( c ) the black - hole mass obtained using the mass function of s13 .
the meaning of the curves is the same as in the panel ( a ) . , width=234 ] allowed by this criterion is @xmath23 kpc .
( c ) the black - hole mass obtained using the mass function of s13 .
the meaning of the curves is the same as in the panel ( a ) . , width=234 ] we consider two major ejection events from grs 1915 + 105 which began on 1994 march 19 ( mr94 ) and 1997 october 29 ( f99 ) .
their measured angular velocities were @xmath24 mas d@xmath25 ( as updated in @xcite ) , and @xmath26 mas d@xmath25 , and @xmath27 mas d@xmath25 , and @xmath28 mas d@xmath25 , respectively .
in addition , three more events with both @xmath3 and @xmath4 measured are listed by @xcite ( see also @xcite ) .
the implied values of the inclination are consistent with being the same for all five events , though the additional three have much larger uncertainties . in particular , the values implied by the radial velocities for 1994 march and 1997 october are @xmath29 , @xmath30 and @xmath31 , at @xmath32 , 10 and 9 kpc , respectively .
note that the stated uncertainties on @xmath9 , and thus on @xmath33 , are significantly lower for the former event than for the latter .
s13 argued , based on the results of @xcite ( who corrected a numerical error by a factor of 50 in the alignment time scale given by @xcite ) , that the axes of the jet and the binary plane in grs 1915 + 105 are likely aligned , which assumption is also commonly adopted in other estimates of the binary parameters of grs 1915 + 105 .
still , the possibility of a misalignment should be kept in mind , given the lack of full understanding of the alignment process .
assuming alignment , the jet is not precessing , and its inclination angle remains the same for all the observations .
thus , we calculate the weighted average of the inclination and its standard error , though we note that it is strongly dominated by the two major events .
[ binary](a ) shows the values of @xmath1 and its uncertainty for the two major events separately and for the average as a function of the distance .
we then consider the implications for the binary parameters .
we use the recent results of s13 , who obtained the orbital period of @xmath34 d , the velocity semi - amplitude of @xmath35 km s@xmath25 , and the rotational broadening of the donor of @xmath36 km s@xmath25 . assuming corotation and the donor filling the roche lobe
, we have the standard formulae for the mass ratio , @xmath37 ( e.g. , @xcite ) , and the radius of the donor , @xmath38 , @xmath39 respectively , where the roche lobe radius ratio for @xmath40 of @xcite has been used . the equation for @xmath41 yields a cubic - polynomial solution , which gives @xmath42 ( s13 ) . on the other hand
, @xmath38 is a function of @xmath1 , which we show in fig .
[ binary](b ) .
the black hole mass is @xmath43 and @xmath44 .
s13 used the inclination range of f99 .
as we note above , it is significantly less accurate than that of mr94 .
also , s13 gave their mass estimate including the entire distance - related inclination uncertainty of f99 .
[ binary](c ) shows instead the range of @xmath45 as a function of the distance and using the weighted average of the inclination from the available mass - ejection events .
we then consider the distance .
the maximum kinematic distances for the 1994 and 1997 events are @xmath46 kpc , @xmath47 kpc , respectively , and thus the latter gives the current upper limit .
the distance implied by the radial systemic velocity is @xmath48 kpc ( s13 ) . @xcite
considered the distance implied by the theoretical k - star surface brightness compared to the observed magnitude , see their equation ( 1 ) .
this yields the angular diameter of the secondary as @xmath490.0220 mas .
the corresponding physical radius of the secondary is @xmath50 .
the minimum @xmath38 as a function of @xmath6 allowed by the minimum @xmath51 above is plotted by the magenta dashed line in fig .
[ binary](b ) .
we see that the distance compatible with this constraint is @xmath0 kpc .
finally , the new parallax distance from radio vlba measurements is @xmath52 kpc @xcite . combining the above determinations , we find the likely distance range as @xmath5310 kpc .
this is somewhat less than the distance of 11 kpc preferred by @xcite because now both the current measurement of @xmath54 of s13 is lower than the previous one by @xcite and the determination of the systemic - velocity distance by s13 is lower than the previous one by @xcite .
our current range of @xmath6 is still compatible with other observational constraints listed in @xcite .
also , a value of @xmath6 significantly less than 11 kpc is implied by considering the jet kinetic power , as we show below .
the values of @xmath45 for our preferred values of @xmath6 are then somewhat higher than @xmath55 given as the best estimate by s13 , see fig . [ binary](c ) .
the reason is simply our values of @xmath6 are lower than 1012 kpc used by s13 to obtain their range of @xmath45 .
( solid curves ) and @xmath22 ( dashed curves ) on @xmath6 .
the blue and red curves are for the 1994 ( mr94 ; @xcite ) and 1997 ( f99 ) flares .
the dotted curves give the ranges of the uncertainties .
, width=238 ] radio spectrum of grs 1915 + 105 observed on 1994 march 24 @xcite .
the spectrum is fitted by a power law ( @xmath56 ) with @xmath57 , shown by the dashed line .
, width=181 ] mr94 gave rather accurate estimates of @xmath7 and @xmath1 for their estimated @xmath58 kpc , but , as pointed out by @xcite , @xmath7 strongly depends on @xmath6 .
this is illustrated in fig .
[ gamma_delta ] , which shows the dependencies of @xmath7 ( increasing with @xmath6 ) for the best - fit values of the radial velocities of the two data sets .
[ gamma_delta ] also shows the dependencies of the doppler factor @xmath22 , decreasing with @xmath6 .
pair jet , respectively .
the lower ( red ) and upper ( blue ) curves correspond to the cases of the emitting region in the jet being a uniform sphere or a spherical region with gaussian distributions of the densities , respectively .
the spread between those two models reflects systematic uncertainties of the modelling .
the dot - dashed line give the eddington luminosity , @xmath59 , at our distant - dependent estimate of the black hole mass , equation ( [ mbh ] ) .
, height=207 ] pair jet , respectively .
the lower ( red ) and upper ( blue ) curves correspond to the cases of the emitting region in the jet being a uniform sphere or a spherical region with gaussian distributions of the densities , respectively .
the spread between those two models reflects systematic uncertainties of the modelling .
the dot - dashed line give the eddington luminosity , @xmath59 , at our distant - dependent estimate of the black hole mass , equation ( [ mbh ] ) .
, height=207 ] , @xmath60 and @xmath61 on @xmath6 .
the gaussian model is shown for the 1994 ( blue curves ) and 1997 flare ( red curves ) .
( a ) the minimum @xmath62 for both jet and the counterjet . the dot - dashed curve corresponds to @xmath63 for the distance - dependent black - hole mass estimate ; the accretion rate at the accretion luminosity @xmath64 is @xmath65 divided by the efficiency , @xmath66 .
( b ) the total ion mass in both the approaching and receding ejection at the minimum jet power .
the corresponding values for a pure e@xmath67-pair jet are @xmath68 times lower .
( c ) the total energy content in both particles and field corresponding to the minimum jet power .
the solid and dashed curves correspond to electron - ion and pair plasmas , respectively.,width=200 ] , @xmath60 and @xmath61 on @xmath6 .
the gaussian model is shown for the 1994 ( blue curves ) and 1997 flare ( red curves ) .
( a ) the minimum @xmath62 for both jet and the counterjet .
the dot - dashed curve corresponds to @xmath63 for the distance - dependent black - hole mass estimate ; the accretion rate at the accretion luminosity @xmath64 is @xmath65 divided by the efficiency , @xmath66 .
( b ) the total ion mass in both the approaching and receding ejection at the minimum jet power .
the corresponding values for a pure e@xmath67-pair jet are @xmath68 times lower .
( c ) the total energy content in both particles and field corresponding to the minimum jet power .
the solid and dashed curves correspond to electron - ion and pair plasmas , respectively.,width=200 ] , @xmath60 and @xmath61 on @xmath6 .
the gaussian model is shown for the 1994 ( blue curves ) and 1997 flare ( red curves ) .
( a ) the minimum @xmath62 for both jet and the counterjet . the dot - dashed curve corresponds to @xmath63 for the distance - dependent black - hole mass estimate ; the accretion rate at the accretion luminosity @xmath64 is @xmath65 divided by the efficiency , @xmath66 .
( b ) the total ion mass in both the approaching and receding ejection at the minimum jet power .
the corresponding values for a pure e@xmath67-pair jet are @xmath68 times lower .
( c ) the total energy content in both particles and field corresponding to the minimum jet power .
the solid and dashed curves correspond to electron - ion and pair plasmas , respectively.,width=200 ] the measured peak flux of the 1994 flare was 660 mjy at @xmath69 ghz on 1994 march 24 @xcite .
the observed 1.415 ghz spectrum is well fitted by a power law with the spectral index of @xmath57 [ @xmath70 , see fig .
[ sp1915 ] , corresponding to the electron power - law index of @xmath71 .
@xcite give the total fluxes , but the flux of the approaching component dominates ( mr94 ) , and thus we use the total values .
the measured peak flux of the 1997 flare was 550 mjy at @xmath72 ghz on 1997 october 29 ( f99 ) .
the data for that day are at 2.3 ghz , 8.3 ghz and 15 ghz .
the 2.38.3 ghz spectrum has @xmath73 ( f99 ) , though this slope under - predicts ( at @xmath74120 mjy ) the average 15 ghz peak flux on that day , see fig . 4 of f99 , which would imply a lower average @xmath20 .
that flux also shows strong qpos , implying a sequence of new ejections .
still , we adopt @xmath75 for consistency with f99 , corresponding to @xmath76 .
we assume the spectrum is emitted from @xmath77 ghz to @xmath78 ghz ( observer frame ) , consistent with the data .
this photon energy range spans two decades , corresponding to one decade of the electron lorentz factor .
this is a conservative assumption , which yields good estimates of the _ minimum _ jet power .
we assume that the only electrons , or e@xmath67 , are those in in a relativistic power - law distribution yielding the synchrotron power law from @xmath79 to @xmath80 .
we consider two cases of the jet composition , one with cold ions and the h abundance of @xmath81 , and one with pure pair plasma .
we adopt the angular source size ( perpendicular ) at the measured peak of the 1994 flare of 20 mas ( mr94 ) , which corresponds to the physical radius of @xmath82 cm . for the 1997 flare ,
we adopt the size estimate of f99 of @xmath83 cm independent of @xmath6 based on the jet rise time . in the calculations ,
we consider either a uniform spherical source , or one with smooth gaussian density distributions , see z14 .
we assume the blob has a spherical symmetry in its frame and thus we do not apply the relativistic transformation of the length along the jet .
we use then the method minimizing the jet power for a given optically - thin synchrotron frequency range , see z14 .
we use equations ( [ incl_gamma][lum ] ) above to determine the jet - frame @xmath7 and @xmath84 as functions of the distance .
the behaviour of @xmath22 and @xmath7 shown in fig .
[ gamma_delta ] leads to a fast decline of @xmath85 with decreasing @xmath6 . at @xmath58 kpc
, we obtain the the minimum jet power of @xmath86 erg s@xmath25 of the 1994 flare for the uniform model , which is a factor of @xmath87 lower than the corresponding estimate by @xcite ( multiplying their value by two to include the counterjet ) .
the main cause of the difference is our more limited range of the electron lorentz factors , which is @xmath88 $ ] , compared to @xmath89 of them .
also , our definition of the jet power is different , see z14 , and the method different , as they assumed a fixed frequency - integrated synchrotron luminosity . taking into account those differences , our results are fully consistent with those of @xcite .
they claimed the 1994 flare had to be strongly super - eddington .
however , as we discuss above , the current limit on @xmath6 is @xmath47 kpc ( f99 ) , and @xmath85 decreases rather fast with decreasing @xmath6 , which we show below . thus , their results are no longer applicable to grs 1915 + 105 .
we then compare our results for the 1997 flare with those of f99 . at @xmath32 kpc ,
we obtain @xmath90 erg s@xmath25 with the uniform model , which is a factor of @xmath91 higher than the corresponding estimate of f99 ( multiplying the power given in their table 3 by 2 , see their section 4.1.3 ) .
one reason for the discrepancy is their method being incorrect , as discussed in z14 .
furthermore , f99 divided the energy content in electrons by the flare rise time , which gives the power averaged over that time rather than the instantaneous power at the moment of the measurement of the peak synchrotron flux .
[ power1915](a ) and ( b ) show @xmath92 for the 1994 and 1997 flares , respectively , for the uniform and gaussian models with the ionic and e@xmath67 compositions ( z14 ) , comparing them with the eddington luminosity at our estimated @xmath6-dependent black - hole mass estimate .
we see that @xmath85 for the pair model is a factor of a few less than for the ionic one
. however , as argued , e.g. , by @xcite and @xcite ( though , as a caveat , see @xcite ) , jets in x - ray binaries likely contain hadrons rather than pure pairs . the minimum jet+counterjet power and the corresponding magnetic field strength using the gaussian model ( used hereafter ) for the 1994 flare at 10 kpc are @xmath93 erg s@xmath25 , @xmath94 g for the cosmic abundances and @xmath95 erg s@xmath25 , @xmath96 g for e@xmath67 pairs . for the 1997 flare at 10 kpc
, we obtain @xmath97 erg s@xmath25 , @xmath98 g for the cosmic abundances and @xmath99 erg s@xmath25 , @xmath100 g for e@xmath67 pairs . at the cosmic abundances ,
we find the ratio of the jet power in electrons to that in ions of @xmath101 , 0.08 for the 1994 and 1997 flare , respectively .
these estimates are thus insensitive to the assumed source radius , which is the case for the ionic power dominating ( z14 ) .
indeed , the values of the critical radius below which the solution becomes independent of @xmath102 ( z14 ) are @xmath103 cm , @xmath104 cm for the 1994 and 1997 flares , respectively , which are @xmath105 our values of @xmath102 .
we note that a possible clumpiness of the jet material , as quantified by the clump volume filling factor , @xmath106 , can reduce the minimum jet power by a factor of @xmath107 in the case of the dominant ionic power , and @xmath108 for the case of pair plasma ( z14 ) .
if , e.g. , @xmath109 , the minimum power can be reduced by a factor of several .
[ power1915 ] shows that the kinetic jet power critically depends on the actual distance to grs 1915 + 105 .
we can compare it to the bolometric luminosity of grs 1915 + 105 , dominated by accretion .
the results of @xcite and @xcite show that the jet kinetic power in agns is strongly correlated with both the accretion rate and the bolometric luminosity , @xmath110 .
in particular , @xcite find the correlation of @xmath111 , where @xmath59 is the eddington luminosity .
thus , we expect @xmath112 at @xmath113 .
furthermore , results of @xcite and @xcite show that the same underlying relationship between jets and accretion flow is present in both agns and black - hole binaries .
indeed , the jet power of cyg x-1 estimated by @xcite satisfies the correlation of @xcite .
furthermore , the jet power of cyg x-3 in its bright state has been estimated based on a model of its @xmath114-ray emission and found @xmath115 , consistent with the above correlation @xcite .
the _ rxte_/asm monitoring shows the x - ray state during the 1997 event ( f99 ) was among the brightest ones observed from this source , which implies that its maximum bolometric luminosity was then moderately super - eddington , @xmath116 @xcite .
thus , based on the above considerations , we may expect ( but not prove ) the jet kinetic power to be approximately sub - eddington .
[ power1915 ] shows that the 1994 and 1997 jets are sub - eddington provided @xmath117 kpc and 8.8 kpc , respectively , using the ionic model with the gaussian density distribution . in the case of pair plasma ,
the respective jets are sub - eddington for @xmath118 kpc and 9.9 kpc .
the minimum ion mass flow rate , @xmath119 , for the gaussian model are shown in fig .
[ mass](a ) .
we see that if @xmath120 kpc , @xmath121 .
the corresponding masses for the ionic model ( in the jet frame ) are shown in fig .
[ mass](b ) . in the case of the pair model ,
both the minimum @xmath62 and @xmath60 are lower by @xmath122 .
[ mass](c ) shows the total energy content corresponding to the minimum jet power for both ionic and pair cases .
the ejecta at the time of the 1994 measurements were optically thin to synchrotron self - absorption ; the radial optical depth , @xmath123 , at @xmath77 ghz is @xmath124 , and @xmath125 at @xmath126 ghz .
( similar values are found for the 1997 flare . )
the peak 8.6-ghz flux , corresponding to @xmath127 , was thus achieved earlier than on 1994 march 24 , and the minimum power at that time was higher . on the other hand ,
the blob expanded after march 24 , which caused the electrons to cool adiabatically , and also reduced the magnetic field strength .
the synchrotron luminosity should then decrease , and the fluxes in table 1 of mr94 are consistent with this .
when we apply the minimum jet power method to those later stages , we find the minimum total power decreasing . on the other hand
, there was no evidence for deceleration in those data ( mr94 ; see also @xcite ) , which implies that the power in ions ( dominating the power in particles in our case ) kept constant .
thus , the further evolution of the source had to be characterized by a decrease of the jet equipartition parameter of @xmath128 ( @xmath129 for the minimum power at the initial observation ) .
this shows that the actual jet power is likely to be higher than the minimum one .
we have calculated the minimum jet power for the major mass ejection events of grs 1915 + 105 in 1994 and 1997 .
we have taken into account the power component associated with the ion bulk motion , which we have found to dominate , and which was neglected in an earlier estimate .
we have found the jet power becomes very high close to the maximum allowed kinematically for a given ejection event . on the other hand ,
it becomes much lower and in agreement with the average relationship between the jet power and the bolometric luminosity for accreting black holes at a distance @xmath0 kpc .
we also have calculated the mass flow rate in the jets and their total mass and energy content .
we have also updated the distance estimate based on a theoretical relationship of the donor radius and its magnitude for k stars @xcite using the @xmath54 measurement of s13 .
we obtained @xmath120 kpc , in full agreement with the estimate based on the jet power .
in addition , we have presented a relatively accurate estimate of the black - hole mass as a function of the distance using the weighted average of the inclination from the observed mass ejections together with the optical measurements of s13 .
i am grateful to p. pjanka for detailed checking of the results of this work , f. mirabel , j. mikoajewska and m. sikora for valuable discussions , the referee for valuable comments and suggestions , and l. f. rodrguez for a clarification regarding the size estimates in mr94 .
this research has been supported in part by the polish ncn grants 2012/04/m / st9/00780 and 2013/10/m / st9/00729 . | we apply a recently developed technique of calculating the minimum jet kinetic power to the major mass ejections of the black - hole binary grs 1915 + 105 observed in radio wavelengths in 1994 and 1997 .
we derive for them the distance - dependent minimum power , and the corresponding mass flow rate and the total energy and mass content .
we find that a fast increase of the jet power with the increasing distance combined with the jet power estimates based on the bolometric luminosity imply the source distance is @xmath0 kpc .
if the jet in grs 1915 contains ions , their bulk motion dominates the jet power , which was either neglected or not properly taken into account earlier .
we also reconsider the parameters of the binary , and derive the current best estimates of the distance - dependent black - hole mass and the inclination based on existing measurements combined with the kinematic constraints from the mass ejections .
we also find the measurement of the donor radius of steeghs et al
. implies the distance to the system of @xmath0 kpc , in agreement with the estimate from the jet power .
[ firstpage ] acceleration of particles ism : jets and outflows radiation mechanisms : non - thermal radio continuum : stars stars : individual : grs 1915 + 105 . |
the dispersion relation for the the single band tight - binding simple cubic lattice with nearest and next - nearest hopping is given by @xmath70 , \end{split}\ ] ] where @xmath71 @xmath72 labels the three cartesian directions of the nearest neighbor bonds of the cubic lattice .
the bandwith @xmath73 \text { or } [ 0,\pi,\pi]}-\varepsilon_{[0,0,0]}$ ] is given by @xmath74 as mentioned in the main text , we define the energy unit by fixing @xmath75 .
this fixes the value of @xmath76 for the different @xmath56 .
our database contains data for @xmath21 $ ] .
we also tested our predictive power by using a lattice with @xmath77 .
we show in fig .
[ fig : n0 ] . the density of states @xmath78 of these five lattices to show the effect of next neighbor hopping .
( black solid curve ) , @xmath79 ( blue dash - dot curve ) , @xmath65 ( black dashed curve ) , @xmath80 ( red dot curve ) and @xmath81 ( magenta solid curve and filled black circle ) . ]
in exact diagonalization , the bath is replaced by a finite number of sites ( @xmath82 ) , each characterized by an onsite energy @xmath83 and hybridization with the impurity @xmath84 .
therefore , the hybridization function is given by @xmath85 which represents in real frequency the approximation of replacing a continuous function by a sum of poles and strengths .
these poles and strengths @xmath86 have to be found by fitting eq . to the target function @xmath87 obtained via the lattice green s function @xmath88 .
this is achieved by defining a distance function and minimizing it . @xmath89 where the function @xmath90 is chosen to give more weight to some frequencies if wanted .
we use @xmath91 to have a better fit of the low frequencies .
@xmath92 is the maximum number of frequency used to define @xmath93 .
what is important in its choice is that @xmath94 .
finally , @xmath95 is the inverse non - interacting green s function of the hamiltonian with a finite number of bath sites and is written as @xmath96 therefore , we need to find the set of parameters @xmath86 that minimizes @xmath93 .
this is a problem of unconstrained optimization in several variables . in dmft
, this is done as many time as necessary to converge the solution .
we can define a bare hybridization function , for @xmath97 , a function containing the information about the crystal structure and chemistry of the problem . for simplicity , we choose to fix @xmath22 . for the non - interacting case , @xmath98 and the lattice green s function
is given by the band dispersion only @xmath99 we can therefore define the bare hybridization function @xmath100 this can be fitted to @xmath101 using eq . and gives a representation of the crystal of size equal to @xmath102 where @xmath82 is the number of bath sites .
the function in this case is thus represented in this basis as @xmath103 .
it could perhaps be seen as a general and compact way to describe the lattice for ml , irrespective of how the database is constructed ( using ed or not ) .
after all , what is needed for creating a @xmath25 is a unique way to describe a problem and @xmath104 $ ] provides one .
this is true for both for model hamiltonians and real materials as well as for solutions obtained from ed , quantum monte carlo etc .
the database contains 1783 converged dmft standard metal solutions , 218 converged critical metal solutions and 494 mott insulating solutions , obtained from exact diagonalization solutions of the dmft equations . in the mott insulating state , physically , any choice of chemical potential in the gap should lead to the same solution with a shifted zero point of the frequency axis .
however , the bath discretization of the ed method means that for otherwise identical parameters different values of @xmath9 lead to different insulating solutions .
for this reason we need to include many ( here 494 ) different mott insulators in the database .
we train insulators using all these solutions .
in kernel ridge regression ( krr ) , there are two free parameters or hyperparameters . in our case , for @xmath43 , we use the weighted exponential kernel @xmath105 where @xmath106 is the manhattan distance between the two parameter sets in descriptors space and @xmath107 gives the radius of effect that a particular point of the data set @xmath44 will have in the prediction process .
therefore , the two free parameters are @xmath107 ( entering the kernel function ) and @xmath49 , the regularization parameter used in the cost function that is minimized in krr .
to fix them , we use cross - validation . cross validation proceeds by first creating a large number of pairs @xmath108 $ ] .
then , for each pair we randomly split the database so that the testing set contains about ten examples . for each of these ( @xmath109 ) tests , we predict only the first five legendre polynomials coefficients @xmath110 and calculate the total mean absolute error .
note that the actual metric to calculate the error is not really important , we are just looking for the pair that minimizes an error .
we look for a pair @xmath108 $ ] that gives as small as possible error , then this @xmath108 $ ] is used to learn all the necessary legendre polynomials coefficients and not only the first five . as an example
, we show the result for the training of the metallic impurity green s function as a contour plot in fig .
[ fig : cross_val ] when the descriptor uses ed representation for the non - interacting hybridization function and fig .
[ fig : cross_val_legendre ] when the legendre representation for the non - interacting hybridization function is chosen .
the white color indicates regions where no data are available .
we see that if the width @xmath107 of the kernel is too small ( ml will use not enough solutions in descriptors spaces ) or too large ( ml will use too many far away solutions in descriptors spaces ) , the error is the largest . for the training / testing using solved problems from our database ,
the best possible @xmath107 would be around but smaller than @xmath111 for fig .
[ fig : cross_val ] and around @xmath112 for fig .
[ fig : cross_val_legendre ] .
the results also show that the value of @xmath49 is not extremely important .
practically we chose the pair @xmath108 $ ] that produced the smallest error among the ones tested while we increase the @xmath107 to prevent overfitting in the case of out of database prediction(see section iii.b below ) .
$ ] in cross - validation for the impurity green s function when the descriptor is chosen to be @xmath113 $ ] . ]
$ ] in cross - validation for the impurity green s function when the descriptor is chosen to be @xmath114 $ ] . ] here we discuss in detail the prediction of the dmft solution for the case @xmath65 , @xmath68 and @xmath115 which is a case completely outside our database .
let us first use the width of the kernel that gave the lowest error during cross - validation using @xmath113 $ ] as the descriptor which is around @xmath111 as shown is section iii.a .
the result for the imaginary part of the reconstructed impurity green s function is shown in fig .
[ fig : ml_g_predic_outside_supp]-(a ) while fig . [ fig : ml_g_predic_outside_supp]-(b ) shows the ml predicted coefficients @xmath59 .
we see a systematic discrepancy in the prediction of the reconstructed @xmath116 , but it is interesting to realize that this is due to a systematic discrepancy of @xmath117 only on the even @xmath59 coefficients at low order .
this is true for every possible example at @xmath65 ( we calculated a total of 295 different examples with this @xmath56 ) .
the discrepancy is not due to a failing of ml but solely on a too tight choice of kernel width ( over - fitting ) which for example might not use enough of the solutions for @xmath79 and/or @xmath80 .
we trained our machine again , but with a larger width of @xmath118 which , according to fig . [
fig : cross_val ] is not the optimal value but still give a pretty acceptable error .
the result is fig . 5 of the main text and is reproduced here as fig .
[ fig : ml_g_predic_outside_supp]-(c ) showing that we must be careful with over - fitting and if this is properly taken into account , we can predict new dmft solutions .
the specific choice of @xmath107 is not well defined since out of database predictions means no comparison with exact results in principle .
a sensible approach is to start from the @xmath107 given by cross - validation of section iii.a and search for a larger value .
to do so , contour plots of figs .
[ fig : cross_val ] and [ fig : cross_val_legendre ] are essential .
we need a value to the right of the cross - validated one since we are looking for a larger radius of effect for database points .
how large will be given by the condition of still keeping a low enough error for intra - database prediction .
hence by looking at figs .
[ fig : cross_val ] and [ fig : cross_val_legendre ] we see that values of @xmath107 somewhere between @xmath119 and @xmath120 respect these conditions .
we chose @xmath120 , but we could have tested many other values .
$ ] as compared to the exact results ( red dots ) for @xmath68 , @xmath65 , @xmath66 ( a ) prediction of @xmath116 with a too small ( over - fitted ) width of kernel ( b ) prediction of @xmath59 with a too small ( over - fitted ) width of kernel ( c ) prediction of @xmath116 with a right width of kernel.,title="fig : " ] $ ] as compared to the exact results ( red dots ) for @xmath68 , @xmath65 , @xmath66 ( a ) prediction of @xmath116 with a too small ( over - fitted ) width of kernel ( b ) prediction of @xmath59 with a too small ( over - fitted ) width of kernel ( c ) prediction of @xmath116 with a right width of kernel.,title="fig : " ] $ ] as compared to the exact results ( red dots ) for @xmath68 , @xmath65 , @xmath66 ( a ) prediction of @xmath116 with a too small ( over - fitted ) width of kernel ( b ) prediction of @xmath59 with a too small ( over - fitted ) width of kernel ( c ) prediction of @xmath116 with a right width of kernel.,title="fig : " ]
we expand the output solution in terms of legendre polynomials . considering the result as a function of imaginary time @xmath121 ( @xmath122 is the inverse temperature ) we have for correlation functions @xcite @xmath123 once the coefficients have been obtained , not only the function in imaginary time
can be reconstructed , but the fourier transform to @xmath32 can also be done analytically@xcite to give @xmath137 where @xmath138 and @xmath139 are the spherical bessel functions .
in the main text we claimed that the results we presented hold irrespective if the chosen representation of the bare hybridization function is from an ed - like fitting @xmath101 or as in term of legendre polynomials @xmath141 expansion .
we show here two examples to support our statement . in fig .
[ fig : z_ml_supp]-(a ) and ( b ) , we reproduce fig . 2-(a ) and fig
. 3-(a ) of the main text .
in addition , we add the predictions ( blue x ) as obtained from the descriptor of the form @xmath114 $ ] .
the predictions for the quasi - particle weight are very close to the case when the machine is trained with @xmath113 $ ] .
other optimizations could be done to get even closer to the exact solutions and thus for intended purpose it can be considered as being equal .
the case of the lattice density is similar , the two representations give again practically the same answer . in conclusion ,
the two choices of representation for the bare hybridization function in the descriptor give essentially the same results and therefore its choice is a matter of which one can be obtained and how easy to calculate it is for a particular situation .
we present exactly how we calculated the median relative difference for the quasi - particle weight .
the same approach is taken for the density .
its meaning is the following : 1 ) a size @xmath142 of learning set is chosen .
2 ) from the database , we select randomly one example that will be the testing system .
3 ) from the remaining examples in the database , we randomly take @xmath142 solutions and train a machine ( calculate the @xmath143 matrix of krr ) .
4 ) we predict the self - energy for the testing example of 2 ) using our trained machine . from it
, we obtain @xmath52 and can calculate the relative difference with the exact answer @xmath144 .
we then repeat steps 3 ) and 4 ) twenty times to obtain the prediction of the same example from many different trained machines and thus assure we have large distributions of learning sets .
finally , we go back to step 2 ) and start again with a new example and do it fifty time in total to have a large distribution of testing examples
. we therefore have one thousand relative difference .
if a randomly chosen learning set only contains examples very far from the testing example we are predicting , the relative difference will be large , but in that case , not because ml is bad , but rather because the learning set was badly picked . therefore to have a good idea of how well a learning set of size @xmath142 does
, we argue that a good choice is to take the median value of the one thousand predictions . for the largest possible size of learning set 1782 , we instead calculated the relative difference for every example in the database choosing the reminder 1782 examples as learning set .
then we randomly choose fifty of these relative difference and calculated the median .
l. boehnke , h. hafermann , m. ferrero , f. lechermann , and o. parcollet , phys . rev . * b 84 * , 075145 ( 2011 ) .
arsenault , a. lopez - bezanilla , o. anatole von lilienfeld and a. j. millis , phys .
b * 90 * , 155136 ( 2014 ) . n. hale and a. townsend , siam j. sci .
comput . * 36 * , a148 ( 2014 ) .
t. a. driscoll , n. hale , and l. n. trefethen ( eds . ) , _ chebfun guide _ ( pafnuty publications , oxford , 2014 ) . | machine learning methods for solving the equations of dynamical mean - field theory are developed .
the method is demonstrated on the three dimensional hubbard model .
the key technical issues are defining a mapping of an input function to an output function , and distinguishing metallic from insulating solutions .
both metallic and mott insulator solutions can be predicted .
the validity of the machine learning scheme is assessed by comparing predictions of full correlation functions , of quasi - particle weight and particle density to values directly computed .
the results indicate that with modest further development , machine learning approach may be an attractive computational efficient option for real materials predictions for strongly correlated systems .
the quantum many - body problem of predicting properties of systems containing electrons or other fermionic entities has challenged physicists and chemists for decades .
this is so because the minus sign associated with fermionic exchange creates a host of difficulties including long - ranged entanglement and a monte - carlo sign problem .
the net effect is to place the generic fermion many - body problem in the class of problems whose full solution is exponentially hard .
although new developments such as matrix product and tensor network methods may provide solutions to ground - state properties with only power - law cost , search for efficient approximate methods to handle a wide range of phenomena at a wide range of temperatures remains a key goal of condensed matter physics and quantum chemistry . in this work ,
we investigate the use of machine learning ( ml ) @xcite to leverage existing results and provide an efficient approximate solution to a generic class of problems in quantum many - body physics .
ml is in essence a way to use a database of known solutions to infer information about a new problem . in the condensed matter physics context
it has been used as an intermediate step in molecular dynamics calculations@xcite , to predict density functionals ( so far only in the 1d context)@xcite , to obtain transmission coefficients for electron transport @xcite and , very recently , to predict the fermi level density of states of weakly correlated solids @xcite and find formation energies of materials @xcite .
these applications relate to classical physics and to single - particle quantum mechanics . in the quantum chemistry context
, ml has been successfully applied to predict energies and other scalar properties of molecules @xcite .
in addition , non - ml ideas from data science have been recently proposed as ways to help the solution of the non - equilibrium many - body problem @xcite .
we propose to use machine learning methods to solve true quantum many - body problems .
a technical issue arises : in many applications of machine learning , including most of the ones referred to above , the goal is to infer a scalar property ( e.g. an energy ) of a model specified by a modest number of scalar parameters . however , the generic quantum many - body problem is the solution of a functional equation relating an input function ( for example a bare electron green s function ) to an output function ( e.g. an electron self - energy ) .
here we build on previous work @xcite which involved learning a function specified by a modest number of input parameters , to develop a formalism capable of solving the more general problem of mapping a function to a function . in independent contemporaneous work ,
questions related to learning functions have been studied in the context of density functional theory where one seeks to learn the relation between a position - dependent charge density and an exchange - correlation potential@xcite .
the context for our work is dynamical mean - field theory ( dmft)@xcite , a widely used approximate method for determining the properties of materials with strong electronic correlations .
dmft approximates the solution to an interacting fermion system in terms of the solution of an auxiliary quantum impurity problem . the impurity model , although simpler than the full problem , is still a quantum many - body problem .
the impurity model is specified by a hybridization function @xmath0 ; the many - body physics by a local green function @xmath1 or self - energy @xmath2 . the hybridization function itself is obtained from a self - consistency condition which involves @xmath2 and an initial band structure which encodes the chemistry and crystal structure of the material in question and may be parametrized as a bare or initial hybridization function @xmath3 . in standard applications ,
the self - consistency condition is solved by iteration .
one may imagine a ml process to solve the impurity model ( relating @xmath4 and @xmath5 ) or a ml process to solve the entire dmft self - consistency loop ( relating @xmath6 and @xmath7 ) . in this paper
, we only present results for the full solution of dmft .
use of ml as impurity solver may also be valuable as an intermediate step , enabling the rapid construction of a database of solved problems in the real - materials context .
our formalism is general enough to apply to this possibility .
we test our methods using the hubbard model defined on a three dimensional cubic lattice with first and second - neighbor hoppings , with hamiltonian @xmath8 . here
@xmath9 is the chemical potential and @xmath10 $ ] .
the bare hybridization function is @xmath11 .
we define energy units such that the full non - interacting bandwidth @xmath12 where @xmath13 if @xmath14 and @xmath15 if @xmath16 . varying the ratio @xmath17 changes the structure in the density of states , in particular shifting the location of the density of states peaks relative to the band center ( see section i of the supplemental material for examples ) .
we seek a machine that enables us to map a @xmath6 to an output local green function or self - energy .
dmft admits two classes of solutions : metallic ones with a non - vanishing density of states at the fermi level and a smooth self - energy , and mott insulating solutions with a gap at the fermi level due to coulomb repulsion and ( in many cases ) a self - energy with a pole near the chemical potential .
we have found it advantageous to introduce a binary classification step that identifies a given solution as metallic or insulating and to use two different machines to determine the properties of the two kinds of solutions . for classification , we use the entire database minus one as training and the one remaining as the testing problem .
we then repeat for all members of the database .
we tested three different ml for classification : simple support vectors machine svm@xcite with @xmath18 accuracy , neural networks@xcite with @xmath19 accuracy and decision forests@xcite with @xmath20 accuracy .
the only misplaced problems are critical metals extremely close to the transition .
we only kept the decision forest as it outperformed the two others .
once the state of a new problem has been decided , the kernel ridge regression ( krr ) method @xcite ( more details follow ) is employed to determine the solution using the sub - databases containing only metal or mott insulating solutions .
the full ml process for dmft is shown in fig .
[ fig : ml_dmft_scheme ] while some details about the parameters are explained later in the text .
the first step in implementing machine learning is to generate a database of initial conditions , in other words a set of bare hybridization functions that span a range of physically reasonable possibilities .
we consider the set of hybridization functions defined by @xmath21 $ ] ( the case of positive @xmath17 could be accounted for by considering electron doping ) and @xmath22 .
sections i and ii.a of the supplemental material give more details .
we then obtain the database of solved problems by using the exact diagonalization ( ed ) method @xcite to solve the single - site dynamical mean field approximation for interaction strengths in the range @xmath23 and densities in the range of @xmath24 .
particularities of the ed database are discussed in section ii.b of the supplemental material . can also be extracted . ]
the second step in implementing machine learning is the construction of a representation of the information to be learned and of the _ descriptor _ @xmath25 , a unique identifier of a problem .
our input and output data are functions .
functions may be specified as a vector of coefficients in a space of basis functions @xmath26 ( e.g. @xmath27 ) .
our previous work@xcite , following work by boehnke _
et al_.@xcite found that legendre polynomials were a very efficient choice of basis , so we adopt this representation here .
the legendre representation is most naturally formulated in imaginary time @xmath28 with @xmath29 the inverse temperature and hence a correlation function is @xmath30@xcite , where @xmath31 are the legendre polynomials .
the fourier transform to @xmath32 can be done analytically@xcite .
the representation is general , we fit either the local green s function or the self - energy as shown in fig .
[ fig : ml_dmft_scheme ] ( or even the hybridization function ) .
see section iv of the supplementary material for details .
the descriptor consists of the input function ( hybridization function ) plus a few scalar parameters ; we denote the expansion coefficients of the function as @xmath33 and the scalar parameters @xmath34 ( interaction strength ) and @xmath9 ( chemical potential ) such that @xmath35 $ ] ( see fig . [
fig : ml_dmft_scheme ] ) . note that both the full dmft problem and the impurity solving part are the same problem as far as ml is concerned , the only difference being what database one chooses .
the exact diagonalization method used here provides a representation of the input hybridization function in terms of bath level energies @xmath36 and hybridization parameters @xmath37 ( @xmath38 labels entries in the database and @xmath39 labels the different bath energies and hybridization parameters for a given entry in the database ) so in practice we use these for the @xmath40 .
we have also implemented machine learning using the representation of the input function in terms of legendre polynomials , with essentially identical results ( see section v of the supplementary material ) .
section ii.a of the supplemental material shows how the bare ed parameters are obtained from a known band structure .
machine learning then estimates the solution @xmath41 of a new problem in terms of an interpolation between known solutions .
we use krr , an expansion in the abstract multidimensional space of descriptors ( each point @xmath25 of this space represents a unique problem and the distance between two points is the distance metric ) , obtaining @xmath42 where @xmath39 labels points in the dataset , @xmath38 labels entries in the output vector and the kernel @xmath43 is a function whose main characteristic is to weight most heavily the contributions of @xmath39 for which @xmath44 is close to @xmath25 . as in @xcite , we use the weighted exponential kernel , and use the manhattan distance between @xmath44 and @xmath25 ( both are defined in section iii.a of the supplemental material ) .
the expansion coefficients @xmath45 are @xmath46@xcite , where @xmath47 is a matrix containing all the @xmath45 , @xmath48 is the kernel matrix and @xmath49 is a regularization parameter . @xmath49 and
the free parameter of the kernel are chosen using standard cross - validation , see section iii.a of the supplemental material .
in particular , as also found in @xcite , we found that the actual value chosen for @xmath49 is not really important .
this formalism is very general and could be applied to the learning of other types of functions . as first two tests of our predictive power we present scalar properties , the quasi - particle weight @xmath50 and the lattice density of electrons @xmath51 as predicted from reconstructed correlation functions with ml obtained legendre polynomial coefficients .
we estimate @xmath52 from a quadratic fit to the values of the reconstructed self - energies at the three lowest matsubara frequencies(see for example @xcite for why @xmath52 can be estimated on imaginary axis ) . as easily seen from fig .
11 of @xcite , values of @xmath53 ( or @xmath54 ) for the first few @xmath55 are given solely by the first few coefficients of the expansion in legendre polynomials .
hence , the prediction of @xmath52 shows how well the first few coefficients are learned .
the results are shown in fig . [
fig : z_ml ] for typical values of interaction from weak to correlated metals and for different @xmath56 .
the predictions for these specific @xmath25 from the database are obtained by using all other examples as the training set .
the predictions are in general very good with a slightly worst predicting power for larger correlation close to half - filling where @xmath52 is close to zero . to study the error in a more rigorous way , we present in inset of fig . [ fig : z_ml ] what we call the median relative difference ( mrd ) for @xmath52 as a function of the learning set size ( see supplemental material section vi for details ) .
this shows the median value of predictions for fifty different random examples each re - calculated with twenty different random learning set . as shown in the inset of fig .
[ fig : z_ml ] , the mrd of @xmath52 is slightly below @xmath57 for a small size of 500 and gets to around @xmath58 for the largest learning set . a predictive power of smaller than @xmath57 error even for a small database is very interesting especially since choosing completely random datasets is the worst case scenario . in the case of the lattice density of electrons as a function of chemical potential , the ml path from fig .
[ fig : ml_dmft_scheme ] is the one where we learn the @xmath59 s of the expansion of the local lattice green s function then reconstruct it in imaginary time . since @xmath60 for all @xmath39 , the density is @xmath61 .
therefore , contrary to the case of @xmath52 , the prediction of the density uses all predicted coefficients of the expansion .
we show results in fig .
[ fig : density_ml ] for typical parameters , yet different than those presented for @xmath52 . to improve readability , we shifted curve ( 3 ) by 0.2 .
once again the results are in good agreement with slightly worst predictions for @xmath62 .
this region tends also to be more problematic for @xmath52 .
this is not fundamental but rather because our dmft database is not as well constructed there . in the inset of fig .
[ fig : density_ml ] we show the mrd calculated the same way as for @xmath52 .
ml does even better in this case where the mrd is at worst @xmath63 .
we now show in fig .
[ fig : g_imp_wn]-(a ) and -(b ) the prediction of the imaginary part of the impurity green s function in matsubara frequency for two typical set of parameters .
as can be seen , ml does a very good job at predicting both the metal and the insulator , although the number of insulating solutions in the database is not very large . in the inset of fig .
[ fig : g_imp_wn]-(a ) , we present the average relative difference ( ard ) for the metallic case as a function of the size of the learning set . the ard was defined in @xcite as a way to measure on average the accuracy of the prediction of a full function using only one number .
the values are obtained by averaging predictions for many random test sets .
the global average prediction of a full function in the metallic case has an error in the worst case of @xmath64 which shows the predictive power of our ml scheme . , @xmath65 , @xmath66 .
] we finally analyse the question of prediction of a totally new problem and the importance of training . because our database is very homogeneous , for out of database predictions
, we chose to use a width ( arbitrarily set to be @xmath67 ) larger than the actual lowest possible error in the cross - validation training used for previous results to avoid overfitting . in the supplemental material ( section iii.b . ) , we show how overfitting influences the predictive power of our ml approach .
we show in fig .
[ fig : ml_g_predic_outside ] that indeed we can very well predict dmft solution for new problems sharing no equal values of @xmath34 , @xmath56 and @xmath9 in the database by choosing as an example @xmath65 , @xmath68 and @xmath9 such that @xmath69 .
we trained a machine with the full database of 1783 metallic solutions , the metal being well predicted by decision tree classification . by this process ( larger width ) , we loose some of the predictive precision we had for intra - database testing sets , but this is of no consequences as once we prove that we can accurately train a machine , what really matters is the predictive power for out of database unsolved problems . in this paper , we have investigated how machine learning can be used in many - body physics as a method to predict correlation functions .
we applied the scheme to dmft and showed that we can accurately predict its solutions .
our approach maps input functions to output functions and can be applied without any changes as an impurity solver for dmft rather than to learn the fully converged solution or for any other cluster embedding theory with a self - consistency relation .
impurity solving might be the best way to use ml for real materials predictions since accuracy depends largely on having a large database .
it is also general enough to be applied to other problems where learning a function is important .
the learning of a function using kernel ridge regression might be improved by adding a simple form of constraints in the minimization problem at small computational cost@xcite . in real materials
applications a more complicated system has to be taken into account ; hunds coupling , multi - band etc .
however , our presented approach for ml is general enough to be adapted .
l .- f.a . was supported by the office of science of the u.s .
department of energy under subcontract no .
3f-3138 and a.j.m . by doe fg - er04169 .
l .- f.a . thanks ara go for numerous discussions on implementing an exact diagonalization code and alejandro lopez - bezanilla for helpful discussions .
b. schlkopf and a. j. smola , _ learning with kernels _ ( mit press , 2002 ) .
bobby g. sumpter and donald w. noid , chem .
. lett . * 192 * , 455 ( 1992 ) .
s. lorenz , a. gross and m. scheffler , chem .
phys .
lett . * 395 * , 210 ( 2004 ) .
s. manzhos and t. carrington , jr . ,
j. chem . phys . * 125 * , 084109 ( 2006 ) .
j. behler and m. parrinello , phys .
rev .
lett . * 98 * , 146401 ( 2007 ) .
albert p. bartk , mike c. payne , risi kondor and gbor csnyi , phys .
rev .
lett . * 104 * , 136403 ( 2010 ) .
zhenwei li , james r. kermode and alessandro de vita , phys .
rev .
lett . * 114 * , 096405 ( 2015 ) .
j. c. snyder , m. rupp , k. hansen , k - r .
mller and k. burke , phys .
rev .
lett .
* 108 * , 253002 ( 2012 ) .
a. lopez - bezanilla and o. a. von lilienfeld , phys .
rev .
b * 89 * , 235411 ( 2014 ) .
k. t. schtt , h. glawe , f. brockherde , a. sanna , k .-
r .
mller and e. k. u. gross , phys .
rev .
b * 89 * , 205118 ( 2014 ) .
f. faber , a. lindmaa , o. a. von lilienfeld and r. armiento , int .
j. quantum chem . in print 2015 .
j. hachmann , r. olivares - amaya , s. atahan - evrenk , c. amador - bedolla , r. s. snchez - carrera , a. gold - parker , l. vogt , a. m. brockway , and a. aspuru - guzik , the jour . of phys .
chem .
lett .
* 2 * , 2241 ( 2011 ) m. rupp , a. tkatchenko , k .-
r .
mller and o. a. von lilienfeld , phys .
rev .
lett . * 108 * , 058301 ( 2012 ) .
g. montavon , m. rupp , v. gobre , a. vazquez - mayagoitia , k. hansen , a. tkatchenko , k .-
r .
mller and o. a. von lilienfeld , new journal of physics * 15 * , 095003 ( 2013 ) .
r. ramakrishnan , p. o. dral , m. rupp and o. a. von lilienfeld , j. chem .
theory comput .
* 11 * , 2087 ( 2015 ) .
p. o. dral , o. a. von lilienfeld and w. thiel , j. chem .
theory comput .
* 11 * , 2120 ( 2015 ) .
o. a. von lilienfeld , r. ramakrishnan , m. rupp and a. knoll , int . j. quantum chem . in print ( 2015 ) .
r. ramakrishnan and o. a. von lilienfeld , chimia int .
j. for chem . * 69 * , 182 ( 2015 ) .
t. bereau , d. andrienko and o. a. von lilienfeld , j. chem .
theory comput . in print ( 2015 ) .
r. ramakrishnan , m. hartmann , e. tapavicza and o. a. von lilienfeld , arxiv:1504.01966v1 j. k. freericks , b. k. nikoli , and o. frieder , int . j. mod
. phys .
b * 28 * , 1430021 ( 2014 ) .
l .- f .
arsenault , a. lopez - bezanilla , o. anatole von lilienfeld and a. j. millis , phys .
rev .
b * 90 * , 155136 ( 2014 ) .
l. li , j. c. snyder , i. m. pelaschier , j. huang , u .- n .
niranjan , p. duncan , m. rupp , k .-
r .
mller and k. burke , arxiv:1404.1333v2 .
k. vu , j. snyder , l. li , m. rupp , b. f. chen , t. khelif , k .-
r .
mller and k. burke , int .
j. quantum chem . in print ( 2015 ) .
a. georges , g. kotliar , w. krauth , and m. j. rozenberg rev .
mod .
phys . * 68 * , 13 ( 1996 ) .
e. dagotto , rev .
mod .
phys .
* 66 * , 763 ( 1994 ) .
m. caffarel and w. krauth , phys .
rev .
lett . * 72 * , 1545 ( 1994 ) .
corinna cortes and vladimir vapnik , machine leaming * 20 * , 273 ( 1995 ) .
j. hertz , a. krogh and r.g .
palmer , introduction to the theory of neural computation , addison - wesley , reading , 1996 .
antonio criminisi , jamie shotton and ender konukogly , foundations and trends in computer graphics and vision * 7 * , 81 ( 2011 ) .
l. boehnke , h. hafermann , m. ferrero , f. lechermann , and o. parcollet , phys .
rev .
b * 84 * , 075145 ( 2011 ) .
l .- f .
arsenault , p. smon and a .- m.s .
tremblay , phys .
rev .
b * 86 * , 085133 ( 2012 ) .
l .- f .
arsenault and a. j. millis , to be published . * supplemental material for machine learning for many - body physics : efficient solution of dynamical mean - field theory * |
the antennae ( ngc 4038/39 ) are the nearest example of a major merger involving two gas - rich disk galaxies of comparable mass .
they allow us to study processes of dissipational galaxy assembly from close up , thus providing a valuable glimpse of what must have been more frequent events in the early universe .
the antennae have been observed extensively at all wavelengths ( e.g. , x - rays : @xcite ; uv : @xcite ; optical : @xcite ; ir : @xcite ; and 21-cm line : @xcite ) and have also been repeatedly modeled via n - body and hydrodynamical simulations ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
an accurate distance to this prototypical merger is of obvious importance in assessing its structure and dynamics , as well as for determining the physical properties of its myriad of stars , star clusters , peculiar objects such as ulxs ( ultraluminous x - ray sources ) , and gas clouds .
traditionally , the distance to ngc 4038/39 has most often been derived from the systemic recession velocity and an adopted hubble constant @xmath8 , with or without corrections for deviations of the hubble flow from linearity due to various attractors .
a frequently used modern value for the distance is 19.2 mpc @xcite .
this value is based on a systemic recession velocity relative to the local group of @xmath9 km s@xmath1@xcite , a linear hubble flow , and @xmath10 km s@xmath1 mpc@xmath1 .
updated to the _ hst _ key project s derived @xmath11 km s@xmath1 mpc@xmath1@xcite , this often - used distance to the antennae becomes @xmath12 mpc .
a significantly shorter distance of @xmath2 mpc has recently been determined by @xcite via photometry of the trgb in a region near the tip of the southern tidal tail .
this photometry performed on new , deep images obtained with _ hst _ s advanced camera for surveys ( acs)seems to support the short distance derived earlier from _
images of the same region by the same method @xcite .
if the short distance is correct , the antennae `` diminish '' in physical size , mass , and luminosity , as do their stars , clusters , and gas clouds .
however , their recession velocity then deviates from the best large - scale - flow model ( * ? ? ?
* hereafter tbad00 ) by about 500 km s@xmath1or 2.7@xmath13 @xcite .
clearly , new measurements of the distance to the antennae by independent methods are highly desirable . in the present paper
we derive a new distance from observations of the type ia sn 2007sr , which appeared in the midsection of the southern tail in 2007 december .
section [ sec2 ] presents our observations , data analysis , and new distance .
section [ sec3 ] discusses problems with the short distance and points out a likely error in the identification of the true trgb by @xcite .
finally , [ sec4 ] presents our conclusions and recommendation .
supernova 2007sr was discovered by the catalina sky survey on 2007 december 18.53 ut @xcite , about 4.2 days after @xmath14 maximum @xcite .
since it was of type ia @xcite , it was followed by the carnegie supernova project ( csp ) as part of their low - z campaign @xcite , with observations beginning on 2007 december 20.33 ut .
figure [ fig01 ] shows the sn on a stacked @xmath15 image obtained with the swope 1 m telescope at las campanas observatory .
the csp measured light curves for the sn using the filter set @xmath16 from 6 days until approximately 130 days after @xmath14 maximum . appendix [ appa ] gives some details about the photometric system and calibration , while its table [ tab : photometry ] lists all measured magnitudes and their uncertainties .
it is now well known that supernovae of type ia ( sn ia ) can serve as excellent standard candles and have been used to both discover the presence @xcite and constrain the nature of dark energy @xcite .
once corrected for extinction and the well - known correlation between intrinsic luminosity and decline rate of the light curve @xcite , sne ia have an intrinsic dispersion of approximately @xmath17 mag .
we can therefore use sn 2007sr to constrain the distance to ngc 4038/39 to this level . because sn 2007sr was caught after maximum
, it is necessary to use sn ia light - curve templates to estimate the peak magnitude in each passband as well as the decline - rate parameter , defined as the change in @xmath14 magnitude between maximum and 15 days later in the rest frame of the sn .
this parameter will allow us to correct for the intrinsic luminosity of the sn and also to predict its intrinsic colors , from which we can derive the host galaxy extinction .
more formally , we model the light curve in each passband @xmath18 with the following formula : @xmath19 where @xmath20 is the light - curve template for passband @xmath18 , @xmath21 is the time of maximum for the @xmath14 light curve , @xmath22 is the absolute magnitude in passband @xmath18 of a sn ia with @xmath23 mag , @xmath24 is the slope of the luminosity relationship , @xmath25 is the @xmath26-correction in passband @xmath18 , and @xmath27 is the true distance modulus .
the extinctions from the milky way and host galaxy are @xmath28 and @xmath29 , respectively .
the color excess due to the milky way foreground extinction is taken from @xcite , which for the position of sn 2007sr corresponds to @xmath30 mag .
the values of the parameters @xmath22 and @xmath24 are determined from the set of sne ia observed by the csp in the first - year campaign by assuming a fixed cosmological model ( @xmath31 km s@xmath1 mpc@xmath1 , @xmath32 , and @xmath33 ; @xcite ) , computing distance moduli from the observed redshifts of the host galaxies , correcting for milky - way and host - galaxy extinction , and fitting the resulting absolute peak magnitudes in each passband to the formula @xmath34 the details of this procedure are given in folatelli et al .
( 2008 , in preparation ) .
in performing this calibration , it has become apparent that for whatever reason a standard reddening law with the typical milky way value of @xmath35 does not provide a good fit to this sample of sne ia .
instead , we obtain a value @xmath36 ( folatelli et al .
2008 , in preparation ) .
other groups have also found surprisingly low values of @xcite .
this may be due to the sne ia being embedded in an environment that differs significantly from a typical ism , or perhaps there is another , intrinsic color
magnitude effect at play that has been inadvertently combined with @xmath37 .
whatever the reason , we are interested in a distance determination and therefore appeal to the empirical result @xmath38 .
however , we note that changing in equation ( [ eq : lc_model ] ) does not only affect the extinction term , but also changes the values of @xmath22 because the calibrating sample has a non - zero average extinction .
indeed , only a sn with an extinction equal to the average extinction of the calibrating sample would be insensitive to changes in . in the case of sn 2007sr
, the extinction is slightly larger than for the calibrating sample , whence an increase in would slightly decrease the distance modulus .
specifically , changing from @xmath39 to @xmath40 would change the estimated true distance modulus of this sn by @xmath410.06 mag .
for sn 2007sr .
for clarity , magnitudes in the different passbands have been offset by amounts indicated in the legend .
_ lines _ represent the best - fit templates .
error bars are all smaller than the data points .
epoch 0 corresponds to peak brightness in the @xmath14 band on 2007 december 14.3 ut .
[ fig02 ] ] the light - curve templates @xmath20 are constructed in a manner similar to that used by @xcite and further described in burns et al .
( 2008 , in preparation ) .
essentially , the highest - quality optical light curves from the csp , which have clearly observed maxima , are fit with a cubic spline , from which , @xmath21 , and the peak magnitude are extracted .
these light curves then define a sparsely sampled 2d surface in @xmath42@xmath43 space . a template of any given
can then be generated by 2d interpolation of this surface . at present
, there are an insufficient number of near - infrared ( nir ) light curves with which to construct templates in this way .
fortunately , the decline - rate parameter for sn 2007sr ( @xmath44 mag ) is nearly identical to that of sn 2006x ( @xmath45 mag ) , for which the csp has obtained high - quality nir photometry well before and after @xmath14 maximum .
we therefore use spline fits of the @xmath46 and @xmath47 light curves of sn 2006x as templates for fitting the @xmath46 and @xmath47 data of sn 2007sr .
the @xmath48 light curves of sn 2007sr are shown in fig .
[ fig02 ] along with the best - fit models for the optical data given by equation ( [ eq : lc_model ] ) .
the extracted light - curve parameters of interest are the true distance modulus of sn 2007sr , @xmath49 mag , the color excess due to the host galaxy , @xmath50 mag , and the decline - rate parameter @xmath51 mag .
the value of this parameter is a typical value for sne ia , and the color excess is reasonably low as one would expect for a source far from the host center .
the uncertainties are derived from the covariance matrix of the fit to the light curves , scaled so that @xmath52 . to be conservative
, we add to the uncertainty in @xmath27 the following systematic errors : ( 1 ) @xmath53 mag for the intrinsic dispersion in sn ia luminosities , ( 2 ) uncertainties in the calibration parameters of @xmath54 , @xmath55 mag , and @xmath56 , and ( 3 ) an uncertainty in the hubble constant of @xmath57 km s@xmath1 mpc@xmath1 . with this error budget
, we arrive at a final distance estimate to sn 2007sr of @xmath58 mpc ( @xmath59 mag ) using the optical data . as a check on this distance
we also fitted the @xmath46- and @xmath47-band data with templates generated from sn 2006x in order to estimate the peak magnitudes : @xmath60 and @xmath61 .
we then used the calibration by @xcite , @xmath62 mag and @xmath63 mag , to compute an average distance modulus of @xmath64 mag .
this modulus , based on nir data and an independent calibration , is nearly insensitive to assumptions about the reddening law of the dust in the host galaxy and agrees with the formal modulus of @xmath65 mag from the optical data to within the combined errors , thus supporting the derived distance of @xmath3 mpc .
as we demonstrate below , the new sn ia distance to ngc 4038/39 agrees well with the distance inferred from the system s recession velocity and the large - scale flow model by tbad00 .
if the short distance of @xmath2 mpc measured by @xcite were to be correct , it would create at least three serious problems : ( 1 ) ngc 4038/39 would have an exceptionally large peculiar recession velocity of about @xmath66522 km s@xmath1 ; ( 2 ) sn 2007sr would have had a peak luminosity differing by @xmath671.1 mag , or about 7@xmath68 , from the mean peak luminosity of sne ia ; and ( 3 ) the antennae would lose their membership in tully s ( 1988 ) group 221 of 13 galaxies .
we now discuss these three problems in turn and then point out a likely error in the derivation of the short distance by @xcite . a careful assessment of the many radial velocities measured for ngc 4038/39 and compiled in the nasa / ipac extragalactic database ( ned ) yields a systemic heliocentric velocity of @xmath69 km s@xmath1 .
figure [ fig03 ] compares this systemic velocity with recession velocities predicted by the large - scale flow model , which is based on surface - brightness fluctuation distances to 300 early - type galaxies ( tbad00 ) .
the antennae are plotted at both the short distance and the new sn iadistance , while the solid curve represents the predicted large - scale flow velocity in their direction .
the undulation in the model curve reflects the gravitational influences of the virgo cluster , whose center lies @xmath70 away from the antennae , and of the great attractor .
as the figure illustrates , with the new sn ia distance the antennae fall within @xmath71 km s@xmath1 or @xmath660.1 of the predicted flow velocity , where @xmath72 km s@xmath1 is the best - fit `` thermal '' ( or random ) radial - velocity dispersion of the model .
if instead we choose the short distance , the peculiar velocity becomes @xmath66522 ( @xmath73 ) km s@xmath1 or @xmath662.8 , clearly an exceptionally high value .
we conclude that the newly measured sn ia distance is in significantly better agreement with the best available large - scale flow model than the short distance is . .
plotted at the new distance based on type ia sn 2007sr ( _ filled dot _ ) , the recession velocity of ngc 4038/39 falls well within 1@xmath74 ( _ dotted lines _ ) of the large - scale flow , where @xmath74 is the cosmic random radial velocity .
in contrast , when plotted at the short distance based on the trgb ( _ square _ , * ? ? ?
* ) , the recession velocity of ngc 4038/39 lies 522 km s@xmath1 or 2.8@xmath13 above the flow . for details , see [ sec3.1 ] .
[ fig03 ] ] our determination of a distance of @xmath75 mpc to sn 2007sr has been based on the assumption that this sn reached the _
mean _ absolute magnitude @xmath76 typical for its decline rate .
the quoted distance error includes a term reflecting the intrinsic dispersion of the corrected peak absolute magnitudes @xmath77 of sne iaaround the mean value @xmath76 , which is @xmath78 mag for csp observations of similar quality ( [ sec2 ] ) . _
if _ the short distance of 13.3 mpc were to be correct , sn 2007sr would have reached a peak absolute magnitude that differed by @xmath79 mag , or about 7@xmath68 , from the mean value for sne ia .
this seems quite unlikely , given the spectral normality of sn 2007sr @xcite . to illustrate this normality , fig .
[ fig04 ] shows an optical spectrum obtained by the csp on 2007 december 27.33 ut with the baade 6.5 m telescope and imacs spectrograph , corresponding to an epoch of 13 days after @xmath14 maximum . for comparison
, we include the sn ia template spectrum by @xcite at + 13 days .
although there are small differences between the two spectra , the overall resemblance is extremely close , demonstrating that sn 2007sr was both photometrically and spectroscopically a completely normal sn ia .
hence , its extraordinarily low peak luminosity ( by 7@xmath68 ! ) , were it to lie at 13.3 mpc from us , is a second good reason for distrusting the short distance .
identifications of the major absorption features are taken from @xcite .
the spectrum has been corrected for reddening produced in both the milky way and the antennae , and has been shifted to a rest wavelength scale .
the flux scale is plotted in ab magnitudes as defined by @xcite . for comparison , the sn ia template spectrum by @xcite for an epoch of + 13 days
is also shown .
[ fig04 ] ] based on their position in the sky and recession velocity , @xcite assigned the antennae to group 221 in the crater cloud .
this group contains 13 galaxies , of which four besides the antennae have measured distances .
none of these distances is less than 20 mpc @xcite .
two early - type galaxies of the group have had their distances measured via both the surface - brightness fluctuation method and the fundamental - plane method @xcite .
the average distance for ngc 4024 ( s0 ) is a rather uncertain @xmath80 mpc , while that for ngc 4033 ( e6 ) is a more concordant @xmath81 mpc .
both galaxies lie within @xmath671.0 from the antennae . were the antennae to lie at the short distance of 13.3 mpc
, it would be difficult to understand why two nearby postulated group members with very similar recession velocities ( @xmath6652 and @xmath4124 km
s@xmath1 relative to the antennae , respectively ) should lie at much larger distances . again
, the evidence favors both the large - scale flow distance of @xmath5 mpc ( see [ sec4 ] ) and our new sn ia distance of @xmath3 mpc to the antennae .
why is the trgb distance of @xmath2 mpc determined by @xcite so discrepant ?
the answer seems to lie in a misidentification of what constitutes the true trgb of the mixed - age stellar populations near the tip of the southern tidal tail . the area imaged by @xcite with _
hst_/acs contains several regions @xcite and a very blue , dwarf - like feature dubbed the `` s78 region '' and now recognized to be the bent -rich tip of the southern tail @xcite . in their figure 3 , @xcite display two luminosity functions ( lf ) for the rgb , one derived for the s78 region and the other for a `` ne region '' that is free of and also , supposedly , of relatively young stars .
the lf of the s78 region shows a relatively sharp rise at @xmath82 mag ( where @xmath83 is an instrumental magnitude close to the kron
cousin @xmath84 magnitude ) , which @xcite identify as the trgb but which given the and blue stellar content of the s78 region seems equally likely to be caused by relatively young agb stars .
in contrast , the lf of the gas - free ne region shows only a minor sharp rise there , but features a more significant `` discontinuity at @xmath85 mag '' @xcite that falls within @xmath860.2 mag of the trgb expected for a distance of 22 mpc .
while @xcite dismiss this discontinuity as possibly `` due to the rgb of a younger , @xmath67200 myr population , '' we suggest that its stars being _ fainter _ than the claimed trgb actually mark the true location of the trgb . to check on this suggestion , we downloaded the acs / wfc ( wide field camera ) frames obtained under proposal go-10580 ( p.i . :
ivo saviane ) from the _
hst_archive , reprocessed them , and extracted a calibrated color magnitude diagram ( cmd ) for the entire tidal - tail tip region .
our search and photometry algorithms , which use dolphot ( for technical details , see * ? ? ?
* ) , detected about 30,000 stellar objects brighter than @xmath87 mag .
figure [ fig05 ] ( left panel ) shows the cmd in a new form proposed by @xcite , in which the traditional @xmath84 ( or extinction - corrected @xmath88 ) magnitude is replaced by the magnitude for the red giant branch ( rgb ) . this magnitude is a color - corrected @xmath84-band magnitude specifically designed to be insensitive to varying metallicity and is defined by @xmath89\ , , \label{eq : trgbdef}\ ] ] where the color slope @xmath90 is chosen to track the known , metallicity - induced run of the @xmath88 magnitude of the trgb as a function of @xmath91 , and @xmath92 mag is the fiducial color of reference @xcite . by construction , in any @xmath91 diagram of an old stellar population with a range of metallicities , the trgb lies at some constant value of . to locate the trgb objectively , we selected all stars in the color range @xmath93 and used a sobel edge - detection kernel @xcite to filter their numbers ( in 0.03 mag bins ) as a function of .
the resulting filter output , expressed as a ratio between the filter response and local poisson noise , is shown in the right panel of fig .
[ fig05 ] . in this panel ,
a peak at a normalized filter response of 2.0 indicates an edge detection at a 2@xmath13-level of significance .
figure [ fig05 ] shows a very clear , 5.4@xmath13 detection of the trgb at @xmath94 mag ( marked `` trgb '' ) , or about 0.9 mag _ fainter _ than the trgb position of @xmath95 claimed by @xcite .
the latter position ( at @xmath96 ) corresponds approximately to the broad @xmath673.5@xmath13 `` hump '' seen centered on @xmath97 mag in the normalized filter response of fig .
[ fig05 ] .
we conclude that @xcite probably misidentified youngish agb stars for the true trgb , in the process overestimating its brightness by about @xmath410.9 mag .
this , then , is the likely source of their discrepant @xmath2 mpc distance .
our newly determined @xmath94 mag yields a true distance modulus to the antennae of approximately@xmath98 mag , where @xmath99 mag is the absolute @xmath84 magnitude for the trgb of @xmath100 cen @xcite and holds at the fiducial color of reference @xcite .
this distance modulus , which is corrected for milky way foreground extinction via [ equation ( [ eq : trgbdef ] ) ] , but not for ( varying ) extinction within the tidal - tail tip region , corresponds to a linear distance of @xmath6 mpc .
we postpone a more elaborate determination of , including a careful assessment of local extinction and any possible small biases due to the presence of young stellar populations , to a future paper dedicated to that subject
. however , note that @xmath6 mpc agrees with both @xmath3 mpc ( [ sec2 ] ) and @xmath101 mpc ( [ sec4 ] ) to within the combined errors , while it strongly disagrees with the short distance of @xmath2 mpc derived by @xcite .
thus , also supports the new sn ia distance .
we have presented @xmath48 light curves of sn 2007sr beginning 6 days after @xmath14 maximum and obtained during four months by the carnegie supernova project .
analysis of these light curves in conjunction with sne ia data from the first - year campaign by the project yields a formal true distance modulus of @xmath102 mag . with all known systematic errors and uncertainties included , this distance modulus becomes @xmath59 mag , corresponding to a distance of @xmath58 mpc to ngc 4038/39 .
( differential depth effects between sn 2007sr and the center of mass of the galaxies are negligible ) .
this new sn ia distance agrees well with the distance @xmath103 mpc estimated from the systemic recession velocity of the antennae and the large - scale flow model by tbad00 , where the quoted error reflects the cosmic random radial velocity ( @xmath72 km s@xmath1 ) of the model . on the other hand ,
disagrees strongly with the short distance of @xmath104 mpc estimated by @xcite from the trgb .
we have discussed three serious problems with such a short distance , and have pointed out the likely misidentification of the trgb by these authors as the cause of the short distance .
reprocessing their _ hst_/acs frames and using an improved method of trgb detection , we have located what we believe to be the true trgb at @xmath105 mag , fully 0.9 mag below ( i.e. , fainter than ) the trgb claimed by @xcite analyzing photometry of this new , fainter trgb we have derived a preliminary distance of @xmath6 mpc . clearly
, additional distance estimates to the antennae ( e.g. , from cepheids , planetary nebulae , etc . ) will be very valuable .
however , for the moment we see no reason for adopting any distance shorter than 20 mpc . given the concordant new @xmath3 mpc and recession - velocity based @xmath5 mpc , both supported by our new , preliminary @xmath6 mpc , we recommend using a conservative , rounded value of @xmath7 mpc as the best currently available distance to the antennae .
we thank david murphy for his efforts in building instruments for the csp , john tonry for his flow - model software and advice , and claudia maraston for helpful discussions .
this research has made use of the nasa / ipac extragalactic database , which is operated by the jet propulsion laboratory , california institute of technology , under contract with nasa .
m.h . and g.f .
ackowledge support from the millennium center for supernova science through grant p06 - 045-f funded by `` programa bicentenario de ciencia y tecnologa de conicyt '' and `` programa iniciativa cientfica milenio de mideplan . ''
. acknowledges additional support from centro de astrofsica fondap 15010003 and fondecyt through grant 1060808 .
observations made under the carnegie supernova project were supported in part by the nsf through grants ast03 - 06969 and ast06 - 07438 .
the magnitudes for sn 2007sr presented in this paper are on the csp natural system .
details of the data reduction , calibration , and filter functions and color terms defining this natural system can be found in @xcite and contreras et al .
( 2008 , in preparation ) , as well as on the csp website : http://csp1.lco.cl/~cspuser1 . in this appendix
, we briefly outline the natural system and present a table of the photometry for sn 2007sr .
the @xmath106 magnitudes are calibrated via the @xcite standard stars .
a color sequence of these stars is used to establish color terms that transform our instrumental magnitudes to the standard system of @xcite . however , since sne ia have spectra that are very different from those of normal stars and evolve with age , these color terms can not be used to transform the sn magnitudes to the standard system .
instead , we use the color terms in reverse to transform the standard magnitudes to our natural system , and then calibrate our sn observations using these natural magnitudes .
it is in this system that we present the photometry of sn 2007sr and perform our entire analysis .
the @xmath107 magnitudes are treated in a similar way , though using the system of @xcite standard stars .
the @xmath108 magnitudes are calibrated using the @xcite standards , for which the instrument and filters used were nearly identical to those employed by the csp and , therefore , were equivalent to our natural system .
lcccccccc 1454.84 & & & & & & & 13.776(0.019 ) & 13.723(0.005 ) + 1455.76 & 14.133(0.013 ) & 12.999(0.004 ) & 12.924(0.005 ) & 13.664(0.006 ) & 13.165(0.005 ) & 12.856(0.005 ) & & + 1456.76 & 14.241(0.014 ) & 13.051(0.004 ) & 12.983(0.006 ) & 13.733(0.007 ) & 13.251(0.007 ) & 12.904(0.006 ) & & + 1457.80 & 14.345(0.009 ) & 13.101(0.004 ) & 13.047(0.005 ) & 13.826(0.005 ) & 13.310(0.006 ) & 12.947(0.005 ) & & + 1458.84 & 14.429(0.012 ) & 13.168(0.008 ) & 13.136(0.009 ) & 13.942(0.010 ) & 13.390(0.007 ) & 13.000(0.008 ) & & + 1461.82 & & & & & & & 14.912(0.017 ) & 13.878(0.004 ) + 1462.80 & 14.936(0.010 ) & 13.484(0.005 ) & 13.406(0.005 ) & 14.190(0.006 ) & 13.810(0.006 ) & 13.265(0.006 ) & & + 1463.84 & & & & & & & 14.971(0.006 ) & 13.837(0.004 ) + 1464.84 & 15.180(0.012 ) & 13.665(0.007 ) & 13.498(0.007 ) & 14.216(0.009 ) & 14.040(0.008 ) & 13.389(0.005 ) & & + 1465.87 & & & & & & & 14.959(0.007 ) & 13.767(0.007 ) + 1467.77 & & 13.910(0.006 ) & 13.540(0.004 ) & 14.130(0.004 ) & 14.367(0.007 ) & 13.537(0.006 ) &
& + 1468.77 & 15.662(0.018 ) & 13.986(0.006 ) & 13.556(0.006 ) & 14.107(0.007 ) & 14.470(0.006 ) & 13.591(0.006 ) & & + 1469.80 & 15.819(0.016 ) & 14.083(0.005 ) & 13.544(0.004 ) & 14.049(0.006 ) & 14.571(0.006 ) & 13.640(0.006 ) & & + 1470.79 & 15.898(0.017 ) & 14.161(0.006 ) & 13.552(0.006 ) & 14.021(0.006 ) & 14.671(0.007 ) & 13.686(0.006 ) & & + 1471.87 & & & & & & & 14.708(0.007 ) & 13.534(0.007 ) + 1476.78 & 16.478(0.013 ) & 14.696(0.006 ) & 13.690(0.004 ) & 13.906(0.006 ) & 15.206(0.006 ) & 14.011(0.006 ) & & + 1477.84 & & & & & & & 14.313(0.006 ) & 13.469(0.005 ) + 1484.78 & 16.924(0.029 ) & 15.252(0.008 ) & 14.206(0.007 ) & 14.331(0.009 ) & 15.728(0.010 ) & 14.574(0.010 ) & & + 1485.82 & & & & & & & 14.630(0.009 ) & 13.880(0.008 ) + 1488.73 & 17.086(0.048 ) & 15.407(0.008 ) & 14.429(0.006 ) & 14.583(0.007 ) & 15.863(0.011 ) & 14.734(0.006 ) & & + 1490.79 & 17.150(0.013 ) & 15.466(0.005 ) & 14.517(0.004 ) & 14.699(0.004 ) & 15.917(0.008 ) & 14.814(0.005 ) & & + 1491.88 & & & & & & & 15.186(0.009 ) &
14.205(0.007 ) + 1494.85 & 17.260(0.015 ) & 15.553(0.007 ) & 14.657(0.006 ) & 14.873(0.008 ) & 15.983(0.009 ) & 14.941(0.007 ) & & + 1497.78 & 17.251(0.035 ) & 15.603(0.006 ) & 14.762(0.006 ) & 15.004(0.007 ) & 16.027(0.009 ) & 15.025(0.006 ) & & + 1498.82 & & & & & & & 15.674(0.011 ) & 14.527(0.008 ) + 1499.77 & & 15.630(0.005 ) & 14.830(0.005 ) & 15.074(0.006 ) & 16.057(0.007 ) & 15.076(0.005 ) & & + 1503.82 & & & & & & & 16.021(0.013 ) & 14.764(0.008 ) + 1508.86 & & 15.768(0.006 ) & 15.123(0.007 ) & 15.444(0.008 ) & 16.164(0.010 ) & 15.312(0.006 ) & & + 1511.79 & & & & & & & 16.527(0.016 ) & 15.083(0.012 ) + 1512.79 & & 15.834(0.006 ) & 15.261(0.007 ) & 15.585(0.008 ) & 16.207(0.009 ) & 15.407(0.007 ) & & + 1515.76 & & 15.870(0.007 ) & 15.360(0.005 ) & 15.700(0.007 ) & 16.237(0.009 ) & 15.490(0.007 ) & & + 1517.76 & & 15.896(0.006 ) & 15.417(0.006 ) & 15.759(0.007 ) & 16.260(0.010 ) & 15.530(0.007 ) & & + 1523.79 & & 15.996(0.006 ) & 15.620(0.005 ) & 15.973(0.007 ) & 16.349(0.008 ) & 15.686(0.006 ) & & + 1527.77 & & 16.047(0.006 ) & 15.737(0.008 ) & 16.115(0.010 ) & 16.396(0.009 ) & 15.784(0.008 ) & & + 1532.70 & & & & & & 15.915(0.020 ) & & + 1538.75 & & 16.214(0.005 ) & 16.089(0.004 ) & 16.470(0.005 ) & 16.571(0.006 ) & 16.051(0.004 ) & & + 1540.68 & & 16.218(0.006 ) & 16.131(0.008 ) & 16.521(0.009 ) & 16.607(0.010 ) & 16.140(0.012 ) & & + 1542.66 & & 16.276(0.007 ) & 16.217(0.008 ) & 16.612(0.010 ) & 16.633(0.010 ) & 16.138(0.008 ) & & + 1549.71 & & 16.396(0.009 ) & 16.447(0.011 ) & 16.819(0.014 ) & 16.725(0.019 ) & 16.298(0.011 ) & & + 1558.65 & & 16.531(0.006 ) & 16.723(0.008 ) & 17.098(0.011 ) & 16.885(0.012 ) & 16.483(0.007 ) & & + 1566.76 & & 16.663(0.007 ) & 16.964(0.008 ) & 17.305(0.011 ) & 17.017(0.012 ) & 16.648(0.009 ) & & + 1571.66 & & 16.747(0.007 ) & 17.096(0.009 ) & 17.422(0.013 ) & 17.099(0.012 ) & 16.750(0.009 ) & & + 1578.70 & & 16.870(0.010 ) & 17.291(0.008 ) & 17.600(0.009 ) & 17.208(0.009 ) & 16.935(0.059 ) & & | traditionally , the distance to ngc 4038/39 has been derived from the systemic recession velocity , yielding about 20 mpc for @xmath0 km s@xmath1 mpc@xmath1 . recently
, this widely adopted distance has been challenged based on photometry of the presumed tip of the red giant branch ( trgb ) , which seems to yield a shorter distance of @xmath2 mpc and , with it , nearly 1 mag lower luminosities and smaller radii for objects in this prototypical merger . here
we present a new distance estimate based on observations of the type ia supernova ( sn ) 2007sr in the southern tail , made at las campanas observatory as part of the carnegie supernova project .
the resulting distance of @xmath3 mpc [ @xmath4 mag ] is in good agreement with a refined distance estimate based on the recession velocity and the large - scale flow model developed by tonry and collaborators , @xmath5 mpc .
we point out three serious problems that a short distance of 13.3 mpc would entail , and trace the claimed short distance to a likely misidentification of the trgb . reanalyzing
_ hubble space telescope _
( _ hst _ ) data in the archive with an improved method , we find a trgb fainter by 0.9 mag and derive from it a preliminary new trgb distance of @xmath6 mpc .
finally , assessing our three distance estimates we recommend using a conservative , rounded value of @xmath7 mpc as the best currently available distance to the antennae . |
in the m / g/1 queueing model , customers arrive at a single server with independent interarrival times , @xmath2 , distributed according to the @xmath3 distribution . here , @xmath4 is the arrival rate , hence @xmath2 has density function @xmath5 for @xmath6 and @xmath7 otherwise .
they are served with independent service times @xmath8 , which have a @xmath9 distribution .
( our mcmc approach can be generalized to other service time distributions . )
we do not observe the interarrival times , only the interdeparture times , @xmath10 .
the goal is to infer the unknown parameters of the queueing model , @xmath11 , using observed interdeparture times , @xmath12 .
we assume that the queue is empty before the first arrival .
the process of interdeparture times can be written as @xmath13 where @xmath14 is the time from the arrival of customer @xmath15 ( or from 0 when @xmath16 ) to the arrival of customer @xmath17 , @xmath18 is the arrival time of the @xmath19-th customer , and @xmath20 is the departure time of the @xmath19-th customer , with @xmath21 defined to be @xmath7 .
we take the arrival times , @xmath22 , to be latent variables .
the @xmath22 evolve in time as a markov process .
viewed this way , the queueing model can be summarized as follows : @xmath23 for convenience , set @xmath24 .
the joint density of the @xmath22 and the @xmath10 can be factorized as follows @xmath25 let @xmath26 be a prior for @xmath27 .
the posterior distribution @xmath28 of @xmath27 is then @xmath29 the integral ( [ eq : vxpdf ] ) can not be computed analytically .
hence to sample from @xmath28 we need to include the @xmath22 in the mcmc state and sample from the joint posterior distribution of the @xmath22 and @xmath27 given the data , which is @xmath30 taking the values of @xmath27 from each draw from ( [ eq : vxpost ] ) and ignoring the @xmath22 will yield a sample from the posterior distribution of @xmath27 .
our mcmc procedure will be based on combining five different updates .
the first is a gibbs update that draws new values for each @xmath22 , given values of the other @xmath22 , @xmath27 , and the data .
the second is a simple metropolis update for @xmath27 , given values of the @xmath22 and the data .
the last three updates are novel one metropolis `` shift '' update and two metropolis - hastings - green `` scale '' updates that propose to simultaneously change components of @xmath27 and all of the @xmath22 .
we first work out how to apply standard gibbs sampling updates for the @xmath22 , for which we need to derive the full conditional density for each @xmath22 given the other @xmath22 and @xmath27 .
we denote all of the @xmath31 except @xmath22 as @xmath32 .
we consider three cases , when @xmath33 , when @xmath34 and when @xmath35 .
for the case @xmath33 , we have @xmath36)}{\theta_{2 } - \theta_{1}}\nonumber \\ & & \times \ \theta_{3}e^{-\theta_{3 } ( v_{2 } - v_{1 } ) } i(v_{1 } \leq v_{2 } ) \nonumber \\ & \propto & i(v_{1 } \in [ \max(0 , x_{1 } - \theta_{2 } ) , \min(v_{2 } , x_{1 } - \theta_{1})])\end{aligned}\ ] ] so , conditional on the parameters and the observed data , the distribution of @xmath37 is @xmath38 .
when @xmath34 , we have @xmath39)}{\theta_{2 } - \theta_{1}}\nonumber \\ & & \times \ \theta_{3}e^{-\theta_{3 } ( v_{i + 1 } - v_{i})}i(v_{i } \leq v_{i+1 } ) \notag \\ & \propto & i(v_{i } \in [ v_{i-1 } , v_{i+1}])i(y_{i } \in [ \theta_{1 } + \max(0 , v_{i } - x_{i - 1 } ) , \theta_{2 } + \max(0 , v_{i } - x_{i - 1 } ) ] ) \ \ \ \ \ \ \label{eq : innercase}\end{aligned}\ ] ] to simplify this expression , note that @xmath40 , and first consider the case @xmath41 . when this is so
, we must have @xmath42 , hence , in this case @xmath22 will have a uniform distribution on @xmath43 $ ] . now consider the case @xmath44 .
we rewrite expression ( [ eq : innercase ] ) as @xmath45)i(y_{i } \in [ \theta_{1 } + \max(0 , v_{i } - x_{i - 1 } ) , \theta_{2 } + \max(0 , v_{i } - x_{i - 1})])i(v_{i } \leq x_{i-1 } ) } \nonumber \\ & + & i(v_{i } \in [ v_{i-1 } , v_{i+1}])i(y_{i } \in [ \theta_{1 } + \max(0 , v_{i } - x_{i - 1 } ) , \theta_{2 } + \max(0 , v_{i } - x_{i - 1})])i(v_{i } > x_{i-1 } ) \nonumber \\ & & = \ i(v_{i } \in [ v_{i-1 } , x_{i+1 } ] ) + i(v_{i } \in ( x_{i-1 } , \min(v_{i+1 } , x_{i } - \theta_{1 } ) ] ) \nonumber \\ & & = \ i(v_{i } \in [ v_{i-1 } , \min(v_{i+1 } , x_{i } - \theta_{1})])\end{aligned}\ ] ] we see that for @xmath44 , @xmath22 will have a uniform distribution on @xmath46 $ ] .
finally , for the case @xmath35 , we have @xmath47)\end{aligned}\ ] ] from this it follows that @xmath48 will have an exponential distribution , truncated to @xmath49 $ ] when @xmath50 and to @xmath51 $ ] when @xmath52 .
all of the above distributions can be easily sampled from by using the inverse cdf method .
the case @xmath35 deserves a small note .
if we suppose the truncation interval of @xmath48 is @xmath53 $ ] , then the cdf and inverse cdf will be @xmath54 so we can sample @xmath48 as @xmath55 where @xmath56 is uniform@xmath57 .
we use simple metropolis updates ( metropolis , et al ( 1953 ) ) to sample from the conditional posterior distribution of @xmath27 given values of the @xmath22 and the data , @xmath58 .
when doing simple metropolis updates of @xmath27 given the arrival times and the data , we used the 1-to-1 reparametrization @xmath59 . a simple metropolis update for @xmath60 that leaves @xmath61 invariant proceeds as follows .
we choose a symmetric proposal density @xmath62 , for which @xmath63 .
given the current value @xmath60 , we generate a proposal @xmath64 .
we compute the acceptance probability @xmath65 and let the new state , @xmath66 , be @xmath67 with probability @xmath68 and otherwise reject the proposal , letting @xmath69 .
we use a normal proposal with independent coordinates centered at the current value of @xmath60 , updating all components of @xmath60 at once .
if the proposed value for @xmath70 or @xmath71 is outside its range , the proposal can be immediately rejected . to prevent overflow or underflow , all mcmc computations use the logarithm of the posterior ( [ eq : vxpost ] ) , which ( up to an additive constant ) simplifies to @xmath72 whenever the following constraints are satisfied @xmath73 otherwise , @xmath74 .
the dominant operation when computing the log likelihood for a data set of length @xmath75 is checking the constraints ( [ eq : c1 ] ) to ( [ eq : c5 ] ) , which requires time proportional to @xmath75 . storing the constraints on @xmath76 and @xmath77 given by ( [ eq : c2 ] ) to ( [ eq : c5 ] ) during evaluation of the log likelihood at @xmath78 allows us to evaluate the log likelihood for another @xmath79 and the same @xmath80 in constant time .
this allows us to do @xmath81 additional metropolis updates for less than @xmath81 times the computational cost it takes to do a single update , which is likely to make sampling more efficient . a similar improvement in sampling should be possible for models with other service time distributions that have low - dimensional sufficient statistics . as well , doing additional simple metropolis updates makes an imperfect choice of proposal distribution have less of an effect on sampling efficiency .
the gibbs and simple metropolis updates are sufficient to give an ergodic mcmc scheme . however , as we will see , these updates are sometimes very inefficient when used on their own . in this section , we introduce our novel shift and scale updates , which can make mcmc sampling much more efficient .
shift updates and scale updates are used to sample from the joint posterior distribution of the parameters and the latent variables , @xmath82 .
the shift update takes the form of a standard metropolis update , while the scale updates are metropolis - hastings - green ( mhg ) updates .
in general , an mhg update proceeds as follows .
we introduce an extra variable @xmath83 .
given a current value @xmath84 and a density @xmath85 we generate @xmath86 .
we then propose @xmath87 .
the function @xmath88 is the inverse of itself , with a jacobian @xmath89 which is nowhere zero or infinite .
we compute the acceptance probability @xmath90 and let the new state , @xmath91 be @xmath92 with probability @xmath68 and let @xmath93 otherwise . for more on the mhg algorithm ,
see geyer ( 2003 ) .
the motivation for the shift updates and scale updates is that conditioning on given values of the arrival times constrains the range of parameter values for which the posterior density is non - zero , preventing us from changing the parameters by a significant amount .
this can lead to inefficient sampling when @xmath27 is updated given the @xmath22 .
in contrast , our new shift and scale updates change the latent variables and the parameters simultaneously , in accordance with their dependence structure , thus allowing for much greater changes to the parameters .
hard constraints on @xmath27 given @xmath94 are nt necessary for these updates to be beneficial .
if the service time density were non - zero for all positive values , the distribution of @xmath27 given @xmath94 might still be much more concentrated than the marginal posterior for @xmath27 .
we first consider updates that shift both the minimum service time , @xmath76 , and all the arrival times , @xmath22 , keeping the range of service times , @xmath95 , fixed .
note that for @xmath33 and for all @xmath19 when @xmath96 , we have @xmath97 .
hence the values of one or more of the @xmath22 will constrain how much we can propose to change @xmath76 any proposal to change @xmath76 that violates these constraints must be rejected . a shift update
addresses the presence of these constraints as follows .
we first draw a shift @xmath98 from any distribution symmetric around @xmath7 . then
, we propose new values @xmath99 for all @xmath19 and @xmath100 .
so , a proposal to increase the minimum service time is coupled with a simultaneous proposal to decrease all of the arrival times in a way that keeps the constraints between the arrival times and the minimum service time satisfied .
a shift proposal will be rejected if it proposes a value for @xmath37 or @xmath76 that is less than @xmath7 . otherwise ,
the acceptance probability for a shift update is @xmath101 next , we consider range scale updates that propose to simultaneously change the latent variables and the range of service times @xmath95 , which is also constrained by the latent variables specifically , for @xmath37 and for all @xmath102 where @xmath103 , we have @xmath104 .
these updates propose to scale the `` gaps '' @xmath105 ( gaps between the current arrival time and the latest possible arrival time ) , while at the same time proposing to scale @xmath95 by the same amount , keeping @xmath76 fixed . in particular , we first fix ( or draw from some distribution ) a scale factor @xmath106 .
we then draw @xmath107 .
the function @xmath108 proposes to change @xmath22 to @xmath109 and @xmath95 to @xmath110 .
we set @xmath111 .
the jacobian @xmath112 . a range scale proposal will be rejected if it proposes a set of @xmath22 for which some @xmath113 or for which @xmath114 . otherwise , the mhg acceptance probability for a range scale update is @xmath115 finally , we consider rate scale updates that propose to simultaneously change both the interarrival times @xmath116 , and the arrival rate @xmath4 .
we motivate these updates as follows .
we expect the distribution of @xmath4 given the interarrival times @xmath2 to be concentrated around @xmath117 .
consequently , we would expect the joint posterior of @xmath4 and the @xmath2 to have a high density along a ridge with values @xmath118 and @xmath119 for some @xmath120 , as long as @xmath118 and @xmath119 do not lie in region with low probability .
so , proposing to change @xmath2 to @xmath118 and @xmath4 to @xmath119 for some @xmath120 potentially keeps us in a region of high density .
we perform these updates in terms of @xmath121 . in detail
, this update proceeds as follows .
we first fix ( or draw from some distribution ) a scale @xmath122 .
we then draw @xmath107 .
the function @xmath123 proposes to change all @xmath2 to @xmath124 and @xmath125 to @xmath126 .
we set @xmath111 .
the jacobian @xmath127 .
we reject the proposal immediately if it violates the following constraints : for @xmath19 where @xmath41 the proposed arrival time @xmath128 must be in @xmath129 $ ] , and for @xmath19 where @xmath44 , we must have @xmath130 . otherwise , the mhg acceptance probability for a rate scale update is @xmath131 although a scheme consisting of the shift , range scale , and rate scale updates changes all of the latent variables and parameters , it is not ergodic . to see this ,
consider updating the arrival times @xmath22 , or equivalently , interarrival times @xmath116 .
shift updates do not change the interarrival times .
range scale updates with scale factor @xmath132 change each interarrival time as @xmath133 , and rate scale updates with scale factor @xmath134 change each interarrival time as @xmath135 . if we are in a situation where @xmath136 and @xmath137 for one or more @xmath138 , then all subsequent updates will keep @xmath136 .
note that this is true even if @xmath132 and @xmath134 are randomly selected at each update .
hence , our novel updates must still be combined with simple gibbs sampling updates of the @xmath22 s to get an ergodic sampling scheme .
we also still do simple metropolis updates for @xmath27 .
although this is not essential for ergodicity , it makes sampling a lot more efficient .
the goal of our empirical study is to determine when using the novel updates improves sampling efficiency .
we generate three simulated data sets that are representative of a range of possible scenarios , sample from the posterior distributions of @xmath27 for each of these data sets using various sampling schemes , and compute autocorrelation times to compare sampling efficiency .
the sort of data arising from the observation of a queueing system can be roughly classified into one of three scenarios .
the first scenario is when the interarrival times are , on average , smaller than the service times , that is , arrivals are relatively frequent . in this case
the queue is generally full , and empty only early on .
hence , most observed interdeparture times will tend to come from the service time distribution , and only a small number of interdeparture times will be relevant for inferring the arrival rate .
consequently , we expect there to be large uncertainty in @xmath4 , but less for @xmath76 and @xmath77 .
the second scenario is when interarrival times are on average slightly larger than the service times .
the queue is then sometimes empty , and sometimes contains a number of people . in this case
, we expect the data to be informative about all parameters .
this is also the case of most practical interest , as it corresponds to how we may expect a queueing system to behave in the real world .
the third scenario is when arrivals happen rarely , while the service time is relatively small . in this case
, the queue will usually be empty .
interarrival times are then informative for inferring @xmath4 , as most of them will come from the @xmath3 distribution with a small amount of added @xmath139 noise .
however , inference of the service time bounds @xmath76 and @xmath77 will now be based on how significantly the distribution of the interarrival times ( for a given @xmath4 ) deviates from the exponential distribution
. this difference may be rather small , and so we expect that the posterior for the service time bounds will be rather diffuse .
the three data sets we generated each consisted of @xmath140 interdeparture times , corresponding to each of the three scenarios above . for the first scenario we take @xmath141 , for the second , @xmath142 , and for the third , @xmath143 .
the plots in figure [ fig : datasets ] of the number of people in the queue up until the last arrival against time for each of the three data sets demonstrate that the data sets have the desired qualitative properties .
0.3 0.3 0.3 numerical values of the simulated interdeparture times are presented in table [ table : datasets ] .
.simulated interdeparture times @xmath144 [ cols=">,>,>",options="header " , ] from table [ table : tauest ] , we see that using just the shift , or just a scale update ( either a range or a rate scale update ) improves performance for sampling parameters that are changed by one of these updates . for a scheme which uses all updates ,
the greatest improvement in performance for sampling @xmath125 is in the case of frequent arrivals ( efficiency gain of 179 times ) , while perfomance improvement when sampling @xmath70 and @xmath71 is greatest in the case of rare arrivals ( efficiency gains of 58 and 61 times ) . in other cases , performance neither increases nor decreases significantly .
these results are in approximate agreement with the estimates of the standard errors of the posterior mean estimates in table [ table : meanest ] . in an additional run (
not used to compute the autocorrelation times shown here ) of the basic plus rate scheme , for the rare arrivals scenario , we found that the sampler got stuck for a while in a region of the parameter space .
the basic and basic + shift methods ( when initialized to a state from this `` stuck '' region ) also stayed in this `` stuck '' region for a while before visting other regions .
the basic plus range and basic plus all methods , when initialized with the stuck state , quickly returned to sampling other regions .
so , autocorrelation time estimates for the rare arrivals scenario , for methods other than basic plus all and basic plus rate , probably underestimate actual autocorrelation times .
we illustrate the performance of the samplers with trace plots in figures [ fig : tracefreq ] , [ fig : traceinter ] , and [ fig : tracerare ] . to produce these figures , we ran the samplers for an equal amount of computation time and thinned each run to @xmath145 points .
the black line on each plot is the true parameter value .
0.45 0.45 0.45 0.45 0.45 0.45 0.45 0.45 0.45 0.45 0.45 0.45 0.45 0.45 0.45 0.45 0.45 0.45 in the case of frequent arrivals , the marginal posterior of @xmath125 is diffuse but concentrated given all @xmath80 and the data , so simultaneously updating all @xmath80 and @xmath125 makes sampling more efficient .
in the case of rare arrivals , the marginal posteriors of both @xmath70 and @xmath71 are diffuse , while they are concentrated given all @xmath80 and the data . simultaneously updating all @xmath80 and
either @xmath70 or @xmath71 then leads to a noticeable gain in efficiency . in the other cases ,
the data is more informative about the parameters , so additional knowledge of the latent variables does not change the concentration of the posterior by as much . as a result , sampling efficiency is not affected as significantly by changing the latent variables and the parameters simultaneously .
in this paper , we have shown how bayesian inference with mcmc can be performed for the m / g/1 queueing model .
as mentioned earlier , fearnhead and prangle ( 2010 ) used abc for inference in the m / g/1 queueing model .
fearnhead and prangle ( 2010 ) also used abc for inference in the ricker model of population dynamics , in which we assume that a population process is observed with poisson noise .
the basis of all abc methods is the ability to simulate observations from a given stochastic model .
the simulation process for a set of observations is always driven by a latent process of sampling and transforming certain random variables .
the distributions of these random variables then determine the distribution of the final observed quantities .
if we consider the latent variables which drive the simulation process jointly with the observations , then we can think of the observations as coming from a model with some latent structrure .
these latent variables and the observed data will sometimes have a tractable joint density , which makes doing bayesian inference with mcmc possible , at least in principle .
in an earlier work , shestopaloff and neal ( 2013 ) , we used the same approach as in this paper to do bayesian inference for the ricker model , i.e. including additional latent variables in the mcmc state and sampling for them as well as model parameters .
we compared a basic mcmc scheme with several `` ensemble mcmc '' schemes for bayesian inference in the ricker model , and showed that using the ensemble schemes leads to a significant improvement in efficiency when compared to the basic mcmc scheme . like for the m / g/1 queueing model
, we have shown that bayesian inference with mcmc is possible for the ricker model , but requires more sophisticated mcmc methods for sampling to be efficient .
it would be interesting to apply alternative inference methods for models with a time series structure to the m / g/1 queue , for example particle filters and particle markov chain monte carlo ( pmcmc ) ( andrieu , doucet , and holenstein ( 2010 ) ) .
as mentioned earlier , it should be possible to extend the mcmc method in this paper to service time distributions other than the uniform one used in this paper .
the most direct extension would be to consider location - scale service time distributions .
besides the simple metropolis updates , we can then do additional updates for the location parameter ( or some 1-to-1 function of it ) using a shift update and additional updates for the scale parameter ( or some 1-to-1 function of it ) using a range scale update .
computation would not be impacted so long as ( [ eq : logpost ] ) can be written in terms of low - dimensional sufficient statistics .
this research was supported by the natural sciences and engineering research council of canada .
a. s. is in part funded by an nserc postgraduate scholarship .
r. n. holds a canada research chair in statistics and machine learning .
0.2 in fearnhead , p. , prangle , d. ( 2012 ) `` constructing summary statistics for approximate bayesian computation : semi - automatic approximate bayesian computation '' , _ journal of the royal statistical society b _ , vol .
74 , pp . 1 - 28
. metropolis , n. , rosenbluth , a.w . ,
rosenbluth , m.n . ,
teller , a.h . , and teller , e. ( 1953 ) .
`` equation of state calculations by fast computing machines '' . _ journal of chemical physics _ , vol . 21 , pp . 1087 - 1092 .
neal , r. m. , beal , m. j. , and roweis , s. t. ( 2004 ) `` inferring state sequences for non - linear systems with embedded hidden markov models '' , in s. thrun , et al ( editors ) , _ advances in neural information processing systems 16 _ , mit press .
neal , r. m. ( 2003 ) `` markov chain sampling for non - linear state space models using embedded hidden markov models '' , technical report no .
0304 , department of statistics , university of toronto , http://arxiv.org/abs/math/0305039 . | we introduce an efficient mcmc sampling scheme to perform bayesian inference in the m / g/1 queueing model given only observations of interdeparture times .
our mcmc scheme uses a combination of gibbs sampling and simple metropolis updates together with three novel `` shift '' and `` scale '' updates . we show that our novel updates improve the speed of sampling considerably , by factors of about @xmath0 to about @xmath1 on a variety of simulated data sets .
this paper proposes a new approach to computation for bayesian inference for the m / g/1 queue ( markovian arrival process / general service time distribution/1 server ) .
inference for this model using abc ( approximate bayesian computation ) was previously considered by bonassi ( 2013 ) , fearnhead and prangle ( 2012 ) , and blum and francois ( 2010 ) .
abc , in general , does not yield samples from the exact posterior distribution .
we use the strategy of considering certain unobserved quantities as latent variables , allowing us to use markov chain monte carlo ( mcmc ) , which converges to the exact posterior distribution . |
laser dynamics is commonly studied considering the electric field as a scalar variable , since in most systems the polarization state is imposed by anisotropies of the cavity .
for instance , brewster windows or gratings , generally used in gas lasers to close the laser tube or to select a vibro - rotational transition , impose a linearly polarized state of the laser emission .
however , in perfectly cylindrical laser cavities without any elements to select a preferred polarization , the study of the dynamics includes the necessity of considering the vector nature of the electric field .
several theoretical works have been devoted to the study of the polarization dynamics of the quasi - isotropic laser , showing the important role played by the material variables .
in particular , the degeneracy of the angular momentum states of the laser transition sublevels has been considered as the coupling source between different polarization states .
initial studies considered only stationary solutions @xcite , but more recently dynamical models have been developed to explore the role of anisotropy due to the laser medium , that from now on we will call intrinsic anisotropy @xcite .
these models predict a rich dynamics even for the simplest transition , @xmath1 .
later versions of these models include also linear and circular cavity anisotropies that we call extrinsic @xcite , or different level structure @xcite
. however , up to now few experiments have been performed on this subject .
experiments carried out on noble gas lasers reveal that in some systems it is necessary to consider the dynamics of the matter variables to fully understand the polarization features of an isotropic system @xcite . in other cases ,
the observations could be explained by a nonlinear coupling of the modes and residual cavity anisotropies @xcite . in this work we present experimental and theoretical results for the polarization competition dynamics in the transient state of a quasi - isotropic low - pressure co@xmath0 laser .
we show that the polarization dynamics during the switch - on is well described by means of a model including optical coherences ( intrinsic anisotropy ) and extrinsic linear anisotropies .
these results extend our previous work in which the coherences were simplified as a parametric cross coupling between the matter polarization and electric fields @xcite , and provide a quantitative comparison between the amount of anisotropy due to the coupling of molecular angular momenta and that induced by an intracavity window .
furthermore , the decay rate of the coherences , defined in ref .
@xcite , is shown explicitly to be closer to the decay rate of the population inversion rather than to the decay rate of the induced polarization .
the experiment has been performed using an unpolarized fabry - perot cavity as shown in fig .
[ setup ] .
a total reflective flat mirror ( m@xmath2 , reflectivity r@xmath2=1 ) and an outcoupler mirror ( m@xmath0 ) with a reflectivity @xmath3 set the cavity length at @xmath475 cm . a piezo translator ( pzt ) is used to select the p(20 ) laser emission line and to adjust the laser detuning .
the active medium , a mixture of @xmath5 ( 82@xmath6 ) , n@xmath0 ( 13.5@xmath6 ) and co@xmath0 ( 4.5@xmath6 ) at a pressure of 25 mbar , is pumped by a dc discharge fixed at 6.1 ma when the threshold current is 3 ma .
as we are interested in the transient dynamics , an intracavity chopper is used to induce a switch - on event at a repetition rate of 200 hz . in order to control the linear anisotropies of the cavity ,
a znse anti reflection coated window ( w ) has been introduced , which can be tilted accurately as shown in fig.[setup ] .
the polarization state of the laser emission has been analyzed by means of a wire grid polarizer ( pol ) , which has the property of reflecting one linear polarization of the incident radiation and transmitting the orthogonal one , with an extinction ratio of 1:180 .
the reflected and transmitted parts of the beam are directed to two hgcdte fast detectors ( d@xmath2 , d@xmath0 , 100 mhz bandwidth ) , whose sensitive areas ( @xmath7 @xmath8m@xmath9 ) are much smaller than the beam size .
both signals are recorded on a digital oscilloscope ( lecroy lt423l ) with @xmath10 mhz bandwidth . in the stationary regime
, we observe that the laser has two possible linear polarization directions which are orthogonal as far as we can measure . in fig .
[ setup ] these cavity eigendirections are called h and v respectively .
the window w can rotate by an angle @xmath11 around the axis v. when @xmath12 , the polarization direction is determined by the detuning of the cavity , as usual @xcite .
this is illustrated in fig.[detuning_scan](a ) , where we plot the laser intensity along both eigendirections when the detuning is varied around the line center .
a residual hysteresis around the transition is masked by the limited resolution of pzt . at the transition point near the line center
there is not a preferred polarization direction and the laser flips spontaneously from one to other as shown in fig .
[ flipflop ] .
our goal is to study the polarization behavior in this bistable region . in particular , we analyze of the switch - on transient state , in which the total intensity presents relaxation oscillations . in our system
the pump strength imposes a relaxation frequency of 55 khz .
when the two polarized components are separated , they show oscillations in relative antiphase , which do not appear on the total intensity ( fig.[monomode ] ) .
the amplitude of these oscillations depends on the angle between the polarizer axes with respect to h - v , reaching a maximum when the analysis is performed at 45@xmath13 . in fig.[monomode](a )
we report an example of these polarization oscillations , when the cavity is tuned at the center of the line .
the total laser output intensity ( thick solid line ) is displayed together with its two orthogonally polarized components analyzed at 45@xmath13 ( thin solid and dashed lines ) .
these oscillations are always damped until they disappear , with a damping rate depending on the cavity detuning .
precisely the oscillations are more persistent the closer is the detuning to the bistable region .
this fact points to the competitive origin of the oscillations . in fig.[monomode ] ( c ) we show an example of the transient dynamics when the cavity detuning is slightly moved from resonance .
it is worth to note that the system is so sensible to noise that the duration of the oscillations suffers slight variations , without apparent changes in the conditions .
successively , we studied the response of the system to small linear anisotropies driven by the tilt angle @xmath11 of the intracavity window w around the vertical axis .
several parameters of the system are simultaneously affected by this action . as a consequence of the common increasing of the optical path for both polarization eigendirections , the line center moves with respect to the isotropic condition @xmath12 . in order to keep the bistable condition ,
the cavity length is varied to recover the resonance . in fig.[detuning_scan]((b ) and ( c ) ) we show the intensity of both polarization components for different @xmath11 values , showing that the gain profile is not significantly distorted with respect to the isotropic condition @xmath12 ( fig.[detuning_scan](a ) ) .
the detuning anisotropies are measured to be smaller than 5@xmath6 of the total change .
the other effect of the tilt angle @xmath11 is to increase the cavity losses . in fig.[angle ] ( b ) we report the dependence of the laser intensity on the tilt angle @xmath11 which is in agreement with the expected theoretical behavior @xcite .
moreover , since the gain profiles for the two polarization directions do not change due to the tilt angle @xmath11 ( fig .
[ detuning_scan ] ) , we can assume that the induced loss anisotropies remain a small effect ( @xmath14 ) compared to the total loss change
. the effect of the losses anisotropy can be observed in the residual modulation in the total intensity ( figs.[monomode ] ( a),(c ) ) .
the dynamical effect observed when @xmath11 is increased is the growing of the frequency of the polarization oscillations .
measurements made near the resonance condition show that the frequency rises as the tilt angle increased(fig.[angle ] ( a ) ) .
it can be seen that a tilt angle of 12@xmath13 is enough to rise the frequency from 100 khz to 400 khz .
our theoretical approach is based on the theory of the isotropic laser developed in ref .
@xcite where the optical coherences between upper levels are considered with the .
this theory was developed for the simplest case ( j=1 @xmath15 j=0 ) , while the transition involved in our system is much more complicated ( j=19 @xmath15 j=20 ) .
however , this theory has been showed to predict also the behavior of lasers with a different level structure , as shown in ref .
@xcite . only first order coherences ( @xmath16 ) will be considered . therefore , independently of the number of sublevels
, there are only two kind of possible transitions , which generate a split of the population in two ensembles , in such a way that an anisotropy is induced in the active medium @xcite .
furthermore we introduce an extrinsic linear anisotropy as done in ref.@xcite .
the field will be decomposed in a circularly polarized base .
just losses and linear detuning anisotropies along the principal axes of the system will be included , but not circular asymmetries since our system does not show signs of dicroism .
it reads as @xmath17 , \nonumber \\ \dot{p_l } & = & -\gamma _
{ \perp } [ p_l- d_l e_l - e_r c^*],\\ \dot{c } & = & -\gamma _ c c - \frac{\gamma _ \parallel}{4 } ( e_l ^ * p_r + e_r p_l^ * ) , \nonumber \\ \dot{d_r } & = & -\gamma _ { \parallel } [ d_r -r + \frac{1}{2}(e_r p_r^{*}+e_r^ { * } p_r ) + \frac{1}{4 } ( e_l p_l^{*}+e_l^ { * } p_l)],\nonumber \\ \dot{d_l } & = & -\gamma _ { \parallel } [ d_l -r + \frac{1}{2 } ( e_l p_l^{*}+e_l^ { * } p_l)+\frac{1}{4}(e_r p_r^{*}+e_r^ { * } p_r ) ] \nonumber\end{aligned}\ ] ] where @xmath18 are the slowly varying electric fields in the circular basis .
stand for the matter polarization fields , @xmath20 and @xmath21 are the respective population inversions .
the field @xmath22 represents the coherence between the upper sublevels .
we recall that , in a density matrix treatment , the polarization correspond to off diagonal matrix elements between upper and lower level of the radiative transition , whereas c is proportional to the of diagonal matrix elements coupling different angular momentum states of the upper level @xcite .
the parameter @xmath23 represents the detuning between the cavity and the atomic transition frequencies .
the parameters @xmath24 and @xmath25 represent respectively the linear anisotropies in the losses and detuning with respect to the cavity h - v axes , where @xmath26 , @xmath27 , are the losses in the horizontal and vertical axis , and @xmath28 , @xmath29 are the corresponding detunings . in our low pressure
co@xmath0 laser , the polarization decay is @xmath30 s@xmath31 and the inversion decay rate as @xmath32s@xmath31 @xcite . in view of numerical solutions , there is no reason to keep dynamical equations for variables whose decay rates differ by several orders of magnitude . depending on whether @xmath33 or @xmath34 , we can respectively reduce the number of differential equations from 7 to 5 ( for e@xmath35 , e@xmath36,c , d@xmath35 and d@xmath36 ) or 4 ( for e@xmath35 , e@xmath36 , d@xmath35 and d@xmath36 ) . in the first case , p@xmath35 , p@xmath36 are eliminated adiabatically and their values correspond to the stationary solutions of the third and fourth of eqs .
[ mb ] . in the second case ( fast c ) , also the c variable is eliminated adiabatically and it follows the relatively slow variations of the field and the population inversion . whether to choose between five or four equations depends on the numerical value of @xmath37 .
the @xmath38 parameter represents the coherence decay rate , whose value should be chosen between @xmath39 and @xmath40 @xcite
. however , this parameter can not be directly measured , and it will be used as control parameter in order to fit the theory to the experimental results @xcite .
we find that the optimal value is @xmath41 in all cases , which is also consistent with the observation that just linearly polarized states are found in the experiment . indeed
, a higher value of @xmath38 would give rise to a periodic modulation of the total intensity @xcite which has never been observed in our experiments .
hence , the coherence c decays on the same time scale as the population inversion and we must keep the corresponding time dependent equation . thus our experimental system is dynamically modeled as a class b laser @xcite , and eqs .
( 1 ) reduce to the following system of two algebraic equations : @xmath42 and five differential equations : @xmath43 , \nonumber\\ \dot{d_l } & = & -\gamma _ { \parallel }
\left[d_l(1+|e_l|^2 ) -r+\frac{1}{2}\left(d_r|e_r|^2+\frac{3}{2}(e_re_{l}^{*}c^*+e_{r}^{*}e_lc)\right)\right ] .
\nonumber\end{aligned}\ ] ] eqs .
( [ adiabatic ] ) show explicitly that the cross field coupling is due partly to a molecular coupling @xmath1 ( intrinsic ) and partly to the intracavity window w ( extrinsic ) .
[ no_coh ] provides evidence that the main responsible for the alternation is the intrinsic anisotropy rather than the extrinsic one .
figure [ no_coh ] ( a ) refers to the case where both coherences and linear anisotropies are acting . in fig .
[ no_coh ] ( b ) , the extrinsic anisotropy has been removed .
now the oscillation competition is not damped , but the alternation of the two polarization is maintained . finally , in fig .
[ no_coh ] ( c ) just the extrinsic anisotropies are acting ; the competition does not appear any more . from the experimental observations we know that the most important effect of the window tilting is the increase of the total loss , while the losses and detuning anisotropies represent a minor effect .
in addition , as the excitation current does not change during the experiment , the increasing of the losses affects the gain to loss ratio . as a consequence , the total losses @xmath44 and pump parameter @xmath45 are functions of @xmath11 as follows @xcite : @xmath46 where @xmath47 and @xmath48 s@xmath31 are the pump strength and total losses when the tilt angle @xmath12 . here @xmath49 is the speed of light , @xmath50 m the cavity length and @xmath51 the mirrors reflectivity .
the parameters @xmath52 and @xmath53 are related to the reflectivity of the window w , and are chosen to fit the experimental results . in fig .
[ monomode ] ( b ) and ( d ) the numerically generated intensity profiles are compared with their experimental counterparts reported in fig [ monomode ] ( a ) and ( c ) , respectively .
it can be observed that the intensity profiles show antiphase oscillations for both polarization components , while the total intensity remains unmodulated .
when @xmath23=0 , the polarization oscillations remain undamped for any degree of anisotropy . for @xmath54 , the oscillations are still undamped only in perfect cavity symmetry conditions @xmath55 . in fig.[monomode ] ( d ) it can be seen that for @xmath56 , a detuning or losses anisotropy of 0.5 @xmath57 is sufficient to damp the oscillation in a few hundred microseconds as observed in the experiment . in the experimental system
unavoidable residual anisotropies break the cylindrical symmetry , and therefore the polarization oscillations are always damped . choosing the resonant condition @xmath58
, we observe that the frequency of the oscillations presents the same dependence on the total losses and pump strength found in the experiment . in fig .
[ angle ] we compare the experimental and numerical angular dependences of the polarization oscillation frequency ( fig .
[ angle ] ( a),(c ) ) and the total intensity ( fig .
[ angle ] ( b),(d ) ) , showing good agreements .
in this work we have reported an experimental and theoretical study of the transient polarization dynamics of a quasi - isotropic co@xmath0 laser . a competition between two modes with orthogonal polarization directions is observed .
the competition manifests itself as an oscillation on the polarized laser field , whose features depend on the cavity parameters . to explain the observed transient polarization alternation , both intrinsic and extrinsic anisotropies must be considered ; the intrinsic ones provide the alternation and the extrinsic ones provide damping and residual ripple of the total intensity .
the numerical results fit well with the experimental observations , pointing out the importance of matter variables in the dynamics of a quasi - isotropic laser . | we study experimentally and theoretically the polarization alternation during the switch - on transient of a quasi - isotropic co@xmath0 laser emitting on the fundamental mode .
the observed transient dynamics is well reproduced by means of a model which provides a quantitative discrimination between the intrinsic asymmetry due to the kinetic coupling of molecules with different angular momenta , and the extrinsic anisotropies , due to a tilted intracavity window .
furthermore , the experiment provides a numerical assignment for the decay rate of the coherence term for a co@xmath0 laser . |
the goal of this paper is to present a fully relativistic calculation of the two - particle two - hole ( 2p-2h ) contributions to the inclusive @xmath0 response functions of nuclei for intermediate to high momentum transfers in a fermi gas model .
consistency with perturbation theory is maintained and all diagrams with one - pion exchange in the nuclear current are considered , constructed by attaching a photon to all possible lines in the basic one - pion exchange feynman diagram . in this way not only meson - exchange currents ( mec ) arise ( for example , where the photon is attached to the pion ) , but also pionic correlation diagrams , where the virtual photon is absorbed by one of the two interacting nucleons .
both kinds of diagrams are considered in our model , together with the usual virtual @xmath1-isobar electroexcitation and decay .
we are motivated by previous work presented in @xcite , where only the mec were included in the 2p-2h transverse ( @xmath2 ) response , together with earlier work both in non - relativistic @xcite and relativistic @xcite regimes .
the contribution found from the 2p-2h excitations is small at the quasielastic ( qe ) peak , and increases with energy transfer , being more important in the dip region , where it is dominated by the @xmath1 current . at the non - relativistic level attempts
were also made to evaluate the 2p-2h contribution of mec in the @xmath2-response for finite nuclei in a shell model @xcite .
the mec are not the only two - body operators able to induce 2p-2h excitations .
the correlation operators arising from feynman diagrams where the photon is attached to a nucleon line , exchanging a pion with another nucleon , are of the same order as the mec in the perturbative expansion and should be included to be consistent @xcite .
these diagrams , however , present the problem of giving an infinite answer in a fermi gas model .
the reason is that there is a nucleon propagator that can be on - shell in the region of the quasielastic peak .
since the response function is the square of the amplitude , the resulting double pole gives an infinite result after integration . in dealing with this problem , in
@xcite a prescription was followed by keeping the lines with a nucleon propagator strictly off the mass shell .
a different approach was taken in @xcite by subtracting from the proper self - energy its value on the mass shell , with the unphysical shortcoming of obtaining negative results for the 2p-2h responses to the left of the qe peak .
finally , in @xcite a nucleon self - energy in the medium was introduced in the nucleon propagator . in dealing with the seven - dimensional integrals appearing in the 2p-2h responses , some of the previous calculations have resorted to the approximation of setting the two hole momenta both equal to zero in some of the diagrams @xcite or by taking into account only an average nucleon momentum @xcite . in this work
we revisit the double - pole problem to analyze the nature of the divergence of the resulting contributions . by isolating the divergent terms we are able to link them to the infinite extension of the fermi gas system .
in fact the double - pole term can be related to the probability of one - nucleon emission followed by nucleon re - scattering off another nucleon , with the final ejection of two particles .
this probability is infinite , since it is proportional to the propagation time of a real nucleon in a fermi gas .
this fact was pointed out in @xcite where it was cured , as mentioned above , by introducing a nucleon self - energy with an imaginary part giving it a finite lifetime for collisions . in this paper
we use a similar procedure by introducing a finite imaginary part @xmath3 in the nucleon propagator , but with a new meaning for the free parameter @xmath4 . instead of being an imaginary part of the nucleon self - energy for collisions , we relate it to the time @xmath2 that a nucleon can travel across the nucleus before leaving it . hence this term accounts for the finite size of a real nucleus in contrast to an infinite system like the fermi gas , where @xmath2 is infinite .
the value of @xmath4 can be estimated to be roughly about 200 mev , appreciably larger than the usual values of the nucleon width for collisions .
the structure of this work is as follows . in sect . ii we present our model and define the 2p-2h response functions and the two - body current operators .
we discuss in depth the divergence of the correlation diagrams and the need to introduce the parameter @xmath4 in sect .
iii ( details of the numerical calculation are given in the appendices ) . in sect .
iv we present results for the 2p-2h longitudinal and transverse response functions . in the case of the correlation
diagrams we present results for several values of the parameter @xmath4 . finally , in sect .
v we present our conclusions .
we consider an electron that scatters off a nucleus transferring four - momentum @xmath5 , with @xmath6 the energy transfer and @xmath7 the momentum transfer .
we follow closely the notation of @xcite . assuming plane waves for the electron , working in the laboratory system and taking the @xmath8 direction along the momentum transfer
, the inclusive cross section is written as @xmath9 \
, , \ ] ] where @xmath10 is the mott cross section , @xmath11 and @xmath12 are the lepton kinematic factors , and the relevant quantities are the longitudinal @xmath13 and transverse @xmath14 response functions , respectively .
these are defined as the following components of the hadronic tensor , @xmath15 where @xmath16 and @xmath17 is the nuclear current operator . in this paper
we take the initial nuclear state as the relativistic fermi gas ( rfg ) model ground state , @xmath18 , with all states with momenta below the fermi momentum @xmath19 occupied .
the sum over final states can be decomposed as the sum of one - particle one - hole ( 1p-1h ) plus two - particle two - hole ( 2p-2h ) excitations plus additional channels . in the impulse approximation
the 1p-1h channel gives the well - known response functions of the rfg .
here we focus on the 2p-2h channel where the final states are of the type @xmath20 , where @xmath21 are momenta of relativistic final nucleons above the fermi sea , @xmath22 , with four - momenta @xmath23 , and @xmath24 are the four - momenta of the hole states with @xmath25
. the spin indices are @xmath26 and @xmath27 .
since we have two species of nucleons , the 2p-2h responses can be further decomposed as the sum of two - proton ( pp ) , two - neutron ( nn ) and proton - neutron ( pn ) emission @xmath28 for the pp channel we write down the l response as ( likewise for the t response ) : @xmath29 where the spin indices are implicit in the matrix elements .
the factor @xmath30 comes from anti - symmetry of the wave functions , to avoid double counting of the final states under the interchange of the indices @xmath31 and @xmath32 . exploiting the anti - symmetry ,
the many - body matrix element of a two - body operator can be written as the direct minus exchange part of the two - body current matrix element @xmath33 to be specified below , @xmath34 going to the thermodynamic limit and integrating over the momentum @xmath35 we obtain @xmath36 where @xmath37 , and the integration limits are @xmath38 , @xmath39 .
we have defined the anti - symmetrized current function @xmath40 with obvious meaning for the abbreviated arguments . expanding the square inside the integral in eq .
( [ integral ] ) , three terms are obtained : @xmath41 changing variables @xmath32 in the second term under the integral , we obtain the first term again .
hence we can finally write for the pp response @xmath42 \nonumber\\ & & \kern -1 cm \mbox{}\times\delta(e'_1+e'_2-\omega - e_1-e_2 ) \theta(p'_2-k_f ) \ , . \label{rl}\end{aligned}\ ] ] note that the factor @xmath43 in front of the sum comes from the anti - symmetry of the particles ( protons ) .
a similar expression is obtained for the nn response @xmath44 . in the case of the pn channel
we subtract the charge exchange contribution without any symmetry term because there are no additional isospin sums , and the result is @xmath45 finally , note that the 2p-2h response is proportional to the volume of the system v which is related to the number of particles @xmath46 ( protons or neutrons ) by @xmath47 .
the mec considered in this work are represented by the feynman diagrams of fig
the pionic four - momenta @xmath48 , @xmath49 are defined via @xmath50 as the four - momenta given to the nucleons 1 and 2 , respectively , by the exchanged pion . assuming pseudo - vector nucleon - pion coupling , the fully relativistic two - body current matrix elements are given by @xcite : * ( a - b ) seagull or contact : @xmath51 * \(c ) pion - in - flight : @xmath52 in the above we use the einstein convention for the sum over a repeated isospin index @xmath53 .
moreover , @xmath54 and @xmath55 are the electromagnetic isovector nucleon and pion form factors , respectively .
the spinors are normalized according to the bjorken and drell convention @xcite and the pion - nucleon coupling constant is @xmath56 . *
( d - g ) delta current : @xmath57 the vector @xmath58 is related to the pion electroproduction amplitude @xmath59 the forward @xmath1 electroexcitation tensor is @xmath60 \gamma_5 \nonumber\\ & - & \theta^{\rho\nu}q_\nu \left [ g_1 \gamma^\mu -g_2 h_1^\mu + g_3q^\mu \right ] \gamma_5\end{aligned}\ ] ] and the backward tensor amplitude is @xmath61 \theta^{\mu\rho } \nonumber\\ & - & \gamma_5 \left [ g_1\gamma^\mu -g_2p'_1{}^{\mu } -g_3 q^\mu \right ] q_\nu\theta^{\nu\rho } \ , .\end{aligned}\ ] ] the tensor @xmath62 is defined by @xmath63 for the @xmath1 propagator we use the usual rarita - schwinger ( rs ) tensor @xmath64 \ , .\end{aligned}\ ] ] in what follows we perform the substitution @xmath65 in the denominator of the propagator to account for the @xmath1 decay probability .
finally , the electromagnetic coupling constants @xmath66 are given by @xmath67 our approach for the @xmath1 follows , as a particular case , from the more general form of the @xmath68 lagrangian of pascalutsa _ et al . _
the @xmath1 coupling constants used here are @xmath69 , @xmath70 , @xmath71 , and @xmath72 .
the correlation current is defined in fig . 2 , and given by @xmath73u({{\bf p}}_2 ) \nonumber\\ & & \kern -2 cm \mbox{}+ ( 1\leftrightarrow2)\ , , \label{correlation}\end{aligned}\ ] ] where @xmath74 is the feynman propagator for the nucleon @xmath75 and @xmath76 is the electromagnetic nucleon vertex , @xmath77
the nucleon form factors @xmath78 and @xmath79 are given by the galster parametrization @xcite .
the isospin sums and isospin matrix elements must be performed separately for each isospin channel .
explicit expressions are given in appendix a.
the response functions computed using the correlation current in eq .
( [ correlation ] ) are divergent in the fermi gas .
there are two sources for this divergence : the first one comes from the double pole of the propagator when taking the square of the current .
this divergence can be shown to behave as @xmath80 plus principal value terms going as @xmath81 .
the second source is related to the behavior of the principal values arising from the double and single poles near the rfg boundary of the quasielastic peak , where the principal values present a logarithmic divergence . to illustrate the mathematical structure of this divergence we isolate as an example the singularities produced by the diagram of fig .
the corresponding current operator can be written as @xmath82 where @xmath83 is the on - shell energy .
we have explicitly extracted the divergent part of the denominator , with a pole for @xmath84 in the limit @xmath85 .
the above equation is equivalent to the quasielastic condition for emission of an on - shell nucleon with four - momentum @xmath86 .
in fact , for a given value of @xmath87 , eq .
( [ condition ] ) holds when the angle between @xmath88 and @xmath7 is given by @xmath89 since the condition @xmath90 defines the boundary of the quasielastic peak , the pole can always be reached in that region . to study the behavior of the response functions due to this pole , it is convenient to change the variable @xmath91 to a new variable defined by @xmath92 in the integral over @xmath88 in eq .
( [ rl ] ) .
then the components of the total current matrix element can be written as a function of @xmath93 in the general form @xmath94 where the first term comes from diagram 2(a ) and the function @xmath95 comes from the sum of the remaining diagrams , and is finite for @xmath96 . since the current appears squared in the response function , we are dealing with the integral of a function of the kind @xmath97 when integrating this function over @xmath93 , and taking the limit @xmath85 , the first term has a double pole for @xmath96 , while the third one has a single pole . to deal with the single pole we use the usual plemeli relation , @xmath98 to apply a similar relation for the double pole term , we add and subtract the on - shell value @xmath99 . taking the limit @xmath85 we can use relations which are valid for any function @xmath100 @xmath101 and @xmath102 \psi(0 ) \sim \frac{\pi}{\epsilon}\psi(0 ) \ , .\ ] ] then eq .
( [ square ] ) can be written in the form @xmath103 the last @xmath104 term in eq .
( [ expansion ] ) provides the dominant contribution to the response function , being infinite for @xmath85 . due to the @xmath105 function
, that term does not contribute outside the quasielastic - peak region , where @xmath93 is different from zero .
the principal values present in eq .
( [ expansion ] ) also diverge in the particular case in which one of the limits of integration is zero . in that case
the principal value in eq .
( [ dirac ] ) should be computed instead using @xmath106 and it gives a @xmath107 term if @xmath108 or @xmath109 is zero .
that situation in fact occurs throughout the quasielastic region , and in particular at the boundary of the quasielastic peak .
therefore one expects an additional divergence @xmath110 .
the meaning of the term @xmath111 is explained in what follows .
diagram 2(a ) , when the intermediate nucleon is on shell , gives the probability of a 1p-1h electroexcitation times the probability of quasielastic nucleon scattering .
since the interaction probability is proportional to the interaction time @xmath2 , the probability of this re - scattering process is proportional to @xmath112 . therefore the cross section is proportional to @xmath2 . in an infinite system such as the fermi gas , the intermediate nucleon never leaves the nucleus and therefore @xmath113 .
however , in a finite nucleus one expects no divergence because a high - energy nucleon will leave the nucleus in a finite time
. therefore the interaction time is finite .
the relation between @xmath4 and @xmath2 can also be obtained by inspection of the momentum - space propagator in quantum field theory @xcite , computed as the vacuum expectation value of time - ordered fermion fields .
the value @xmath4 in the denominator of the propagator can be seen as a regularization parameter in the fourier transform of the time step function for a particle with four - momentum @xmath114 @xmath115 where @xmath116 and @xmath117 . for a real particle , @xmath118 ,
the left - hand side of the above equation is @xmath119 , and the right - hand side is @xmath80 .
therefore @xmath120 this can be obtained alternatively by replacing the on - shell value of the propagator in eq .
( [ propagator ] ) as a delta function @xmath121 and using the integral representation @xmath122 in this paper we cure the divergence of the correlation diagram by a regularization procedure , using a finite value for @xmath4 to account for the finite propagation time of a high - energy nucleon in a nucleus before leaving it . to estimate the value of @xmath4 for a nucleus such as @xmath123c
, we assume that the nucleon moves at the velocity of light and it has to cross a distance equal to the nuclear radius @xmath124 fm .
then @xmath125 note that this value , @xmath126 mev , is very different from the nucleon width @xmath127 mev which is usually obtained in nuclear matter as the width for nuclear inelastic interaction . in practice
the value of @xmath4 can be taken as a parameter to be fitted to data . in the next section
we perform a study of the dependence of our results upon @xmath4 . unless otherwise specified we assume @xmath128 mev . at this point
we should mention that the use of eq .
( [ square ] ) to compute the 2p-2h response functions becomes impractical due to complications in the numerical calculation of principal values in multidimensional integrals including the four diagrams of fig . 2 ( and the corresponding exchange parts ) .
since we are forced to use a finite value of @xmath4 , it becomes more convenient to keep from the beginning the @xmath3 term in the denominator of the nucleon propagator in eq .
( [ correlation ] ) .
here we present results for the longitudinal and transverse response functions for inclusive two - particle emission .
we compute the 2p-2h response functions in the rfg model as the 9-dimensional integrals given by eqs .
( [ rl],[rlpn ] ) . the energy delta function can be used to integrate over @xmath129 , fixing the energy @xmath130 of the first particle .
more details are given in appendix b. by rotational invariance considerations , one of the the azimuthal angles can be fixed , multiplying at the same time the result by a factor @xmath131 .
we choose @xmath132 . at the end
we have a 7-dimensional integration to be performed numerically .
the usual procedure is to use a multi - dimensional montecarlo ( mc ) integration .
since the pole structure of the integrand is numerically delicate , in this work we use instead a mixed montecarlo - simpson integration procedure .
the simpson algorithm is used for integration over the angles of the two holes @xmath133 and of the first particle @xmath134 . the remaining 4-dimensional integral over the hole momenta @xmath135 and their angles @xmath136
is made by montecarlo . to keep the cpu times manageable we use a number of mc points of the order of @xmath137 for @xmath138 gev / c .
for other values of the momentum transfer the number of mc points is modified linearly with @xmath139 .
we have performed a study of the stability of the results with the number of mc points and have found that the error from the integration procedure is within a few percent . a pion - nucleon form factor is included in the 2-body currents : @xmath140 , with @xmath141 gev .
we use the same value for the @xmath142 form factor in the delta current .
the electromagnetic form factors are those of galster for the nucleon , and those used in @xcite for the mec .
to make contact with previous work , we apply our model to compute the 2p-2h longitudinal and transverse response functions for the nucleus @xmath143fe , and for momenta @xmath144 and @xmath145 mev / c .
the results are presented in figs . 36 , where the separate contributions of the correlation and mec currents to the 2p-2h responses are also shown .
the 1p-1h responses produced by the one - body ( ob ) current in the relativistic fermi gas ( rfg ) without interaction are also shown .
a critical input for our model is the value of the parameter @xmath4 in the nucleon propagator , introduced to cure the divergence of the double pole . to see how the responses for the correlation contribution depend on @xmath4
we show results for three different values : @xmath146 and 300 mev . for @xmath147 mev
, the correlation 2p-2h contribution presents a shape with a maximum in the region of the quasielastic peak , but with a long tail extended to high transferred energies .
the maximum is reminiscent of the pole structure of the nucleon propagator , and therefore a resonance appears for kinematics corresponding to the quasielastic condition in eq .
( [ quasielastic ] ) . a shift to higher energies ( of the order of @xmath148 mev )
is seen in the case of @xmath144 mev / c ( figs . 3 , 5 ) . indeed for this value of @xmath139 the phase space for two - particle emission causes a suppression of the low - energy side of the response function .
the resonant structure produced by the 2p-2h correlation contribution diminishes significantly with increasing values of the parameter @xmath4 .
notice that for @xmath149 mev there is no maximum located at the qe peak .
for an even lower value of the escape width , say @xmath150 mev , the magnitude of the resonant peak is of the same size as the ob response function .
this correction coming from 2p-2h states is obviously too large to be compatible with experimental data that are already of the order of the 1p-1h response at the region of the qe peak .
it should be mentioned that , although the 2p-2h contribution should be added to the 1p-1h one , the latter should be first corrected for final - state interaction ( fsi ) contributions not included in the bare rfg results shown in the figures . in fact
fsi contribute importantly to one - nucleon emission through the coupling of 1p-1h to 2p-2h states in the final nucleus @xcite .
these processes involve , in particular , two - pion exchange , and are therefore of the same order as the 2p-2h response in the perturbative series , since it is the square of one - pion exchange matrix element .
the inclusion of such contributions is out of the scope of the present study .
the dependence of the correlation responses on the parameter @xmath4 is better appreciated in figs . 7 and 8 , where we show its contribution for the three chosen values of @xmath4 in the same plot .
in the qe region the height of the responses approximately reduces to one half when @xmath4 doubles .
this behavior follows because of the leading @xmath80 dependence in eq .
( [ expansion ] ) , coming from the pole in the propagator . for high @xmath6
the results are more similar and they are almost independent of @xmath4 in the high - energy tail . in this case , _
i.e. _ large @xmath6 , there is no pole in the integrand and the contribution from the propagator is less sensitive to the precise value of @xmath4 .
let us return now to figs . 36 , where the mec separate contribution is also shown .
the transverse response ( figs . 3 , 4 ) has a large peak with a maximum around @xmath151 , that comes from the @xmath1 propagator appearing in the @xmath1-current .
it has the same resonant structure as the correlation current , but located in the region of the @xmath1 peak , where the real pion emission cross section has a maximum .
we do not include the pion emission channel in our calculation .
both channels should be summed up to obtain the total inclusive @xmath0 cross section .
the @xmath1 peak is very small in the longitudinal response presented in figs . 5 and 6 .
this is consistent with the predominant transverse character of the @xmath1 current , hence providing a small contribution to the longitudinal channel . for @xmath144
mev the mec 2p-2h contribution is large ( small ) in the t ( l ) response .
however , for @xmath152 mev ( figs . 4 , 6 ) we find a larger effect in @xmath153 coming from the mec seagull and pionic at large energy transfer .
indeed in a non - relativistic expansion in powers of @xmath154 the time component of the mec is of higher order than the transverse one . however , for @xmath152 mev , @xmath154 is larger than one , and the relative l component of the mec , compared to the t one , starts to increase . in the case of the correlation
current , we observe that its contribution , compared with the ob responses , is similar in the t and l channels .
note that in the correlation current ( fig .
2 ) the photon couples directly to a nucleon with the same interaction vertex @xmath155 as the ob current . the other side of the diagram with a pion coupled to a second nucleon is independent of the particular component of the current .
the separate effects of the different currents contributing to the 2p-2h transverse responses are shown in figs .
as shown , the seagull plus pionic ( spp ) currents alone give a small effect compared with the contributions from the @xmath1 and correlations .
in fact , for @xmath128 mev the correlation response is much larger ( by a factor 2 or 3 ) than the spp response function ( middle panels in figs . 9 , 10 )
. we also observe that the separate seagull contribution is larger in magnitude than the pionic one , which is negligible for @xmath152 mev / c .
note that the two currents interfere destructively and partially cancel when both are considered in the spp responses ( bottom panels ) . in fig .
11 we show the transverse response obtained by adding the total 2p-2h contributions to the ob response .
a word of caution should be raised when analyzing these results .
first , we have not added the correlation nor mec corrections to the 1p-1h response . moreover ,
the two - pion - exchange interaction generates self - energy corrections to the ob current that lead to interference effects of the same order in the expansion as the corrections included here . as an example , fsi are known to redistribute the strength of the responses , producing a hardening , a reduction of the maximum and an increase of the high - energy tail @xcite . recently also
a large effect from both mec and fsi has been found in the 1p-1h channel for high momentum transfer @xcite , which should be added to the present results .
finally , the process of real pion emission ( not included here ) gives also a contribution in the transverse response located mainly the region of the delta peak .
so far we have presented results for intermediate to high momentum transfer .
results for lower values of @xmath156 and 410 mev / c are shown in figs .
12 and 13 for the t and l response functions .
this allows us to compare the present results with previous non - relativistic calculations @xcite . in fig
12 the structure function @xmath157 is presented , to allow a direct comparison with figs .
11 and 12 of @xcite .
the separate mec and correlation contributions to the 2p-2h t response shown in fig .
12 are similar to the ones presented in @xcite .
the mec produces a tail above the qe peak that increases with the energy transfer .
the presence of correlations lead to an additional , significant raise of the tail .
note that our correlation results are obtained for @xmath128 mev . in @xcite
another prescription to deal with the nucleon pole was adopted . from our results
we conclude that both prescriptions are compatible numerically .
the ob response of @xcite included rpa correlations producing a reduction and hardening of the ob response .
the 2p-2h longitudinal responses were not computed in @xcite , since the time components of the mec are of higher order in the non - relativistic reduction and hence , they were expected to be very small . however , our prediction for the correlation 2p-2h contribution in the l response , presented in fig . 13 , shows a similar effect as in the t response , _
i.e. , _ a tail also appears for high energy transfer in the l response coming from correlations .
contrary to the t channel , mec give no contribution in the l response .
since the 2p-2h excitation is produced in this work by one - pion exchange , the results are strongly dependent on the details of this particular interaction .
this is illustrated in fig .
14 where we show how the results depend on the strong @xmath158 form factor for @xmath152 mev / c .
the results without a form factor , _
i.e. , _ with @xmath159 , are about three times as large as the results with the form factor .
this is different from the findings at low momentum transfer @xcite , where the pion form factor can be safely ignored .
another issue is the dependence of the results on the @xmath1 form factors used in this work , both the electromagnetic and the strong ones , which are somewhat different from the parametrization used in the 2p-2h mec calculation of @xcite .
calculations done with both sets of parameters are compared in fig . 15 .
our calculation gives a larger contribution for the t response than the one of @xcite .
hence the use of the same form factors reduces the discrepancy between the two calculations .
some of the remaining differences could be linked to other details of the models , in particular to the different lagrangian chosen for @xmath1 electroexcitation .
we should note that the two models are fully independent .
while all the spin sums are performed analytically in @xcite resulting in thousands of terms to be numerically integrated , in this work we first compute the spin matrix elements of the current and later we evaluate the squares and perform the sums numerically . before concluding
, we would like to stress that the 2p-2h responses in the present model are crucially dependent on details of the pion interaction .
a critical ingredient of the model is the value of the parameter @xmath4 , identified with an escape width of a high - energy nucleon from the nucleus .
we have proven that a value @xmath160 mev leads to results in agreement with the previous calculation of @xcite .
this parameter @xmath4 is different from the usual interaction width of particle states , usually associated with matrix elements of the phenomenological imaginary optical potential derived from elastic scattering data @xcite .
it has also been computed in nuclear matter in a semi - phenomenological approach @xcite .
the resulting width for 100 mev nucleons is of the order of 10 mev , which is too small to give reasonable results in our calculation .
this is due to the @xmath80 behavior of the 2p-2h response divergence in the qe region , where the pole is being hit . due to this divergent behaviour
, for @xmath161 mev the results would be almost one order of magnitude larger than the ob responses at the maximum .
we have checked that the the @xmath80 term in the forward diagrams is the main contribution to the 2p-2h correlations in the qe region for @xmath162 mev .
the importance of correlations , for the same value of @xmath4 , increases with the nuclear mass .
we have checked that for the case of @xmath123c where the sizes of the correlation responses , relative to the ob , are about 20% smaller than for @xmath143fe .
this is what one would expect , since the number of correlated pairs increases with @xmath163 .
moreover , since the estimated value of @xmath4 depends on the nuclear radius , eq .
( [ epsilon ] ) indicates that one should use larger @xmath4-values for lighter nuclei , which in turn would reduce even more the size of the correlation responses .
thus we expect an important @xmath164-dependence of correlations on the nuclear responses coming from the @xmath164-dependence of the escape width @xmath4 .
a more deailed study of this issue will be presented in forthcoming work .
in this work we have presented a fully relativistic model of inclusive two - particle emission reactions induced by electrons . starting with the free relativistic fermi gas we have considered all feynman diagrams in a perturbative expansion of the scattering amplitude with one photon and one pion exchange producing 2p-2h excitations .
those diagrams can be classified in two sets , namely mec and correlation currents . in the latter
there is a nucleon propagator that can be put on shell giving a double pole from @xmath165 when taking the square of the current matrix element .
the corresponding 2p-2h response function diverges as @xmath80 when @xmath85 plus additional @xmath107 terms . giving a physical meaning to @xmath4 as the escape width of the nucleus , namely , twice the inverse of the nucleon propagation time
, the fact that the corresponding response is infinite is related to the infinite extension of the fermi gas . using a finite value of @xmath4 we account for the finite size of the nucleus , hence getting a finite result . having no way to compute @xmath4 in a fermi gas ,
we take it as a parameter . estimating in a crude way a value around @xmath166 mev , we have made an exploratory study of the results as a function of @xmath4 .
the correlation effects decrease with increasing @xmath4 .
our analysis shows that the assumption @xmath167 mev is not unreasonable , whereas for smaller @xmath4-values the correlation contribution increases significantly in the qe region . within this framework
we have studied the properties and effects of the different 2p-2h contributions and other ingredients of the model on the transverse and longitudinal response functions of @xmath143fe for intermediate to high momentum transfer .
the mec give rise to a wide peak in the region of the @xmath1 resonance that dominates the t response . in the l channel
the mec are small for low momentum transfer , but they importantly increase for high momentum above the qe peak where their contribution is of the same size as the ob longitudinal response . concerning the correlations , they add to the mec in the high - energy tail and are of the same order of magnitude .
the contribution of the correlations is similar in the l and t responses .
the main goal of this paper has been to study the effect of 2p-2h pion correlations in the l and t response , analyzing the properties of these effects as a function of a single parameter @xmath4 . in future work
we plan to investigate more physically founded ways to `` fine tune '' this parameter , including its dependence on kinematics and nuclear species .
finite - size calculations in conjunction with the use of semi - phenomenological fits of the nucleon spreading width or fits to existing @xmath0 data will also be explored .
jea thanks e. ruiz - arriola for useful discussions .
this work was partially supported by dgi ( spain ) : fis2008 - 01143 , fpa2006 - 13807-c02 - 01 , fis2008 - 04189 , fpa2007 - 62216 , by the junta de andaluca , by the infn - mec collaboration agreement , projects fpa2008 - 03770-e - infn , aci2009 - 1053 , the spanish consolider - ingenio 2000 programmed cpan ( csd2007 - 00042 ) , and part ( twd ) by u.s .
department of energy under cooperative agreement de - fc02 - 94er40818 .
in the model used in this work we compute explicitly the isospin matrix elements of the current operator in the different channels pp ( two protons ) , nn ( two neutrons ) and pn ( proton - neutron ) emission .
we first consider the channel in which we eject a pn pair . in this case
there is no symmetry in the wave function and we assume that the first hole is a proton and the second is a neutron , _
i.e. _ , the initial isospin wave function is @xmath168 .
the final state can be @xmath168 or @xmath169 depending on if there is or is not charge exchange . in the case of the mec seagull and pion - in - flight , fig .
c ) , this is the only channel which contributes .
the isospin operator is @xmath170 where repeated indices are meant to be summed .
the relevant isospin matrix element is obtained by operating over a pn state @xmath171 in the case of the correlation current we find four isospin operators for the diagrams of fig .
2 , including the isospin dependence in the single nucleon current @xmath155 , namely @xmath172 operating over the initial @xmath168 state we obtain @xmath173 in the case of the @xmath1 current , diagrams of fig . 1 ( d g ) , we find the following isospin operators @xmath174 where @xmath175 are the @xmath176 isospin transition operators verifying @xmath177 for instance we have @xmath178_z \label{iso } \\
t_3^{(1 ) } t_a^{\dagger(1 ) } \tau^{(2)}_a & = & \frac{2}{3 } \tau_z^{(2 ) } + \frac{i}{3 } \left[{\mbox{\boldmath $ \tau$}}^{(1 ) } \times { \mbox{\boldmath $ \tau$}}^{(2 ) } \right]_z \label{iso2}\end{aligned}\ ] ] and similarly changing @xmath179 .
operating over the initial @xmath168 state we obtain @xmath180 in the case of two proton emission only the @xmath1 and correlation diagrams contribute . in the case of the correlations ,
the isospin operators over the initial @xmath181 state give @xmath182 and exactly the same result for the remaining three operators .
in the case of the @xmath1 we have @xmath183 and exactly the same answer for the remaining three operators . once more only the @xmath1 and correlation diagrams contribute . in the case of the correlations ,
we have @xmath184 and the same for the remaining three operators . finally , for the @xmath1 we have @xmath185 and the same answer again for the remaining three operators .
the 9-d integral for the 2p-2h response functions is of the type @xmath186 where @xmath37 .
the delta function allows us to perform one integration analytically imposing energy conservation .
therefore , for fixed values of the two hole momenta @xmath187 and for fixed values of the two emission angles @xmath188 of particle 1 , we can integrate over the momentum @xmath129 , fixing the energy of the first particle .
to this end we change variables @xmath189 .
taking into account that both energies @xmath130 and @xmath190 depend on @xmath129 to compute the jacobian of the transformation , we obtain @xmath191 where the momentum of the final nucleon for fixed emission angles @xmath188 is obtained by solving the energy conservation equation .
this is a second degree equation with two solutions given explicitly by @xmath192 where @xmath193 @xmath194 is the final total energy , @xmath195 is the final total momentum and @xmath196 is the angle between @xmath197 and @xmath198 . to compute the integral
we add the contributions from these two solutions , corresponding to two possible final states compatible with energy - momentum conservation .
j. e. amaro , m. b. barbaro , j. a. caballero , t. w. donnelly and a. molinari , phys .
rept . * 368 * , 317 ( 2002 ) .
j. e. amaro , m. b. barbaro , j. a. caballero , t. w. donnelly and a. molinari , nucl .
phys . * a 723 * , 181 ( 2003 ) . | two - particle two - hole contributions to electromagnetic response functions are computed in a fully relativistic fermi gas model .
all one - pion exchange diagrams that contribute to the scattering amplitude in perturbation theory are considered , including terms for pionic correlations and meson - exchange currents ( mec ) .
the pionic correlation terms diverge in an infinite system and thus are regularized by modification of the nucleon propagator in the medium to take into account the finite size of the nucleus .
the pionic correlation contributions are found to be of the same order of magnitude as the mec . |
metamaterials , as synthetic composite materials with a structure such that they exhibit properties not usually found in natural materials , now form a major emerging research area that barely existed before 2000 ; in fact , the term `` metamaterial '' itself was first used in 1999 . since then
, the area has grown extensively and shows little sign of slowing down .
the key point is that materials can be designed to have , say , a negative refractive index as predicted by @xcite and later by@xcite ; and subsequently fabricated by @xcite , which is impossible in naturally occurring materials .
the first metamaterials were developed in optics and electromagnetism and relied upon having simultaneously negative permittivity and permeability , this was made physically possible using a microstructured periodic medium consisting of sub - wavelength resonators such as split - ring resonators @xcite .
it has subsequently been realised that these ideas can also be profitably utilised to create acoustic or elastic ( negative density and negative shear or bulk modulus ) metamaterials @xcite . similarly , although metamaterials were initially developed for bulk media , one can also create microstructured surfaces that act as metasurfaces @xcite with , again , most of the activity centered around electromagnetic waves and surface plasmons .
given that the array of resonators we introduce here modifies the surface wave properties we choose to call this a metasurface rather than a metamaterial although there is no strict definition , as yet .
much more recently seismic metamaterials have begun to be considered , although here the challenges are substantial : not only are the rayleigh waves , which are of primary interest , surface waves , but the underlying system of equations are the full vector equations of elasticity .
nonetheless there have been attempts to modify the local properties of the ground through the addition of inclusions , or resonators , of a different material at a sub - wavelength scale .
the different types of inclusions , resonant or non - resonant , determine the properties and the performance of the structured medium or metamaterial/ metasurface . turning to the non - resonant case first
, @xcite show with both large - scale experiments and theory that the soil properties can be critically affected by periodic arrangements of boreholes with a spacing of about a metre ; in the bragg scattering regime these induce bandgaps that can be used for seismic protection .
more recently , resonant sub - wavelength scatterers placed on top of an elastic substrate have been investigated and we are motivated by recent geophysical experiments @xcite that have demonstrated that natural forest trees can act as a metasurface for frequencies between 30 and 100 hz .
quite remarkably these experiments show a distinct reduction in the transmission of waves over a broad range of frequencies ; these frequencies appear to lie in bandgaps created by local resonances between trees and elastic waves in the substrate .
these large - scale experiments can be interpreted through laboratory - scale experiments and theory @xcite that have , for simplicity , utilised elastic plates , not the full elastic system ; notably , these papers neglect the flexural deformations of the resonators .
in contrast , in the present paper , we account for both the flexural and compressional deformations of the resonators and demonstrate that the flexural resonances significantly impact the spectrum of the metamaterial in certain regimes .
although we have focused upon the geophysical setting , elastic waves are also very important in ultrasonics and surface acoustic wave devices , and the bragg scattering aspects of surface arrays as phononic crystals have been explored , i.e. , in , say , @xcite .
locally resonant structures have also been shown to give rise to so - called _ super - wide pseudo - directional _ band gaps in platonic crystals @xcite ; such band gaps are formed by the coalescence of bragg and resonance band gaps . given the interest in this emerging area , and the developing applications such as surface to bulk elastic wave filters @xcite , there is a need for the fundamental solutions to canonical problems involving periodic arrays of sub - wavelength resonators atop elastic substrates . here
we provide the required analytical and theoretical background for these seismic metasurfaces by considering arrays of these resonators attached to elastic substrates , thereby considering the interaction with elastic rayleigh waves and for comparison and completeness we also consider the elastic plate .
we do so in a two - dimensional setting as the essential concepts are uncluttered by excessive algebra , and the solutions we obtain allow for interpretation and are , on occasion , completely explicit .
the mathematical tools for treating periodic arrays are extensive : linear partial differential operators , quasi - periodic green s functions , and integral transforms based upon fourier series and we use a variety of these approaches freely to obtain the cleanest solutions ; these theoretical approaches are complimented by illustrative numerical calculations based on the spectral element method .
we begin in section [ sec : plates ] by considering the array of resonators attached to an elastic plate , this is complementary to @xcite , and we obtain explicit dispersion relations and explore the relative importance of flexural and compressional resonances in the array and how they interact with the plate waves .
this naturally leads in the full elastic situation of an array atop an elastic half - space , section [ sec : full_elastic ] investigates this and again explicit results emerge that are used to interpret and predict metamaterial / metasurface phenomena .
finally , we draw together some concluding remarks in section [ sec : conclude ] .
( -7 , 0 ) ( -6 , 0 ) ; ( -7 , 0 ) ( -7 , 1 ) ; at ( -6,0 ) @xmath0 ; at ( -7,1 ) @xmath1 ; ( -5,0 ) ( 5,0 ) decorate[decoration = random steps , segment length=2,amplitude=1 ] ( 5,-1 ) ( -5,-1 )
decorate[decoration = random steps , segment length=2,amplitude=1 ] ( -5,0 ) ; ( -0.1 , 0 ) rectangle ( 0.1 , 5 ) ; ( 4.4 , 0 ) rectangle ( 4.6 , 5 ) ; ( -4.4 , 0 ) rectangle ( -4.6 , 5 ) ; ( 0.4 , 0 ) ( 0.4 , 5 ) ; ( 0.3 , 5 ) ( 0.5 , 5 ) ; at ( 0.4 , 2.5 ) @xmath2 ; ( -2.25 , -1.5 ) ( -2.25 , 5.5 ) ; ( -2.25 , -1.25 ) ( 2.25 , -1.25 ) ; at ( 0 , -1.25 ) @xmath3 ; ( 2.25 , -1.5 ) ( 2.25 , 5.5 ) ; we begin by considering an infinite periodic array of beams ( which support both flexural and compressional waves , whereas rods only support compressional waves ) on a thin elastic plate , as shown in figure [ fig : plate - schematic ] ; the boundary of the elementary cell is indicated by the dashed lines .
we use cartesian coordinates @xmath4 orientated along the thin elastic plate , and perpendicular to it with @xmath5 being the associated components of the displacement vector .
it is convenient to introduce two sets corresponding to the two different components within each cell .
the first defines the flexural foundation @xmath6 , whilst the second corresponds to the resonator @xmath7 for @xmath8 .
we note that @xmath9 and @xmath10 .
the equations of motion for time - harmonic waves of radian frequency @xmath11 are [ eq : plate - eom ] @xmath12 where @xmath13 , @xmath14 , @xmath15 is the young s modulus of the foundation , @xmath16 is the area moment of inertia of the foundation , and @xmath17 is the thickness of the foundation ; the subscript @xmath18 denotes the corresponding properties for the resonators .
the amplitudes of the horizontal and vertical forces and moments at the base of the resonators are denoted by @xmath19 , @xmath20 , and @xmath21 respectively ; the dirac delta function , @xmath22 , and its derivative , @xmath23 , are understood in the distributional sense .
the tops of the resonators are free leading to boundary conditions of the form @xmath24 imposing continuity of forces and moments at the base of the resonators yields @xmath25 for @xmath26 where @xmath27 is the cross - sectional area of the resonators .
these boundary conditions are then supplemented with kinematic equations corresponding to continuity of displacements and rotations at the base of the resonator @xmath28 where @xmath29 belongs to the closure of @xmath30 and @xmath31 belongs to the closure of @xmath32 .
the negative signs in the first and final continuity conditions are a result of the choice of coordinate system and forcing orientation . introducing the fourier transform and its inverse as @xmath33 respectively , together with the bloch - floquet quasi - periodicity conditions on the forces and moments @xmath34 where @xmath35 , the fields are expressed in the form @xmath36 @xmath37e^{in(\xi - k)}e^{ikx_1}}{k^4 - \beta^4\omega^2 } { \mathrm{d}}k,\ ] ] where we have restricted ourselves to @xmath38 and the second argument is therefore omitted . making use of the poisson summation formula @xcite ,
@xmath39 the fields can be further reduced to infinite sums . setting @xmath40 leads to a @xmath41 linear system for the displacements and rotations , the solvability condition for which yields the dispersion equation @xmath42 where @xmath43 is the identity matrix and @xmath44 with @xmath45 and @xmath46 being the dispersion equations for a uniform infinite membrane and plate respectively , [ eq : plate - forces - moments ] @xmath47 @xmath48 - \sinh\left [ ( 1+i)\beta_r\ell\sqrt{\omega}\right]}{1+\cos\left(\beta_r\ell\sqrt{\omega}\right)\cosh\left(\beta_r\ell\sqrt{\omega}\right)},\ ] ] @xmath49 @xmath50 and @xmath51 - \sinh\left [ ( 1+i)\beta_r\ell\sqrt{\omega}\right]}{1+\cos\left(\beta_r\ell\sqrt{\omega}\right)\cosh\left(\beta_r\ell\sqrt{\omega}\right)}.\ ] ] it now remains to evaluate the infinite sums appearing in the matrix problem .
we start by considering the simplest summation which , using poisson summation , can be expressed as @xmath52 the integrand has four simple poles located at @xmath53 and @xmath54 . in order to evaluate the integral
we take the usual semi - circular contour closing it in the upper half - plane for @xmath55 and the lower half - plane otherwise . for @xmath55
the contour is indented such that the pole at @xmath56 lies outside the contour and @xmath57 is enclosed by the contour .
the converse is true for @xmath58 .
computing the residues we find that @xmath59 for the remaining summations we find that @xmath60 @xmath61 and finally @xmath62 0.49 the dispersion curves for the plate system .
panel ( a ) shows the case when the flexural interactions of the resonators are accounted for ; panel ( b ) shows the curves when we neglect these flexural interactions and consider the compressional resonator modes only .
the solid black lines show the solutions of the dispersion equation , the thin solid grey lines indicate the compressional and flexural dispersion curves for the @xmath63 and @xmath64 modes of an elastic plate without resonators .
the dashed grey lines indicate the flexural resonances of the resonators , whilst the thick solid grey lines denote the compressional resonances of the resonators.,title="fig : " ] 0.49 the dispersion curves for the plate system .
panel ( a ) shows the case when the flexural interactions of the resonators are accounted for ; panel ( b ) shows the curves when we neglect these flexural interactions and consider the compressional resonator modes only .
the solid black lines show the solutions of the dispersion equation , the thin solid grey lines indicate the compressional and flexural dispersion curves for the @xmath63 and @xmath64 modes of an elastic plate without resonators .
the dashed grey lines indicate the flexural resonances of the resonators , whilst the thick solid grey lines denote the compressional resonances of the resonators.,title="fig : " ] now that the @xmath65 matrix @xmath66 is expressed in a finite number of terms , it is straightforward to compute the determinant and hence obtain the dispersion equation @xmath67 the coefficients @xmath68 are given in [ app : dispersion - plate ] and contain the information about the material and geometrical properties of the resonators .
the first parenthesised term in equation has no positive real roots for @xmath11 and therefore can be ignored ; the second term is the dispersion equation for flexural waves in a uniform thin plate .
the remaining parenthesised term in equation is a cubic polynomial in @xmath69 and therefore has exact closed - form solutions .
an example of a typical dispersion diagram is shown in figure [ fig : plate - disp ] , the corresponding material and geometrical parameters are detailed in table [ tab : parameters ] .
the dispersion curves for compressional and flexural waves in the foundation , without resonators , are also shown in figure [ fig : plate - disp ] .
the grey dashed horizontal lines are associated with the flexural resonances of the beams ; the corresponding boundary value problem is that of an euler - bernoulli beam with one end clamped at @xmath38 and the remaining end free at @xmath70 . in this case , the resonances satisfy the transcendental equation @xcite @xmath71 we note that the density of the resonances reduces with increasing frequency , tending toward the expected constant value @xcite .
the compressional resonances , indicated by the thick solid grey lines in figure [ fig : plate - disp ] , are associated with the spectrum of longitudinal waves in a clamped - free thin elastic rod ; the natural frequencies are thus @xmath72 .
.[tab : parameters ] the geometrical and numerical parameters used to produce the dispersion curves for the plate system shown in figure [ fig : plate - disp ] . [ cols="^,^,^ " , ]
figure [ fig : half - plane - disp ] shows the dispersion curves for a typical configuration ; the material parameters are detailed in table [ tab : elastic - parameters ] .
the dispersion curves , which are solutions of , are shown as solid black lines .
these curves correspond to combinations of frequency , @xmath11 , and wavenumber , @xmath73 , for which a surface waves exists ; that is , waves that propagate over the surface of the half - space and decay exponentially into the bulk .
these waves arise as a result of the periodicity of the array present on the boundary of the half - space and are distinct from the usual rayleigh waves that exist on free surfaces ; it is natural , therefore , to refer to such waves as rayleigh - bloch waves @xcite .
in addition to the dispersion curves , we also indicate the compressional resonances of the resonators , shown as the horizontal dashed lines in figure [ fig : half - plane - disp ] .
the remaining dashed lines of increasing slope are the curves corresponding to rayleigh waves , bulk shear waves , and bulk compressional waves in the homogeneous elastic half - space respectively ; we call these dispersion lines the rayleigh / shear / compressional wave sound - line by analogy with the terminology used in electromagnetism ( see , for example , @xcite ) . previous investigations of similar systems were based on purely numerical simulations @xcite ; with the dispersion equation in hand , we can now examine several interesting features of the dispersion curves offering additional physical insight . as noted in the earlier papers @xcite on elastic plates and from the numerical simulations of @xcite for a half - space ,
the onset of band gaps coincide with the longitudinal resonances of the resonators . here
we observe that the upper boundary of the band gaps coincide with the intersection of the dispersion curves with the shear sound - line .
these points are indicated by the red circles in figure [ fig : half - plane - disp ] and correspond to the case when @xmath74 ( @xmath75 ) , in which case , equation reduces to @xmath76 which has the solutions @xmath77,\quad \text{for}\quad n\in\mathbb{n}.\ ] ] the longitudinal resonances of the resonators lie at @xmath72 .
thus , in this regime where the pass bands are bounded from below by the resonances of the resonators and from above by the intersections of the dispersion curves with the shear sound - line , the band gaps have a constant width defined by @xmath78 .
\label{eq : elastic - band - width}\ ] ] since @xmath79 , the argument of @xmath80 in equation is real and positive ; hence @xmath81 and therefore band gaps always exist for finite parameters .
equation provides a convenient formula for tuning the width of the band gap : in order to maximise the band gap , the factor @xmath82 and the argument of @xmath80 in should be made as small as possible simultaneously .
the converse will minimise the width of the stop band . in rough terms , short resonators with high compressional wave - speeds
will maximise the stop band width .
however , as the material parameters of the resonators also appear in the argument of @xmath80 , some care is required when tuning the system .
thus far we have studied infinite periodic structures .
however , metamaterials , metasurfaces , and phononic structures often find application as filters , lenses , and polarisers . as has been shown in the electromagnetic @xcite , elastodynamic @xcite , and thin plate @xcite literature , the dispersive properties of an infinite system can be used to design finite structures that posses interesting properties such as negative refraction , flat lenses , filtering , and cloaking .
to illustrate the potential of rayleigh waves in this regard , we consider their interaction with a finite array of resonators . to this end , and given that we are concerned with the control of surface waves on elastic bodies , the intersection of the dispersion curves with the rayleigh wave sound - line for an homogeneous elastic half - space without resonators ( indicated by the blue dot on figure [ fig : half - plane - disp ] ) is of interest . for combinations of wavenumber and frequency corresponding to this point of intersection , a rayleigh wave incident on the array of resonators from the elastic half - space can generate a rayleigh - bloch wave , that is a surface wave whose behaviour is affected by the periodicity , in the array .
conversely , at this point , a rayleigh - bloch wave leaving the array can couple into a pure rayleigh wave in the ambient half - space . with the exception of embedded surface waves , which we do not consider here , the homogeneous half - space does not support localised rayleigh waves corresponding to points above the rayleigh wave sound - line ; in practical terms , this means that the region between the lower red dot and the blue dot on figure [ fig : half - plane - disp ] is an `` effective stop band '' for rayleigh waves incident on the array from a homogeneous half - space . the point of intersection between the dispersion curves and the rayleigh sound - line can be obtained from by setting @xmath83 , where @xmath84 is the rayleigh wave speed ( see @xcite , among others ) .
the frequency of intersection is then @xmath85 ^ 2}{\sqrt{\gamma^2 - r^2 } } - 4\gamma^2\sqrt{\gamma^2 - 1}\right)\right]\right\ } , \label{eq : rayleigh - intersection}\ ] ] where @xmath86 .
the `` effective width '' of the first band gap is then @xmath87 ^ 2}{\sqrt{\gamma^2 - r^2 } } - 4\gamma^2\sqrt{\gamma^2 - 1}\right)\right]\right\}.\ ] ] we now illustrate the dispersive properties of the resonant array atop an elastic half - space using the results from time domain spectral element ( sem ) simulations . we consider a linearly elastic , isotropic , homogeneous medium with a linear array of resonators , as depicted in figure [ fig : half - line - fig0 ] ; the material and goemetrical properties of the system
are specified in table [ tab : elastic - parameters ] .
the simulations are performed with specfem2d @xcite , a well known parallel code widely used by the seismological community .
the mesh is constructed using quadrilateral elements and the commercial software cubit .
perfectly matched layers @xcite ( pml ) are applied on the bottom and vertical boundaries of the computational domain containing the half - space ; the remaining boundaries ( top surface and resonators ) are traction free .
the accuracy of this method has been thoroughly tested in a previous study @xcite against experimental measurements .
thus far we have considered an infinite array of resonators , however for the numerical simulations ( figs .
[ fig : half - line - fig1 ] and [ fig : half - line - fig2 ] ) , a large , finite , number is more convenient .
we take an array of 30 resonators and place a point source 60 m away on the surface of the half - space , as shown in figure [ fig : half - line - fig0 ] .
the source takes the form of a sinusoidal vertical force of angular frequency @xmath11 . to avoid spurious effects at the onset of time
, the amplitude of the sinusoidal force is initially modulated by a ramp function that progressively increases in amplitude over the first eight periods ( see the inset of figure [ fig : half - line - fig0 ] ) , @xmath88 the distance between the source and the resonant array is such that only rayleigh waves scatter on the resonators ( e.g. figs .
[ fig : half - line - fig1 ] and [ fig : half - line - fig2 ] ) ; the remaining classes of waves are absorbed by the pml .
once the ramp function has reached its final value , the simulations continue in a stationary state where only a monochromatic signal with constant amplitude propagates . the snapshots in figs . [
fig : half - line - fig1 ] and [ fig : half - line - fig2 ] have been taken when the simulation is in this regime .
figures [ fig : half - line - fig1 ] and [ fig : half - line - fig2 ] demonstrate the metamaterial properties of the resonant array in various regimes . here
we have chosen not to present the deformation of the resonators in order to highlight the wavefield in the half - space ; the full videos associated with these figures are provided in the supplementary materials ] .
figure [ fig : half - line - fig1]c shows the transmission of a surface wave through the array at a frequency corresponding to the lower branch of the dispersion curves , figure [ fig : half - line - fig1]a .
we also note the decrease in wavelength inside the array corresponding to the slower effective wavespeed inside the metasurface as compared with the homogeneous half - space .
the shorter wavelength is accompanied by an increase in amplitude of the waves inside the array as well as an increased rate of decay into the bulk .
in contrast , figure [ fig : half - line - fig1]b shows the case when the frequency of the incident wave almost coincides with a resonance of the resonators leading to virtually no transmission . moving on to figure [ fig : half - line - fig2 ] , the two panels ( b ) and ( c ) correspond to the second branch of the dispersion curves , as indicated .
it is important to note that this upper branch transitions from the rayleigh sound - line to the shear sound - line , that is , the rayleigh - bloch wave in the resonant array preserves the rayleigh - like surface wave properties at higher frequencies but will evolve into a shear - like wave as the frequency decreases .
the hybrid nature of this branch , sitting between the shear and rayleigh sound - lines , is responsible for the mode conversion of rayleigh waves into shear waves and used in the design of so - called resonant _ metawedge _
systems @xcite . the point of intersection ( marked in red ) between the hybrid branch and the shear sound - line is of particular interest .
at this frequency , waves can propagate past the metasurface but they must only do so as shear waves ( note the polarization ) as is illustrated in figure [ fig : half - line - fig2](c ) for a frequency very close to this point . in this regime the transition between rayleigh waves propagating in the free half - space ( to the right hand side of the array ) and the hybrid mode in the array happens abruptly resulting in an apparent bandgap " for surface waves .
therefore the wavefield beneath the resonators , represented in figure [ fig : half - line - fig2](c ) , is characterised by shear waves that propagate away from the surface and into the bulk leaving the left hand side of the array almost untouched . as the frequency increases ( figure [ fig : half - line - fig2](b ) ) the wavefield reveals more clearly its hybrid nature .
the waves propagating under the resonators are grazing the surface but they are not yet exponentially decaying as rayleigh waves . in this regime waves can propagate in the resonator but the transmission is still weak .
the hybrid mode progressively turns into a rayleigh wave as the frequency increases toward the crossing point ( blue dot in figure [ fig : half - plane - disp ] ) .
metamaterials barely existed fifteen years ago , but now form a major research area : although initially developed for optics , the field of metamaterial research has expanded rapidly and now includes the development of novel materials for applications in acoustics and elasticity .
more recently , the applicability of metamaterials to seismology has sparked the interest of geophysicists in the development of novel methods to control surface waves . given the interest in this emerging area ,
there is a need to study the properties of the solutions to fundamental canonical problems . in this paper
we have provided the theoretical and analytical framework necessary for the rigorous study of these seismic metasurfaces that adjust the surface wave behaviour .
the framework is introduced through the study of two canonical problems associated with the control of mechanical surface waves .
initially , we study the propagation of surface bloch waves through an array of resonators on a thin elastic plate . in this case
, the resonators are thin elastic rods supporting both compressional and flexural waves .
explicit exact solutions are developed and used to examine the behaviour of the system .
in particular , we study the coupling between the symmetric and anti - symmetric modes in the substrate and the resonances of the rods .
interestingly it is found that , in the frequency range of interest and for sufficiently rigid plates , the flexural resonances of the resonators couple very weakly into the substrate with the dominant effect coming from the compressional modes of resonators , which open up band gaps associated with the compressional resonances . for more flexible plates , the flexural deformations of the resonators become important .
the plate system is used as the motivation for the far more challenging problem considered in section [ sec : full_elastic ] where we examine the propagation of surface bloch waves through an array of resonators resting on a fully - elastic half - space . here
we develop closed form expressions for the dispersion equation and wave - fields in the deep sub - wavelength regime of interest . as for the plate problem , the compressional resonances of the resonators create band gaps in the dispersion curves and analytical expressions for their position and width are provided . in particular , it is shown that band gaps will always exist regardless of the material or geometrical parameters of the half - space and resonators . the notion of `` effective band gaps '' for surface waves
is also introduced and discussed . in section [ sec : full_elastic ] , we also examine the scattering problem associated with a half - line of resonators and use the dispersive properties of the array to examine , the filtering effects of the resonant array . with
the formal framework of seismic metasurfaces now established , the time is ripe to exploit these results and explore the possibilities of extending the existing results for electromagnetism and acoustics to seismology .
indeed , the theoretical framework developed in the present paper is already being used in the design and development of the so - called _ meta - wedge _ that is capable of mode - converting destructive seismic surface waves into mainly harmless bulk shear waves @xcite .
the authors thank the epsrc ( uk ) for their support through the programme grant ep / l024926/1 .
p.r . acknowledges the financial support of anr under project labex osug@2020 .
the dispersion equation for bloch - waves propagating through the plate can be cast as a cubic polynomial in @xmath69 with the coefficients depending on angular frequency @xmath11 as well as the material and geometrical parameters .
these coefficients are cumbersome and are therefore stated here , rather than disturb the flow of the main text . with reference to equation ,
the polynomial coefficients are as follows @xmath89 @xmath90 + 4\beta^3 \alpha\omega^{3/2}\cosh\beta\sqrt{\omega } \right .
- \alpha\left [ v + \beta^2 m_\theta\omega \right ] \sin\beta\sqrt{\omega } + 2 \beta^3 f_u\sqrt{\omega}\sin\alpha\omega \right .
+ \alpha\left [ v - \beta^2 m_\theta\omega\right]\sinh\beta\sqrt{\omega } \right\},\end{gathered}\ ] ] @xmath91 \right .
\\ + \sqrt{\omega}\cosh\beta\sqrt{\omega}\left [ \alpha\left(8\beta^4\omega^2 - m_\theta v\right)\cos\beta\sqrt{\omega } + 8\beta^4\alpha\omega^2\cos\alpha\omega \right .
- 2\beta\alpha\sqrt{\omega}\left(v + \beta^2\omega m_\theta\right)\sin\beta\sqrt{\omega } + 4\beta^3 f_u\omega\sin\alpha\omega \right ]
\\ - \beta\sin\beta\sqrt{\omega}\left [ \left ( f_uv + \beta^2\omega\left\ { f_um_\theta - f_\theta m_u \right\}\right)\sin\alpha\omega \right .
+ 2\alpha\omega\left ( v + \beta^2\omega m_\theta \right)\cos\alpha\omega \right ] + \beta\left [ 2\alpha\omega\left ( v - \beta^2\omega m_\theta \right)\left(\cos\beta\sqrt{\omega } + \cos\alpha\omega \right ) \right .
+ f_u v + \beta^2\omega\left\ { f_\theta m_u - f_um_\theta\right\}\sin\alpha\omega \right]\sinh\beta\sqrt{\omega},\end{gathered}\ ] ] @xmath92\cos\alpha\omega + \left [ 16\beta^4\omega^2 f_u \right.\right . \\ \left.\left .
+ v\left(f_\theta m_u - f_um_\theta\right)\right]\sin\alpha\omega + 2v\sec\beta\sqrt{\omega}\operatorname{sech}\beta\sqrt{\omega}\left [ 2\alpha m_\theta\omega\cos\alpha\omega \right.\right .
+ \left(f_um_\theta - f_\theta m_u\right)\sin\alpha\omega\right ] + v\operatorname{sech}^2\beta\sqrt{\omega}\left [ \sin\alpha\omega\left(m_uf_\theta \right.\right.\right .
\\ \left.\left.\left .
- f_um_\theta\right ) - 2\alpha\omega m_\theta\cos\alpha\omega \right ] - 4\beta\sqrt{\omega}\left [ 2\alpha\omega\left ( v + \beta^2\omega m_\theta \right)\cos\alpha\omega \right.\right . \\
+ \left ( f_u v + \beta^2\omega \left\ { f_u m_\theta - f_\theta m_u \right\}\right)\sin\alpha\omega\right]\tan\beta\sqrt{\omega } \right .
+ 4 \beta \sqrt{\omega}\left [ 2\alpha\omega\left ( v - \beta^2\omega m_\theta\right)\cos\alpha\omega + \left ( f_uv + \beta^2\omega\left\ { f_\theta m_u \right.\right.\right.\right . \\ \left.\left.\left.\left .
- f_u m_\theta\right\}\right)\sin\alpha\omega \right]\tanh\beta\sqrt{\omega } + v\left [ \sin\alpha\omega\left ( f_\theta m_u - f_um_\theta \right)\sin\alpha\omega \right.\right .
\\ \left.\left .
- 2\alpha m_\theta\omega\cos\alpha\omega \right]\tanh^2\beta\sqrt{\omega}\right\}.\end{gathered}\ ] ]
the dispersion equation for flexural waves propagating through a plate with rod - like resonators can be expressed as quadratic polynomial in @xmath69 with coefficients depending on the spectral properties of the plate and resonators .
these coefficients are @xmath93 @xmath94 + v\left[\sinh\beta\sqrt{\omega } - \sin\beta\sqrt{\omega}\right],\ ] ] @xmath95\cos\beta\sqrt{\omega}\cosh\beta\sqrt{\omega}.\ ] ] achaoui , y. , khelif , a. , benchabane , s. , robert , l. , laude , v. , 2011 .
experimental observation of locally - resonant and bragg band gaps for surface guided waves in a phononic crystal of pillars .
phys . rev .
b 83 ( 10 ) , 10401 .
colquitt , d. , jones , i. , movchan , n. , movchan , a. , 2011 . dispersion and localization of elastic waves in materials with microstructure .
proceedings of the royal society a : mathematical , physical and engineering science 467 ( 2134 ) , 28742895 .
dubois , m. , farhat , m. , bossy , e. , enoch , s. , guenneau , s. , sebbah , p. , 2013 .
flat lens for pulse focusing of elastic waves in thin plates . applied physics letters 103 ( 7 ) , .
http://scitation.aip.org/content/aip/journal/apl/103/7/10.1063/1.4818716 roux , p. , rupin , m. , colombi , a. , craster , r. v. , gunneau , s. , kuperman , w. a. , lemoult , f. , lerosey , g. , williams , e. g. , in press .
new trends toward locally - resonant metamaterials at the mesoscopic scale . in : maier , s. , shamonina , k. , gunneau , s. , hess , o. , aizpurua , j. ( eds . ) , handbook of metamaterials and nanophotonics .
world scientific , singapore .
yoritomo , j. y. , weaver , r. l. , roux , p. , rupin , m. , williams , e. g. , 2016 . on band gap predictions for multiresonant metamaterials on plates .
the journal of the acoustical society of america 139 ( 3 ) , 12821284 . | we consider the canonical problem of an array of rods , which act as resonators , placed on an elastic substrate ; the substrate being either a thin elastic plate or an elastic half - space . in both cases the flexural plate , or rayleigh surface , waves in the substrate interact with the resonators to create interesting effects such as effective band - gaps for surface waves or filters that transform surface waves into bulk waves ; these effects have parallels in the field of optics where such sub - wavelength resonators create metamaterials , and metasurfaces , in the bulk and at the surface respectively . here
we carefully analyse this canonical problem by extracting the dispersion relations analytically thereby examining the influence of both the flexural and compressional resonances on the propagating wave . for an array of resonators atop an elastic half - space
we augment the analysis with numerical simulations . amongst other effects
, we demonstrate the striking effect of a dispersion curve that transitions from rayleigh wave - like to shear wave - like behaviour and the resultant change in displacement from surface to bulk waves .
= 1 [ firstpage ] |
parametric decomposition of galaxy images has become a popular tool to estimate the structural parameters of different galactic components , particularly bulges and discs . through this sort of analysis , one is also able to determine the relative importance of the bulge component , with parameters such as the bulge - to - total luminosity ratio b / t , one of the major features that define the @xcite sequence .
it thus provides indispensable means to investigate the formation and evolution of galaxies , and the origin of the hubble sequence , some of the key subjects in current astrophysical research .
recently , the focus of studies using image decomposition shifted from relatively small samples of very nearby galaxies , where the fits can be done on a more careful , individual basis ( e.g. , ( * ? ? ?
* de jong 1995 ) ) , to include large samples of more distant galaxies , using automated procedures , that allow solid statistical analyses ( e.g. , ( * ? ? ?
* allen 2006 ) ) .
so far , most studies have ignored other components , such as bars ( but see ( * ? ? ?
* ; * ? ? ?
* laurikainen 2004 , 2005 ) ) , although the majority of disc galaxies host bars that usually contain a significant fraction of the galaxy total stellar mass .
it is unclear how the parameters of bulge and disc are affected when bars are ignored .
a similar issue concerns galaxies that host bright , type 1 agn : if the light contribution from the agn is not modelled , the structural parameters obtained might be incorrect .
it is also not clear if the low physical spatial resolution usually found in images of more distant galaxies introduces any bias in the parameters recovered . to shed light on these issues , i performed multi - component image decomposition of a sample of very nearby galaxies and their artificially redshifted images , and verified the effects of removing the bar and agn components from the models .
the results of these tests are shown in the next section [ the reader is referred to @xcite for full details on this work ] .
the images used here are from @xcite . these are v and r images of 17 galaxies at @xmath1 , most of them hosting bars and agn .
image decomposition is done using budda v2.1 ( see ( * ? ? ? * de souza , gadotti & dos anjos 2004 ) ) .
all components are modelled as sets of concentric ellipses , with constant ellipticity and position angle .
bulges are described with a srsic luminosity profile and discs are exponential .
bars are described with boxy ellipses , also with a srsic profile , and the agn contribution is modelled using a circular moffat profile .
when one takes advantage of the full capabilities of the code ( i.e. , all necessary components ) on these images , the results show a fairly good agreement with other studies .
in addition , several known scaling relations are reproduced .
0.3 cm when one does not include the bar component in the models , the major effect is seen in the bulge models recovered : bulges get bigger , in a way to accommodate light from the bar . as a result
, b / t is overestimated , on average , by 50% , but this overestimation can reach a factor of 2 .
disc models are also systematically affected , but to a lesser extent : d / t is overestimated , on average , by 10% , with a maximum overestimation of 30% ( see fig . [ nobar ] ) .
similar effects ( albeit restricted to the bulge component ) result when bright , type 1 agn are not considered in the models : b / t and the srsic index of the bulge @xmath2 can be overestimated by factors of 2 and 4 , respectively . to verify if these results hold when the physical spatial resolution of the images is relatively low , i degraded the original images , thus artificially redshifting them to @xmath3 .
the resolution in these images is 1.5 kpc , which is the typical resolution of sdss at this redshift .
such a resolution is also typical in other works as well , and is what can be achieved at @xmath4 with hst . as with the original images ,
the redshifted images were first used to perform a complete decomposition , and then to test the effects of removing the bar and agn components from the models .
figure [ hz ] shows that , in general , structural parameters can be reliably retrieved through image fitting even in a low resolution regime .
this is particularly true for the disc parameters .
however , bulge parameters are prone to errors if its effective radius @xmath5 is small compared to the seeing radius , and b / t might suffer from systematic effects .
b / t seems to be systematically overestimated , on average , by 0.05 , i.e. , 5% of the total galaxy luminosity , even when @xmath5 is similar to the seeing radius .
when bars are excluded from the models to fit the redshifted images , one still sees similar effects as with the original images , namely , a significant overestimation of b / t and @xmath5 , if bars are prominent .
the change in @xmath5 is even more pronounced in the redshifted images , likely because the geometrical properties of the bulge are substantially smoothed by the seeing , making it more difficult to constrain its properties . for less prominent bars
these effects are reduced in strength .
the effects caused by not taking into account bright agn , i.e. , the overestimation of b / t and @xmath2 , do _ not _ occur with the redshifted images , as the agn contribution is smeared out by the seeing .
when bars are not taken into account , the amount of mass in stars in bulges and discs is overestimated , and the excess is an indication of the amount of mass in stars that reside in bars .
@xcite estimated the @xmath6 stellar mass content in classical bulges and discs through image decomposition of @xmath7 galaxies , with a spatial resolution similar to that of the redshifted images above .
bars are not taken into account in the fitted models , but these authors made a thorough quality control , removing poor fits .
however , it is usually the case that , even when there is no bar in the model , when fitting a barred galaxy , one gets an acceptable ( though wrong ) fit , essentially because the bulge model is distorted . hence , with reasonable assumptions , one can use their results to obtain a _ rough _ estimate of what can be the stellar content in bars , assuming that the biases produced in a low resolution regime and by ignoring bars , as found above , can be used in this case to obtain the true bulge and disc luminosity fractions
thus , the stellar content in classical bulges and discs is found to be @xmath8 and @xmath9 , respectively . and
the stellar content in bars is @xmath10 .
although these are rough estimates , that need to be confirmed by further studies , they open up the possibility of bars being as important as bulges in what concerns their stellar mass content in the local universe .
are disc galaxies ) from the sdss .
bulges show a remarkably different distribution as compared to ellipticals , with the family of pseudo - bulges ( i.e. , those with srsic index less than about 2 ) standing out clearly.,width=340 ] figure [ ndistr ] shows the distribution of the srsic index of bulges and ellipticals in a sample of about 200 disc galaxies and 100 ellipticals from the sdss , obtained using budda v2.1 ( see ( * ? ? ?
* gadotti & kauffmann 2007 ) ) .
one clearly sees the dichotomy between classical bulges , with @xmath11 , and pseudo - bulges , with @xmath12 .
this figure indicates that the fraction of pseudo - bulges at @xmath6 is @xmath13 .
i would like to thank martin bureau and the organisers for this very fruitful meeting .
it is a pleasure to thank guinevere kauffmann for her active role in this work , with insightful comments and suggestions , and ronaldo de souza , for his help in the making of the new budda version .
dag is supported by the deutsche forschungsgemeinschaft priority program 1177 ( `` witnesses of cosmic history : formation and evolution of galaxies , black holes and their environment '' ) , and the max planck society .
2006 , _ mnras _ 371 , 2 1995 , phd thesis , university of groningen , the netherlands 2004 , _ apjs _ 153 , 411 2007 , _ apjl _ , in press ( astro - ph/0701728 ) 2007 , _ mnras _ , submitted 2006 , _ apjs _ 163 , 270 2007 , in : stellar populations as building blocks of galaxies , iau symp .
241 , a. vazdekis , r. peletier ( eds . ) , in press ( astro - ph/0702054 ) 1926 , _ apj _ 64 , 321 2004 , _ mnras _ 355 , 1251 2005 , _ mnras _ 362 , 1319 | image decomposition of galaxies is now routinely used to estimate the structural parameters of galactic components . in this work ,
i address questions on the reliability of this technique .
in particular , do bars and agn need to be taken into account to obtain the structural parameters of bulges and discs ? and to what extent can we trust image decomposition when the physical spatial resolution is relatively poor ? with this aim , i performed multi - component ( bar / bulge / disc / agn ) image decomposition of a sample of very nearby galaxies and their artificially redshifted images , and verified the effects of removing the bar and agn components from the models .
neglecting bars can result in a overestimation of the bulge - to - total luminosity ratio of a factor of two , even if the resolution is low .
similar effects result when bright agn are not considered in the models , but only when the resolution is high .
i also show that the structural parameters of more distant galaxies can in general be reliably retrieved , at least up to the point where the physical spatial resolution is @xmath0 1.5 kpc , but bulge parameters are prone to errors if its effective radius is small compared to the seeing radius , and might suffer from systematic effects .
i briefly discuss the consequences of these results to our knowledge of the stellar mass budget in the local universe , and finish by showing preliminary results from a large sdss sample on the dichotomy between classical and pseudo - bulges . |
in the past decade , a series of experiments , in particular the wilkinson microwave anisotropy probe ( wmap ) @xcite , have measured the anisotropy in the cosmic microwave background ( cmb ) radiation with extraordinary precision .
the data from these experiments have helped to usher in the era of precision cosmology .
most of the cosmological results that have been obtained from these experiments have been derived from the microwave anisotropy at relatively large angular scales .
the remarkable agreement between the angular power spectrum of these data and the predictions of the adiabatic inflationary scenario have established the empirical success of inflationary cosmology .
cosmic strings , though ruled out as the origin of cosmological structure , have recently enjoyed a renaissance .
this renewed popularity has been brought about by the recognition that a variety of string theory - motivated and hybrid models for inflation generically predict the formation of cosmic string networks @xcite .
strings are limited to producing less than about 10% of the primordial cmb anisotropy @xcite , though it was shown in @xcite that cmb data can actually favor a contribution from strings if the inflationary spectrum is exactly harrison - zeldovich ( @xmath3 ) ; this corresponds to a string tension ( @xmath4 ) between @xmath5 and @xmath6 for a standard set of string network parameters .
the b ( i.e. , curl ) mode polarization in the cmb caused by the active perturbations of a cosmic string network has a spectrum distinct from those expected either from inflationary gravity waves or the lensing of e ( i.e. , gradient ) mode polarization into b mode by large scale structure . when the above bound is marginally saturated , this b mode polarization should be measurable @xcite , providing a powerful test of the presence of cosmic strings .
another consequence is the power spectrum of perturbations that strings source at large @xmath7 ( small angular size ) .
cosmic string networks continually generate cmb anisotropies , both primordially through active density perturbations and subsequent to recombination through the lensing of the primary cmb light the kaiser - stebbins effect ( ks ) @xcite .
the cmb anisotropy due to the ks effect alone at large @xmath7 is expected to decrease only as @xmath8 @xcite .
this rate of decrease is much slower than that expected for inflationary perturbations ( which fall off exponentially as a function of @xmath7 due to silk damping , which is due to radiative diffusion ) .
if @xmath9 is not too small , this large @xmath7 power spectrum may be measurable .
it should appear as an excess above the prediction from inflationary perturbations .
in particular , for @xmath10 , the power created on small angular scales , @xmath2 , by cosmic strings will actually dominate over that created by the primary inflationary perturbations .
this range of angular scales in the cmb is presently being measured by a number of experiments , so that this prediction of cosmic string networks will be tested soon . in this note , we present the large @xmath7 power spectrum due to cosmic strings .
we use cmbact @xcite , a modified version of cmbfast @xcite , to produce the string sourced anisotropy spectra .
the model , described in refs .
@xcite , is based on representing the cosmic string network as a collection of moving straight string segments . in brief , there are two important length scales in this model : @xmath11 , the length of a string segment , which represents the typical length of roughly straight segments in a full network ; and @xmath12 , the typical length between two string segments , which sets the number density of strings in a given volume ( @xmath13 ) .
we use the velocity - dependent one - scale model ( vos ) @xcite to evolve the network length - scales and density . in the simplest and most relevant cases , @xmath14 . with these parameters
set , we average over a set of randomized realizations of the network approximated as a set of straight string segments . this model was introduced in @xcite , based on the approach suggested in @xcite , and was developed into its present form in @xcite .
the overall normalization of the spectrum has a simple dependence on the string tension and number density : c_^strings n_s ( g)^2 ( ) ^2 .
lower reconnection probability @xcite for cosmic strings will rescale the amplitude of this spectrum , but will not change its shape . although this model assumes a single tension cosmic string network , the results described here should generally apply for more complex multi - tension string network models of the sort that may be produced in the aftermath of brane inflation @xcite .
[ f1 ] shows our result . the dominance of vector modes ( the blue dashed line in the left panel ) for @xmath15 is an effect that is a distinctive signature of cosmic strings .
as we show with the magenta , dash - dotted line , the string spectrum at @xmath16 where vector - sourced perturbations start to dominate is well approximated by a @xmath17 fall off .
this can be captured by a simple fitting formula : ( ( + 1)/ 2 ) c_^tt[strings ] 500 ( ) ^1.5 ( ) ^2 for 1000 .
we find that towards the very high @xmath7 end of the considered range , i.e. at @xmath18 , the fall off is better described by @xmath19 , in agreement with the pure ks contribution analytically predicted in @xcite . at smaller @xmath7 ,
the residual fluctuations from the last - scattering surface are non - negligible leading to a @xmath20 fall off in the @xmath21 range . the right panel in fig .
[ f1 ] shows the high-@xmath7 power sourced by strings relative to the inflationary contribution .
the string spectrum s amplitude is set by saturating the observational bound : it accounts for @xmath22% of power for @xmath0 . ; average coherence length @xmath23 ( measured in units of the horizon size ) ; rms velocity near @xmath24 ; and average wiggliness parameter @xmath25 .
the solid black line is the total power .
the red dotted line is the power due to scalar perturbations , the blue dashed line represents the vector mode perturbations .
the ( green ) long dashed line shows the ( negligible ) contribution from tensor modes .
note that vector modes dominate above @xmath26 , a distinctively stringy effect .
the magenta dash - dotted line shows the @xmath17 fitting formula .
_ right : _
contributions from strings ( solid red ) , inflation ( blue dot ) ( including lensing , from camb @xcite ) , and their combination ( black dash ) at high-@xmath7 .
the string contribution satisfies the @xmath22% bound on total power imposed on large angular scales.,title="fig:",width=288 ] ; average coherence length @xmath23 ( measured in units of the horizon size ) ; rms velocity near @xmath24 ; and average wiggliness parameter @xmath25 .
the solid black line is the total power .
the red dotted line is the power due to scalar perturbations , the blue dashed line represents the vector mode perturbations .
the ( green ) long dashed line shows the ( negligible ) contribution from tensor modes .
note that vector modes dominate above @xmath26 , a distinctively stringy effect .
the magenta dash - dotted line shows the @xmath17 fitting formula .
_ right : _
contributions from strings ( solid red ) , inflation ( blue dot ) ( including lensing , from camb @xcite ) , and their combination ( black dash ) at high-@xmath7 .
the string contribution satisfies the @xmath22% bound on total power imposed on large angular scales.,title="fig:",width=288 ] the spectrum in fig .
[ f1 ] should not be taken as the unique prediction of the @xmath27 spectrum from strings , but as a representative example of what one can get for a reasonably motivated string network .
it corresponds to particular values of string model parameters , such as wiggliness , coherence length , and rms velocity .
it also relies on the moving segments " approximation used in cmbact .
this model is designed to describe statistical properties of scaling string networks .
for instance , it can be used to calculate cmb power spectra , but can not make a sky map of string network effects .
real string simulations that capture more of the physics of networks such as their curvature and loops are too computationally costly to be used over many expansion times . the segments model has been shown to match cmb spectra from full simulations reasonably well over the scales where they can be compared ( for a fuller discussion , see refs .
@xcite ) . while varying the string tension , @xmath4 , simply renormalizes the spectrum , changing the other parameters can re - distribute the power between the large and small scales .
generally , smaller velocities and smaller coherence lengths enhance the power in vector modes at high @xmath7 @xcite .
more wiggliness tends to make strings move more slowly .
this suppresses all contributions to the anisotropic stress , including the vector modes @xcite . for this plot
, we used a model with @xmath28 with an average string coherence length of @xmath29 of the horizon size , . ] a rms velocity of @xmath30 , and an average wiggliness parameter @xmath31 of @xmath32 .
these values are fairly close to those seen in numerical simulations ( e.g. @xcite ) , but are slightly different from those used in the default version of cmbact ; the values chosen here are those which slightly enhance the string - sourced power at high @xmath7 .
fractionally greater ( or attenuated ) power at small angular scales can be achieved with different model parameters , but the overall behavior at high @xmath7 is a generic feature of string networks .
all cosmological parameters , _
i.e. _ @xmath33 , @xmath34 , etc , are those of the latest wmap best fit @xcite .
cosmic strings produce power on small angular scales because they are _ active _ sources that continue contributing to the anisotropy after the last scattering . for that reason they evade silk damping , _
i.e. _ the erasure of anisotropies on small scales due to the finite thickness of the last scattering surface . a network of cosmic strings with a tension near the present observational bound of @xmath35 can dominate the power spectrum of cmb fluctuations in the strongly silk - damped regime ( @xmath36 ) of the microwave background anisotropy , creating an apparent excess of power over what is expected from an inflationary adiabatic perturbation spectrum .
this high @xmath7 regime is accessible to existing fine scale resolution experiments like the cosmic background imager ( cbi ) @xcite and the arcminute cosmology bolometer array receiver ( acbar ) @xcite now , and will soon be measured very accurately by experiments like the south pole telescope @xcite and the atacama cosmology telescope @xcite .
it is interesting to note that these experiments published data already show some hints of excess power in the high - multipole range ; see fig .
[ withacbar ] for a rough comparison with the data from acbar .
cmb anisotropy in the acbar range , plotted against the latest data from the acbar experiment .
the dotted blue line is the inflationary prediction alone ( including lensing , from the software package camb @xcite ) ; the dashed black line includes the contribution from a cosmic string network .
the amplitude fitting has been done roughly , by eye , to match the two theoretical curves with the second acbar data point ; the cosmological parameters are all those of the wmap best fit . ,
width=288 ] although we have not done a statistical analysis , the preliminary results presented here suggest that cosmic strings with @xmath37 could contribute enough power to account for the excess over inflation suggested by the acbar data at @xmath38 without exceeding observational bounds on string contributions at @xmath39 .
excess power in the high-@xmath7 cmb can also be generated by other physical phenomena , including the sunyaev - zeldovich effect @xcite and tangled primordial magnetic fields @xcite .
using the sz effect to account for any substantial excess measurable by today s experiments is problematic , however , because the amount of small - scale gravitational clustering ( measured by the parameter @xmath40 ) required to generate a large excess over the inflationary prediction via the sz effect is in some conflict with values determined by other experiments .
tangled magnetic fields , on the other hand , are not meaningfully constrained by competing experiments , but their existence at the necessary epoch is by no means accounted for .
discovery of significant high @xmath7 excess power in cmb that can not be explained by more conventional means could be taken as evidence for existence of cosmic strings with tensions near the observational bound .
on the other hand , if no excess is seen at large @xmath7 , non - observation of this effect will provide a useful bound on the properties of any cosmic string network .
strings with tensions in this range can be searched for by some other means , such as gravitational lensing and microlensing @xcite , gravitational radiation bursts @xcite , pulsar timing or non - gaussian step - like fluctuations in cmb temperature @xcite .
an especially promising signature of such strings would be a substantial b mode polarization @xcite ; for @xmath4 around @xmath41 , the b mode polarization fluctuations from strings could exceed expected power from e to b conversion by gravitational lensing by factors of a few .
we thank j. richard bond , anthony readhead , and mark hindmarsh for discussions leading to the exploration of this question ; we also thank david chernoff for conversations and jonathan sievers for comments on the draft .
lp s work is supported by a discovery grant from the national sciences and engineering research council of canada .
shht s work is supported by the national science foundation ( nsf ) under grant phy-0355005 .
iw s work is supported by the nsf under grant phy-0555216 .
the work of m. w. at the perimeter institute is supported by the government of canada through industry canada and by the province of ontario through the ministry of research & innovation .
j. dunkley _ et al . _
[ wmap collaboration ] , arxiv:0803.0586 [ astro - ph ] . s .-
h. tye , arxiv : hep - th/0610221 ; + j. polchinski , arxiv:0707.0888 [ astro - ph ] ; + m. sakellariadou , arxiv:0802.3379 [ hep - th ] .
a. villenkin and e.p.s .
shellard , , cambridge university press , 2000 .
copeland , a.r .
liddle , d.h .
lyth , e.d .
stewart , and d. wands , phys rev . *
d49 * ( 1994 ) 6410 .
r. jeannerot , arxiv : hep - ph/0610295 .
n. jones , h. stoica and s .- h .
h. tye , jhep * 0207 * ( 2002 ) 051 , hep - th/0203163 .
g. dvali and a. vilenkin , jcap * 0403 * , 010 ( 2004 ) [ arxiv : hep - th/0312007 ] .
e. j. copeland , r. c. myers and j. polchinski , jhep * 0406 * , 013 ( 2004 ) [ arxiv : hep - th/0312067 ] .
n. barnaby , a. berndsen , j. m. cline and h. stoica , jhep * 0506 * , 075 ( 2005 ) [ arxiv : hep - th/0412095 ] .
h. firouzjahi and s .- h .
h. tye , jcap * 0503 * , 009 ( 2005 ) [ arxiv : hep - th/0501099 ] ; + s. e. shandera and s .- h .
h. tye , jcap * 0605 * , 007 ( 2006 ) [ arxiv : hep - th/0601099 ] ; + r. bean , s. e. shandera , s .- h .
h. tye and j. xu , jcap * 0705 * , 004 ( 2007 ) [ arxiv : hep - th/0702107 ] .
m. wyman , l. pogosian and i. wasserman , phys .
d * 72 * , 023513 ( 2005 ) [ erratum - ibid .
d * 73 * , 089905 ( 2006 ) ] [ arxiv : astro - ph/0503364 ] ; + m. landriau and e. p. s. shellard ,
d * 69 * 023003 ( 2004 ) , astro - ph/0302166 ; + l. pogosian , m. wyman and i. wasserman , jcap 09 ( 2004 ) 008 , astro - ph/0403268 ; + e. jeong and g. f. smoot , astro - ph/0406432 ; + j .- h .
p. wu , astro - ph/0501239 .
l. pogosian and t. vachaspati , phys .
d * 60 * , 083504 ( 1999 ) [ arxiv : astro - ph/9903361 ] .
u. seljak and m. zaldarriaga , astrophys .
j. * 469 * , 437 ( 1996 ) [ arxiv : astro - ph/9603033 ] . l. pogosian , i. wasserman and m. wyman , arxiv : astro - ph/0604141 . c.j.a.p .
martins and e.p.s .
shellard , phys . rev . *
d54 * 2535 ( 1996 ) ; + c.j.a.p .
martins and e.p.s .
shellard , phys .
rev . * d65 * , 043514 ( 2002 )
. g. r. vincent , m. hindmarsh and m. sakellariadou , phys .
d * 55 * ( 1997 ) 573 [ arxiv : astro - ph/9606137 ] .
a. gangui , l. pogosian and s. winitzki , phys .
d * 64 * , 043001 ( 2001 ) [ arxiv : astro - ph/0101453 ] .
m. g. jackson , n. t. jones and j. polchinski , jhep * 0510 * , 013 ( 2005 ) , hep - th/0405229 .
h. tye , i. wasserman and m. wyman , phys .
d * 71 * , 103508 ( 2005 ) [ erratum - ibid .
d * 71 * , 129906 ( 2005 ) ] [ arxiv : astro - ph/0503506 ] .
l. leblond and m. wyman , phys .
d * 75 * , 123522 ( 2007 ) [ arxiv : astro - ph/0701427 ] . http://lambda.gsfc.nasa.gov/toolbox/tb_camb_form.cfm ; + a. challinor and a. lewis , phys . rev .
d * 71 * , 103010 ( 2005 ) [ arxiv : astro - ph/0502425 ] .
l. pogosian and m. wyman , phys .
d * 77 * , 083509 ( 2008 ) [ arxiv:0711.0747 [ astro - ph ] ] .
b. allen and e. p. s. shellard , phys .
lett . * 64 * , 119 ( 1990 ) .
a. c. s. readhead _ et al .
_ , astrophys .
j. * 609 * , 498 ( 2004 ) [ arxiv : astro - ph/0402359 ] .
a. lewis , phys .
d * 70 * , 043011 ( 2004 ) [ arxiv : astro - ph/0406096 ] ; + t. r. seshadri and k. subramanian , phys .
d * 72 * , 023004 ( 2005 ) [ arxiv : astro - ph/0504007 ] ; + subramanian , k. and barrow , j. d. , mnras , 335 ( 2002 ) l57 .
k. j. mack , d. h. wesley and l. j. king , arxiv : astro - ph/0702648 ; + d. f. chernoff and s. h. h. tye , arxiv:0709.1139 [ astro - ph ] ; + k. kuijken , x. siemens and t. vachaspati , arxiv:0707.2971 [ astro - ph ] .
t. damour and a. vilenkin , phys .
d * 64 * , 064008 ( 2001 ) [ arxiv : gr - qc/0104026 ] ; + t. damour and a. vilenkin , phys . rev .
d * 71 * , 063510 ( 2005 ) [ arxiv : hep - th/0410222 ] ; + x. siemens , v. mandic and j. creighton , phys .
lett . * 98 * , 111101 ( 2007 ) [ arxiv : astro - ph/0610920 ] .
a. a. fraisse , c. ringeval , d. n. spergel and f. r. bouchet , arxiv:0708.1162 [ astro - ph ] ; + s. amsel , j. berger and r. h. brandenberger , arxiv:0709.0982 [ astro - ph ] . | cosmic string networks generate cosmological perturbations actively throughout the history of the universe .
thus , the string sourced anisotropy of the cosmic microwave background is not affected by silk damping as much as the anisotropy seeded by inflation .
the spectrum of perturbations generated by strings does not match the observed cmb spectrum on large angular scales ( @xmath0 ) and is bounded to contribute no more than @xmath1 of the total power on those scales .
however , when this bound is marginally saturated , the anisotropy created by cosmic strings on small angular scales @xmath2 will dominate over that created by the primary inflationary perturbations .
this range of angular scales in the cmb is presently being measured by a number of experiments ; their results will test this prediction of cosmic string networks soon . |
the dynamic phase transition ( dpt ) in a ferromagnetic system below its critical temperature was first observed in numerical solutions of a mean - field model exposed to an oscillating magnetic field @xcite .
it was then studied further , both in mean - field models @xcite and in kinetic monte carlo ( kmc ) simulations @xcite . a review of this early work can be found in ref .
more recently , the study of the dpt has expanded to include varying ( and often more physical ) model geometries .
these include mean - field studies of domain - wall motion in an anisotropic xy model in one dimension @xcite , kmc simulations of a three - dimensional ising system @xcite , and kmc simulations of a uniaxially anisotropic heisenberg system in an off - axial field @xcite , an elliptically polarized applied field @xcite , and with the effect of a thin - film surface energy @xcite .
the phenomenon has also been observed in simulations of co oxidation under oscillating co pressure @xcite .
further simulation studies of the dpt in the two - dimensional kinetic ising model have appeared @xcite , as well as analytical studies of the dpt @xcite . here , we concentrate on the dpt in the two - dimensional kinetic ising model . it was observed in simulations of this model that there exists a singularity at a critical period of the applied oscillating field @xcite , and that the critical exponents @xmath6 and @xmath7 ( and , with less accuracy , @xmath8 ) are consistent with the universality class of the equilibrium two - dimensional ising transition in zero field @xcite @xmath9 . in those studies ,
the techniques of finite - size scaling were extended to the study of the dynamic order parameter ( @xmath0 , defined in sec .
[ s : model ] ) in the non - equilibrium steady state .
this provided evidence for a diverging correlation length at a critical value of the period . in particular , because the field conjugate to @xmath0 and a fluctuation - dissipation relation were not known , a susceptibility could not be measured directly , and the scaled variance @xmath10 , where @xmath11 is the linear system size , was used as a proxy . an analytical argument ,
based on the correspondence of the two - dimensional kinetic ising model and the continuous , two - dimensional ginzburg - landau model at the equilibrium critical point , provided an effective hamiltonian for the non - equilibrium system and confirmed that the dpt is in the ising universality class @xcite .
these findings are consistent with earlier symmetry arguments that any continous phase transition in a stochastic cellular automaton that preserves the ising up - down symmetry should be in the equilibrium ising universality class @xcite .
recently , experiments were performed on a @xmath12_3 $ ] multilayer film with strong uniaxial anisotropy @xcite , whose equilibrium behavior is known to be ising - like @xcite .
the film was exposed to an oscillating ( sawtooth ) applied field with varying period , in the presence of constant ` bias ' magnetic fields @xmath1 of varying strength and sign .
( the bias field is defined explicitly in sec .
[ s : model ] . ) the behaviors of the dynamic order parameter and its variance , as functions of the applied field period and the bias field , provided strong evidence for the presence of the dpt in this experimental system .
the observed behavior of the order parameter with respect to the bias field supported previous conjectures that the conjugate field could include the period - averaged magnetic field as an important component , and stimulated the numerical investigations in this paper .
this paper is organized as follows . in sec .
[ s : model ] , we describe the two - dimensional kinetic ising model and our computational methods . in sec .
[ s : delta.scaling ] , we verify directly the scaling of the dynamic order parameter with respect to the period - averaged magnetic field at the critical period , with scaling exponent @xmath13 , in agreement with the equilibrium ising transition . in secs .
[ s : finite.size ] and [ s : first.compare ] , we derive the expected asymptotic scaling functions in a finite - size scaling analysis of the dynamic phase transition with non - zero period - averaged bias field , and then compare the expected scaling of the dynamic order parameter to our numerical results . in sec . [ s : fluct.relation ] , we present numerical data assessing the applicability of a fluctuation - dissipation relation ( fdr ) to this far - from - equilibrium system .
we then in sec .
[ s : second.compare ] compare the expected scaling of the susceptibility ( of the dynamic order parameter ) to our numerical data , using the results of sec . [ s : fluct.relation ] to reconcile our findings with previous results on the scaling of the fluctuations of the dynamic order parameter .
finally , we present a summary of our results in sec . [ s : conclusion ] .
in order to facilitate comparison with previous results , we employ the same model and computational method as in ref .
specifically , we perform kinetic monte carlo ( kmc ) simulations of a two - dimensional periodic square lattice of ising spins @xmath14 , which can take only the values @xmath15 the hamiltonian of the model is @xmath16 where @xmath17 is the ferromagnetic exchange interaction , @xmath18 runs over all nearest - neighbor pairs , @xmath19 runs over all @xmath20 lattice sites , and @xmath21 is an oscillating , spatially uniform applied magnetic field .
the form of @xmath21 is here taken as a square wave with amplitude @xmath22 and period @xmath23 , measured in monte carlo steps per spin ( mcss ) .
the square - wave form not only allows for more efficient kmc simulation , but also reduces the critical period and the finite - size effects for the dpt @xcite .
other symmetric field shapes , such as sinusoidal @xcite and sawtooth @xcite , yield essentially the same results , but with a larger critical period and with stronger finite - size effects .
the glauber single - spin - flip mc algorithm with updates at randomly chosen sites is used , in which each attempted spin flip is accepted with probability @xmath24 where @xmath25 is the energy change that would result from acceptance of the spin flip , and @xmath26 is the absolute temperature in energy units ( i.e. , with boltzmann s constant set to unity ) .
all simulations were performed at @xmath27 , where @xmath28 is the equilibrium critical temperature of the square - lattice ising ferromagnet in zero applied field @xcite .
the system responds to the oscillating field via the time - dependent magnetization per site , @xmath29 the dynamic order parameter is defined as the average of @xmath30 over a given field cycle @xmath31 @xcite : @xmath32 we define the bias field , so named because it measures the shift ( or ` bias ' ) of the periodic field toward either negative or positive field values , as the period - averaged magnetic field , @xmath33 this definition applies generally to any periodic magnetic field @xmath21 . in this paper ,
the applied field consists of a square wave with period @xmath23 superposed with a constant magnetic field . applying ( [ bias.def ] ) ,
since the period - average of the square - wave field is zero , the bias field @xmath1 in this case is simply equal to the superposed constant magnetic field .
in the two - dimensional equilibrium ising model , the critical isotherm is given ( in the thermodynamic limit , i.e. , as @xmath34 as @xmath35 where the critical exponent @xmath36 @xcite .
for finite systems , this relationship breaks down when the infinite - system correlation length , @xmath37 , which diverges as @xmath38 , becomes comparable to the linear system size @xmath11 .
the relationship also naturally breaks down at larger fields away from the critical region .
therefore , a plot of @xmath39 vs @xmath40 for a given system size @xmath11 will follow the power law ( [ equ.powerlaw ] ) for a range of @xmath40 near the critical value @xmath41 , with this range extending to smaller @xmath40 as @xmath11 is increased @xcite .
we can determine directly whether the non - equilibrium system exhibits a similar relationship , @xmath42 in an analogous way . in fig .
[ delta.scaling ] , we plot @xmath43 vs @xmath1 at the critical value of the period , @xmath44 . in previous work ,
the reversal time for the magnetization , following instantaneous reversal of the uniform magnetic field @xmath40 at @xmath45 , was found as @xmath46 mcss @xcite .
the critical scaled half - period for the square waveform was determined to be @xmath47 @xcite .
this yields @xmath48 mcss , and in our simulations and analysis in this paper we use @xmath49 mcss .
a power - law dependence is indeed seen to hold in fig .
[ delta.scaling ] , within a range which extends to lower values of @xmath1 as @xmath11 is increased .
we fit the @xmath50 data between the points labeled a and b in fig . [ delta.scaling ] , finding a statistically significant fit with power - law exponent @xmath51 . as including points with @xmath52
was found to greatly reduce the statistical significance of the fit , the value @xmath53 serves as a boundary of the scaling region at @xmath44 .
this result is consistent with an exponent @xmath54 , suggesting that the bias field @xmath1 , for these parameters and the square waveform , is the dominant component of a conjugate field which exhibits the same scaling exponent in the dpt as does the applied magnetic field in the equilibrium ising transition .
to provide more complete evidence that @xmath1 is the dominant component of the field conjugate to @xmath43 , in the next several sections we will demonstrate data collapse onto a two - parameter finite - size scaling function for the system - size dependent quantity @xmath55 at points @xmath56 , for lattice sizes @xmath11 = 90 , 128 , 180 , and ( in several cases ) @xmath50 , using the critical exponents for the equilibrium ising system . in this section ,
we briefly review the theory of finite - size scaling as it applies to this system .
we then determine the expected asymptotic forms of the scaling functions , which are compared in later sections of the paper to our computational data .
the theory of finite - size scaling @xcite states that near a continuous phase transition , the singular part of the free - energy density for a @xmath57-dimensional system of linear size @xmath11 can be written as @xmath58 where @xmath59 , @xmath40 ( in units of @xmath60 ) is the field conjugate to the order parameter , @xmath8 is the critical exponent for the correlation length , @xmath61 is the exponent for the order parameter , @xmath62 is the exponent for the critical isotherm , and @xmath63 are scaling functions above ( + ) and below ( @xmath64 ) the critical point .
this yields for the order parameter at finite @xmath11 @xmath65 where the exponent for @xmath11 in the prefactor is obtained by using the hyperscaling relation @xmath66 and the exponent equality @xmath67 .
further differentiation yields the susceptibility , @xmath68 where the exponent for @xmath11 in the prefactor is obtained by using the exponent equality @xmath69 .
it has previously been shown analytically that the dpt for a sinusoidal applied field , which is symmetric under @xmath70 and so which can safely be assumed to have @xmath71 , has an effective ginzburg - landau free - energy density in the same universality class as the equilibrium ising model @xcite .
it therefore appears reasonable to write corresponding scaling functions for the dynamic order parameter @xmath43 and its associated susceptibility @xmath72 , @xmath73 and @xmath74 where @xmath75 , and @xmath76 is the ( as yet unknown ) field conjugate to @xmath43 . in this paper
we express @xmath76 ( and @xmath1 ) in units of the exchange constant , @xmath77 , so that the second scaling parameter in eqs .
[ eq : qscale ] and [ eq : chihatscale ] is dimensionless .
the specific form , @xmath78 , with which @xmath76 is assumed to enter the second scaling parameter needs more theoretical investigation , and could conceivably change as the theory of the dpt is further developed .
however , this should not affect our conclusions @xcite .
computational results for sinusoidal and square - wave fields , which both are symmetric under @xmath70 and so presumably have @xmath79 , have previously confirmed the scaling behavior with respect to @xmath80 .
the exponent values were determined as @xmath81 , @xmath82 , and @xmath83 @xcite , consistent with the exact values for the two - dimensional equilibrium ising model , @xmath84 , @xmath85 , and @xmath86 .
we now determine the expected asymptotic forms of the scaling functions @xmath87 and @xmath88 , where we emphasize that the @xmath89 subscript indicates that the scaling functions refer to the range @xmath90 , and where the scaling parameters are @xmath91 and @xmath92 .
we expect @xmath93 ( independent of @xmath94 ) and @xmath95 , where @xmath96 was used to obtain the exponent for @xmath11 in @xmath55 .
we expect that @xmath97 and @xmath98 ( both independent of @xmath99 ) , where @xmath100 was used to obtain the exponent for @xmath11 in @xmath101 .
+ thus , the asymptotic forms of the scaling functions are expected to be @xmath102 and @xmath103
in fig . [ f+.vs.y1](a ) , using the equilibrium values @xmath104 and @xmath105 in calculating @xmath106 , we present a plot of the scaling function @xmath107 vs @xmath99 for different values of @xmath94 , for lattice sizes @xmath108 , 128 , and 180 . here and for the remainder of the paper , exponents with the subscripts ` d ' and ` e ' refer to the behavior of the nonequilibrium system ( with a dynamic phase transition ) and the equilibrium system , respectively .
the scaling function exhibits a power - law dependence in the regime @xmath109 , which is consistent with eq .
( [ eq : fscal ] ) . at progressively
larger values of the constant @xmath94 , the power - law scaling can be seen to begin at increasing values of @xmath99 , as would be expected . a best - fit line to the final five points of the @xmath110 data at @xmath111 yields an estimate of the scaling exponent @xmath112 in eq .
( [ eq : fscal ] ) .
this is consistent with the previous results for @xmath79 cited above @xcite , and it supports the hypothesis that @xmath113 . in fig .
[ f+.vs.y1](b ) , we present just the data for @xmath111 , including additional data points at @xmath114 and @xmath115 .
the data deviate from the power - law behavior for @xmath116 at @xmath117 , for @xmath118 at @xmath119 , and for @xmath120 at @xmath121 .
this locates the boundary of the scaling regime ( for @xmath111 ) at @xmath122 . in fig .
[ f+.vs.y2.smally2 ] , again using @xmath104 and @xmath105 in calculating @xmath107 , we plot the scaling function @xmath107 vs @xmath94 at different values of @xmath99 , in order to examine the scaling behavior for @xmath109 . for the constant values @xmath123 and 149 , power - law scaling can be observed in the regime @xmath109 .
a best - fit line to the @xmath124 data for the five points from @xmath111 to @xmath125 yields a scaling exponent of @xmath126 , which is consistent with the value of @xmath127 expected from eq .
( [ eq : fscal ] ) . in order to investigate the scaling of @xmath107 in the asymptotic limit @xmath128 , we plot in fig .
[ f+.vs.y2.largey2 ] the scaling function @xmath87 vs @xmath94 at the critical period @xmath129 ( i.e. , @xmath130 ) , at lattice sizes @xmath131 and 256 . in the range @xmath132 , power - law scaling
is observed for all three lattice sizes . for @xmath133 ,
the data deviate from power - law scaling , with the smallest lattice size deviating first , as expected in a finite - size scaling plot .
a fit of the @xmath50 data in the range from @xmath134 to @xmath125 produces a scaling exponent @xmath135 . since the constant factors @xmath136 and @xmath137 do not affect the fit of the scaling exponent , this is the same exponent found in the fit of @xmath43 vs @xmath1 in fig . [ delta.scaling ] .
the reciprocal of this scaling exponent is thus @xmath51 , which is consistent with the exponent of the critical isotherm , @xmath36 , in the equilibrium ising model .
the comparison of the second scaling function , @xmath138 ) , to numerical data is more clearly presented after the relationship of the susceptibility @xmath101 and the scaled variance @xmath139 has been examined .
therefore , we present in the next section numerical results on the extent of applicability of an fdr between @xmath101 and @xmath139 , before turning in sec . [ s : second.compare ] to the second scaling function .
fdrs , such as the einstein relation , green - kubo relations , etc . , hold a central place in equilibrium statistical mechanics .
this is essentially a consequence of detailed balance and the role of the partition function as a moment - generating function , and thus such relations can not be readily extended to nonequilibrium steady states .
however , it has recently been shown that certain fdrs can be extended to far - from equilibrium steady states by use of an _ effective temperature _ @xcite . here
we will therefore consider whether the nonequilibrium susceptibility and the scaled variance of the dynamic order parameter can be related as @xmath140 with an effective temperature @xmath141 , in a way analogous to the equilibrium fdr , @xmath142 in which @xmath26 is the temperature . as mentioned in secs .
[ s : intro ] and [ s : finite.size ] , this conjecture motivated the use in previous work of the scaled variance @xmath139 as a proxy for @xmath101 in investigating the scaling behavior of the nonequilibrium system near its critical period . to test the extent to which eq .
( [ eq : fdr ] ) holds , we computed values of @xmath101 and @xmath139 for a range of periods from @xmath143 to 250 mcss and a range of bias fields from @xmath144 to an upper limit between @xmath145 and @xmath146 .
( the bias field necessary to ` saturate ' the nonequilibrium system , i.e. , to produce values of @xmath101 and @xmath139 near zero , increases as the period is increased . )
the computations were perfomed at @xmath110 .
the quantity @xmath101 was computed directly as a numerical derivative : @xmath147 the choice of @xmath148 was found to produce sufficiently accurate values of the numerical derivative across the range of bias fields studied .
the results for periods @xmath149 through 190 mcss are shown in fig .
[ fluct.susc ] .
a linear relationship is seen to exist between @xmath101 and @xmath139 , for each value of @xmath23 , over a wide range of @xmath101 values . at each period , the dependence becomes nonlinear below a certain value of @xmath101 , as illustrated for periods @xmath150 , 170 , and 190 mcss in fig .
[ fluct.susc.lowchi ] . since low values of the susceptibility @xmath101 correspond to large values of the bias field @xmath1
, we interpret this breakdown of linearity as an indication that the fdr in eq .
( [ eq : fdr ] ) holds only in a scaling regime around the critical point , i.e. , for a limited range of @xmath1 around @xmath151 .
the relationship between @xmath139 and @xmath101 at the higher periods , @xmath152 and @xmath153 mcss , is more complicated , as shown in fig .
[ fluct.susc2 ] . at @xmath152 mcss , following the nonlinear regime at low @xmath101 , there is a linear relationship with slope @xmath154 up to @xmath155 , followed by a second distinct linear dependence with slope @xmath156 above @xmath157 . at @xmath153 mcss
, the initial nonlinear dependence is again present .
then the first linear regime has @xmath158 up to @xmath155 , and is followed by a regime which can be characterized as either linear with very gentle slope @xmath159 , or as an effective ` saturation ' of @xmath139 past @xmath157 . in fig .
[ teff.theta ] we plot the best - fit slopes from figs .
[ fluct.susc ] and [ fluct.susc2 ] , which according to eq .
( [ eq : fdr ] ) represent estimates of @xmath141 , vs the scaling parameter @xmath160 .
we have included in the plot the slopes of both linear regimes for the values @xmath161 and 0.825 ( @xmath152 and 250 mcss ) . for @xmath80 below 0.4 ( @xmath162 mcss ) , @xmath141 increases with @xmath80 in a way not inconsistent with a linear relationship ( with slope @xmath163 )
. it may be interesting to note that an extrapolation of the linear relationship to @xmath164 ( @xmath44 ) yields the value @xmath165 , which is significantly higher than the critical temperature , @xmath166 , of the equilibrium ising system .
however , one should not put too much emphasis on the numerical values of @xmath141 @xcite , as they could easily be changed .
for instance , if @xmath1 is only _ proportional to _ the full conjugate field @xmath76 , with a proportionality constant different from unity , this would trivially change @xmath141 in eq .
( [ eq : fdr ] ) . the important result , which we have demonstrated to hold in the critical region , is the linear relationship between @xmath139 and @xmath101 .
we can thus characterize the extent of applicability of an fdr to the dpt above the critical period as follows .
for @xmath167 , an fdr holds outside of a small nonlinear regime at low @xmath101 ( high @xmath1 ) , with an effective temperature @xmath141 which increases approximately linearly with @xmath80 . for @xmath80 above 0.4 ,
two linear relationships appear to exist between @xmath139 and @xmath101 in separate regimes , making it impossible to define a unique @xmath141 at a given value of @xmath80 .
an understanding of the nonlinear regime , which is present at low @xmath101 for all periods examined , as well as of the complication of the fdr above @xmath168 , would be highly desirable .
we hope that these numerical results can stimulate the development of , as well as test the accuracy of , a theoretical description of the non - equilibrium steady states produced in the presence of non - zero @xmath1 for this dpt .
we now test the asymptotic scaling forms for @xmath169 in eq .
( [ eq : gscal ] ) .
first , we note that in performing least - squares fits , one normally requires the goodness - of - fit parameter @xmath170 , i.e. , the probability that ( assuming the fit relationship were true ) random error alone could produce the observed data , to be greater than @xmath171 to consider the fit reasonable . within this section , however , and in the captions to figs .
[ gvsy1.h2 ] through [ gvsy1.h0.xabsq ] , it will be useful for descriptive purposes to refer to scaling exponents resulting from attempts at least - squares fits with @xmath170 values below this acceptable range .
we will refer to the results of such unsuccessful fitting attempts as ` nominal ' scaling exponents , and for clarity will report the value of the parameter @xmath170 for each scaling exponent presented in this section . in fig .
[ gvsy1.h2 ] , we plot the scaling function @xmath169 vs @xmath99 for @xmath134 at @xmath120 , using @xmath172 and @xmath105 to calculate @xmath169 , and evaluating @xmath101 numerically according to eq .
( [ eq : num.deriv ] ) .
in addition , we plot in the same figure the scaling function @xmath173 vs @xmath99 , for the same values of @xmath94 and @xmath11 .
a fit to all four @xmath169 data points yields a scaling exponent @xmath174 ( @xmath175 ) , while a fit to the last three @xmath169 data points yields a scaling exponent @xmath176 ( @xmath177 ) . we will provide evidence in the next paragraph that only the last three @xmath169 data points , and not the first , satisfy the asymptotic condition @xmath109 .
thus , these data are consistent with power - law scaling of @xmath101 with exponent @xmath178 .
attempts to fit the @xmath179 data to all four and the last three data points yield nominal scaling exponents @xmath180 ( @xmath181 ) and @xmath182 ( @xmath183 ) , respectively .
thus , while in fig . [ gvsy1.h2 ] it appears that the power - law relationships with these nominal scaling exponents give respectable visual fits to the @xmath179 data , there are variations in the data which , while small , are larger than the statistical error bars , and which prevent a statistically significant fit .
we will describe the causes of these variations later in this section .
we present in fig .
[ gvsy1.h0 ] a plot of @xmath169 and @xmath179 vs @xmath99 at @xmath184 , again for @xmath120 , for a larger range from @xmath185 to 477 . with @xmath184
, we expect that @xmath185 ( and indeed , essentially any nonzero value of @xmath99 ) should satisfy the asymptotic scaling condition @xmath109 .
an attempt to fit to all six @xmath169 data points in fig .
[ gvsy1.h0 ] gives a nominal scaling exponent @xmath186 ( @xmath187 ) , while a fit to the first five points ( @xmath185 through 280 ) gives a scaling exponent @xmath188 ( @xmath189 ) .
excluding the first data point at @xmath185 has little effect on either fit .
this supports the assumption that with @xmath184 , the asymptotic scaling condition @xmath109 holds for @xmath185 , while for @xmath134 , as used in fig .
[ gvsy1.h2 ] , the asymptotic scaling condition does not hold for @xmath185 . attempts to fit all of , and the first five of , the @xmath179 data points to power - law scaling again yield only nominal scaling exponents @xmath190 ( @xmath181 ) and @xmath191 ( @xmath181 ) , respectively .
we considered that the low statistical significance of the fits to the @xmath179 data could be caused by underestimation of the error bars on @xmath139 .
these error bars were calculated by ( i ) finding the correlation time in the numerical data series @xmath192 from the simulation , and sampling data at intervals of twice the correlation time ; ( ii ) dividing this sampled data into @xmath193 groups and calculating the value of @xmath139 within each group ; ( iii ) finding the mean and standard error of this collection of @xmath139 values . as a check on self - consistency
, we performed several independent calculations of @xmath139 by this method , and found that the standard error of these values ( corrected for small sample size ) was comparable to the standard error found within each calculation .
thus , we have strong evidence that the error bars for @xmath139 ( and @xmath179 ) are accurate .
these scaling results can be understood in light of the observations in sec .
[ s : fluct.relation ] on the relationship between @xmath139 and @xmath101 .
we can reasonably assume that the breakdown in scaling of @xmath169 past @xmath114 ( @xmath194 mcss , @xmath195 ) in fig .
[ gvsy1.h0 ] occurs because this is the boundary of the critical region .
the small variations of the @xmath179 data for @xmath196 mcss in fig .
[ gvsy1.h0 ] from a scaling relationship with exponent @xmath197 then have three main causes .
the first cause is the multiplication of the accurately scaling function @xmath169 by the @xmath80- and @xmath99-dependent value @xmath141 , according to eq .
( [ eq : fdr ] ) .
however , such a variation would also occur in an analogous plot for the scaling of @xmath198 vs @xmath199 in the equilibrium ising model , since the susceptibility @xmath200 scales with exponent @xmath201 , and @xmath202 is related to @xmath203 by the @xmath204-dependent temperature @xmath26 according to eq .
( [ eq : fdr0 ] ) .
this effect is small enough to be neglected in equilibrium critical scaling , and , since the change in @xmath141 from @xmath205 to @xmath168 is comparable to the change in @xmath26 from @xmath206 to @xmath207 in the equilibrium transition , it can also be neglected here .
the second cause is the presence of the nonlinear regimes in the plots of @xmath139 vs @xmath101 at low @xmath101 , resulting in non - zero @xmath139-intercepts in the application of eq .
( [ eq : fdr ] ) to figs .
[ fluct.susc ] and [ fluct.susc2 ] .
because of this , division of the @xmath139 data by the appropriate @xmath141 values ( given in the caption of fig
. [ fluct.susc ] ) does not quite reproduce the corresponding @xmath101 data , and ( even below @xmath168 ) does not quite result in scaling consistent with @xmath208 with statistical signficance .
the third and most signficant cause of the variations of the @xmath179 data is the ` doubly linear ' behavior observed in fig .
[ fluct.susc2 ] for @xmath209 , which prevents identification of a unique @xmath141 in this range .
the assumption that @xmath139 can be used as a proxy for @xmath101 is thus fairly well justified close to the critical period , where @xmath141 varies over a limited range and the more complicated effects observed at @xmath209 are not relevant .
this is supported by fig .
[ gvsy1.h2 ] , where the data points cluster closely around the line corresponding to power - law scaling with exponent @xmath210 .
however , because of the first two causes just described , there are small systematic variations in the @xmath179 data which prevent a statistically significant fit to a pure power - law relationship as a function of @xmath99 . in order to clarify the relationship of these scaling results to those in previous work , we also plot in fig .
[ gvsy1.h0.xabsq ] data of @xmath211 vs @xmath99 at @xmath184 , again using the equilibrium values @xmath172 and @xmath105 to calculate @xmath212 .
this can be directly compared to fig .
11(d ) in ref .
@xcite , in which the quantity we call @xmath213 was called @xmath139 . attempted fits to all five data points and to the last four data points of @xmath212 in fig .
[ gvsy1.h0 ] produce nominal scaling exponents @xmath214 and @xmath215 ( both with @xmath181 ) .
the agreement in fig .
11(d ) of ref .
@xcite of the line with slope @xmath216 with the data for @xmath217 must therefore be viewed as qualitative .
the method used in ref .
@xcite to numerically estimate @xmath218 , however , which involves finite - size scaling at the critical period , is fully consistent with the results of this paper , since at each period with @xmath167 we have found that the fdr in eq .
( [ eq : fdr ] ) holds to a very good approximation . finally , in fig .
[ g+.vs.y2 ] , we plot @xmath179 vs @xmath94 at @xmath129 , to study its scaling in the regime @xmath128 . as just noted , the use of @xmath139 as a proxy for @xmath101 is well justified at @xmath44 by our results .
power - law scaling is perhaps suggested in the range @xmath132 for @xmath120 , and it is clearly obeyed from @xmath219 to @xmath220 for @xmath221 .
the scaling exponent was determined as @xmath222 , which is consistent with the corresponding equilibrium value @xmath223 .
in this article , we have continued the computational study of the dynamic phase transition ( dpt ) in the two - dimensional kinetic ising model exposed to a periodically oscillating field , which was begun in refs .
we have established two distinct but related results about the field conjugate to the dynamic order parameter .
first , we have identified the period - averaged magnetic field , or ` bias field ' , @xmath1 as an important component of the full conjugate field .
this claim is supported by numerical evidence that the dynamic order parameter and its susceptibility follow critical scaling with respect to @xmath1 .
in particular , the scaling exponent @xmath224 of the conjugate field was determined for the first time , and found by finite - size scaling analysis of large - scale kinetic monte carlo simulations to be equal to the critical - isotherm exponent for the equilibrium ising transition , @xmath225 .
in agreement with previous results @xcite , the dynamic scaling exponents @xmath226 , @xmath227 , and @xmath228 were also found to equal their equilibrium ising counterparts , @xmath229 , @xmath230 , and @xmath231 . these results further strengthen previous numerical @xcite and analytical @xcite claims that the dpt in a periodically driven two - dimensional kinetic ising model belongs to the universality class of the equilibrium two - dimensional ising model . however , with respect to the direct applicability of the symmetry arguments of refs .
@xcite , we caution the reader that what is claimed in the present paper ( as well as in ref .
@xcite ) is only equivalence of the _ phase transitions _ in the driven kinetic ising model and the equilibrium ising model . _
outside _ the critical region , it is neither clear how closely @xmath232 and @xmath1 play the roles of @xmath233 and the ordinary magnetic field , respectively , nor how closely the dynamic order parameter , @xmath0 , corresponds to the average equilibrium magnetization . from our discussion of the fdr in sec .
[ s : fluct.relation ] , it appears likely that one or more of these relations break down outside the critical region .
much theoretical work remains to be done in this area .
the second main result of this article is that a fluctuation - dissipation relation ( fdr ) , that is , a proportionality relation between the scaled fluctuations @xmath234 and the susceptibility @xmath101 with a slope we have called @xmath141 , holds for a range of periods above @xmath235 and for a range of bias fields around @xmath144 .
we stress again that we have found the fdr of eq .
( [ eq : fdr ] ) to hold _ only in the critical region _ in this nonequilibrium system , in contrast to the equilibrium fdr of eq .
( [ eq : fdr0 ] ) which follows directly from the partition function , and which thus holds everywhere .
we note that , for the parameters used in our computation at least , the critical region in which the nonequilibrium fdr holds ( @xmath236 mcss ) is somewhat smaller than the critical region in which power - law scaling is obeyed ( @xmath196 mcss ) . in previous work , when the conjugate field had not been identified , the scaled fluctuations @xmath139 were used as a proxy for the ( then unknown ) quantity @xmath101 .
the evidence for the fdr presented here shows this assumption to be fully justified at the critical period ( see fig .
[ g+.vs.y2 ] ) , and to be a very good approximation nearly as good as the use of the scaled fluctuations as a proxy for the susceptibility in the equilibrium ising model in the critical region where the fdr holds .
there are at least three further computational projects suggested by the progress reported here .
the first is to investigate whether the field @xmath1 functions as the conjugate field , with scaling exponents consistent with the equilibrium ising transition for periods @xmath237 , below the critical period . in the equilibrium system
, the study of critical scaling in nonzero field for @xmath238 is complicated by the long time correlations and strong finite - size effects which accompany the bimodal distributions of magnetization below @xmath239 .
similar effects would complicate the investigation of scaling with respect to @xmath1 in the dpt for @xmath237 , but the advanced techniques @xcite which make the equilibrium simulations tractable do not extend obviously to the nonequilibrium case .
the second computational project suggested is to determine the nature of the full conjugate field @xmath76 .
the third project would be to study the fdr at different values of the temperature , @xmath26 , and the amplitude , @xmath240 , of the driving field .
finally , we remark that it would be very desirable to extend the current understanding of the theory of nonequilibrium steady states to include the conjugate field @xmath1 , the fdr found in the critical region , and the scaling of @xmath1 .
research at florida state and mississippi state universities was supported by nsf grant no .
dmr-0444051 , and at clarkson university by nsf grant no . dmr-0509104 .
this research also used resources of the national center for computational sciences at oak ridge national laboratory , which is supported by the office of science of the u.s .
department of energy under contract no .
de - ac05 - 00or22725 .
until the theory of the dpt , e.g. in ref .
@xcite , is extended to include @xmath76 , it is unclear in what form it should enter the dimensionless scaling parameter , i.e. as @xmath241 , @xmath242 , or otherwise .
here we have chosen the form @xmath78 . since our numerical data
is all taken at a single temperature , @xmath243 , a change in this form would at most multiply values of @xmath94 by a constant factor , and would not affect the scaling relations reported in this paper . to avoid misunderstandings , note that it is certainly possible to define other sensible effective temperatures for an ising - like model , so that one should be clear about the definition being used in a given context .
for example , the nonequilibrium fdr proposed in refs .
@xcite connects _ two - time _ response and correlation functions , and applies to critical ageing in a two - parameter family of ising - like models .
this family includes many models without detailed balance ( e.g. , the noisy voter and majority models ) , and one model the equilibrium ising model with glauber dynamics with detailed balance .
in contrast , our eq .
( [ eq : fdr ] ) concerns _ single - time _ response and correlations , and the kinetic ising model driven by an oscillating field is not included in the two - parameter family of models considered in refs .
@xcite . , vs bias field , @xmath1 , at @xmath44 for @xmath244 and 256 .
a least - squares fit to power - law scaling of the @xmath50 data , in the range between the labels a and b above , produced a statistically significant fit with scaling exponent @xmath245 ( corresponding to @xmath51 ) .
the dotted line corresponds to the scaling exponent @xmath246 .
a reference line representing scaling with the equilibrium ising exponent , @xmath36 ( @xmath247 ) , is shown as the dashed line .
, width=432 ] vs @xmath99 for lattice sizes @xmath248 and @xmath249 .
the values of @xmath99 plotted are 4.00 , 17.2 , 30.3 , 43.4 , 69.7 , 147 , and ( in * ( b ) * ) 280 and 477 . * ( a ) * the data are shown for the values of @xmath94 listed on the plot .
the best - fit line for the last five points of the @xmath110 data at @xmath111 , with slope @xmath250 , is included along with a reference line with the slope @xmath251 . *
( b ) * the data for @xmath111 with two additional @xmath99 values illustrates the boundary of the regime of power - law scaling .
, title="fig:",width=336 ] vs @xmath99 for lattice sizes @xmath248 and @xmath249 .
the values of @xmath99 plotted are 4.00 , 17.2 , 30.3 , 43.4 , 69.7 , 147 , and ( in * ( b ) * ) 280 and 477 . * ( a ) * the data are shown for the values of @xmath94 listed on the plot .
the best - fit line for the last five points of the @xmath110 data at @xmath111 , with slope @xmath250 , is included along with a reference line with the slope @xmath251 . * ( b ) * the data for @xmath111 with two additional @xmath99 values illustrates the boundary of the regime of power - law scaling .
, title="fig:",width=336 ] vs @xmath252 for lattice sizes @xmath253 128 , and 180 , for the constant values of @xmath99 labeled in the plot .
the values of @xmath94 used are @xmath254 and @xmath255 .
the dotted line represents the best fit to the first five points of the @xmath110 data at @xmath124 , and has a slope of @xmath126 .
the dashed line shows the slope value of 1 , expected from eq .
( [ eq : fscal ] ) . ,
width=432 ] vs @xmath256 for lattice sizes @xmath257 180 , and 256 , at the critical period @xmath235 , where @xmath130 . in the @xmath221 data ,
near the values @xmath258 and 335 , two closely spaced data points are actually plotted .
the best - fit line to the @xmath50 data in the range @xmath259 , shown as a dotted line in the plot , corresponds to a scaling exponent @xmath245 .
a reference line corresponding to scaling exponent @xmath260 is also shown .
these results are in complete agreement with those shown in fig . [ delta.scaling ] .
, width=432 ] of the dynamic order paramater plotted vs its susceptibility @xmath101 to the bias field @xmath1 , calculated at @xmath120 , for periods @xmath261 and 190 mcss .
the quantity @xmath101 was calculated using the numerical derivative in eq .
( [ eq : num.deriv ] ) .
the best - fit lines shown , whose slopes increase monotically with the period @xmath23 of the data to which they were fit , were calculated as @xmath262 , @xmath263 , @xmath264 , @xmath265 and @xmath266 , respectively . , width=432 ] , showing the relationship of @xmath139 and @xmath101 at low values of @xmath101 , which correspond to large values of the bias field @xmath1 .
for @xmath267 , 170 and 190 mcss , data have been taken ( and are shown ) down to very low values of @xmath101 , where the breakdown of the linear relationship between @xmath139 and @xmath101 can be clearly seen .
the dashed lines are the same best - fit lines shown in fig .
[ fluct.susc ] . , width=432 ] of the dynamic order paramater plotted vs its susceptibility @xmath101 to the bias field @xmath1 , calculated for lattice size @xmath120 , at periods @xmath152 and 250 mcss . at each period , the data were fit ( purely phenomenologically ) to two linear relationships . for @xmath152 mcss ,
the fits were calculated as @xmath268 at low @xmath101 , and @xmath269 at high @xmath101 . for @xmath153 mcss ,
the fits were @xmath270 at low @xmath101 , and @xmath271 at high @xmath101 .
, width=432 ] , obtained as the slopes of the linear fits to the data in figs . [ fluct.susc ] and [ fluct.susc2 ] , plotted vs @xmath272 . for the values @xmath161 and 0.825 ( @xmath152 and 250 mcss ) ,
the slopes of both linear regimes fit in fig .
[ fluct.susc2 ] are plotted as @xmath273 values , using filled squares and diamonds rather than filled circles .
the straight line is a weighted least - squares fit to the data below @xmath274 ( @xmath275 mcss ) , and has a slope of @xmath163 .
, width=432 ] and @xmath276 vs @xmath99 , over the range @xmath185 through 149 , for @xmath134 at @xmath120 .
the relatively small error bars on each data point can be seen inside the larger symbols .
the solid and dash - dash - dotted lines are fits to all four and the last three @xmath169 data points , respectively , and correspond to scaling exponents @xmath174 and @xmath176 .
the dotted and dash - dotted lines are the result of attempts to fit all four and the last three @xmath277 data points , respectively .
they correspond to nominal scaling exponents @xmath180 and @xmath182 . the dashed line is a reference line corresponding to scaling exponent @xmath278 . , width=432 ] and @xmath276 vs @xmath99 , over the range @xmath185 through 477 , for @xmath184 at lattice size @xmath120 .
the relatively small error bars on each data point can be seen inside the larger symbols .
the solid and dash - dash - dotted lines show the result of attempts to fit all six and the first five @xmath169 data points , and correspond to a nominal scaling exponent @xmath186 , and a statistically significant scaling exponent @xmath188 , respectively .
the dotted and dash - dotted lines are the result of attempts to fit all six and the first five @xmath277 data points , and correspond to nominal scaling exponents @xmath190 and @xmath191 , respectively . the dashed line is a reference line corresponding to scaling exponent @xmath278 . , width=432 ] vs @xmath99 , over the range @xmath185 through 149 , for @xmath184 at lattice size @xmath120 .
the relatively small error bars on each data point can be seen inside the larger symbols .
the solid and dotted lines are the results of attempts to fit all five and the last four data points with a power - law relationship , and correspond to nominal scaling exponents of @xmath279 and @xmath280 , respectively .
the dashed line is a reference line corresponding to a scaling exponent of @xmath278 .
, width=432 ] vs @xmath252 for lattice sizes @xmath257 180 , and 256 , at the critical period @xmath235 , where @xmath130 .
the best - fit line to the @xmath50 data in the range @xmath259 , shown as a dotted line in the plot , corresponds to a scaling exponent @xmath281 . a reference line ( dashed ) , corresponding to a scaling exponent @xmath223 ,
is also shown .
, width=432 ] | the two - dimensional kinetic ising model , when exposed to an oscillating applied magnetic field , has been shown to exhibit a nonequilibrium , second - order dynamic phase transition ( dpt ) , whose order parameter @xmath0 is the period - averaged magnetization .
it has been established that this dpt falls in the same universality class as the equilibrium phase transition in the two - dimensional ising model in zero applied field . here
we study for the first time the scaling of the dynamic order parameter with respect to a nonzero , period - averaged , magnetic ` bias ' field , @xmath1 , for a dpt produced by a square - wave applied field .
we find evidence that the scaling exponent , @xmath2 , of @xmath1 at the critical period of the dpt is equal to the exponent for the critical isotherm , @xmath3 , in the equilibrium ising model .
this implies that @xmath1 is a significant component of the field conjugate to @xmath0 .
a finite - size scaling analysis of the dynamic order parameter above the critical period provides further support for this result .
we also demonstrate numerically that , for a range of periods and values of @xmath1 in the critical region , a fluctuation - dissipation relation ( fdr ) , with an effective temperature @xmath4 depending on the period , and possibly the temperature and field amplitude , holds for the variables @xmath0 and @xmath1 .
this fdr justifies the use of the scaled variance of @xmath0 as a proxy for the nonequilibrium susceptibility , @xmath5 , in the critical region . |
le _ profil isoprimtrique _ dune varit riemannienne @xmath0 est la fonction @xmath1 qui , en @xmath2 , vaut la borne infrieure @xmath3 des volumes des bords des domaines de volume gal @xmath4 .
il y a peu de varits riemanniennes do nt on connaisse exactement le profil ( voir @xcite pour un tat de lart ) .
parmi tous les exemples connus , seule la sphre ronde a un profil lisse .
toute varit possde - t - elle une mtrique riemanienne de profil lisse ?
dans @xcite , on construit sur toute varit de dimension 2 ayant suffisamment de bout des mtriques lisses de profil lisse .
il est probable que des exemples similaires existent sur de nombreuses varit compactes . en revanche , il semble plus
difficile de construire des exemples analytiques rels .
on sait ( @xcite ) que le maximum de la courbure scalaire intervient dans le dveloppement asymptotique du profil prs de 0 .
une condition suffisante pour que le profil soit lisse au voisinage de 0 est que ce maximum soit atteint en un seul point , et que celui - ci soit un point critique non dgnr , voir @xcite .
sous cette hypothse supplmentaire , nous montrons que demander un profil lisse est trs restrictif .
[ main]soit @xmath0 une varit riemannienne compacte analytique relle . on fait les hypothses suivantes . 1 . toutes les solutions du problme isoperimtrique dans @xmath0 sont lisses .
2 . la courbure scalaire atteint son maximum en un seul point , et celui - ci est un point critique non dgnr .
le profil isoprimtrique de @xmath0 est lisse .
alors @xmath0 est homomorphe une sphre .
lhypothse ( i ) est toujours satisfaite si @xmath5 .
ignorons si on peut se passer de lhypothse ( ii ) .
si le profil est lisse , il est analytique .
les solutions lisses du problme isoprimtrique forment alors un espace analytique rel de dimension finie .
par hypothse , les solutions de petit volume forment un unique arc , une famille un paramtre ( on peut prendre le volume comme paramtre ) de domaines proches des petites boules centres au maximum de la courbure scalaire .
cet arc analytique se prolonge analytiquement , et ne peut se terminer que par la famille des domaines do nt les prcdents sont les complmentaires . @xmath0
est donc partage en deux domaines diffomorphes des boules fermes .
the _ isoperimetric profile _ of a riemannian manifold @xmath0 is the function @xmath1 which maps @xmath2 to the infimum @xmath3 of boundary volumes of domains of volume equal to @xmath4 .
very few profiles of compact riemannian manifolds are exactly known ( @xcite contains a state of the art ) . among those
, only the round sphere has a smooth profile .
we address here the differentiability question : does every compact manifold admit a metric with smooth profile ? in @xcite ,
smooth metrics with smooth profiles are constructed on 2-dimensional manifolds , provided they have enough ends .
it is likely that similar examples can be found on many compact manifolds .
however , it seems much harder to produce real analytic metrics with smooth profiles .
it is known ( @xcite ) that the asymptotics of the isoperimetric profile near 0 is determined by the maximum of scalar curvature . a sufficient condition for the profile to be smooth near 0 is that this maximum be achieved at only one point , which is a nondegenerate critical point , see @xcite . under this assumption , we show that having a smooth profile is very restrictive . [ main]let @xmath0 be a compact real analytic riemannian manifold .
assume the following . 1
. all solutions of the isoperimetric problem in @xmath0 are smooth .
2 . the scalar curvature function achieves its maximum at a unique point , and this point is a nondegenerate critical point .
the isoperimetric profile of @xmath0 is smooth
. then @xmath0 is homeomorphic to a sphere .
assumption ( i ) is always satisfied if @xmath5
. we do not know wether assumption ( ii ) can be removed .
we collect more or less standard material on real analytic spaces .
we express thanks to jean - jacques risler for his help with these matters .
the following definitions are borrowed from h. hironaka , @xcite , sections 1 and 5 .
[ ringed ] _ ( h . hironaka , @xcite , definition 1.1 ) .
_ a _ ringed space _ is a topological space equipped with a sheaf of commutative algebras over @xmath6 .
a _ morphism _ of ringed spaces @xmath7 is the data of a continuous map @xmath8 together with a homomorphism of sheaves @xmath9 .
the stalk @xmath10 at each point is a local ring .
its krull dimension ( the length of the longest chain of prime ideals in the maximal ideal ) is called the _ dimension of @xmath11 at @xmath12_. [ exringed ] let @xmath13 be an open set , let @xmath14 be a real analytic map
. then @xmath15 equipped with the sheaf of restrictions ( to open subsets of @xmath15 ) of real analytic functions on open subsets of @xmath16 , is a ringed space .
such ringed spaces are called _ local models_. [ defsmooth ] _ ( h . hironaka , @xcite , definition 5.6 ) . _
a local model is _ smooth _ at a point @xmath12 if the defining map @xmath17 is a submersion at @xmath12 .
[ raspace ] _ ( h . hironaka , @xcite , definition 1.5 ) . _
a _ real analytic space _ is a ringed space in which every point has a neighborhood isomorphic to some local model .
its dimension is the supremum of dimensions at points .
smooth points are those where the local model is smooth .
[ filtrationbis ] let @xmath11 be an @xmath18-dimensional paracompact real analytic space . then @xmath11 admits a locally finite filtration @xmath19 by closed real analytic subspaces @xmath20 such that @xmath21 is smooth of dimension @xmath22 .
the proof is an adaptation of proposition 5.8 of @xcite .
@xmath11 admits a complexification @xmath23 .
this is a complex analytic space containing @xmath11 equipped with an antiholomorphic involution @xmath24 whose fixed point set equals @xmath11 .
the set @xmath25 of points of @xmath23 where @xmath23 is not smooth and @xmath18-dimensional is defined in local models by holomorphic equations .
therefore @xmath25 is a complex analytic subspace .
let @xmath26 .
then @xmath27 is smooth of dimension @xmath18 .
define other @xmath20 s recursively .
[ extract ] let @xmath11 be a paracompact real analytic space .
let @xmath12 be a smooth point where @xmath11 has dimension 1 .
there exists a closed real analytic subset @xmath28 which contains a neighborhood of @xmath12 and whose dimension at each point is 1 .
put @xmath29 . by construction of @xmath30 , the open set @xmath31 of points where @xmath11 is smooth of dimension 1
is contained in @xmath32 .
sullivan , @xcite , page 166 , example a ) ) .
_ let @xmath11 be a 1-dimensional real analytic space .
let @xmath33 be a singular point .
then a punctured neighborhood of @xmath12 in @xmath11 consists of an even number of arcs .
they match pairwise into real analytic arcs .
[ ends ] let @xmath11 be a real analytic space whose dimension at each point is 1 .
assume that the singular set of @xmath11 is compact .
then each connected component of @xmath11 has a finite , even number of ends .
let @xmath34 denote the singular set of @xmath11 .
since @xmath35 is a closed real analytic subspace , @xmath35 is discrete , and therefore finite .
since the property of being a real analytic space is local , and local models are locally connected , one can assume that @xmath11 is connected .
@xmath36 is a 1-dimensional manifold , thus a collection of arcs and circles .
since @xmath11 is connected , either @xmath35 is empty , @xmath11 is a circle ( no ends ) or there are no circles .
then each arc has at least one endpoints in @xmath35 , thus there are finitely many arcs
. ends correspond to ends of arcs which do not belong to @xmath35 .
let @xmath37 denote the number of ends of @xmath11 and @xmath38 the number of pairs @xmath39 where @xmath40 and @xmath41 is an arc with @xmath42 as an endpoint .
then @xmath43 is twice the number of arcs , @xmath38 is even ( according to proposition [ even ] ) , thus @xmath37 is even .
[ top ] fix @xmath44 .
let us define the @xmath45-topology on the space of domains with smooth boundary as follows : as neighborhoods of a smooth domain @xmath46 , take all domains @xmath35 whose boundary is the graph , in normal exponential coordinates , of a @xmath45-small function on @xmath47 .
[ mean ] let @xmath0 be a compact real analytic riemannian manifold .
let @xmath48 be a positive real analytic function on the interval @xmath49 .
the set @xmath50 of pairs @xmath51 where @xmath46 is a smooth domain whose boundary has constant mean curvature equal to @xmath52 and satisfying @xmath53 has a natural structure of a finite dimensional real analytic space .
the proof is essentially contained in section 4 of @xcite .
let @xmath54 belong to @xmath50 . by definition ,
a neighborhood of @xmath46 consists of domains @xmath55 whose boundaries are graphs in normal coordinates of @xmath45 functions @xmath56 which satisfy the constant mean curvature equation @xmath57 , @xmath58
. since the linearized operator at @xmath59 , @xmath60 , need not be invertible , one solves instead @xmath61 where @xmath62 is the orthogonal projection onto the @xmath63-orthogonal complement @xmath64 of the kernel of @xmath60 in @xmath65 .
then @xmath66 is a real analytic map , whose linearization at 0 is @xmath67 . by construction , it is onto with a finite dimensional kernel .
therefore the set of solutions is a finite dimensional smooth real analytic submanifold @xmath68 of @xmath69 in a neighborhood of @xmath46 .
on @xmath68 , @xmath70 is a real analytic map , whose zero set , denoted by @xmath71 , is a neighborhood of @xmath46 in @xmath50 . a choice of coordinates for @xmath68 maps @xmath71 to a local model .
the sheaf @xmath72 is defined on small enough open sets by restricting to @xmath71 analytic functions on @xmath69 .
different choices of @xmath46 yield local embeddings of @xmath50 into banach spaces which differ only up to a change of chart for the ( infinite dimensional ) manifold structure on the space of domains . since constant mean
curvature hypersurfaces in real analytic riemannian manifolds are real analytic , these changes of charts are real analytic diffeomorphisms between open sets of @xmath45 spaces .
this yields the structure of a real analytic space on @xmath50 .
[ compact]let @xmath0 be a compact riemannian manifold with isoperimetric profile @xmath1 .
assume that all solutions of the isoperimetric problem in @xmath0 are smooth .
then for each @xmath73 , the subset @xmath74 is compact in @xmath45 topology .
the proof is essentially the same as in lemma @xmath75 of @xcite .
it depends on the fact that the @xmath45 topology is equivalent to the flat topology of geometric measure theory on the space of smooth isoperimetric domains .
compactness follows the compactness theorem for currents of bounded mass and boundary mass , provided one assumes that flat limits of smooth minimizers ( they are again minimizers ) are smooth .
by definition , elements of @xmath76 coincide with solutions of the isoperimetric problem . we know from @xcite that elements of @xmath76 of small volume are nearly round balls emanating from the unique maximum of scalar curvature ( see also @xcite ) .
they form a smooth connected 1-dimensional manifold . in other words , for @xmath77 small enough , @xmath76 is the union of @xmath78 and of two smooth one - ended arcs , @xmath79 > from proposition [ compact ] , it follows that @xmath76 has exactly two ends . according to @xcite , @xmath1 is semi - analytic on @xmath80 $ ] .
if smooth on @xmath49 , it must be analytic on the same interval .
consider the subset @xmath81 consisting of domains diffeomorphic to the @xmath18-ball .
since @xmath82 is open and closed in @xmath76 , it inherits the structure of a finite dimensional real analytic space .
let @xmath83 be the filtration provided by proposition [ filtrationbis ] .
the subset @xmath84 is a real analytic space of dimension 1 .
let @xmath85 denote the connected component of @xmath84 which contains
@xmath86 . by corollary [ ends ] , @xmath85 has a finite even number of ends .
since it contains one of the two ends of @xmath76 , it must contain the other .
in other words , there exists a domain @xmath46 in @xmath0 which is diffeomorphic to an @xmath18-ball , and whose complement is diffeomorphic to an @xmath18-ball as well .
this implies that @xmath0 is homeomorphic to a sphere .
vincent bayle , proprits de concavit du profil isoprimtrique .
thse , universit de grenoble ( 2003 ) .
olivier druet , sharp local isoperimetric inequalities involving the scalar curvature .
soc . 130(8):23512361
( 2002 ) dennis sullivan , combinatorial invariants of analytic spaces , proceedings of liverpool singularities symposium i ( 1969/70 ) , pp .
165168 , lecture notes in math .
192 , springer , berlin , 1971 .
renata grimaldi , pierre pansu , calibrations and isoperimetric profiles .
129 , 315350 ( 2007 ) .
renata grimaldi , stefano nardulli , pierre pansu , semianalyticity of isoperimetric profiles , to appear in differential geom .
heisuke hironaka , introduction to real - analytic sets and real - analytic maps .
quaderni dei gruppi di ricerca matematica del consiglio nazionale delle ricerche .
istituto matematico `` l. tonelli '' delluniversit di pisa , pisa , 1973 .
stefano nardulli , the isoperimetric profile of a compact riemannian manifold for small volumes .
rugang ye , foliation by constant mean curvature spheres , pacific j. math . 147(2):381396 , 1991 . | we show that smooth isoperimetric profiles are exceptional for real analytic riemannian manifolds .
for instance , under some extra assumption , this can happen only on topological spheres .
_ to cite this article : renata grimaldi , stefano nardulli , pierre pansu , c. r. acad .
sci .
paris , ser .
i * * * ( 200 + ) . _ 0.5 * rsum * 0.5 * differentiabilit du profil isoprimtrique et topologie des varits riemanniennes analytiques relles * on montre que la differentiabilit du profil isoprimtrique est une condition trs contraignante pour les varits riemannniennes analytiques relles .
par exemple , sous une hypothse supplmentaire , ce nest possible que si la varit est homomorphe une sphre .
_ pour citer cet article : renata grimaldi , stefano nardulli , pierre pansu , c. r. acad .
sci .
paris , ser .
i * * * ( 200 + ) .
_ differential geometry , , received * * * * * ; accepted after revision + + + + + + presented by - |
the recent discoveries of the massive neutron stars psr j@xmath0 @xcite and psr j@xmath1 @xcite have brought new challenges for theories of dense matter beyond the nuclear saturation density .
recently the radio timing measurements of the pulsar psr j@xmath0 and its white dwarf companion have confirmed the mass of the pulsar to be in the range of @xmath7 m@xmath8 at @xmath9 or @xmath10 m@xmath8 at @xmath11 confidence @xcite .
this is only the second neutron star(ns ) with a precisely determined mass around 2m@xmath8 , after psr j@xmath1 and has a 3@xmath6 lower mass limit @xmath12 m@xmath8 higher than the latter .
it therefore provides the tightest reliable lower bound on the maximum mass of neutron stars .
compact stars provide the perfect astrophysical environment for testing theories of cold and dense matter .
densities at the core of neutron stars can reach values of several times of @xmath13 . at such high densities ,
the energies of the particles are high enough to favour the appearance of exotic particles in the core .
since the lifetime of neutron stars are much greater than those associated with the weak interaction , strangeness conservation can be violated in the core due to the weak interaction .
this would result in the appearance of strange particles such as hyperons .
the appearance of such particles produces new degrees of freedom , which results in a softer equation of state ( eos ) in the neutron star interior .
the observable properties of compact stars depend crucially on the eos . according to the existing models of dense matter
the presence of strangeness in the neutron star interior leads to a considerable softening of the eos , resulting in a reduction of the maximum mass of the neutron star @xcite .
therefore many existing theories involving hyperons can not explain the large pulsar masses @xcite .
most relativistic models obtain maximum neutron star masses in the range @xmath14 @xcite , when hyperons are included .
some authors have tackled this problem by including a strong vector repulsion in the strange sector or by pushing the threshold for the appearance of hyperons to higher densities @xcite . in several studies
the maximum neutron star masses were generally found to be lower than @xmath15 @xcite which is in contradiction with observed pulsar masses .
however , neutron stars with maximum mass larger than @xmath16 have been obtained theoretically .
bednarek @xcite achieved a stiffening of the eos by using a non - linear relativistic mean field ( rmf ) model with quartic terms involving the strange vector meson .
lastowiecki @xcite obtained massive stars including a quark matter core .
taurines @xcite achieved large neutron star masses including hyperons by considering a model with density dependent coupling constants .
the coupling constants were varied nonlinearly with the scalar field .
bonanno and sedrakian @xcite also modeled massive neutron stars including hyperons and quark core using a fairly stiff eos and vector repulsion among quarks .
authors in ref .
@xcite incorporated higher order couplings in the rmf theory in addition to kaonic interactions to obtain the maximum neutron star mass .
agrawal @xcite have optimized the parameters of the extended rmf model using a selected set of global observables which includes binding energies and charge radii for nuclei along several isotopic and isotonic chains and the iso - scalar giant monopole resonance energies for the @xmath17zr and @xmath18pb nuclei .
weissenborn @xcite investigated the vector meson - hyperon coupling , going from su(6 ) quark model to a broader su(3 ) , and concluded that the maximum mass of a neutron star decreases linearly with the strangeness content of the neutron star core independent of the nuclear eos .
on the other hand , h. dapo @xcite found that for several different bare hyperon - nucleon potentials and a wide range of nuclear matter parameters the hyperons in neutron stars are always present .
the parameters of the rmf model are fitted to the saturation properties of the infinite nuclear matter and/or the properties of finite nuclei . as a result extrapolation to higher densities and asymmetry
involve uncertainties .
three of these properties of the infinite nuclear matter are more precisely known : ( a ) the saturation density , ( b ) the binding energy and ( c ) the asymmetry energy , compared to the remaining ones - the effective nucleon mass and the compression modulus of the nuclear matter .
the uncertainty in the dense matter eos is basically related to the uncertainty in these two saturation properties .
it has been seen that to reproduce the giant monopole resonance ( gmr ) in @xmath18pb , accurately fitted non - relativistic and relativistic models predict compression modulus in the symmetric nuclear matter ( @xmath19 ) that differ by about @xmath20 .
the reason for this discrepancy being the density dependence of the symmetry energy .
moreover , the alluded correlation between @xmath19 and the density dependence of the symmetry energy results in an underestimation of the frequency of oscillations of neutrons against protons , the so - called isovector giant dipole resonance ( ivgdr ) in @xmath18pb .
fsugold is a recently proposed accurately calibrated relativistic parameterization .
it simultaneously describes the gmr in @xmath17zr and @xmath18pb and the ivgdr in @xmath18pb without compromising the success in reproducing the ground - state observables @xcite .
the main virtue of this parameterization is the softening of both the eos of symmetric nuclear matter and the symmetry energy .
this softening appears to be required for an accurate description of different collective modes having different neutron - to - proton ratios . as a result ,
the fsugold effective interaction predicts neutron star radii that are too large and a maximum stellar mass that is too small @xcite .
the indiana university - florida state university ( iufsu ) interaction , is a new relativistic parameter set , derived from fsugold .
it is simultaneously constrained by the properties of finite nuclei , their collective excitations and the neutron star properties by adjusting two of the parameters of the theory - the neutron skin thickness of @xmath21 and the maximum neutron star mass @xcite . as a result
the new effective interaction softens the eos at intermediate densities and stiffens the eos at high density .
as it stands now , the new iufsu interaction reproduces the binding energies and charge radii of closed - shell nuclei , various nuclear giant ( monopole and dipole ) resonances , the low - density behavior of pure neutron matter , the high - density behavior of the symmetric nuclear matter and the mass - radius relationship of neutron stars . whether this new eos can accommodate the hyperons inside the compact stars , with the severe constraints imposed by the recent observations of @xmath22 pulsars , needs to be explored . in this work
we plan to make a detailed study of such a possibility . for this purpose
we have extended the iufsu interaction by including the full baryon octet .
a new eos is constructed to investigate the neutron star properties with hyperons . [
cols="^,^,^,^,^,^,^,^,^ " , ] the paper is organized as follows . in section 2 ,
we briefly discuss the model used and the resulting eos . in the next section we use this eos to look at static and rotating star properties .
we give a brief summary in section 4 .
one of the possible approaches to describe neutron star matter is to adopt an rmf model subject to @xmath2 equilibrium and charge neutrality . for our investigation of nucleons and hyperons in the compact star matter
we choose the full standard baryon octet as well as electrons and muons .
contribution from neutrinos are not taken into account assuming that they can escape freely from the system . in this model ,
baryon - baryon interaction is mediated by the exchange of scalar ( @xmath6 ) , vector ( @xmath23 ) , isovector ( @xmath24 ) and the strange vector ( @xmath25 ) mesons .
the lagrangian density we consider is given by @xcite @xmath26{\psi}_{b } + \frac{1}{2}\partial_{\mu}\sigma\partial^{\mu}\sigma - \frac{1}{2 } m_\sigma^2\sigma^2 \nn\\ & & - \frac{\kappa}{3!}(g_{\sigma n}\sigma)^3 -\frac{\lambda}{4!}(g_{\sigma n}\sigma)^4 - \frac{1}{4}f_{\mu\nu}f^{\mu\nu } + \frac{1}{2}m_\omega^2\omega_\mu\omega^\mu + \frac{\zeta}{4!}(g^{2}_{\omega n}\omega_\mu\omega^\mu)^2 + \frac{1}{2}m_\rho^2\vec{\rho}_{\mu}\cdot\vec{\rho}^{\mu } - \frac{1}{4}\vec{g}_{\mu\nu}\vec{g}^{\mu\nu}\nn\\ & & + \lambda_{v}(g^{2}_{\rho n}\vec{\rho}_{\mu}\cdot\vec{\rho}^{\mu})(g^{2}_{\omega n}\omega_\mu\omega^\mu ) + \frac{1}{2}m_\phi^2\phi_\mu\phi^\mu -\frac{1}{4}h_{\mu\nu}h^{\mu\nu } + \sum_{l}\bar{\psi}_{l}[i\gamma^{\mu}\partial_{\mu } - m_{l}]{\psi}_{l}\end{aligned}\ ] ] where the symbol b stands for the baryon octet ( @xmath27 , @xmath28 , @xmath29 , @xmath30 , @xmath31 , @xmath32 , @xmath33 , @xmath34 ) and @xmath35 represents @xmath36 and @xmath37 .
the masses @xmath38 , @xmath39 , @xmath40 , @xmath41 and @xmath42 are respectively for baryon , @xmath6 , @xmath23 , @xmath24 and @xmath25 mesons .
the antisymmetric tensors of vector mesons take the forms @xmath43 = @xmath44 , @xmath45 = @xmath46 $ ] and @xmath47 = @xmath48 . the isoscalar meson self - interactions ( via @xmath49 , @xmath50 and @xmath51 terms ) are necessary for the appropriate eos of the symmetric nuclear matter @xcite .
the new additional isoscalar - isovector coupling ( @xmath52 ) term is used to modify the density dependence of the symmetry energy and the neutron - skin thickness of heavy nuclei @xcite .
the meson - baryon coupling constants are given by @xmath53 , @xmath54 , @xmath55 and @xmath56 .
all the nucleon - meson parameters used in this work are shown in table [ para ] .
the saturation properties of the symmetric nuclear matter produced by iufsu are : saturation density @xmath57 @xmath58 , binding energy per nucleon @xmath59 mev and compression modulus @xmath60 mev .
the hyperon - meson couplings are taken from the su(6 ) quark model @xcite as , @xmath61 = 0 , @xmath62 = @xmath63 = @xmath64 @xmath65 = @xmath66 = @xmath67 = @xmath68 @xmath69 = @xmath70 = @xmath71 = @xmath72 the scalar couplings are determined by fitting the hyperonic potential , @xmath73 where y stands for the hyperon and @xmath74 , @xmath75 are the values of the scalar and vector meson fields at saturation density @xcite .
the values of @xmath76 are taken from the available hypernuclear data .
the best known hyperonic potential is that of @xmath29 , having a value of about @xmath77 = -30 mev @xcite . in case of @xmath3 and @xmath4 hyperons ,
the potential depths are not as clearly known as in the case of @xmath29 .
however , analyses of laboratory experiments indicate that at nuclear densities the @xmath29-nucleon potential is attractive but the @xmath78 -nucleon potential is repulsive @xcite .
therefore , we have varied both @xmath79 and @xmath80 in the range of -40 mev to + 40 mev to investigate the properties of neutron star matter . for neutron star matter , with baryons and charged leptons ,
the @xmath2-equilibrium conditions are guaranteed with the following relations between chemical potentials for different particles : @xmath81 and the charge neutrality condition is fulfilled by @xmath82 where @xmath83 is the number density of the _ ith _ particle . the effective chemical potentials of baryons and leptons can be given by @xmath84 @xmath85 where @xmath86 is the baryon effective mass and @xmath87 is the fermi momentum of the lepton ( e , @xmath88 ) .
the eos of neutron star matter can be given by , @xmath89 @xmath90 where @xmath91 and @xmath92 stand for energy density and pressure respectively and @xmath93 is the baryon spin - isospin degeneracy factor . 1.5 cm + + + + in fig . [ eos ]
we plot the eos for different values of the hyperonic potentials .
the upper branch is for the usual nuclear matter which does not contain any strange particle .
the middle and lower branches are for full baryon octet , leptons and @xmath6 , @xmath23 , @xmath24 mesons .
in addition , the middle branch contains the @xmath25 meson . in the left panel , _
i.e. _ in fig .
[ eos]a , we keep @xmath80 fixed at -18 mev , this value is generally adopted from hypernuclear experimental data @xcite .
for the middle and lower branches we vary the @xmath3 potential from -40 mev to + 40 mev in steps of 20 mev .
the lower branch shows that for an attractive @xmath3 potential the eos gets stiffer as @xmath94 increases .
however as @xmath94 becomes positive the eos seems to become independent of @xmath94 .
we see from fig .
[ eos]a that for @xmath95 mev the eos remains identical to that for @xmath96 mev .
however , once we add @xmath25 meson to the system , the eos continues to get stiffer as @xmath94 moves to more positive side ( middle branch of fig . [ eos]a ) .
+ + + + we then fix @xmath79 and vary @xmath80 .
this is represented in fig .
[ eos]b , where we have fixed the value of @xmath97 mev ( adopted from hypernuclear experimental data @xcite ) .
we vary @xmath80 from -40 mev to + 40 mev .
we see that for the lower branch , _
i.e _ the case without the @xmath25 meson , the eos gets stiffer with the increase in @xmath4 potential up to @xmath98 mev .
however , for positive values of @xmath80 the eos remains unchanged .
adding an extra repulsion to the system by including the @xmath25 meson changes the scenario altogether .
the eos becomes totally independent of the @xmath4 potential ( middle branch of fig .
[ eos]b ) . from figures
1a and 1b one can generally conclude that the inclusion of @xmath25 meson makes the eos stiffer , however , hyperonic eos is much softer than the usual nuclear matter eos .
[ t ] in fig .
[ pf1 ] we have plotted the particle fractions for an attractive @xmath3 potential @xmath99 mev and a repulsive potential @xmath97 mev keeping @xmath80 fixed at -18 mev , with and without @xmath25 in each case . from fig .
[ pf1]a , when @xmath25 is not present , we see that all the hyperons contribute to the particle fractions for an attractive @xmath3 potential whereas for repulsive @xmath79 there is no @xmath3 present in the matter ( fig .
[ pf1]b ) .
the appearance of @xmath29 is also pushed to higher density compared to the case of an attractive potential .
when @xmath25 is included in the system @xmath100 and @xmath78 appear with @xmath29 for @xmath99 mev ( fig .
[ pf1]c ) . however , for @xmath97 mev ( fig .
[ pf1]d ) , the threshold of @xmath78 is pushed to higher density compared to the case of @xmath99 mev , @xmath100 disappears and @xmath101 appears in the system .
we also note that in the case of attractive @xmath3 potential , @xmath78 is always the first hyperon to appear in the system . for repulsive
@xmath79 , @xmath101 appears before others in the `` @xmath102 '' case and @xmath29 is the the first hyperon to appear in case of `` @xmath103 '' . from fig .
[ pf1 ] we see that for negative values of @xmath79 , the @xmath3 s are bound in matter and the effective mesonic interaction would be more attractive as the potential gets deeper . as a result , the eos gets softer with more attractive @xmath79 ( see fig . [ eos]a ) . for @xmath104 ,
@xmath3 s are no longer bound to matter and the effective mesonic interaction becomes more and more repulsive with increasing @xmath79 .
this should , in principle , stiffen the eos .
however , for the `` @xmath102 '' case , up to neutron star densities , _
i.e _ about @xmath105 , @xmath3 s are not present in the matter when the potential is repulsive and hence the eos up to these densities becomes insensitive to @xmath79 . in fig .
[ pf2 ] the particle fractions are plotted for an attractive @xmath4 potential @xmath106 mev and a repulsive potential @xmath107 mev keeping @xmath79 fixed at + 30 mev .
we see that in the first case _
i.e. _ when @xmath25 is not present and the potential is attractive ( fig .
[ pf2]a ) , all the hyperons except @xmath3 s are present in the system and the @xmath29 hyperon dominates . when the @xmath4 potential becomes positive ( fig .
[ pf2]b ) @xmath108 disappears and the threshold for appearance of @xmath101 shifts to much higher density .
however @xmath78 is present in matter in this potential and it appears before @xmath101 .
when @xmath25 is introduced in the system , for an attractive @xmath4 potential ( fig .
[ pf2]c ) , again @xmath78 and @xmath101 are present along with @xmath29 .
however , the difference from fig .
i.e _ `` @xmath102 '' case and @xmath1090 is that , here @xmath101 appears much before @xmath78 . in the last case ( fig .
[ pf2]d ) , we see that as a result of the combined effects of inclusion of @xmath25 and repulsive potentials , only the @xmath29 and @xmath78 are present in the system .
from both figures [ pf1 ] and [ pf2 ] , we see that , inclusion of @xmath25 meson decreases the density of hyperons .
since @xmath25 is a strange particle , further strangeness is suppressed and as a result the hyperon densities are reduced compared to the `` @xmath102 '' case .
in this section we are going to discuss the properties of static and rotating axisymmetric stars using the eos which we have studied in the last section . the eos without @xmath25 meson is softer compared to that with @xmath25 meson .
so we do not discuss the eos without @xmath25 as it results in less maximum mass .
the stationary , axisymmetric space - time used to model the compact stars are defined through the metric @xmath110 where @xmath111 , @xmath112 , @xmath24 and @xmath23 are the gravitational potentials which depend on r and @xmath113 only . in this work
we adopt the procedure of komatsu @xcite to look into the observable properties of static and rotating stars .
einstein s equations for the three gravitational potentials @xmath112 , @xmath24 and @xmath23 can be solved using green s function technique .
the fourth potential @xmath111 can be determined using these three potentials .
once these potentials are determined one can calculate all the observable quantities using those .
the solution of the potentials and hence the determination of physical quantities is numerically quite an involved process . for this purpose the `` rns '' code @xcite
is used in this work .
this code , developed by stergoilas , is very efficient in calculating the rotating star observables .
we discuss the properties of static stars first . in fig .
[ static ] we have plotted the mass - radius curves of static stars using the eos with `` @xmath103 '' . a plot for the pure nuclear matter case
is also given for comparison ( uppermost curve of both the panels ) .
the maximum mass of pure nuclear matter star in the static case is @xmath114 with a radius of @xmath115 km .
we have found that the mass of hyperonic star becomes maximum for @xmath116 mev and @xmath117 mev . hence in fig .
[ static ] and fig .
[ mr3 ] we have shown the effect of these potentials on the maximum mass of neutron stars by fixing one of the potentials at + 40 mev and varying the other .
the left panel , _
i.e. _ fig . [ static]a , corresponds to @xmath118 mev and @xmath119 varying from -40 mev to + 40 mev . in the right panel , _
i.e. _ in fig .
[ static]b , it is the other way round . from fig .
[ static]a one can see that the maximum mass of the star increases with @xmath120 . for @xmath121
mev the maximum mass is @xmath122 with a radius of @xmath123 km .
the central energy density of such a star is @xmath124 .
this is a reflection of the eos shown in fig .
[ eos]a , which shows that the eos becomes stiffer with increase in @xmath120 .
however , as seen from fig .
[ static]b , the maximum mass of static stars is insensitive to @xmath125 , which should be obvious from fig .
[ eos]b as the eos is independent of the cascade potential . furthermore , from fig .
[ pf2]d one can see that there is no cascade present in the medium
. so the insensitivity of the eos and hence the maximum mass , towards the cascade potential is expected .
one should note that the maximum mass we obtain for the static stars is less than the observed mass of psr j@xmath0 .
so the static stars with hyperons in the iufsu parameter set can not incorporate a maximum mass @xmath22 .
this result is consistent with the findings in ref .
however , since both of the observed @xmath22 stars are pulsars , it would be a better idea to compare the observations with results from the rotating stars , which we do in the next part . in fig .
[ mr3 ] we plot the mass - radius curves for stars rotating with keplerian velocities , for two cases . in fig .
[ mr3]a we fix the cascade potential at @xmath126mev and vary @xmath79 from @xmath127mev to @xmath128mev . in fig .
[ mr3]b it is the other way round . the pure nuclear matter case is also shown in the uppermost curve .
the maximum mass for the pure nucleonic star is @xmath129 with a radius of @xmath130 km .
we see that the maximum mass obtained for a rotating star with hyperonic core is @xmath5 with a radius of @xmath131 km in the keplerian limit with angular velocity @xmath132 , for @xmath133 mev and @xmath134 . as in the case of static sequence
, we see that the maximum mass for the rotating case also increases with @xmath79 as we go towards more positive values of this potential . at @xmath135
mev we get a maximum mass of @xmath136 whereas for @xmath133 mev the maximum mass is @xmath137 .
the effect of @xmath80 is much less significant on the maximum mass . from @xmath138 mev to @xmath126 mev mass
is changed only by @xmath139 . in order to have a look at the composition of the maximum mass star , we have plotted the particle densities as a function of radius along the equator in fig .
[ pd ] . for @xmath98 and @xmath133 mev
, we see that a fair amount of hyperons are present in the core .
there are @xmath29 , @xmath78 and @xmath101 present .
another interesting observation is that near the core , the density of @xmath29 is much more compared to that of protons and it continues up to a distance of about 5 km from the center .
to summarize , we have studied the static and rotating axisymmetric stars with hyperons using iufsu model .
the original fsugold parameter set has been very successful in describing the properties of finite nuclei . with the discovery of
highly massive neutron stars the reliability of this model was questioned .
it was then revised in the form of iufsu to accommodate such highly massive stars leaving the low density finite nuclear properties unchanged . in this work
we have studied this new parameter set in the context of the possibility of having a hyperonic core in such massive stars .
we have included the full octet of baryons in iufsu .
the eos gets softened due to the inclusion of hyperons whereas the inclusion of the @xmath25 meson makes the eos stiffer .
we have also investigated the influence of @xmath3 and @xmath4 potentials on the eos . for static stars with hyperonic core
we get a maximum mass of @xmath122 .
so iufsu with hyperons can not reproduce the observed mass of static stars . however , as the observed @xmath22 neutron stars are both pulsars , we compare the results in the rotating limit . in the keplerian limit
we get a maximum mass of @xmath137 , which is within the 3@xmath6 limit of the mass of psr j@xmath0 and 1@xmath6 limit of the earlier observation of psr j@xmath1 .
we have looked at the particle densities inside the star having the maximum mass and found that a considerable amount of hyperons are present near the core .
therefore , our results are consistent with the recent observations of highly massive pulsars confirming the presence of hyperons in the core of such massive neutron stars .
to conclude , iufsu model , which reproduces the properties of finite nuclei quite successfully also reproduces the recent observations of @xmath140 stars , in case of stars having exotic core and rotating in the keplerian limit .
it will be interesting to see whether such a star can hold a quark core .
related work is in progress .
this work is funded by the university grants commission ( rfsms , dskpdf and drs ) and department of science and technology , government of india .
99 j. antoniadis _
_ _ science _ * 340 * , ( 2013 ) 6131 .
p. b. demorest , t. pennucci , s. m. ransom , m. s. e. roberts and j. w. t. hessels , _ nature _ * 467 * , ( 2010 ) 1081 .
e. massot , j. margueron1 and g. chanfray , _ _ * 97 * , ( 2012 ) 39002 .
m. baldo , g. f. burgio and h .- j .
schulze , _ phys .
c _ * 61 * , ( 2000 ) 055801 .
i. vidana , a. polls , a. ramos , l. engvik and m. hjorth - jensen , _ phys .
c _ * 62 * , ( 2000 ) 035801
. h. apo , b .-
schaefer and j. wambach , _ phys .
c _ * 81 * , ( 2010 ) 035803
. j. m. lattimer and m. prakash , _ from nuclei to stars : festschrift in honor of gerald brown , _
p.*275 * , world scientific , singapore , ( 2011 ) .
n. k. glendenning _ astrophys .
j. _ * 293 * , ( 1985 ) 470 .
n. k. glendenning and s. a. moszkowski , _ phys .
lett . _ * 67 * , ( 1991 ) 2414 .
r. knorren , m. prakash and p. j. ellis , _ phys .
c _ * 52 * , ( 1995 ) 3470
. s. balberg and a. gal , _
nucl.phys . a _ * 625 * , ( 1997 ) 435 .
s. pal , m. hanauske , i. zakout , h. st@xmath141ecker and w. greiner , _ phys .
c _ * 60 * , ( 1999 ) 015802 .
m. hanauske , d. zschiesche , s. pal , s. schramm , h. st@xmath141ecker and w. greiner , _ astrophys .
j. _ * 537 * , ( 2000 ) 958 .
s. schramm and d. zschiesche , _
g _ * 29 * , ( 2003 ) 531 .
w. h. long , b. y. sun , k. hagino and h. sagawa , _ phys .
c _ * 85 * , ( 2012 ) 025806 . h. huber , m. k. weigel and f. weber , z. _ naturforsch .
_ * 54a * , ( 1999 ) 77 .
f. hofmann , c. m. keil and h. lenske , _ phys .
c _ * 64 * , ( 2001 ) 034314 .
j. rikovska - stone , p. guichon , h. matevosyan and a. thomas , _ nucl .
phys . a _ * 792 * , ( 2007 ) 341 .
s. k. dhiman , r. kumar and b. k. agrawal , _ phys .
c _ * 76 * , ( 2007 ) 045801
. v. dexheimer and s. schramm , _ astrophys .
j. _ * 683 * , ( 2008 ) 943 .
i. bombaci , p. k. panda , c. providencia and i. vidana , _ phys .
d _ * 77 * , ( 2008 ) 083002 .
r. cavagnoli , d. p. menezes and c. providencia , _ phys .
c _ * 84 * , ( 2011 ) 065810 .
m. baldo , g. f. burgio and h .- j .
schulze , _ phys .
c _ * 58 * , ( 1998 ) 3688 .
s. nishizaki , t. takatsuka and y. yamamoto , _ prog . theor
_ * 108 * , ( 2002 ) 703 .
schulze , a. polls , a. ramos and i. vidana , _ phys .
c _ * 73 * , ( 2006 ) 058801 .
schulze and t. rijken , _ phys .
c _ * 84 * , ( 2011 ) 035801 .
d. logoteta , i. vidana , c. providencia , a. polls and i. bombaci , _ j. phys .
_ * 342 * , ( 2012 ) 012006 .
i. bednarek , p. haensel , j. l. zdunik , m. bejger and r. maka , _ astron .
astrophys . _ * 543 * , ( 2012 ) a157 .
r. lastowiecki , d. blaschke , h. grigorian and s. typel , _ acta phys .
* 5 * , ( 2012 ) 535 .
a. r. taurines , c. a. z. vasconcellos , m. malheiro and m. chiapparini , _ mod .
* 15 * , ( 2000 ) 1789 .
l. bonanno and a. sedrakian , _ astron .
astrophys . _ * 539 * , ( 2012 ) a16 . n. gupta and p. arumugam , _ c _ * 85 * , ( 2012 ) 015804
. b. k. agrawal , a. sulaksono and p. -g .
reinhard , _ a _ * 882 * , ( 2012 ) 1
. s. weissenborn , d. chatterjee and j. schaffner - bielich , _ phys . rev .
c _ * 85 * , ( 2012 ) 065802
. b. g. todd - rutel and j. piekarewicz _ phys .
lett . _ * 95 * , ( 2005 ) 122501 . c. wu and z. ren , _ c _ * 83 * ( 2011 ) 025805 .
f. j. fattoyev , c. j. horowitz , j. piekarewicz and g. shen , _ _ * 82 * , ( 2010 ) 055803 .
g. a. lalazissis , j. k@xmath141nig and p. ring , _ phy .
_ * c 55 * , ( 1997 ) 540 . c. b. dover and a. gal , _ prog .
* 12 * , ( 1985 ) 171 .
j. schaffner , c. b. dover , a. gal , c. greiner , d. j. millener and h. st@xmath141ecker , _ annals of physics _ * 235 * , ( 1994 ) 35 .
d. j. millener , c. b. dover and a. gal , _ phys .
c _ * 38 * , ( 1988 ) 2700 ; j. schaffner , h. st@xmath141ecker and c. greiner , _ phys .
c _ * 46 * , ( 1992 ) 322 .
j. mares , w. friedman , a. gal and b. k. jennings , _ nucl .
a _ * 594 * , ( 1995 ) 311 .
j. schaffner - bielich and a. gal , _
c _ * 62 * , ( 2000 ) 034311 .
h. komatsu , y. eriguchi and i. hachisu , _ monthly notices of royal astronomical society _ * 237 * , ( 1989 ) 355 .
n. stergioulas and j. h. friedman , _ astrophys . j. _ * 444 * , ( 1995 ) 306 .
b. k. agrawal , a. sulaksono , p. -g .
reinhard , _ nucl .
phys . a_*882 * ( 2012 ) 1 . | the recent discoveries of massive neutron stars , such as psr j@xmath0 and psr j@xmath1 , have raised questions about the existence of exotic matter such as hyperons in the neutron star core . the validity of many established equations of states ( eos s ) like the gm1 and fsugold are also questioned .
we investigate the existence of hyperonic matter in the central regions of massive neutron stars using relativistic mean field ( rmf ) theory with the recently proposed iufsu model .
the iufsu model is extended by including hyperons to study the neutron star in @xmath2 equilibrium . the effect of different hyperonic potentials , namely @xmath3 and @xmath4 potentials , on the eos and hence the maximum mass of neutron stars has been studied .
we have also considered the effect of stellar rotation since the observed massive stars are pulsars .
it has been found that a maximum mass of @xmath5 , which is within the 3@xmath6 limit of the observed mass of psr j@xmath0 , can be obtained for rotating stars , with certain choices of the hyperonic potentials .
the said star contains a fair amount of hyperons near the core .
0.2 cm |
optimal olfactory searches are essential for the survival of insects and animals in search for food and mating . these searches are based on the capacity of these organisms to detect the chemical signals emitted by the target , usually at distances that are further away to be visually located . in recent years several olfactory search strategies
have been proposed to model the way in which such organisms are capable of finding their targets @xcite . to accomplish a successful search the strategy has to be reliable to reach the target most of the times , robust against errors and efficient in the sense of exhibiting reasonably short search times . from a mathematical point of view ,
an optimal strategy is one that minimizes the search time , failure probability or any appropriate cost function .
the complexity of an olfactory search is in general a consequence of the complexity of the transport of the chemical signals through the environment .
it is natural to expect that organisms using optimal strategies will have a survival advantage with respect to other competitors that do not .
however , biological success in one environment will be unsuccessful in others as the optimal search will depend on the environment and on how limited the information about the properties of the transport is . at low reynolds numbers , molecular diffusion dominates , and the concentration of the chemical signals used in the search are smooth and more importantly , static . in this regime , searches up - gradient in the concentration field such as chemotaxis , are successful .
this completely exploitative strategy is used by bacteria in search for food supplies , phages to localize bacteria and euchariotic cell organisms in general . at larger reynolds numbers the environment becomes turbulent , the concentration field breaks into complex filamentous structures and
concentration decays rapidly with the distance from the source . in this regime
, chemotaxis is ineffective since any local maxima , will attract the searcher , keeping it from ever reaching the global maxima that marks the position of the target . in turbulent environments the searcher will need to deal with very limited dynamically evolving information to solve the task .
few years ago , vergassola , _ et al _ , proposed a strategy for olfactory search called infotaxis @xcite .
they analyzed the search process from an information theory point of view , as an inverse problem in which the searcher needs to decode partial and noisy messages to find the position of the emitter .
as the searcher moves , the information accumulated by the detections of the emitted chemicals is used to iteratively refine the instantaneous posterior distribution function for the probability to find the target at a given position @xmath0 , and to decide the further movements of the searcher as those that maximize the local gain of information .
infotaxis has shown to be successful and robust , leading to narrow distributions for the search time , even if the knowledge of the environment is not exact @xcite . in this paper
we are concerned about the fluctuations of the search time for infotaxis .
an optimal search minimizing the fluctuations is more reliable as then one would expect that any search trajectory would be completed in a time similar to the mean search time .
we will show for infotaxis , that when the search process is defined on a bounded domain , the fluctuations of the search time strongly depend on the geometry of the domain and the topology of the lattice on which the searcher evolves . at the beginning of the search ,
when no detections have yet occurred , one would expect that the searcher dynamics correspond to a random explorer .
however , here we show that the first movements of the searcher are not stochastic but fully determined by the geometry and the topology of the domain . leading to a complicated dependence of the search time on the initial searcher s position .
the domain is partitioned in terms of the initial searcher s behaviour and we find that the fluctuations of the search time are largely enhanced when the search starts along the interface between partitions of different behaviour .
the paper is organized as follows : for completeness , in section [ sec : infotaxis ] we briefly review the infotaxis search algorithm and our implementation . in section [ sec : inipos ] the dependence of the search time on the initial position of the searcher is analyzed , showing that the domain is decomposed in regions according to the first deterministic step .
furthermore , in section [ sec : geom ] we show that the boundary of these regions depends not only on the geometry of the domain but also on the topology of the lattice and the parameters of the transport process .
the geometry - induced fluctuations are studied in section [ sec : pw ] in terms of the simultaneity of the first passage process .
our findings are summarized in section [ sec : concl ] .
infotaxis is a bayesian olfactory search strategy in which the information released by the target source is transported through a turbulent environment @xcite .
an infotactic searcher initially explores the space collecting information , which is encoded in the trace of detections of molecules emitted by the source @xmath1 .
this information is used to reconstruct , at each time @xmath2 , the posterior probability distribution @xmath3 for the unknown position of the source @xmath0 .
both @xmath1 and @xmath3 are time - varying quantities that are constantly updated .
the posterior probability distribution , referred to as the _ belief function _ , depends on the rate of detections @xmath4 _ i.e. _ , the expected number of detections at position @xmath5 , given that the emitting source is located at @xmath0 .
this function @xmath4 incorporates our modelling of the natural transport process , and it can be either postulated from rather general hypothesis , or else adapted dynamically as the searcher learns from the environment . as it is customary , in this paper we choose the first option , and assume that @xmath4 is the solution of an advection - diffusion partial differential equation @xcite , which in two spatial dimensions takes the following explicit form : @xmath6 where @xmath7 is the rate of emission of volatile particles , @xmath8 is their isotropic effective diffusivity , @xmath9 is the characteristic size of the detector , @xmath10 is the modified bessel function of order 0 , and @xmath11 the _ correlation length _ , given by @xmath12 where @xmath13 the finite lifetime of the emitted molecules , and @xmath14 the mean current or wind ( which blows , without loss of generality , in the negative @xmath15-direction ) .
the correlation length @xmath11 can be interpreted as the mean distance travelled by a volatile particle before it decays . in this paper
we set the values of @xmath9 and @xmath8 to unity .
we stress that no perfect model of the environment is realistically available in concrete applications and that the rate function @xmath4 will only be an approximation to the actual transport process . however , the performance of the infotaxis strategy is still acceptable despite an incorrect modelling of the medium , which happens , for instance , if the parameters used by the searcher in its model of the environment differ from the real parameters of the transport process .
the simplest model for the time correlation of successive detections is to assume that they are independent . in a real turbulent transport process
there are clearly spatio - temporal correlations between detections , as when a plume of odor particles hits the searcher , it is more likely that in a short time after it will have a higher chance of detection .
however , ignoring this fact and assuming no correlations works surprisingly well for the bayesian searcher to update its belief function of the source s position @xcite . assuming independence of the detections , the stochastic process is poissonian and therefore the probability distribution at time @xmath2 posterior to experiencing a trace @xmath1 is given by @xmath16 where @xmath17 and @xmath18 is the total number of detections registered by the searcher at successive times @xmath19 . assuming that the detections are uncorrelated
, the evolution of the probability map can be written as a markovian update @xmath20 where @xmath21 is the number of detections during the short time interval @xmath22 and @xmath23 is a normalization constant with respect to @xmath24 .
so far , this is a purely bayesian approach to infer the position of the source from the trace of detections .
what is characteristic of infotaxis is the criterion for the searcher s motion : rather than moving towards the most probable source position , at each time the searcher moves so that the expected information gain is maximized . in what follows
we assume that the searcher moves on a regular lattice with given topology .
suppose the searcher is at position @xmath25 at time @xmath2 with belief function @xmath26 . at time @xmath27 its new position @xmath28 can be any of the neighbouring lattice sites or @xmath29 ( no motion ) . for each of the possible motions @xmath28 infotaxis estimates the shannon entropy associated to the updated belief function @xmath30 and chooses from them the one for which the decrease in entropy is highest .
the expected variation of entropy upon moving to @xmath28 is @xmath31\ ] ] where @xmath32 is the probability of having @xmath33 detections during the time @xmath34 , where @xmath35 is the mean number of detections at position @xmath28 .
taking @xmath36 in allows to estimate the updated belief function at each of the sites @xmath28 provided @xmath33 detections were made in the last time interval .
the shannon entropy of each of these probability densities is denoted by @xmath37 and @xmath38 appearing in is the difference between the shannon entropies of the estimated @xmath39 and the current @xmath40 .
the first term on the right hand side of corresponds to the event of finding the source at the neighboring @xmath28 point .
the second term takes into account the entropy decrease if the source is not found at @xmath28 .
the function to maximize at each time step therefore considers a balance between exploration and exploitation of the acquired information , as discussed in @xcite .
the algorithm evaluates at each of the possible @xmath28 and chooses to move in the direction where @xmath41 is highest , i.e. where the largest expected decrease in entropy is obtained .
we have numerically solved the infotaxis algorithm described above for a searcher moving on a finite two - dimensional lattice @xmath42 with square , triangular and hexagonal topology .
the lattice is bounded by a subset of lattice points @xmath43 that limit the region in which the search process takes place .
when the searcher is at position @xmath44 it can move to any of the nearest neighbouring points to @xmath25 , unless @xmath45 , in which case the motion is allowed only over the subset of nearest neighbouring points that are contained in @xmath42 .
we set the separation between lattice sites to one .
furthermore , we set the time between two movements of the searcher to @xmath46 . at time @xmath47 , the searcher starts at @xmath48 .
initially the belief function is the uniform distribution over the domain @xmath42 , corresponding to a complete lack of knowledge about the position of the source and therefore of maximal entropy @xmath49 , with @xmath50 the total number of points contained in @xmath42 . during the search process
the searcher moves , registers detections , and updates its belief function according to the rules described above .
as detections occur , the belief function starts to concentrate around the true position of the source .
the search ends at the time @xmath2 at which the shannon entropy reaches a value below a certain threshold , which we take to be @xmath51 .
note that this first hitting time criterion for the end of the search is different than the first passage criterion considered in @xcite . for small values of the emission rate @xmath52 ,
the search time obtained from both criteria is different by at most one unit of time .
however , at high values of @xmath7 the first passage time may diverge while our first hitting time criterion is still able to find the source . in such situations
the search ends and the source is located even if the searcher has not physically reached the source . in the rest of the paper we study the statistics of the search time with respect to the geometry of the boundary @xmath43 and the topology of the lattice . in this paper
we consider only low emission rates and set @xmath53 .
therefore , our present results are also valid for the search time corresponding to the first passage time .
in this section we study the dependence of the mean search time as a function of the initial position of the searcher .
we consider an infotactic searcher in a square lattice of dimensions @xmath54 .
the source is located at the center of the lattice and we consider first the case without advection ( @xmath55 ) .
it is natural to expect that the search time increases monotonically with the initial distance of the searcher to the source @xcite .
this is indeed the case in an infinite lattice . here instead we first show in fig . [ fig : tau ] that due to the finite size of the domain with reflecting boundaries , the search time exhibits a complex dependence with the initial distance to the source . each point in fig .
[ fig : tau ] corresponds to the mean search time @xmath56 , averaged over @xmath57 trajectories starting at the same initial position @xmath58 . as a function of the starting position @xmath58 of the searcher .
the source is located at the center of the lattice @xmath59 and @xmath60 .
the error bars corresponds to the standard deviation.,scaledwidth=40.0% ] from fig .
[ fig : tau ] one observes two separate regions : when the starting position is @xmath61 , the mean search time is @xmath62 and shows a small variation with the distance @xmath63 .
surprisingly , we find that when the searcher starts further away at @xmath64 , the mean search time drops down to @xmath65 .
furthermore , near the transition distance @xmath66 , the fluctuations of @xmath13 are larger .
the key observation comes from realizing that , in the absence of detections , the first step of an infotactic searcher is not random but deterministic . with an initial uniform belief function the searcher moves in the direction
in which the expected gain of information is maximal and , in the absence of information about the source location , the initial direction depends only on the parameters of the transport process and the geometry of the domain , but not on the position of source . and @xmath60 in a @xmath54 square lattice .
right panel : expected entropy decrease @xmath67 after the first step.,title="fig:",scaledwidth=19.5% ] and @xmath60 in a @xmath54 square lattice .
right panel : expected entropy decrease @xmath67 after the first step.,title="fig:",scaledwidth=4.3% ] and @xmath60 in a @xmath54 square lattice .
right panel : expected entropy decrease @xmath67 after the first step.,title="fig:",scaledwidth=21.0% ] at dilute conditions , _
i.e. _ far away from the source , it is clear that the initial step will persist at longer times , until the searcher starts detecting the molecules emitted by the source . as a consequence
, the direction of the first step will determine the region the searcher explores first and thus , will have a considerable effect on the value of @xmath56 .
to understand the consequences of the first step we show in fig .
[ fig : nowind ] ( left panel ) , a color code map for the first step of infotaxis .
the color of each lattice site represents the direction of the first step of the searcher when its initial position is at the same lattice site .
it is noticeable that although standing still is always an option for the infotaxis algorithm , we find that this option is never optimal for the first step . from fig .
[ fig : nowind ] it is evident that the first step in infotaxis is deterministic and the direction is strongly influenced by the geometry of the lattice boundary .
the complex dependence of @xmath56 with the initial distance to the source of fig .
[ fig : tau ] can now be explained from the structure of the first - step decision map . at short distances
@xmath68 the first step is upwards ( away from the source ) . on the contrary , at larger distances
@xmath69 the first step is downwards ( towards the source ) .
depending on the direction of its first step , the searcher will explore different regions of the search domain .
if the first step is away from the source , it is less probable for the searcher to have detections and it will keep moving until it reaches the vicinity of the boundary . this first exploration depends on the size and geometry of the domain and the time invested on it will be added to the total search time .
instead , when the searcher starts close to the boundary , it moves towards the center of the lattice , and since the source is placed there , the searcher will start detecting much sooner than in the previous case and the mean search time is smaller . from the previous explanation , it is important to remark that the dependence @xmath56 on the initial distance @xmath63 shown in fig .
[ fig : tau ] will be different if the source is not located at the center , but the structure of the first - step decision map shown in fig .
[ fig : nowind ] does not depend on the position of the source .
( left panel ) and its standard deviation ( right panel ) as a function of the initial position , averaged over @xmath70 trajectories and the same parameters as in fig .
[ fig : nowind].,scaledwidth=45.0% ] to understand the different regions in fig .
[ fig : nowind ] we need to look at the expression for the expected entropy change . since the initial belief map is the uniform distribution , only the sum inside the brackets in eq .
changes for different choices of @xmath28 .
this quantity depends on the position through the functions @xmath71 and @xmath72 . in the right panel of fig .
[ fig : nowind ] we show for each lattice site , the maximal expected decrease of entropy as given by eq . .
there are four minima on each diagonal of the square domain which act as attractors for the infotaxis , and the the first step of the searcher is determined by them .
the locus of points for which the maximal decrease in entropy occurs for two different directions define the smooth curves separating the different regions in the left panel of fig .
[ fig : nowind ] .
the symmetry of the map reflects the symmetry of the boundary , and in this case of no advection ( @xmath55 ) the isotropy of the transport process .
therefore , the first - step decision map is determined by the geometry of the domain via the way in which the expected maximal entropy decrease . this generates a well defined partition in the search domain into regions with a structure that inherit , in the absence of wind , the same symmetry as that of the boundary . with the source located at the center of the lattice , in fig .
[ fig : nowind _ ] we show , for each initial position in the lattice , the mean search time @xmath56 ( left panel ) and its standard deviation ( right panel ) as a color density plot .
the regions near the midpoints of the edges on the boundary lead to the shortest search times . as discussed above , in the absence of detections , the initial first steps direct the searcher trajectory to the center of the lattice , and therefore , to the region in which a larger number of detections is expected .
more interesting , it is clear from the right panel that the fluctuations of the search time @xmath13 are larger over the partition curves of the first - step decision map .
the fluctuations will be discussed in greater detail in section [ sec : pw ] .
in this section we complete the study of the effects that different boundary geometries , lattice topologies and the parameters of the transport process have on the structure of the first - step decision map , and consequently on the average search time .
we start by computing the first - step decision map on a square lattice with square , circular and triangular boundary .
the results are shown in fig.[fig : boundary ] for a correlation length @xmath60 .
we see how the partition of the search domain is affected by the geometry and symmetry of the boundary .
the conclusions drawn in the previous section apply to all geometries : in the absence of detections , the first step of an infotactic searcher is determined by the geometry of the boundary .
the center of the lattice acts as a repeller , i.e. all searches starting at neighbouring points will have their first step away from the center .
the boundary by construction is also repelling , but we see in fig .
[ fig : boundary ] that there are also attracting points : four in the case of the square boundary , three for the triangular and none for the circular case . in terms of the expected entropy variation
, this means that imposing reflecting boundaries , an infotactic search will always produce maximal entropy decrease in regions that are away from the center and the boundary of the search domain ( see the right panel of fig .
[ fig : nowind ] ) . , @xmath60 and different boundary geometries .
the motion of the searcher is on a square lattice .
, title="fig:",scaledwidth=45.0% ] , @xmath60 and different boundary geometries .
the motion of the searcher is on a square lattice .
, title="fig:",scaledwidth=2.9% ] , and three different values of the correlation length @xmath60 ( left ) , @xmath73 ( middle ) and @xmath74 ( right ) .
, scaledwidth=45.0% ] in the previous figures , the correlation length @xmath60 has the same size as the entire search domain . roughly speaking
@xmath11 is the average distance travelled by the molecules emitted by the source before decaying .
therefore , the statistical inference done by infotaxis has an effect over an area with extension @xmath11 around the searcher . in fig .
[ fig : lambda ] we study the dependence of the first - step decision map on the correlation length for a square lattice . to consider different values of the correlation length we have kept fixed the effective diffusivity @xmath75 and varied the molecules life time @xmath76 . as the correlation length decreases , the central area of the decision map extends towards the boundary .
clearly , the relevant parameter is the ratio between @xmath11 and the size of the lattice . the results in the right panel of fig .
[ fig : lambda ] suggest that in the limit in which the lattice size is much larger than the correlation length , the central repeller dominates the dynamics .
as expected , this happens irrespectively of the geometry of the boundary .
the effect of the boundary becomes relevant only at distance @xmath11 from it .
moreover , assuming that the source is located near the center of the lattice , the searcher will start moving away form the source into regions where the detections become more rare , and this outward motion will only stop when the searcher reaches the neighborhood of the boundary .
in particular , this means that in an infinite domain infotactic searches in the absence of wind will rarely be successful .
besides the effect of boundary discussed above , the topology of the lattice also influences the structure of the decision map .
we illustrate this in fig . [
fig : lambda_hex ] where the first - step decision map is shown for an hexagonal lattice with a square boundary .
the symmetry of the partitions inherit the symmetry of the lattice topology , while the rest of the properties found for the square lattice remain the same . to conclude this section we consider the structure of the decision map for the case in which an advection field exists .
this is shown in fig .
[ fig : wind ] for three different values of the wind velocity @xmath14 .
the presence of wind destroys the isotropy of the problem . a non - zero @xmath14 yields additional information on the region of the search domain where the source may be located .
roughly speaking , at large pclet numbers , when the advection dominates the molecular diffusion , if the searcher does not record detections then the source is most likely located downwards ( recall that the wind is directed along the negative @xmath15-axis ) .
conversely , if detections occur then the searcher will automatically move upwards as the detected molecule is expected to come from that region .
this extra information clearly appears in the decision map . in the absence of detections , the region in which the first step is downwards dominates at larger values of @xmath14 .
the region corresponding to upwards motion shrinks to zero and the regions in which the first move is left or right remain important only near the boundary . , and three different values of the correlation length @xmath60 ( left ) , @xmath73 ( middle ) and @xmath74 ( right).,title="fig:",scaledwidth=45.0% ] , and three different values of the correlation length @xmath60 ( left ) , @xmath73 ( middle ) and @xmath74 ( right).,title="fig:",scaledwidth=2.6% ]
in the last section we focus on the fluctuations of the search time in terms of the statistics of the _ simultaneity _ index of the search as discussed in @xcite . consider two independent searchers that start simultaneously at the same position @xmath77 , with corresponding search times @xmath78 and @xmath79 .
one defines then the random variable @xmath80 such that @xmath81 ranges in the interval @xmath82 $ ] .
the _ simultaneity index _
@xmath81 measures the _ likelihood _ that both searches end simultaneously , and thus it sheds light about the size of the fluctuations of the search time and the robustness of the search in general .
when @xmath83 any typical search will end at similar times , meaning that the most probable value for the search time is close to its statistical mean . on the contrary ,
when @xmath84 or @xmath85 it means that the search will end at very dissimilar times , so that large fluctuations in the search time are expected .
the statistics of the simultaneity index is a quantitative measure of the robustness of a search process when finding the source at reasonably similar times is important .
these statistics have been used in the past in the analysis of random probabilities induced by normalisation of self - similar lvy processes @xcite , of the fractal characterisation of paretian poisson processes @xcite , and of the so - called matchmaking paradox @xcite , and more recently , to quantify sample - to - sample fluctuations in mathematical finances @xcite , chaotic systems @xcite , the analysis of distributions of the diffusion coefficient of proteins diffusing along dnas@xcite and optimal estimators of the diffusion coefficient of a single brownian trajectory.@xcite . in section [ sec : inipos ] , we have observed that the fluctuations of the search time depend on the position from which the search starts .
we showed in the right panel of fig .
[ fig : nowind _ ] that when the search starts over a line separating different regions of the decision map the standard deviation is largest . in this section
, we explore this issue using the statistics of the simultaneity index .
( left ) , @xmath86 ( center ) and @xmath87 ( right ) .
the corresponding values of the correlation length determined by equation are @xmath88 ( left ) , @xmath89 ( center ) and @xmath90 ( right ) .
[ fig : wind],title="fig:",scaledwidth=15.0% ] ( left ) , @xmath86 ( center ) and @xmath87 ( right ) .
the corresponding values of the correlation length determined by equation are @xmath88 ( left ) , @xmath89 ( center ) and @xmath90 ( right ) .
[ fig : wind],title="fig:",scaledwidth=15.0% ] ( left ) , @xmath86 ( center ) and @xmath87 ( right ) .
the corresponding values of the correlation length determined by equation are @xmath88 ( left ) , @xmath89 ( center ) and @xmath90 ( right ) .
[ fig : wind],title="fig:",scaledwidth=15.0% ] ( left ) , @xmath86 ( center ) and @xmath87 ( right ) .
the corresponding values of the correlation length determined by equation are @xmath88 ( left ) , @xmath89 ( center ) and @xmath90 ( right ) .
[ fig : wind],title="fig:",scaledwidth=3.3% ] we have numerically computed the probability distribution @xmath91 for different initial positions of the searcher and the source located at the origin .
the histograms shown in the left panel of fig .
[ fig : pw ] correspond to some of the initial positions considered in fig .
[ fig : tau ] . for @xmath92 , @xmath91
is found to be a unimodal , bell - shaped function with a maximum at @xmath93 .
this means that two different infotactic searchers beginning their trajectory within this region will most likely employ the same time to find the source for the first time . for @xmath64 ,
@xmath91 remains unimodal , but with a larger dispersion than before .
more interestingly , for @xmath94 , when the initial position lays on one of the separatrix curves , @xmath91 exhibits a bimodal , m - shaped form with a local minimum at @xmath95 and two maxima at around @xmath96 and @xmath971 , meaning that most likely , different searchers will find the source at quite different times , which in turn implies that trajectory - to - trajectory fluctuations are expected to be large .
this is exactly what it has been qualitatively observed in fig . [
fig : nowind _ ] one can grasp the origin of these large fluctuations by looking at how the searcher approaches to the source in time . in the right panel of fig .
[ fig : pw ] we show how the distance from the source evolves on average . in agreement with the first - step decision map discussed in section [ sec : inipos ] , the searchers with initial position near the center of the lattice move outwards from the center ( and from the source ) for a time @xmath98 , after which they approach the source monotonically until detection become frequent . in the opposite case
in which the searcher s initial position is far from the center , they rapidly approach the source . as a final remark ,
we have computed the distribution @xmath91 averaged over all initial positions in the lattice ( see the black dashed curve in the left panel of fig .
[ fig : pw ] ) .
the bell - shaped distribution indicates that infotaxis is a robust search algorithm in bounded domains , even when the exact initial position of the search is not accessible . :
@xmath99 ( red solid ) , @xmath100 ( brown dashed ) , @xmath101 ( blue dotted ) and @xmath102 ( orange dotted - dashed ) .
these results were obtained for a square lattice , with @xmath55 and @xmath60 averaged over @xmath57 trajectories .
[ fig : pw ] , title="fig:",scaledwidth=24.0% ] : @xmath99 ( red solid ) , @xmath100 ( brown dashed ) , @xmath101 ( blue dotted ) and @xmath102 ( orange dotted - dashed ) .
these results were obtained for a square lattice , with @xmath55 and @xmath60 averaged over @xmath57 trajectories .
[ fig : pw ] , title="fig:",scaledwidth=24.0% ]
we have studied the infotaxis search strategy in finite domains with a reflecting boundary .
we find that the time at which infotactic searchers locate the source shows a complex dependence on the starting position .
we explain this dependence by noting that in the absence of detections the first step of infotaxis is deterministic , and we study the decision map for this first step .
the search domain is partitioned into regions on which the first step is constant , and the boundaries of these regions are determined by the geometry of the boundary , the topology of the lattice and in general by the parameters of the transport process .
when the starting position falls near a separatrix curve , the fluctuations around the mean search time increase .
these higher fluctuations can also be explained in terms of the initial step .
when the correlation length is much smaller than the domain size , the effectiveness of infotaxis will decrease .
when the correlation length is comparable to the domain size , infotaxis remains on average a robust search strategy .
this work has been supported by grant n. 245986 of the eu - project robots fleets for highly agriculture and forestry management .
jdr was also supported by a picata predoctoral fellowship of the moncloa campus of international excellence ( ucm - upm ) .
the research of dgu has been supported in part by spanish mineco - feder grants mtm2012 - 31714 and fis2012 - 38949-c03 - 01 .
we acknowledge the use of the upc applied math cluster system for research computing ( see http://www.ma1.upc.edu/eixam/index.html ) .
d. boyer , d. s. dean , c. meja - monasterio , and g. oshanin , optimal least - squares estimators of the diffusion constant from a single brownian trajectory , _ eur .
. j. special topics _ * 216 * , 0 57 ( 2013 ) . | in olfactory search an immobile target emits chemical molecules at constant rate .
the molecules are transported by the medium which is assumed to be turbulent . considering a searcher able to detect such chemical signals and
whose motion follows the infotaxis strategy , we study the statistics of the first - passage time to the target when the searcher moves on a finite two - dimensional lattice of different geometries .
far from the target , where the concentration of chemicals is low the direction of the searcher s first movement is determined by the geometry of the domain and the topology of the lattice , inducing strong fluctuations on the average search time with respect to the initial position of the searcher .
the domain is partitioned in well defined regions characterized by the direction of the first movement . if the search starts over the interface between two different regions , large fluctuations in the search time are observed . |
* figure 1 : linearly polarized intensity @xmath29 at 2.3 ghz from s - pass * ( @xmath30 ) .
the thick dashed lines delineate the radio lobes reported in this letter , while the thin dashed lines delimit the @xmath0-ray fermi bubbles@xcite .
the map is in galactic coordinates , centred at the galactic centre with galactic east to the left and galactic north up ; the galactic plane runs horizontally across the centre of the map .
the polarized flux density is indicated by the scale bar given in unit of jy / beam with a beam size of 10.75 ( 1 jy @xmath31 w m@xmath32hz@xmath33 ) .
the lobe edges follow the @xmath0-ray border up to @xmath34 where the radio emission extends beyond .
the three polarized radio ridges discussed in the text are also indicated along with the two limb brightening spurs .
the ridges appear to be the front side of a continuous windings of collimated structures around the general biconical outflow of the lobes ( see text ) .
the galactic centre spur is nearly vertical at low latitude , possibly explained by a projection effect if it is mostly at the front of the northern lobe . at its higher latitudes ,
the galactic centre spur becomes roughly parallel with the northern ridge ( above ) which itself exhibits little curvature ; this is consistent with the overall outflow s becoming cylindrical above 4 - 5 kpc as previously suggested@xcite .
in such a geometry , synchrotron emission from the rear side of each cone is attenuated by a factor @xmath35 with respect to the front - side , rendering it difficult to detect the former against the foreground of the latter and of the galactic plane ( see supplementary information ) .
* figure 2 : lobes polarized intensity and @xmath0ray spurs . *
schematic rendering of the edges of two @xmath0-ray substructures evident in the 2 - 5 gev fermi data as displayed in figure 2 of ref.@xcite which seem to be counterparts of the galactic centre spur and the southern ridge .
the latter appears to be connected to the galactic centre by its @xmath0-ray counterpart . with
the flux densities and polarization fraction quoted in the text we can infer equipartition@xcite magnetic field intensities of @xmath36 g ( 1 @xmath2 g @xmath37 t ) if the synchrotron - emitting electrons occupy the entire volume of the lobes , or @xmath38 g if they occupy only a 300 pc thick skin ( the width of the ridges ) . for the southern ridge , @xmath39 g ; for the galactic centre spur @xmath40 g ; and , for the northern ridge , @xmath41 g .
note the large area of depolarization and small angular scale signal modulation visible across the galactic plane extending up to @xmath42 on either side of the galactic centre ( thin dashed line ) .
this depolarization is due to faraday rotation by a number of shells that match h@xmath28 emission regions@xcite , most of them lying in the sagittarius arm at distances from the sun up to 2.5 kpc , and some in the scutum - centaurus arm at @xmath123.5 kpc .
the small scale modulation is associated to weaker h@xmath28 emission encompassing the same hii regions and most likely associated with the same spiral arms .
thus 2.5 kpc constitutes a lower limit to the lobes near side distance and places the far side beyond 5.5 kpc from the sun ( cf . ref .
@xcite ) . along with their direction in sky
, this suggests the lobes are associated with the galactic bulge and/or centre .
* figure 3 : polarized intensity and magnetic angles at 23 ghz from wmap[@xcite]*. the magnetic angle is orthogonal to the emission polarization angle and traces the magnetic field direction projected on to the plane of the sky .
the three ridges are obvious while traces of the radio lobes are visible ( 2.3 ghz edges shown by the black solid line ) .
the magnetic field is aligned with the ridges and curves following their shape .
two spurs match the lobe edges northwest and southwest of galactic centre and could be limb brightening of the lobes .
a third limb brightening spur candidate is also visible north east of the galactic centre .
the map is in galactic coordinates , centred at the galactic centre .
grid lines are spaced by @xmath43 .
the emission intensity is in brightness temperature , the unit is k. data have been binned in 1@xmath44 pixels to improve the signal - to - noise ratio . from a combined analysis of microwave and
@xmath0-ray data ( see also the supplementary information ) we can derive the following magnetic fields limits ( complementary to the equipartition limits reported in the text and figure 2 ) : for the overall lobes / bubbles @xmath45 g and for the galactic centre spur @xmath46 g @xmath47 g . *
figure 4 : the vertical range of cosmic ray electrons as a function of their kinetic energy . *
( due to geometrical uncertainties , adiabatic losses can not be determined so that the plotted range actually constitutes an upper limit . )
electrons are taken to be transported with a speed given by the sum of the inferred vertical wind speed ( 1100 km / s ) and the vertical component of the alfven velocity in the magnetic field .
the former is inferred from the geometry of the northern ridge : if its footpoint has executed roughly half an orbit in the time the galactic centre spur has ascended to its total height of @xmath48 kpc , its upward velocity must be close to 1,000 km / s@xmath49[@xcite ] , where we have normalised to a footpoint rotation speed of 80 km / s at a radius of 100 pc from the galactic centre@xcite ( detailed analysis gives 1100 km / s : see the supplementary information ) . in a strong , regular magnetic field ,
the electrons are expected to stream ahead of the gas at the alfven velocity@xcite in either the ridges ( @xmath50 g , @xmath51 km / s ; this is a lower limit given that @xmath52 @xmath53 on the basis of the rosat data@xcite ) or in the large - scale field of the lobes ( @xmath54 g , @xmath55 km / s for @xmath56 @xmath53 in the lobes interior as again implied by the data ) .
also plotted are : the characteristic energies of electrons synchrotron radiating at 2.3 and 23 ghz ( for a @xmath57 g field ) and into 1 and 50 gev @xmath0-rays via inverse compton ( ` ic ' ) upscattering of a photon background with characteristic photon energy 1 ev ; and the approximate 7 kpc distance of the top of the northern ridge from the galactic plane . | the nucleus of the milky way is known to harbour regions of intense star formation activity as well as a super - massive black hole@xcite .
recent fermi space telescope observations have revealed regions of @xmath0-ray emission reaching far above and below the galactic centre , the so - called fermi bubbles@xcite .
it is uncertain whether these were generated by nuclear star formation or by quasar - like outbursts of the central black hole@xcite and no information on the structures magnetic field has been reported . here
we report on the detection of two giant , linearly - polarized radio lobes , containing three ridge - like sub - structures , emanating from the galactic centre .
the lobes each extend @xmath1 , bear a close correspondence to the fermi bubbles , are located in the galactic bulge , and are permeated by strong magnetic fields of up to 15 @xmath2 g .
our data signal that the radio lobes originate in a bi - conical , star - formation ( rather than black hole ) driven outflow from the galaxy s central 200 pc that transports a massive magnetic energy of @xmath3 erg into the galactic halo .
the ridges wind around this outflow and , we suggest , constitute a ` phonographic ' record of nuclear star formation activity over at least 10 myr . ettore carretti,@xmath4 roland m. crocker,@xmath5 lister staveley - smith@xmath6 , marijke haverkorn@xmath7 , cormac purcell@xmath8 , b. m. gaensler@xmath8 , gianni bernardi@xmath9 , michael j. kesteven@xmath10 , and sergio poppi@xmath11 csiro astronomy and space science , po box 276 , parkes , nsw 2870 , australia max - planck - institut f " ur kernphsik , p.o .
box 103980 heidelberg , germany research school of astronomy & astrophysics , australian national university , weston creek , act 2611 , australia international centre for radio astronomy research , m468 , university of western australia , crawley , wa 6009 , australia arc centre of excellence for all - sky astrophysics department of astrophysics / imapp , radboud university nijmegen , p.o .
box 9010 , 6500 gl nijmegen , the netherlands leiden observatory , leiden university , p.o .
box 9513 , 2300 ra leiden , the netherlands sydney institute for astronomy , school of physics , the university of sydney , nsw 2006 , australia harvard smithsonian center for astrophysics , 60 garden street , cambridge , ma , 02138 , usa csiro astronomy and space science , po box 76 , epping , nsw 1710 , australia inaf osservatorio astronomico di cagliari , st .
54 loc .
poggio dei pini , i-09012 capoterra ( ca ) , italy we use the images of the recently concluded s - band polarization all sky survey ( s - pass ) that has mapped the polarized radio emission of the entire southern sky with the parkes radio telescope at a frequency of 2307 mhz , with 184 mhz bandwidth , and 9 angular resolution@xcite . the lobes exhibit diffuse polarized emission ( figure [ fig_1 ] ) , an integrated total intensity flux of 21 kjy , and high polarization fraction of 25% .
they trace the fermi bubbles excepting the top western ( i.e. , right ) corners where they extend beyond the @xmath0-ray region .
depolarization by hii regions establish that the lobes are almost certainly associated with the galactic centre ( figure [ lobes : fig ] and supplementary information ) , implying their height is @xmath128 kpc .
archival data of wmap[@xcite ] reveal the same structures at the microwave frequency of 23 ghz ( figure [ wmap : fig ] ) .
the 2.3 to 23 ghz spectral index @xmath13 ( with flux density @xmath14 modelled as @xmath15 ) of linearly - polarized emission interior to the lobes spans the range @xmath13 = [ -1.0 , -1.2 ] generally steepening with projected distance from the galactic plane ( see the supplementary information ) . along with the high polarization fraction , this phenomenology indicates the lobes are due to cosmic ray electrons , transported from the plane , synchrotron - radiating in a partly ordered magnetic field .
three distinct emission ridges that all curve towards galactic west with increasing galactic latitude are visible within the lobes ( figure [ fig_1 ] ) ; two other substructures proceeding roughly north - west and south - west from around the galactic centre hint at limb brightening in the biconical base of the lobes .
these substructures all have counterparts in wmap polarization maps ( figure [ wmap : fig ] ) and one of them@xcite , already known from radio continuum data as the galactic centre spur@xcite , appears to connect back to the galactic centre ; we label the other substructures the northern and southern ridges .
the ridges magnetic field directions ( figure [ wmap : fig ] ) curve following their structures . the galactic centre spur and southern ridges also seem to have gev @xmath0-ray counterparts ( figure [ lobes : fig ] also cf .
ref.[@xcite ] ) .
the two limb brightening spurs at the biconical lobe base are also visible in the wmap map , where they appear to connect back to the galactic centre area .
a possible third spur develops north - east from the galactic centre .
these limb brightening spurs are also obvious in the stokes u map as an x - shape structure centred at the galactic centre ( figure s3 of supplementary information ) .
such coincident , non - thermal radio , microwave and @xmath0-ray emission indicates the presence of a non - thermal electron population covering at least the energy range 1 - 100 gev ( figure [ fig_4 ] ) that is simultaneously synchrotron radiating at radio and microwave frequencies and up - scattering ambient radiation into @xmath0-rays by the inverse compton process .
the widths of the ridges are remarkably constant at @xmath12 300 pc over their lengths .
the ridges have polarization fractions of 25 to 31% ( see supplementary information ) , similar to the average over the lobes . given this emission and the stated polarization fractions , we infer magnetic field intensities @xmath16 g for the lobes and @xmath17 g the ridges ( see figure [ lobes : fig ] and [ wmap : fig ] and the supplementary information ) .
an important question about the fermi bubbles is whether they are ultimately powered by star - formation or by activity of the galaxy s central , super - massive black hole .
despite their very large extent , the @xmath0-ray bubbles and the x - shaped polarized microwave and x - ray structures tracing their limb - brightened base@xcite have a narrow waist of only 100 - 200 pc diameter at the galactic centre .
this matches the extent of the star - forming molecular gas ring ( @xmath18 ) recently demonstrated to occupy the region@xcite . with 5 - 10% of the galaxy s molecular gas content@xcite ,
star - formation activity in this ` central molecular zone ' is intense , accelerating a distinct cosmic ray population@xcite and driving an outflow@xcite of hot , thermal plasma , cosmic rays , and ` frozen - in ' magnetic field lines@xcite .
one consequence of the region s outflow is that the cosmic ray electrons accelerated there ( dominantly energised by supernovae ) are advected away before they lose much energy radiatively in situ@xcite .
this is revealed by the fact that the radio continuum flux on scales up to 800 pc around the galactic centre is in anomalous deficit with respect to the expectation afforded by the empirical far - infrared - radio continuum correlation@xcite . the total 2.3 ghz radio continuum flux from the lobes of @xmath12 21 kjy , however , saturates this correlation normalising to the inner @xmath19 pc diameter region s 60 @xmath2 m flux of 2 mjy[@xcite ] .
together with the morphological evidence , this strongly indicates the lobes are illuminated by cosmic ray electrons accelerated in association with star - formation within this region ( see the supplementary information ) not as a result of black hole activity .
the ridges appear to be continuous windings of individual , collimated structures around a general biconical outflow out of the galactic centre .
the sense of galactic rotation ( clockwise as seen from galactic north ) and angular - momentum conservation mean that the ridges get ` wound - up'@xcite in the outflow with increasing distance from the plane explaining the projected curvature of the visible , front - side of the ridges towards galactic west .
polarized , rear - side emission is attenuated rendering it difficult to detect against the lobes front - side and the galactic plane s stronger emission ( figure [ fig_1 ] and the supplementary information ) . for cosmic ray electrons synchrotron - emitting at 2.3 ghz to be able to ascend to the top of the northern ridge at @xmath12 7 kpc in the time it takes them to cool ( mostly via synchrotron emission itself ) requires vertical transport speeds of @xmath20500 km / s ( for 15 @xmath2 g ; see figure [ fig_4 ] ) . given the geometry of the gc spur , the outflowing plasma is moving at @xmath21 km / s ( figure [ fig_4 ] and the supplementary information ) , somewhat faster than the @xmath22 km / s gravitational escape velocity from the galactic centre region@xcite , implying that 2.3-ghz - radiating electrons can , indeed , be advected to the top of the ridges in their loss time .
given the calculated fields and the speed of the outflow , the total magnetic energy for each of the ridges , @xmath23 erg ( see the supplementary information ) , is injected at a rate @xmath24 erg / s over a few @xmath25 years ; this is very close to the rate at which independent modelling@xcite suggests galactic centre star formation is injecting magnetic energy into the region s outflow .
on the basis of the ridges individual energetics , geometry , outflow velocity , timescales , and plasma content ( see supplementary information ) , we suggest that their footpoints are energised by and rotate with the super - stellar clusters inhabiting@xcite the inner @xmath12100 pc ( in radius ) of the galaxy .
in fact , we suggest that the ridges constitute ` phonographic ' recordings of the last @xmath1210 myr of galactic centre star - formation . given its morphology , the galactic centre spur likely still has an active footprint .
in contrast , the northern and southern ridges seem not to connect to the plane at 2.3 ghz .
this may indicate their footpoints are no longer active though the southern ridge may be connected to the plane by a gamma - ray counterpart ( see fig .
[ lobes : fig ] ) .
unfortunately , present data do not allow us to trace the galactic centre spur all the way down to the plane but a connection between this structure and one ( or some combination ) of the @xmath26-scale radio continuum spurs@xcite emanating north of the star - forming giant molecular cloud complexes sagittarius b and c or the bright , non - thermal ` radio arc'@xcite ( itself longitudinally coincident with the @xmath124 myr quintuplet@xcite stellar cluster ) , seems plausible .
the magnetic energy content of both lobes is much larger than the ridges , @xmath27 erg .
this suggests the magnetic fields of the lobes are the result of the accumulation of a number of star formation episodes .
alternatively , if the lobes field structure were formed over the same timescale as the ridges , it would have to be associated with recent activity of the super - massive black hole , perhaps occurring in concert with enhanced nuclear star - formation activity@xcite .
our data indicate the process of gas accretion on to the galactic nucleus inescapably involves star - formation which , in turn , energises an outflow .
this carries away low angular momentum gas , cosmic rays and magnetic field lines and has a number of important consequences : i ) the dynamo activity in the galactic centre@xcite , likely required to generate its strong@xcite in situ field , requires the continual expulsion of small - scale helical fields to prevent dynamo saturation@xcite ; the presence of the ridges high in the halo may attest to this process .
ii ) the lobes and ridges reveal how the very active star formation in the galactic centre generates and sustains a strong , large - scale magnetic field structure in the galactic halo .
the effect of this on the propagation of high - energy cosmic rays in the galactic halo should be considered .
iii ) the process of gas expulsion in the outflow may explain how the milky way s super - massive black hole is kept relatively quiescent@xcite despite sustained , inward movement of gas
. 1 morris , m. , & serabyn , e. the galactic centre environment , ann .
rev .
astron .
astrophys . , 34 , 645 - 702 ( 1996 ) su , m. , slatyer , t. r. , & finkbeiner , d. p.
giant gamma - ray bubbles from fermi - lat : active galactic nucleus activity or bipolar galactic wind ?
, , 724 , 1044 - 1082 ( 2010 ) su , m. , & finkbeiner , d. p. evidence for gamma - ray jets in the milky way , , 753 , 61 ( 2012 ) zubovas , k. , king , a. r. , & nayakshin , s. the milky way s fermi bubbles : echoes of the last quasar outburst ?
, , 415 , l21-l25 ( 2011 ) crocker , r. m. , & aharonian , f. fermi bubbles : giant , multibillion - year - old reservoirs of galactic center cosmic rays , physical review letters , 106 , 101102 ( 2011 ) crocker , r. m. non - thermal insights on mass and energy flows through the galactic centre and into the fermi bubbles , 423 , 3512 - 3539 ( 2012 ) carretti , e. , galactic foregrounds and cmb polarization , in : the dynamic ism : a celebration of the canadian galactic plane survey , eds .
r. kothes , t. l. landecker , and a. g. willis 6 - 10 june 2010 , naramata , bc , canada , asp conf . ser .
cs-438 , 276 ( 2011 ) hinshaw , g. , weiland , j. l. , hill , r. s. , et al .
five - year wilkinson microwave anisotropy probe observations : data processing , sky maps , and basic results , astrophys .
j.,suppl . , 180
, 225 - 245 ( 2009 ) jones , d. i. , crocker , r. m. , reich , w. , ott , j. , & aharonian , f. a. magnetic substructure in the northern fermi bubble revealed by polarized microwave emission , , 747 , l12-l15 ( 2012 ) sofue , y , reich , w. , & reich , p. the galactic center spur - a jet from the nucleus ?
, , 341 , l47-l49 ( 1989 ) beck , r. , & krause , m. revised equipartition and minimum energy formula for magnetic field strength estimates from radio synchrotron observations , astronomische nachrichten , 326 , 414 - 427 ( 2005 ) bland - hawthorn , j. , & cohen , m. the large - scale bipolar wind in the galactic center , , 582 , 246 - 256 ( 2003 ) molinari , s. , bally , j. , noriega - crespo , a. , et al . a 100 pc elliptical and twisted ring of cold and dense molecular clouds revealed by herschel around the galactic center , , 735 , l33-l39 ( 2011 ) aharonian , f. a. et al .
discovery of very - high - energy @xmath0-rays from the galactic centre ridge , , 439 , 695 - 698 ( 2006 ) crocker , r. m. , jones , d. i. , aharonian , f. , law , c. j. , melia , f. , oka , t. , & ott , j. wild at heart : the particle astrophysics of the galactic centre , , 413 , 763 - 788 ( 2011 ) law , c. j. a multiwavelength view of a mass outflow from the galactic center , , 708 , 474 - 484 ( 2010 ) crocker , r. m. , jones , d. i. , aharonian , f. , et al
. @xmath0-rays and the far - infrared - radio continuum correlation reveal a powerful galactic centre wind , , 411 , l11-l15 ( 2011 ) crocker , r. m. , jones , d. i. , melia , f. , ott , j. , & protheroe , r. j. a lower limit of 50 microgauss for the magnetic field near the galactic centre , , 463 , 65 - 67 ( 2010 ) yun , m. s. , reddy , n. a. , & condon , j. j. radio properties of infrared - selected galaxies in the iras 2 jy sample , , 554 , 803 - 822 ( 2001 ) launhardt , r. , zylka , r. , & mezger , p. g. the nuclear bulge of the galaxy iii .
large - scale physical characteristics of stars and interstellar matter , , 384 , 112 - 139 ( 2002 ) heesen , v. , beck , r. , krause , m. , & dettmar , r .- j .
cosmic rays and the magnetic field in the nearby starburst galaxy ngc 253 iii .
helical magnetic fields in the nuclear outflow , , 535 , a79 ( 2011 ) muno , m. p. , et al .
diffuse x - ray emission in a deep chandra image of the galactic center , , 613 , 326 - 342 ( 2004 ) kulsrud , r. , & pearce , w. p. the effect of wave - particle interactions on the propagation of cosmic rays , , 156 , 445 - 469 ( 1969 ) pohl ,
m. , reich , w. , & schlickeiser , r. synchrotron modelling of the 400 pc spur at the galactic center , , 262 , 441 - 454 ( 1992 ) humann , b. , stolte , a. , brandner , w. , & gennaro , m. the present - day mass function of the quintuplet cluster , astron.astrophys .
, 540 , a57 ( 2012 ) ferrire , k. interstellar magnetic fields in the galactic center region , , 505 , 1183 - 1198 ( 2009 ) brandenburg , a. , & subramanian , k. astrophysical magnetic fields and nonlinear dynamo theory , phys.rep . , 417 , 1 - 209 ( 2005 ) gaustad , j. e. , mccullough , p. r. , rosing , w. , & van buren , d. , a robotic wide - angle h - alpha survey of the southern sky publications of the astronomical society of the pacific , 113 , 1326 - 1348 ( 2001 ) almy , r. c. et al . , distance limits on the bright x - ray emission toward the galactic center : evidence for a very hot interstellar medium in the galactic x - ray bulge , , 545 , 290 - 300 ( 2000 ) is linked to the online version of the paper at www.nature.com/nature .
this work has been carried out in the framework of the s - band all sky survey collaboration ( s - pass ) .
we thank the parkes telescope staff for outstanding support both while setting up the non - standard observing mode and during the observing runs .
rmc thanks felix aharonian , rainer beck , geoff bicknell , david jones , casey law , mark morris , christoph pfrommer , wolfgang reich , andrea stolte , troy porter , and heinz vlk for discussions and the max - planck - institut f " ur kernphsik for supporting his research .
rmc also acknowledges the support of a future fellowship from the australian research council through grant ft110100108 .
bmg and cp acknowledge the support of an australian laureate fellowship from the australian research council through grant fl100100114 .
mh acknowledges the support of the research programme 639.042.915 , which is partly financed by the netherlands organisation for scientifc research ( nwo ) .
the parkes radio telescope is part of the australia telescope national facility which is funded by the commonwealth of australia for operation as a national facility managed by csiro .
ec performed the s - pass observations , was the leader of the project , developed and performed the data reduction package , did the main analysis and interpretation .
rmc provided theoretical analysis and interpretation .
lss , mh , and sp performed the s - pass observations .
mjk performed the telescope special setup that has allowed the survey execution .
lss , mh , bmg , gb , mjk , and sp were co - proposers and contributed to the definition of the project .
cp performed the estimate of the h@xmath28 depolarizing region distance .
ec and rmc wrote the paper together .
all the authors discussed the results and commented on the manuscript .
the authors declare that they have no competing financial interests .
correspondence and requests for materials should be addressed to e.c .
( email : [email protected] ) .
reprints and permissions information is available at npg.nature.com/reprintsandpermissions . |
the determination of cosmographic parameters , such as @xmath5 and @xmath6 , has a long and interesting history in cosmology ( see , e.g. , @xcite ) . in particular
, the evolution of such parameters provides a unique and direct method to map the expansion history of the universe in a model - independent way .
since all evidence we have so far for the current cosmic acceleration are indirect @xcite , extracting the evolution of these two parameters from future redshift surveys constitutes one of the major challenges in observational cosmology may also reduce the so - called smearing effect that makes constraining the dark energy equation of state ( eos ) @xmath1 extremely difficult @xcite . ] . in this paper
, we investigate how well cosmography may provide a model - independent way to check the reality of cosmic acceleration .
specifically , we study the evolution of the deceleration parameter from the @xmath0 data which are to become available by some planned projects . to this end
, we performed a monte carlo ( mc ) simulation based on @xmath1cdm models ( of which the @xmath7cdm model is a special case with @xmath8 , where @xmath9 stands for the dark energy pressure and @xmath10 for its energy density ) .
we generated three samples of @xmath0 with increasing accuracy and derived @xmath2 from numerical differentiation .
we show that a comographic evidence for cosmic acceleration will be possible only with a great ( maybe out of the perspective of current planned surveys ) accuracy in @xmath0 data .
it is worth mentioning that the determination of @xmath2 will also allow us to estimate the transition redshift , @xmath4 , which corresponds to the epoch when the hubble expansion switched from a decelerating to an accelerating phase . since @xmath11 , the same technique
can also be used to map the recent past evolution of @xmath12 in a given model of dark energy .
recently , it has been shown that luminous red galaxies ( lrg s ) can provide us with direct measurements of the expansion rate @xmath0 @xcite ( see also @xcite for a recent review on @xmath0 measurements from different techniques ) at different redshifts will also be possible through measurements of the line - of - sight or radial component of baryonic acoustic oscillations ( bao ) from large redshift surveys with redshift precision of the order of @xmath13 ( see , e.g. , @xcite ) . ] .
this can be done by calculating the derivative of cosmic time with respect to redshift @xmath14 from measurements of age difference between two passively evolving galaxies at different @xmath15 .
simon , verde and jimenez @xcite ( svj ) have demonstrated the feasibility of directly measuring @xmath0 using this differential age method and have provided 9 determinations of the expansion rate at @xmath16 .
more recently , stern _
_ @xcite have updated svj sample to 11 estimates of @xmath0 lying in the redshift interval @xmath17 .
new age - redshift datasets for different galaxy velocity dispersion groups ( 7 lrg s sample ) have also been made available by the sdss collaboration @xcite .
@xcite have also pointed out that in the near future , the atacama cosmology telescope ( act ) is expected to provide observations of over 500 galaxy cluster up to @xmath18 .
this , together with spectra to be acquired in other telescopes in chile and southern african large telescope ( salt ) in south africa , will provide a sample of more than 2000 passively evolving galaxies in the redshift range @xmath19 . from these observations , it will be possible to determine @xmath20 values of the hubble parameter at a @xmath21 accuracy level if the ages of the galaxies are estimated with @xmath22 error . following a similar approach , crawford _ et al .
@xcite examined the observational requirements to estimate @xmath0 to a given precision . in their simulation ,
observations of lrg s are made at two redshifts , namely , @xmath23 and @xmath24 ( average @xmath25 ) and they supposed that uncertainty on ages of individuals galaxies lies in the range @xmath26 gyr .
they estimated that the uncertainty on the mean ages will be 0.10 , 0.05 and @xmath27 gyr and that hubble parameter can be measured with a precision of @xmath22 , 5% and @xmath28 .
a study was also made on the constraints upon the exposure time per galaxy and number of galaxies observed .
they concluded that with a total time of 17 , 72 and 184 hours , observing 80 , 327 and 840 galaxies on the southern african large telescope , will make possible to recover @xmath0 to 10% , 5% and @xmath28 accuracy , respectively @xcite ( private communication ) .
[ fig3 ] [ fig2 ]
in our analyses , we perform monte carlo simulations which provide us with samples of @xmath0 based on the flat @xmath7cdm model . for each `` measurement '' , we assume a gaussian distribution of @xmath0 centered at the value predicted by a flat @xmath1cdm with @xmath29 with a standard deviation corresponding to the percentage accuracy predicted in future experiments
. we shall examine two cases based on the projections made in refs .
@xcite and @xcite . for the former case , we simulated 1000 data points for @xmath0 , uniformly distributed between @xmath30 and @xmath31 with @xmath21 precision . in the latter case , we simulated 15 measurements of the hubble parameter between redshifts 0 and 1.5 and equally spaced ( @xmath32 ) .
we take the errors estimated in ref .
@xcite when recovering @xmath0 from lrg s ages , namely 10% , 5% and @xmath28 .
note that in the last case ( the one explored throughout this paper ) , one needs 8400 lrg s in the whole interval @xmath33 . in all our simulations ,
we have adopted @xmath34 km / s / mpc , which corresponds to the central value given in ref .
@xcite based on differential measurements of cepheids variable observations and @xmath35 , as given by current cmb measurements @xcite .
figure 1a shows one realization of 1000 simulated values of @xmath0 with @xmath21 accuracy according to ref .
@xcite for @xmath15 in the range ( @xmath36 ) .
the binned data points of this realization ( with a bin width @xmath32 ) are shown in figure 1b .
the error bars correspond to the standard deviation around the mean of individual values of @xmath0 within each bin , which contains approximately 67 galaxies . for comparison ,
the observed values given in @xcite are also shown . in figure 1c
we show the results of one mc realization taking the errors estimated in ref .
@xcite with the resulting @xmath0 data points at @xmath28 accuracy . from the above @xmath0 simulated data
, one may derive evolution of the deceleration parameter @xmath2 , defined as @xmath37(1+z)-1.\ ] ] in order to do that one needs to compute the derivative of the @xmath0 with respect to the redshift .
however , while integration tends to smooth out data fluctuations , numerical differentiation tends to magnify errors and the scatter of the points .
thus , depending on the degree of scatter of @xmath0 measurements , the gradient estimation may be useless . for equally spaced points ,
the derivative is calculated using finite difference approximation , that is , @xmath38/2\delta z$ ] , where @xmath39 .
we then substitute this into ( [ q1 ] ) and calculate @xmath40 and the associated uncertainty by using the standard error propagation method . for our first simulation , shown in figures 1a and 1b , we use the binned points since the large scatter of the raw data makes them useless .
the resulting values of the deceleration parameter from the mc realization of figure 1b are shown in figure 2a .
although the points lie near the predicted curve of the @xmath7cdm model ( black curve ) , we can not expect a definitive evidence for a recent cosmic acceleration from these simulated @xmath0 sample with @xmath21 accuracy .
note that the same is also true even when the evolution of @xmath2 is derived from the 15 data points with @xmath28 accuracy following @xcite ( fig .
in this case , although the error of individual points is smaller , the determination of the deceleration parameter is less precise with considerable scattering of points and error bars allowing both a decelerating and accelerating universe today . in order to quantify the above results
, we also performed 1000 mc realizations to calculate the mean of the absolute values of the difference between the calculated and the background model values of @xmath41 , that is , @xmath42 we plot this quantity for both cases discussed above in figures 2c ( 1000 binned points ) and 2d ( 15 points ) , with the error bars representing the standard deviation from the mean . as we see , @xmath43 has a weak dependence on redshift for the first case and in the interval @xmath44 it lies in the range @xmath45 , although the uncertainty is relatively large . from figure 2d
we see that the departure of the calculated @xmath46 from the expected value given by the model is larger than in former case , with @xmath43 varying from 0.2 to 0.4 .
this clearly reflects the large scattering of points shown in figure 2b . in view of the above results
, we also speculated about how precise should future observations be in order to provide a clear evidence for cosmic acceleration from cosmography .
we did that for the method proposed in @xcite and found that it will give very good results if the accuracy is improved to as much as @xmath47 .
this is illustrated in figure 3 . in this case
, the calculated value of @xmath46 is very close to the @xmath7cdm model predictions in almost the entire range of @xmath15 and the deviation from the expected value is as low as 0.07 and always @xmath48 .
naturally , the observational requirements for this case would be much more stringent than the previous ones , with the need of thousand galaxies and more observational time as well .
by considering a dark - energy dominated universe , a step further in the above discussion concerns the reconstruction of its equation - of - state parameter @xmath1 from measurements of @xmath0 . in order to be consistent with the simulations presented earlier , we consider accelerating @xmath1cdm models with @xmath1 lying in the interval [ -1/3 , -3/2 ] . in this context ,
( [ q1 ] ) can be rewritten as @xmath49 where @xmath50 is the current fractional contribution of dark energy to the critical density .
note that such procedure does not involve several integrations of @xmath12 , as in cosmological tests based on age or distance measurements and , therefore , may reduce the smearing effect on @xmath12 determinations discussed earlier ( see also @xcite for a different approach ) .
figure 4 shows the evolution of @xmath1 with redshift calculated from the simulated measurements of @xmath0 using the method proposed of ref .
@xcite with @xmath47 accuracy .
although the input model ( @xmath29 ) is fairly recovered , we note that the estimated errors of the dark energy eos ( @xmath51 ) increase considerably at the upper end of the redshift range .
this can be more easily understood after solving eq .
( [ wz ] ) for @xmath1 and calculating @xmath51 , which depends strongly on a competition between the different terms of eq .
( [ wz ] ) at the @xmath15 interval considered . for the sake of completeness
, we also performed the same analysis for the other two cases ( with 15% and 3% accuracy in @xmath0 ) discussed earlier . in both
, the scattering of the points is large enough to hamper any definitive conclusion on the expected evolution of the cosmic equation of state .
[ fig3 ] cosmography explores the possibility of extracting the maximum amount of information from cosmological measurements , as well as the assumption that the universe can be modeled by the friedmann - robertson - walker line element without assuming any dynamical theory to describe it . in this paper , differently from many analyses that study the current phase of cosmic evolution in the context of a given cosmological model , we have investigated the prospects for a cosmographic mapping of cosmic acceleration using two different projections of age determinations of passively evolving galaxies recently discussed in the literature @xcite .
we have found that a model - independent check of the reality of cosmic acceleration requires very accurated measurements of the expasion rate ( @xmath52 ) , which seems to be far from the perspectives of current planned surveys . as a step further in this analysis
, we have also used the direct relation between @xmath2 and @xmath12 to study possible constraints on the dark energy eos from future observations ( fig .
4 ) . finally
, it is also worth emphasizing that the method discussed here can be applied regardless of the technique used to obtain @xmath0 measurements . in this regard
, we note that @xmath0 measurements from radial bao methods ( despite the systematic errors introduced mainly by distortion effects ) surpasses the age method in precision and can extend the @xmath0 measurements into deeper redshift ranges @xcite ( see also @xcite for expected measurements of expansion rate at @xmath53 by observing the lyman - forest absorption spectra of high-@xmath15 quasars ) .
a detailed analysis involving expected high-@xmath15 estimates of @xmath0 from current planned bao surveys is currently under investigation and will appear in a forthcoming communication .
t. j. zhang and c. ma , `` constraints on the dark side of the universe and observational hubble parameter data , '' arxiv:1010.1307 [ astro-ph.co ] ; c. ma and t. j. zhang , apj lett .
( in press ) .
arxiv:1007.3787 [ astro-ph.co ] . | we investigate the prospects for determining the accelerating history of the universe from upcoming measurements of the expansion rate @xmath0 . in our analyses , we use monte carlo simulations based on @xmath1cdm models to generate samples with different characteristics and calculate the evolution of the deceleration parameter @xmath2 .
we show that a cosmographic ( and , therefore , model - independent ) evidence for cosmic acceleration ( @xmath3 , where @xmath4 is the transition redshift ) will only be possible with an accuracy in @xmath0 data greater than the expected in current planned surveys . a brief discussion about the prospects for reconstructing the dark energy equation of state from the parameters @xmath0 and @xmath2
is also included . |
in the quest to construct large - scale quantum information processors , in particular quantum computers , decoherence is still the main obstacle to realization .
decoherence is the degradation of quantum information due to inevitable interactions with the environment .
early skepticism @xcite concerning the viability of quantum computation ( qc ) in the presence of decoherence was overcome by the discovery of quantum error correcting codes ( qeccs),@xcite that flourished into a comprehensive theory that incorporates all elements of quantum computation.@xcite a qecc relies on an encoding of quantum information into the state of several quantum bits ( qubits ) , and a closed - loop , active error diagnosis and correction procedure.@xcite in principle , it is possible to correct arbitrary errors using sufficiently large qeccs.@xcite in practice , however , this may require a very large overhead , especially in terms of qubit resources.@xcite this is troubling in light of the substantial difficulties associated with generating and controlling systems with very large numbers of qubits .
in addition , the theory of qeccs is rather abstract , in that it presumes that one can execute certain logical operations , but does not refer to the underlying hamiltonians governing specific physical systems .
hence it is of interest to explore alternatives to qeccs , that are more economical in qubit resources , and that are direcly tailored to specific quantum computer proposals .
here we review our recent progress in developing such an alternative , `` encoded recoupling and decoupling '' ( erd),@xcite and report on some new results .
erd is based on a combination of encoding quantum information into a decoherence - free subspace ( dfs ) @xcite ( for a review see ref . ) , and the application of fast and strong dynamical - decoupling ( or `` bang - bang '' , bb ) pulses.@xcite the dfs encoding provides a first layer of protection against decoherence , while the bb pulses are used to efficiently reduce the remaining decoherence .
the encoded `` recoupling '' part of erd refers to the application of control operations that enact universal quantum logic , in a manner that is fully compatible with the dfs encoding and the bb pulses , and takes into account the experimentally available control resources , such as the underlying system hamiltonian @xcite .
the utility of erd as a general method for quantum simulation , universal qc , and decoherence suppression has also been stressed and explored by viola.@xcite our work builds in part on earlier efforts to combine universal qc with dfs encoding ( without bb pulses).@xcite the structure of the paper is as follows .
we begin in section [ decoherence ] with a brief formal summary of decoherence .
we then give in section [ dfs ] a simple example of a dfs encoding , protecting against collective dephasing .
section [ logic ] shows how to perform universal qc on this encoding .
we then briefly and formally review the dynamical decoupling method , in section [ decoupling ] .
the methods are then combined in sections [ createdfs],[leakage - elim ] , where we show how bb pulses can eliminate decoherence sources beyond collective dephasing . in section [ offres ]
we present a new and somewhat surprising result : the elimination of off - resonant transitions , that will typically be induced by bb pulses , via bb pulses .
section [ all ] shows how to combine all the pieces together , by including logic gates with the bb pulses .
concluding remarks are presented in section [ conclusions ] .
the dynamics of an open quantum system coupled to a bath is formally obtained from the time - ordered evolution @xmath0 under the combined system - bath hamiltonian @xmath1 where @xmath2 is the identity operator , @xmath3 ( @xmath4 ) is the internal hamiltonian for the system ( bath ) alone , @xmath5 is an externally applied control hamiltonian , @xmath6 is the system - bath interaction hamiltonian , and the @xmath7 ( @xmath8 ) are operators acting on the system ( bath ) .
the @xmath7 play the role of error generators in qc . making the standard assumption of initially decoupled system and bath , one traces over the bath degrees of freedom in order to obtain the time - evolved reduced system density matrix : @xmath9 , \label{eq : rhot}\ ] ] where @xmath10 [ @xmath11 is the initial density matrix of the system [ bath ] .
as is well - known , @xmath12 is in general a mixed state : @xmath13<1 $ ] , indicating that the environment has decohered the system ( see , e.g. , ref . ) .
we briefly review the simplest example of a dfs .
suppose that a system of @xmath14 qubits ( two - level systems ) is coupled to a bath in a symmetric way , and undergoes a dephasing process .
this can be described by the system - bath hamiltonian @xmath15 where @xmath16 is the pauli-@xmath17 matrix acting on the @xmath18th qubit and @xmath19 is an arbitrary bath operator . under this interaction qubit @xmath18
undergoes the transformation @xmath20 which ( after @xmath21 ) puts a random phase @xmath22 between the basis states @xmath23 and @xmath24 ( eigenstates of @xmath25 with respective eigenvalues @xmath26 and @xmath27 ) .
this can also be described by the matrix @xmath28 acting on the @xmath29 basis .
as implied by ( [ eq : hcol - deph ] ) , we assume that the phase has no space ( @xmath18 ) dependence , i.e. , the dephasing process is invariant under qubit permutations .
since the errors can be expressed in terms of the single pauli spin matrix @xmath30 of the two - level system , this example is referred to as `` collective phase damping '' , or `` weak collective decoherence''.@xcite the more general situation when errors involving all three pauli matrices are present , i.e. , dissipation and dephasing , is referred to as `` strong collective decoherence '' , @xcite or just `` collective decoherence''.@xcite without encoding a qubit initially in an arbitrary pure state @xmath31 will decohere .
this can be seen by calculating its density matrix as an average over all possible values of @xmath22 , @xmath32 where @xmath33 is a probability density , and we assume the initial state of all qubits to be a product state . for a gaussian distribution , @xmath34 ,
it is simple to check that @xmath35 the decay of the off - diagonal elements in the computational basis is a signature of decoherence .
let us now consider what happens in the two - qubit hilbert space .
the four basis states undergo the transformation @xmath36 observe that the basis states @xmath37 and @xmath38 evolves under the dephasing process as @xmath39 and the overall phase thus acquired is clearly unimportant .
this means that the 2-dimensional subspace @xmath40 of the 4-dimensional hilbert space of two qubits is _ decoherence - free _ ( df ) .
the subspaces @xmath41 and @xmath42 are also ( trivially ) df , since they each acquire a global phase as well , @xmath43 and @xmath44 respectively .
since the phases acquired by the different subspaces differ , there is decoherence _ between _ the subspaces . for @xmath45 qubits
a similar calculation reveals that the subspaces @xmath46 and @xmath47 are df , as well the ( trivial ) subspaces @xmath48 and @xmath49 .
more generally , let @xmath50 in a computational basis state ( i.e. , a bitstring ) over @xmath14 qubits .
then it is easy to check that any subspace spanned by states with constant @xmath51 is df , and can be denoted @xmath52 in accordance with the notation above .
the dimensions of these subspaces are given by the binomial coefficients : @xmath53={\binom{k}{\lambda _ { k}}}$ ] and they each encode @xmath54 qubits .
the encoding for the `` collective phase damping '' model discussed here has been tested experimentally .
the first - ever experimental implementation of dfss used the @xmath55 subspace to protect against artifially induced decoherence in a linear optics setting.@xcite the same encoding was subsequently used to alleviate the problem of external fluctuating magnetic fields in an ion trap quantum computing experiment @xcite , and figures prominently in theoretical constructions of encoded , universal qc.@xcite
next we show how to perform universal qc on the encoding @xmath56 .
let @xmath57 denote the standard pauli matrices @xmath58 , acting on the @xmath59th physical qubit ( we will use both notations interchangeably ) . in ref .
it was shown that for the code @xmath60 the encoded logical operations ( involving the first two physical qubits ) are @xmath61 these operations form an @xmath62 algebra ( i.e. , we think of them as hamiltonians rather than unitary operators ) .
we use a bar to denote logical operations on the encoded qubits . in refs .
@xcite these logical operations were denoted by @xmath63 , @xmath64 , and a detailed analysis was given on how to use typical solid - state hamiltonians ( heisenberg , @xmath65 , and @xmath66 models , defined below ) to implement quantum logic operations using this dfs encoding .
e.g. , the term @xmath67 is the spin - spin interaction in the @xmath66 model , and @xmath68 represents a zeeman splitting .
a static zeeman splitting and a controllable @xmath66 interaction can be used to generate a universal set of logic gates .
similar conclusions hold when the @xmath66 interaction is replaced by a heisenberg @xcite or @xmath65 interaction,@xcite or even by a heisenberg interaction that includes an anistropic spin - orbit term.@xcite we remark that , as first shown in refs .
, the various types of exchange interactions can be made universal also without any single - qubit terms ( such as a zeeman splitting ) , by encoding into three or more qubits,@xcite a result that has been termed `` encoded universality''.@xcite the two - qubit gate can be expressed as follows : @xmath69 where @xmath70 and @xmath2 is the identity operator . in the context of trapped - ion qc,@xcite
the phase @xmath71 is the phase of the driving laser at the @xmath59th qubit , while @xmath72 is proportional to the rabi frequency , and can be set over a wide range of values.@xcite introducing the operators @xmath73 ( denoted @xmath74 , @xmath75 respectively in refs . )
we can express @xmath76 where @xmath77 @xmath78 .
it is simple to check that @xmath79 and @xmath80 annihilate the code subspace @xmath81 and have non - trivial action ( as encoded @xmath82 and @xmath83 ) on the orthogonal subspace @xmath84 .
therefore , upon restriction to the dfs we can write : @xmath85 the fact that @xmath86 depends only on the relative phase @xmath87 is crucial in the trapped - ion context : this quantity can be controlled by adjusting the angle between the driving laser and the interatomic axis , as well as by small adjustments of the trap voltages ( which , in turn , control the trap oscillation frequency , and hence the qubit spacing ) , whereas it is much harder to control the absolute phase @xmath71,@xcite and hence also @xmath88 .
this is why the code subspace @xmath89 enjoys a preferred status over the subspace @xmath90 .
the hamiltonians in eq .
( [ eq : bars ] ) generate the logic operations ( [ eq : ubar ] ) : @xmath91 the third line follows from the elementary operator identity @xmath92 which holds for any @xmath62 angular momentum set @xmath93 , i.e. , operators that satisfy the commutation relation @xmath94=2iz$ ] ( and cyclic permutations thereof ) , in particular also the encoded operators @xmath95 . it is a well - known requirement of qc that in order to enact a universal set of logic gates ( that allow any unitary transformation to be implemented ) it is suffient to be able to implement all single - qubit operations ( as in the previous subsection ) , and to entangle pairs of qubits.@xcite in this subsection we discuss the latter requirement . in ref .
the following unitary gate was introduced , which is particularly suitable for qc using trapped ions : @xmath96 this gate can be used to entangle two dfs - qubits .
it involves simultaneous control over two phase differences @xmath97 , and thus control over the motion of two pairs of qubits .
the case @xmath98 was used in ref . to demonstrate entanglement of four trapped - qubit qubits , but this choice is not unique . for general exchange hamiltonians a different method is required .
we review the encoded recoupling method introduced in ref . .
the exchange interaction quite generally has the form @xmath99 where @xmath100 and @xmath101 the isotropic ( heisenberg ) case corresponds to @xmath102 .
the @xmath66 model is the case @xmath103 , @xmath104 .
the @xmath105 model is the case @xmath106 .
a summary of qc proposals that fall into each category can be found in ref . .
the free hamiltonian is @xmath107 where @xmath108 is the single - particle spectrum . in each instance of @xmath109 ( heisenberg , @xmath66 , @xmath65 ) one
typically has control over only one type of parameter out of the set @xmath110 .
let @xmath111 ( @xmath112 ) , and @xmath113 .
let @xmath114 and @xmath115 be two angular momentum operators satisfying @xmath62 commutation relations . then if follows from eq .
( [ eq : identity ] ) that the operation of `` conjugating by @xmath114 '' , @xmath116 causes @xmath115 s sign to be flipped .
consider now an @xmath65-type hamiltonian where the @xmath117 parameters are controllable but @xmath118 and @xmath119 are fixed .
for simplicity let us consider just the case @xmath120 .
then we can rewrite @xmath121 where we have omitted a constant term .
the important point is now that @xmath122 and @xmath123 satisfy @xmath62 commutation relations .
therefore by using recoupling through `` conjugation by @xmath124 '' we can selectively turn on and off the single - encoded - qubit rotation @xmath123 and the encoded - ising interaction @xmath125 .
this example of `` encoded selective recoupling '' establishes that encoded universal computation in the @xmath65 model can be done using control over the @xmath117 parameters alone .
next , consider the @xmath66 model , with controllable @xmath126 , but fixed @xmath127 . to implement encoded single - qubit operations
, we can use the same encoded recoupling method as for the @xmath65 model . as for encoded two - qubit operations
, we now no longer have the @xmath128 terms .
since the @xmath66 model with nearest - neighbor interactions can be shown not to be universal,@xcite we turn on also next - nearest neighbor @xmath126 terms ( these can still be nearest - neighbor in a 2d hexagonal geometry ) .
first note that @xmath129 .
now assume we can control @xmath130 ; then , using conjugation by @xmath131 : @xmath132 .
since @xmath133 is constant on the code subspace it can be ignored . on the other hand
, @xmath134 again acts as @xmath135 , i.e. , as an encoded @xmath136 .
this establishes universal encoded computation in the @xmath66 model .
taken together , the results in this section show how universal qc can be implemented using qubits encoded into a dfs offering protection against collective dephasing , while using only reasonable models of physically controllable hamiltonians .
we now move on to a discussion of how to reduce additional sources of decoherence .
we briefly review the dynamical decoupling method , introduced in ref . , and further developed in refs . .
a set of `` symmetrization '' operations is chosen such that they form a discrete subgroup of the full unitary group of operations on the hilbert space of the system . denote this group @xmath137 where @xmath138 .
the cycle time is @xmath139 , where @xmath140 is the number of symmetrization operations , and @xmath141 is the time that the system evolves freely between operations under @xmath142 .
the symmetrized evolution after a single cycle is given by @xmath143 where the evolution under @xmath144 has been neglected during application of the pulses @xmath145 , which quantifies the sense in which these pulses have to be fast and strong ( for details see , e.g , refs . ) .
@xmath146 denotes the resulting effective hamiltonian , that can be computed to any order using the magnus expansion.@xcite in the bb limit one is interested in the evolution @xmath147 after @xmath148 cycles , such that @xmath149 and @xmath150 while @xmath151 is finite . as shown in ref .
one can then retain only the first order magnus term , approximating @xmath152 by @xmath153 is the projector into the centralizer @xmath154 of @xmath155 , defined as @xmath156 since the pulses @xmath157 are unitary the centralizer equals the commutant in this case : @xmath158=0,\;\forall v_{j}\in \mathcal{g}\}. \nonumber\ ] ] it is clear that @xmath159 commutes with the adjoint action @xmath160 for all @xmath18 , so that if @xmath161 is generated by @xmath162 , the evolution will proceed without the operators @xmath7 affecting the system as errors : the evolution has been _ symmetrized _ with respect to @xmath161.@xcite for a geometric interpretation of this symmetrization see ref . , where it was also pointed out that while the bb pulses should have the effect of rotating @xmath163 to the vertices of a symmetric object , they need in fact not form a group .
one of the important advantages of the dfs encoding @xmath164 is that it is immune to collective dephasing . however , other sources of decoherence inevitably remain . in this and the following section
, we algebraically classify all additional decoherence effects and show how they can be eliminated .
first , let us analyze the effect of breaking the collective dephasing symmetry , by considering a system - bath interaction of the form @xmath165 where @xmath166 are arbitrary bath operators [ compare to eq .
( [ eq : hcol - deph ] ) ] .
this describes a general dephasing interaction on two qubits .
the source of such dephasing during storage can be long wavelength , randomly fluctuating ambient magnetic fields,@xcite that randomly shift the relative phase between the qubit @xmath23 and @xmath24 states through the zeeman effect .
the interaction can be rewritten as a sum over a collective dephasing term @xmath167 and another , differential dephasing term @xmath168 , that is responsible for errors on the dfs : @xmath169 here @xmath170 and @xmath171 . if @xmath172 were zero then there would only be collective dephasing and the dfs encoding would offer perfect protection . however , in general @xmath173 , and the dfs encoding will not suffice to offer complete protection .
the crucial observation is that , since @xmath174 [ recall eq .
( [ eq : bars ] ) ] , the offending term causes _ logical _ errors on the dfs.@xcite as shown in refs .
, then the problem of @xmath175 can be solved using a series of pulses that symmetrize @xmath176 such that only the collective term remains .
to do so note that since the offending term @xmath177 , it anticommutes with @xmath178 . at the same time @xmath179 commutes with @xmath180
this allows us to flip the sign of the offending term by using a pair of @xmath181 pulses in @xmath179 , while leaving only the collective term .
evolution with the flipped sign followed by unaltered evolution leads to cancellation of the offending term .
specifically:@xcite @xmath182 or , in terms of gates : @xmath183 where @xmath184 was defined in eq .
( [ eq : ubar ] ) .
this equation means that the system - bath coupling effectively looks like collective dephasing at the end of the pulse sequence .
thus , the system is periodically ( every @xmath185 ) projected into the dfs . in order for the the procedure described in eq .
( [ eq : sym ] ) to work , the gate @xmath186 must be executed at a timescale faster than the cutoff frequency associated with the fluctuating ( magnetic ) fields causing the differential dephasing term in @xmath187 .
so far we have discussed creation of collective dephasing conditions on a single dfs qubit .
however , it is essential for the reliable execution of an entangling logic gate to have collective dephasing over all four qubits participating in the gate , even if only two are coupled at a time . a procedure for creating collective _ decoherence _ conditions over blocks of @xmath188 and @xmath189 qubits was given in ref .
here we show how to do the same for a block of 4 qubits with collective dephasing .
let us start with a general dephasing hamiltonian on @xmath148 qubits , and rewrite it in terms of nearest - neighbor sums and differences : @xmath190 where @xmath191 .
as noted above , @xmath192 , so that to eliminate all nearest - neighbor differences of the form @xmath193 we can use the collective decoupling pulse @xmath194 : @xmath195 or , in gate terms : @xmath196 e^{-ih_{sb}\tau } \left [ \bigotimes_{j=1}^{n/2}\bar{u}_{2j-1,2j } ( \frac{\pi } { 2},0)\right ] = e^{-i2\tau \sum_{j=1}^{n/2}\left ( z_{2j}+z_{2j-1}\right ) \otimes b_{2j}^{+}}.\ ] ] the next step is to eliminate next - nearest neighbor differential terms . to this end let us rewrite the outcome of the @xmath197 pulse in terms of sums and differences over blocks of four ions : @xmath198 \otimes b_{2j}^{+,+ } \nonumber \\ & + & \sum_{j=1}^{n/2}\left [ ( z_{2j+2}-z_{2j})+(z_{2j+1}-z_{2j-1})\right ] \otimes b_{2j}^{+,-},\end{aligned}\ ] ] where @xmath199 .
the term in the first line contains only the desired block - collective dephasing over @xmath200 ions .
the term in the second line contains undesired differential dephasing terms that we wish to eliminate . but these terms once again have the appearance of encoded @xmath154 operators , between next - nearest neighbor ion pairs .
therefore we need to apply a second collective pulse @xmath201 , that applies encoded @xmath82 operators on these qubit pairs . at this point
we are left just with collective dephasing terms on blocks of @xmath200 qubits , as required : @xmath202 & \times & \nonumber \label{eq : create4 } \\
e^{-i2\tau \sum_{j=1}^{n/2}\left ( z_{2j}+z_{2j-1}\right ) \otimes b_{2j}^{+ } } \left [ \bigotimes_{j=1}^{n/2}\bar{u}_{2j-1,2j+1}(\frac{\pi } { 2},0)\bar{u } _ { 2j,2j+2}(\frac{\pi } { 2},0)\right ] & = & e^{-i4\tau \sum_{j=1}^{n/2}\left ( z_{2j+2}+z_{2j+1}+z_{2j}+z_{2j-1}\right ) \otimes b_{2j}^{+,+}}. \nonumber \\ & & \end{aligned}\ ] ] this pulse sequence is important to ensure that collective dephasing conditions will prevail during the execution of logic gates between dfsqubits .
the reduction of differential dephasing errors , as in the previous subsection , is particularly relevant for storage errors .
however , this is only the first step .
additional sources of decoherence may take place during storage , and in particular during the execution of logic gates .
it is useful to provide a complete algebraic classification of the possible decoherence processes .
this will allow us to see what can be done using decoupling pulses .
to this end let us now write the system - bath hamiltonian on two physical qubits in the general form @xmath203 where @xmath204 where @xmath2 is the identity operator , @xmath205 ( etc . ) , and where @xmath206 means a linear combination of these operators tensored with bath operators .
the @xmath207 operators in eq .
( [ eq : class ] ) form a complete basis for all @xmath208-qubit operators .
this classification , first introduced in ref . , has the following significance .
the operators in @xmath209 either vanish on the dfs , or are proportional to identity on it . in either case
their effect is to generate an overall phase on the dfs , so they can be safely ignored from now on .
the operators in @xmath210 are the _ leakage errors _ : terms that cause transitions between states inside and outside of the dfs .
a universal and efficient decoupling method for eliminating such errors , for arbitrary numbers of ( encoded)qubits was given in ref . .
finally , the operators in @xmath211 have the form of logic gates on the dfs . however , these are undesired logic operations , since they are coupled to the bath , and thus cause decoherence . in the previous subsection we showed how to eliminate the logical error @xmath212 , but we see now that this was only one error in a much larger set . to deal with the additional errors it is useful at this point to introduce a more compact notation for the pulse sequences .
we denote by @xmath213 $ ] a period of evolution under the free hamiltonian , i.e. , @xmath214 $ ] , and further denote @xmath215 thus eq .
( [ eq : sym ] ) can be written as : @xmath216=[\tau , p,\tau , p^{\dagger } ] .\ ] ] as a first step in dealing with the additional errors , note that the symmetrization procedure @xmath217 $ ] can in fact achieve more than just the elimination of the differential dephasing @xmath168 term . since @xmath218 also anticommutes with @xmath219 ,
if such a term appears in the system - bath interaction it too will be eliminated using the same procedure .
so far we have used a @xmath220 pulse .
interestingly , the hamiltonian @xmath218 can also be used to eliminate all leakage errors.@xcite to see this , note that @xmath221 .
this operator anticommutes with _ all _ terms in @xmath222 .
hence it too can be used in a parity - kick pulse sequence , that will eliminate all the leakage errors . at this point
we are left with just a single error : @xmath223 itself , in @xmath224 .
clearly , we can not use a pulse generated by @xmath218 to eliminate this error . instead , to deal with this error
we need to introduce one more pulse pair that anticommutes with @xmath218 , e.g. , @xmath225 .
let us now see how to combine all the decoherence elimination pulses into one efficient sequence .
first we introduce the abbreviations @xmath226 as argued above , the @xmath227 pulse @xmath228 eliminates @xmath222 : @xmath229=[\tau , \pi , \tau , \pi ] .\ ] ] now let us discuss adding the extra pulses needed to achieve full decoherence elimination .
the @xmath230 pulse @xmath231 eliminates @xmath232 and @xmath212 in @xmath224 . combining this with the sequence for leakage elimination we have the sequence of @xmath200 pulses : @xmath233p^{\dagger } [ u(\tau ) \pi u(\tau ) \pi ] p \nonumber \\ & = & [ \tau , \pi , \tau , p,\tau , \pi , \tau , p^{\dagger } ] , \label{eq:4pulses}\end{aligned}\ ] ] ( where we have used @xmath234 , @xmath235 ) .
if we wish to entirely eliminate decoherence then we are left just with getting rid of the logical error due to @xmath236 . to eliminate it we now combine with the @xmath232-direction , @xmath230 pulse , @xmath237 : @xmath238q^{\dagger } [ u(\tau ) \pi u(\tau ) pu(\tau ) \pi u(\tau ) p^{\dagger } ] q \nonumber \\ & = & [ \tau , \pi , \tau , p,\tau , \pi , \tau , p^{\dagger } , q^{\dagger } , \tau , \pi , \tau , p,\tau , \pi , \tau , p^{\dagger } , q]\end{aligned}\ ] ] [ eq:10pul ] which takes ten pulses .
unfortunately it is not possible to compress this further , since @xmath239 and @xmath240 , neither of which can not be generated directly ( in one step ) from the available gate @xmath241 .
one important caveat in our discussion so far is that , because we need very strong and fast pulses , our gate operation may become imperfect .
specifically , off - resonant coupling may become important .
this can cause _ unitary _ leakage errors from the dfs .
these can in turn be reduced using the methods in refs . .
here we present an alternative and new method , that uses bb pulses generated in terms of the system hamiltonian .
consider an @xmath148-level system hamiltonian @xmath242 where the levels @xmath243 denote our qubit . if all parameters are fixed , the evolution @xmath244 is always on .
if we turn on an interaction @xmath245 then the total hamiltonian is @xmath246 which can be written as @xmath247 where the leakage term is @xmath248 now note that @xmath249 using this we now show how to eliminate each of the leakage terms one by one .
first , we eliminate @xmath250 by the bb sequence @xmath251 the new hamiltonian @xmath252 does not contain @xmath250 next , we eliminate @xmath253 by @xmath254where @xmath255 .
we can clearly repeat this procedure so as to eliminate all leakage . _
the crucial point is that since we have used the system hamiltonian to generate bb pulses , off - resonant transitions do not take place _ : @xmath256 has no matrix elements between different levels .
a concern is how we to obtain @xmath257 .
this problem is shared by other methods dealing with the same problem.@xcite in principle it can be solved provided the level spacings are rationally related . note in this context that usually we do not need to eliminate all @xmath258 , since @xmath259 and typically decrease exponentially according to time - independent perturbation theory .
we have assumed that there is no degeneracy .
if there is , then all degenerate transitions will be eliminated simultaneously , so the procedure is simplified .
so far we have discussed computation using the encoded recoupling method ( section [ logic ] ) , and encoded decoupling ( sections [ createdfs],[leakage - elim ] ) .
we now put the two together in order to obtain the full erd picture .
at least two methods are available for combining quantum computing operations with the sequences of decoupling pulses we have presented above . for a general analysis of this issue see ref . . the decoupling pulse sequences given in sec .
[ leakage - elim ] `` stroboscopically '' create collective dephasing conditions at the conclusion of each cycle . as noted above
, this is equivalent to a periodic projection into the dfs .
this property allows for `` stroboscopic '' quantum computation at the corresponding projection times.@xcite here the computation pulses need to be synchronized with the decoupling pulses , and inserted at the end of each cycle . because of the conditions on the validity of the bb method,@xcite the amount of time available for implementation of a logic gate is no more than the bath correlation time @xmath260 .
( an exception to this rule is the case of @xmath261 noise.@xcite ) assuming the dominant decoherence contributions not accounted for by the dfs encoding to come from differential dephasing ( setting the @xmath262 time - scale ) , and given that we already assumed that we can use pulses with interval @xmath263 @xmath262 , it is consistent to assume that we can then also perform logic gates on the same time scale
. there may be an advantage to using fast but weak pulses for the logic gates , while preserving the fast + strong property of the decoupling pulses .
to see how to combine logic gates with decoupling in this case , let us denote by @xmath264 the controllable system hamiltonian that generates the entangling gate @xmath265 [ recall eq .
( [ eq : uij ] ) ] .
suppose first that we turn on this logic - gate generating hamiltonian in a manner that is neither very strong nor very fast , so that the system - bath interaction is not negligible while @xmath3 is on ( this obviously puts less severe demands on experimental implementation ) .
then the corresponding unitary operator describing the dynamics of system plus bath is : @xmath266.\ ] ] now , _ if we choose _
@xmath3 _ _ so that it commutes with the decoupling pulses _ _
, then we can show that after decoupling @xmath267 , \label{eq : ucomp}\ ] ] provided @xmath268 is sufficiently small .
tracing out the bath then leaves a purely unitary , decoherence - free evolution on the system . to prove this , assume we have chosen @xmath269 and the decoupling hamiltonian @xmath270 so that ( i ) @xmath271 , and ( ii ) @xmath272=0 $ ]
. then @xmath273 } \\ & = & e^{-it(h_{s}+h_{sb}+h_{b})}e^{-it(h_{s}-h_{sb}+h_{b } ) } \\ & = & e^{-\{2it(h_{s}+h_{b})+t^{2}([h_{sb},h_{s}]+[h_{sb},h_{b}])+o(t^{3})\}},\end{aligned}\ ] ] where we have used the baker - campbell - hausdorff formula , @xmath274+o(\alpha ^{3})\}$ ] .
let us now show how to efficiently combine logic operations and decoupling pulses . for simplicity consider only the case where we can neglect the @xmath275 error , i.e. , our decoupling sequence is the 4-pulse one given in eq .
( [ eq:4pulses ] ) .
suppose we wish to implement a logical @xmath82 operation , i.e. , @xmath276 .
recall that this involves turning on the hamiltonian @xmath277 between two physical qubits . because the decoupling pulses @xmath278 and @xmath279 are generated in terms of the same hamiltonian
, they commute with @xmath280 while eliminating @xmath6 ( except for the terms in @xmath6 that have trivial action on the dfs ) .
thus the conditions under which eq .
( [ eq : ucomp ] ) were shown to hold are satisfied .
this allows us to insert the logic gates into the four free evolution periods involved in the pulse sequence of eq .
( [ eq:4pulses ] ) .
thus , the full pulse sequence that combines creation of collective dephasing conditions with execution of the logic gate is : @xmath281 with @xmath282 $ ] , and which , using the dfs encoding , is equivalent to the desired @xmath283 .
this involves 8 control pulses , 4 of which are of the fast+strong type ( those involving @xmath231 and @xmath228 ) , and 4 of which must be fast , but need not be so strong that we can neglect @xmath6 .
if we wish to implement logical @xmath83 operation , i.e. , @xmath284 , then we can not now use @xmath231 and @xmath228 , since they anticommute with @xmath285 and will eliminate it .
instead we should use decoupling pulses generated in terms of @xmath285 , which will also have the desired effect of eliminating @xmath222 , as well as @xmath286 and @xmath287 logical errors , while commuting with the @xmath288 logic operations ( and for this reason can of course not eliminate @xmath232 errors ) .
these are just the @xmath237 and @xmath289 pulses defined in eq .
( [ eq : pqetc ] ) . in gate terms
this implies turning on the hamiltonian @xmath290 between two physical qubits .
thus : @xmath291 with @xmath292 $ ] , and which , using the dfs encoding , is equivalent to the desired @xmath293 .
finally , to generate single dfs - qubit rotations about an arbitrary axis we can combine eqs .
( [ eq : x]),([eq : y ] ) according to the euler angles construction.@xcite given that eqs .
( [ eq : x]),([eq : y ] ) each take 8 pulses , the euler angle method will generate an arbitrary dfs - qubit rotation in at most 24 pulses .
concerning gates that entangle two dfs - qubits , the situation may be more involved , since now the next - nearest neighbor pulses in eq .
( [ eq : create4 ] ) , that create the collective dephasing conditions on four qubits , do not all commute with the @xmath294 gate of eq .
( [ eq : u4 ] ) . therefore here we must resort to the strong + fast method of the previous subsection , i.e. , we need to synchronize the @xmath294 pulses with the end of the decoupling pulse sequence .
we do not here analyze the situation with respect to the exchange hamiltonian implementation of conditional logic gates .
taken together , the methods described in this section provide an explicit way to implement universal qc in a manner that offers protection against all sources of qubit decoherence , using a fast + strong ( or fast + weak ) versions of logic gates .
we have proposed a method of encoded recoupling and decoupling ( erd ) for performing decoherence - protected quantum computation .
our method combines an encoding into qubit - pair decoherence - free subspaces ( each pair yielding one encoded qubit ) , with sequences of recoupling and decoupling pulses .
the qubit encoding protects against collective dephasing processes , while the decoupling pulses symmetrize all other sources of decoherence into a collective dephasing interaction .
the recoupling pulses are used to implement encoded quantum logic gates , either during or in between the decoupling pulses .
the dynamical decoupling method requires an exponential number of pulses if the most general form of decoherence is to be suppressed , that can couple arbitrary numbers of qubits to the environment ( total decoherence @xcite ) .
this exponential scaling is avoided here by focusing on decoherence elimination inside blocks of _ finite _ size ( e.g. , at most four qubits ) where arbitrary decoherence is allowed .
however , we have implicitly assumed that there are no decoherence processes coupling different blocks .
this is a reasonable assumption for most quantum computer implementation , provided the different blocks can be kept sufficiently far apart until they need to be brought together in order to execute inter - block logic gates .
when this happens , erd can still be efficiently applied on the temporarily larger block .
it may be questioned whether there is any advantage in using erd compared to methods of active quantum error correcting codes ( qecc ) .
both erd and qecc are capable of dealing with arbitrary decoherence processes , and are fully compatible with universal quantum computation .
there are two main advantages to erd : first , we need only two physical qubits per logical qubit , compared to a redundancy of five to one , in order to handle all single - qubit errors in qecc.@xcite so far the most advanced experiments outside nmr , i.e. , those involving trapped ions , have used up to four qubits,@xcite so that this encoding economy is a distinct advantage for near - term experiments .
second , our method is directly compatible with hamiltonians describing a variety of quantum computer proposals . on the other hand
it is not clear how to directly use qecc given hamiltonians describing specific systems .
these are general features of erd : economy of encoding redundancy and use of only the most easily controllable interactions .
the disadvantage of erd compared to qecc is that there does not exist , at this point , a result analogous to the threshold theorem of fault tolerant quantum error correction .
this means that we can not yet guarantee full scalability of erd as a stand - alone method , because we do not yet know how to compensate for imperfect pulses .
however , in principle it is always possible to concatenate erd with qecc , as done , e.g. , for dfs with qecc in refs .
, and then the standard fault tolerance results apply .
we expect that the theory of composite pulses @xcite will also play a key role in this further development of erd .
we hope that the methods proposed here will inspire experimentalists to implement encoded recoupling and decoupling in the lab , thus demonstrating the possibility of fully decoherence - protected quantum computation .
this material is based on research sponsored by the defense advanced research projects agency under the quist program and managed by the air force research laboratory ( afosr ) , under agreement f49620 - 01 - 1 - 0468
further gratefully acknowledges financial support from pro , nserc , and the connaught fund .
, `` encoded universality in physical implementations of a quantum computer , '' in _ proceedings of the 1st international conference on experimental implementations of quantum computation , sydney , australia _ , r. clark , ed . , p. 257
, rinton , ( princeton , nj ) , 2001 . , `` experimental issues in coherent quantum - state manipulation of trapped atomic ions , '' _ j. of res . of the national inst . of standards and technology _ * 103 * , p. 259 | decoherence is the main obstacle to the realization of quantum computers . until recently it was thought that quantum error correcting codes are the only complete solution to the decoherence problem . here
we present an alternative that is based on a combination of a decoherence - free subspace encoding and the application of strong and fast pulses : `` encoded recoupling and decoupling '' ( erd ) .
this alternative has the advantage of lower encoding overhead ( as few as two physical qubits per logical qubit suffice ) , and direct application to a number of promising proposals for the experimental realization of quantum computers . |
this paper reports on the still preliminary , but already satisfying results of the learning computational grammars ( lcg ) project , a postdoc network devoted to studying the application of machine learning techniques to grammars suitable for computational use .
the member institutes are listed with the authors and also included issco at the university of geneva .
we were impressed by early experiments applying learning to natural language , but dissatisfied with the concentration on a few techniques from the very rich area of machine learning .
we were interested in a more systematic survey to understand the relevance of many factors to the success of learning , esp .
the availability of annotated data , the kind of dependencies in the data , and the availability of knowledge bases ( grammars ) .
we focused on syntax , esp .
noun phrase ( np ) syntax from the beginning .
the industrial partner , xerox , focused on more immediate applications @xcite .
the network was focused not only by its scientific goal , the application and evaluation of machine - learning techniques as used to learn natural language syntax , and by the subarea of syntax chosen , np syntax , but also by the use of shared training and test material , in this case material drawn from the penn treebank . finally , we were curious about the possibility of combining different techniques , including those from statistical and symbolic machine learning .
the network members played an important role in the organisation of three open workshops in which several external groups participated , sharing data and test materials .
this section starts with a description of the three tasks that we have worked on in the framework of this project .
after this we will describe the machine learning algorithms applied to this data and conclude with some notes about combining different system results .
in the framework of this project , we have worked on the following three tasks : 1 . base phrase ( chunk ) identification 2 .
base noun phrase recognition 3 .
finding arbitrary noun phrases text chunks are non - overlapping phrases which contain syntactically related words .
for example , the sentence : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ @xmath0 he @xmath1 $ ] @xmath2 reckons @xmath1 $ ] @xmath0 the current account deficit @xmath1 $ ] @xmath2 will narrow @xmath1 $ ] + @xmath3 to @xmath1 $ ] @xmath0 only @xmath4 1.8 billion @xmath1 $ ] + @xmath3 in @xmath1 $ ] @xmath0 september @xmath1 $ ] . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ contains eight chunks , four np chunks , two vp chunks and two pp chunks .
the latter only contain prepositions rather than prepositions plus the noun phrase material because that has already been included in np chunks .
the process of finding these phrases is called chunking .
the project provided a data set for this task at the conll-2000 workshop @xcite .
it consists of sections 15 - 18 of the wall street journal part of the penn treebank ii @xcite as training data ( 211727 tokens ) and section 20 as test data ( 47377 tokens ) .
a specialised version of the chunking task is np chunking or basenp identification in which the goal is to identify the base noun phrases .
the first work on this topic was done back in the eighties @xcite .
the data set that has become standard for evaluation machine learning approaches is the one first used by ramshaw and marcus .
it consists of the same training and test data segments of the penn treebank as the chunking task ( respectively sections 15 - 18 and section 20 ) .
however , since the data sets have been generated with different software , the np boundaries in the np chunking data sets are slightly different from the np boundaries in the general chunking data .
noun phrases are not restricted to the base levels of parse trees .
for example , in the sentence _ in early trading in hong kong monday , gold was quoted at $ 366.50 an ounce .
_ , the noun phrase @xmath0 $ 366.50 an ounce @xmath1 $ ] contains two embedded noun phrases @xmath0 $ 366.50 @xmath1 $ ] and @xmath0 an ounce @xmath1 $ ] . in the np bracketing task ,
the goal is to find all noun phrases in a sentence .
data sets for this task were defined for conll-99 .
the data consist of the same segments of the penn treebank as the previous two tasks ( sections 15 - 18 ) as training material and section 20 as test material .
this material was extracted directly from the treebank and therefore the np boundaries at base levels are different from those in the previous two tasks . in the evaluation of all three tasks ,
the accuracy of the learners is measured with three rates .
we compare the constituents postulated by the learners with those marked as correct by experts ( gold standard ) .
first , the percentage of detected constituents that are correct ( precision ) .
second , the percentage of correct constituents that are detected ( recall ) . and third , a combination of precision and recall , the f@xmath5 rate which is equal to ( 2*precision*recall)/(precision+recall ) .
this section introduces the ten learning methods that have been applied by the project members to the three tasks : lscgs , allis , lsommbl , maximum entropy , aleph , mdl - based dcg learners , finite state transducers , ib1ig , igtree and c5.0 . *
local structural context grammars * ( lscgs ) @xcite are situated between conventional probabilistic context - free production rule grammars and dop - grammars ( e.g. , bod and scha ) .
lscgs outperform the former because they do not share their inherent independence assumptions , and are more computationally efficient than the latter , because they incorporate only subsets of the context included in dop - grammars .
local structural context ( lsc ) is ( partial ) information about the immediate neighbourhood of a phrase in a parse . by conditioning bracketing probabilities on lsc , more
fine - grained probability distributions can be achieved , and parsing performance increased . given corpora of parsed text such as the wsj , lscgs
are used in automatic grammar construction as follows .
an lscg is derived from the corpus by extracting production rules from bracketings and annotating the rules with the type(s ) of lsc to be incorporated in the lscg ( e.g. parent category information , depth of embedding , etc . ) .
rule probabilities are derived from rule frequencies ( currently by maximum likelihood estimation ) . in a separate optimisation step
, the resulting lscgs are optimised in terms of size and parsing performance for a given parsing task by an automatic method ( currently a version of beam search ) that searches the space of partitions of a grammar s set of nonterminals .
the lscg research efforts differ from other approaches reported in this paper in two respects .
firstly , no lexical information is used at any point , as the aim is to investigate the upper limit of parsing performance without lexicalisation .
secondly , grammars are optimised for parsing performance _ and _ size , the aim being to improve performance but not at the price of arbitrary increases in grammar complexity ( hence the cost of parsing ) .
the automatic optimisation of corpus - derived lscgs is the subject of ongoing research and the results reported here for this method are therefore preliminary . * theory refinement * ( allis ) .
allis ( @xcite , @xcite ) is a inductive rule - based system using a traditional general - to - specific approach @xcite . after generating a default classification rule ( equivalent to the n - gram model )
, allis tries to refine it since the accuracy of these rules is usually not high enough .
refinement is done by adding more premises ( contextual elements ) .
allis uses data encoded in xml , and also learns rules in xml .
from the perspective of the xml formalism , the initial rule can be viewed as a tree with only one leaf , and refinement is done by adding adjacent leaves until the accuracy of the rule is high enough ( a tuning threshold is used ) .
these additional leaves correspond to more precise contextual elements . using the hierarchical structure of an xml document ,
refinement begins with the highest available hierarchical level and goes down in the hierarchy ( for example , starting at the chunk level and then word level ) .
adding new low level elements makes the rules more specific , increasing their accuracy but decreasing their coverage .
after the learning is completed , the set of rules is transformed into a proper formalism used by a given parser . * labelled som and memory based learning * ( lsommbl ) is a neurally inspired technique which incorporates a modified self - organising map ( som , also known as a ` kohonen map ' ) in memory - based learning to select a subset of the training data for comparison with novel items .
the som is trained with labelled inputs . during training ,
each unit in the map acquires a label .
when an input is presented , the node in the map with the highest activation ( the ` winner ' ) is identified . if the winner is unlabelled , then it acquires the label from its input .
labelled units only respond to similarly labelled inputs .
otherwise training proceeds as with the normal som .
when training ends , all inputs are presented to the som , and the winning units for the inputs are noted .
any unused units are then discarded .
thus each remaining unit in the som is associated with the set of training inputs that are closest to it .
this is used in mbl as follows .
the labelled som is trained with inputs labelled with the output categories .
when a novel item is presented , the winning unit for each category is found , the training items associated with the winning units are searched for the closest item to the novel item and the most frequent classification of that item is used as the classification for the novel item . *
maximum entropy * when building a classifier , one must gather evidence for predicting the correct class of an item from its context .
the maximum entropy ( maxent ) framework is especially suited for integrating evidence from various information sources .
frequencies of evidence / class combinations ( called features ) are extracted from a sample corpus and considered to be properties of the classification process .
attention is constrained to models with these properties .
the maxent principle now demands that among all the probability distributions that obey these constraints , the most uniform is chosen . during training
, features are assigned weights in such a way that , given the maxent principle , the training data is matched as well as possible . during evaluation
it is tested which features are _ active _
( i.e. , a feature is active when the context meets the requirements given by the feature ) .
for every class the weights of the active features are combined and the best scoring class is chosen @xcite . for the classifier built here we use as evidence the surrounding words , their pos tags and basenp tags predicted for the previous words . a mixture of simple features ( consisting of one of the mentioned information sources ) and complex features ( combinations thereof ) were used .
the left context never exceeded 3 words , the right context was maximally 2 words .
the model was calculated using existing software @xcite . * inductive logic programming ( ilp ) * aleph is an ilp machine learning system that searches for a hypothesis , given positive ( and , if available , negative ) data in the form of ground prolog terms and background knowledge ( prior knowledge made available to the learning algorithm ) in the form of prolog predicates .
the system , then , constructs a set of hypothesis clauses that fit the data and background as well as possible . in order to approach the problem of
np chunking in this context of single - predicate learning , it was reformulated as a tagging task where each word was tagged as being ` inside ' or ` outside ' a basenp ( consecutive nps were treated appropriately ) .
then , the target theory is a prolog program that correctly predicts a word s tag given its context . the context consisted of pos tagged words and syntactically tagged words to the left and pos tagged words to the right , so that the resulting tagger can be applied in the left - to - right pass over pos - tagged text . * minimum description length * ( mdl )
estimation using the minimum description length principle involves finding a model which not only ` explains ' the training material well , but also is compact .
the basic idea is to balance the generality of a model ( roughly speaking , the more compact the model , the more general it is ) with its specialisation to the training material .
we have applied mdl to the task of learning broad - covering definite - clause grammars from either raw text , or else from parsed corpora @xcite .
preliminary results have shown that learning using just raw text is worse than learning with parsed corpora , and that learning using both parsed corpora and a compression - based prior is better than when learning using parsed corpora and a uniform prior .
furthermore , we have noted that our instantiation of mdl does not capture dependencies which exist either in the grammar or else in preferred parses .
ongoing work has focused on applying random field technology ( maximum entropy ) to mdl - based grammar learning ( see osborne for some of the issues involved ) . *
finite state transducers * are built by interpreting probabilistic automata as transducers .
we use a probabilistic grammatical algorithm , the ddsm algorithm @xcite , for learning automata that provide the probability of an item given the previous ones .
the items are described by bigrams of the format feature : class . in the resulting automata
we consider a transition labeled feature : class as the transducer transition that takes as input the first part ( feature ) of the bigram and outputs the second part ( class ) . by applying the viterbi algorithm on such a model
, we can find out the most probable set of class values given an input set of feature values . as the ddsm algorithm has a tuning parameter
, it can provide many different automata .
we apply a majority vote over the propositions made by the so computed automata / transducers for obtaining the results mentioned in this paper . * memory - based learning * methods store all training data and classify test data items by giving them the classification of the training data items which are most similar .
we have used three different algorithms : the nearest neighbour algorithm ib1ig , which is part of the timbl software package @xcite , the decision tree learner igtree , also from timbl , and c5.0 , a commercial version of the decision tree learner c4.5 @xcite .
they are classifiers which means that they assign phrase classes such as i ( inside a phrase ) , b ( at the beginning of a phrase ) and o ( outside a phrase ) to words . in order to improve the classification process we provide the systems with extra information about the words such as the previous _
n _ words , the next _
n _ words , their part - of - speech tags and chunk tags estimated by an earlier classification process .
we use the default settings of the software except for the number of examined nearest neighbourhood regions for ib1ig ( k , default is 1 ) which we set to 3 .
when different systems are applied to the same problem , a clever combination of their results will outperform all of the individual results @xcite .
the reason for this is that the systems often make different errors and some of these errors can be eliminated by examining the classifications of the others .
the most simple combination method is majority voting .
it examines the classifications of the test data item and for each item chooses the most frequently predicted classification . despite its simplicity ,
majority voting has found to be quite useful for boosting performance on the tasks that we are interested in .
we have applied majority voting and nine other combination methods to the output of the learning systems that were applied to the three tasks .
nine combination methods were originally suggested by van halteren et al . .
five of them , including majority voting , are so - called voting methods .
apart from majority voting , all assign weights to the predictions of the different systems based on their performance on non - used training data , the tuning data .
totprecision uses classifier weights based on their accuracy .
tagprecision applies classification weights based on the accuracy of the classifier for that classification .
precision - recall uses classification weights that combine the precision of the classification with the recall of the competitors . and
finally , tagpair uses classification pair weights based on the probability of a classification for some predicted classification pair @xcite .
the remaining four combination methods are so - called stacked classifiers .
the idea is to make a classifier process the output of the individual systems .
we used the two memory - based learners ib1ig and igtree as stacked classifiers . like van halteren et al . , we evaluated two features combinations .
the first consisted of the predictions of the individual systems and the second of the predictions plus one feature that described the data item .
we used the feature that , according to the memory - based learning metrics , was most relevant to the tasks : the part - of - speech tag of the data item .
in the course of this project we have evaluated another combination method : best - n majority voting @xcite .
this is similar to majority voting except that instead of using the predictions of all systems , it uses only predictions from some of the systems for determining the most probable classifications .
we have experienced that for different reasons some systems perform worse than others and including their results in the majority vote decreases the combined performance .
therefore it is a good idea to evaluate majority voting on subsets of all systems rather than only on the combination of all systems .
apart from standard majority voting , all combination methods require extra data for measuring their performance which is required for determining their weights , the tuning data .
this data can be extracted from the training data or the training data can be processed in an n - fold cross - validation process after which the performance on the complete training data can be measured .
although some work with individual systems in the project has been done with the goal of combining the results with other systems , tuning data is not always available for all results .
therefore it will not always be possible to apply all ten combination methods to the results . in some cases we have to restrict ourselves to evaluating majority voting only .
this sections presents the results of the different systems applied to the three tasks which were central to this this project : chunking , np chunking and np bracketing .
chunking was the shared task of conll-2000 , the workshop on computational natural language learning , held in lisbon , portugal in 2000 @xcite .
six members of the project have performed this task .
the results of the six systems ( precision , recall and f@xmath5 can be found in table [ tab - chunking ] .
belz used local structural context grammars for finding chunks .
djean applied the theory refinement system allis to the shared task data .
koeling evaluated a maximum entropy learner while using different feature combinations ( me ) .
osborne used a maximum entropy - based part - of - speech tagger for assigning chunk tags to words ( me tag ) .
thollard identified chunks with finite state transducers generated by a probabilistic grammar algorithm ( fst ) .
tjong kim sang tested different configurations of combined memory - based learners ( mbl ) .
the fst and the lscg results are lower than those of the other systems because they were obtained without using lexical information . the best result at the workshop was obtained with support vector machines @xcite . .
the chunking results for the six systems associated with the project ( shared task conll-2000 ) .
the baseline results have been obtained by selecting the most frequent chunk tag associated with each part - of - speech tag .
the best results at conll-2000 were obtained by support vector machines . a majority vote of the six lcg systems does not perform much worse than this best result .
a majority vote of the five best systems outperforms the best result slightly ( @xmath6 error reduction ) . [ cols="<,^,^,^",options="header " , ] finding arbitrary noun phrases was the shared task of conll-99 , held in bergen , norway in 1999 .
three project members have performed this task .
belz extracted noun phrases with local structural context grammars , a variant of data - oriented parsing ( lscg ) .
osborne used a definite clause grammar learner based on minimum description length for finding noun phrases in samples of penn treebank material ( mdl ) .
tjong kim sang detected noun phrases with a bottom - up cascade of combinations of memory - based classifiers ( mbl ) .
the performance of the three systems can be found in table [ tab - npbracketing ] . for this task
it was not possible to apply system combination to the output of the system .
the mdl results have been obtained on a different data set and this left us with two remaining systems .
a majority vote of the two will not improve on the best system and since there was no tuning data or development data available , other combination methods could not be applied .
the project has proven to be successful in its results for applying machine learning techniques to all three of its selected tasks : chunking , np chunking and np bracketing .
we are looking forward to applying these techniques to other nlp tasks .
three of our project members will take part in the conll-2001 shared task , ` clausing ' , hopefully with good results .
two more have started working on the challenging task of full parsing , in particular by starting with a chunker and building a bottom - up arbitrary phrase recogniser on top of that .
the preliminary results are encouraging though not as good as advanced statistical parsers like those of charniak and collins .
it is fair to characterise lcg s goals as primarily technical in the sense that we sought to maximise performance rates , esp .
the recognition of different levels of np structure .
our view in the project is certainly broader , and most project members would include learning as one of the language processes one ought to study from a computational perspective like parsing or generation .
this suggest several further avenues , e.g. , one might compare the learning progress of simulations to humans ( mastery as a function of experience )
. one might also be interested in the exact role of supervision , in the behaviour ( and availability ) of incremental learning algorithms , and also in comparing the simulation s error functions to those of human learners ( wrt to phrase length or construction frequency or similarity ) .
this would add an interesting cognitive perspective to the work , along the lines begun by brent , but we note it here only as a prospect for future work .
r. bod and r. scha .
data - oriented language processing . in s.
young and g. bloothooft , editors , _ corpus - based methods in language and speech processing _ , pages 137173 .
kluwer academic publishers , boston .
erik f. tjong kim sang , walter daelemans , herv djean , rob koeling , yuval krymolowski , vasin punyakanok , and dan roth .
2000 . applying system combination to base noun phrase identification . in _ proceedings of the 18th international conference on computational linguistics ( coling 2000)_.
saarbruecken , germany . | this paper reports on the learning computational grammars ( lcg ) project , a postdoc network devoted to studying the application of machine learning techniques to grammars suitable for computational use .
we were interested in a more systematic survey to understand the relevance of many factors to the success of learning , esp .
the availability of annotated data , the kind of dependencies in the data , and the availability of knowledge bases ( grammars ) .
we focused on syntax , esp .
noun phrase ( np ) syntax . |
despite the fact that effect to background ratio is the central problem of all experimental physics , there is a certain class of experiments for which this problem is so crucial that even the possibility of their performance itself depends strongly on the reached background level of the used detectors .
these are so called underground experiments devoted to investigation of the very rare or forbidden decays and processes like , for instance , double beta decay , proton decay , dark matter particles search , solar neutrino study and so on .
the ultimate sensitivity of such experiments is determined mainly ( except the available source strengths ) by the detector background .
the first origin of background is due to cosmic rays and can be eliminated by the proper underground site for the set up .
the second ( and the most crucial for sensitivity ) source of the background is the decays of the nuclides from the radioactive impurities in the detector itself , in the materials used for detector mounting and shielding , and in the surroundings .
therefore it is obvious that simulation of the background and , in particular , simulation of the nuclides decays is the overwhelmingly important part of such kind of research which can allow : i ) to understand and determine the origins of the background and hence to find certain methods to eliminate or suppress the background contributions ; ii ) to build up the background model and response functions of the detector for the effect being sought ( together with the detector energy and efficiency calibrations , resolution , source activities , etc . ) , thus to extract and evaluate searched effect ( or to exclude it ) more precisely .
there are several general programs , which are commonly used for simulation of the particles interactions in the experimental set up as , for example , geant package @xcite or egs4 code @xcite . in any
such a program user should describe initial kinematics of events by using the so called event generator .
the last is an important part of the simulation program providing the information , which particles and how many of them are emitted , what are their energies , directions of movement and times of emission .
existing computer codes radlst @xcite and imrdec @xcite only determine the radiation spectra due to decay of nuclides , and therefore could not be used for the further particles tracking . in attempt to cover this lack ,
the code , named decay4 , was developed for generation of events in the low energy nuclear and particle physics ( double beta decay and decay of radioactive nuclides ) .
this code was elaborated during the last decade , mainly for 2@xmath0 decay research @xcite .
the plan of the paper is as following : first , the overall features of the decay4 generator and used databases are considered , then detail descriptions of the parts associated with the double beta decay of atomic nuclei and decays of the natural and artificial radioactive nuclides are presented . in the last section
several examples illustrating the use of the decay4 generator in the real underground experiments are shown .
the developed program decay4 gives the possibility to generate events of the 2@xmath0 decay of atomic nuclei and of the radioactive decays ( @xmath1 , @xmath2 @xmath3 @xmath4 decay , electron capture ) of all known unstable isotopes .
it is divided into two main sections : \a ) init search and reading of all parameters of the nucleus and its decay needed for decay simulation from the ensdf @xcite ( or nudat @xcite ) , eadl @xcite and other libraries , in order to build up the nuclear and atomic decay schemes ; \b ) gendec a monte carlo events generator itself .
the ensdf database library includes the following information on about 2500 isotopes used for generation of the radioactive decays : a ) decay modes , their probabilities and energy releases , isotopes half - lives ; b ) radiation type , particles energies and intensities ; c ) parameters of nuclear levels ( half - life , spin , parity and excitation energy ) ; d ) parameters of nuclear transitions ( branching ratios , multipolarities , coefficients of internal conversion and mixing ratios ) .
the decay4 code uses also tables with the data on atomic properties of isotopes ( electron binding energies , electron capture ( ec ) subshell ratios , x rays and auger electrons intensities ) from the eadl database @xcite ( as well as from @xcite ) and tables with theoretical hager - seltzer conversion coefficients @xcite . the gendec part of the decay4 generates the energy , time of emission , direction and polarization for the following emitted particles : 1 ) electrons and positrons from single and double @xmath0 decay ; 2 ) @xmath1 particles from @xmath1 decay , protons and neutrons from @xmath5 and @xmath4 decays ; 3 ) @xmath6 quanta from nuclear deexcitation process ; 4 ) conversion electrons ; 5 ) e@xmath7e@xmath8 pairs from the internal pair conversion ; 6 ) @xmath9 quanta from bremsstrahlung in @xmath0 decay and ec ; 7 ) neutrinos ( antineutrinos ) from ec or @xmath0 ( 2@xmath0 ) decay ; 8) x rays and auger electrons from the atomic deexcitation process .
the decay4 describes double beta processes ( 2@xmath10 and 2@xmath11 decays , electron capture with emission of positron @xmath12 and double electron capture 2@xmath13 ) for all nuclides .
2@xmath0 transitions to the ground state as well as to excited 0@xmath8 and 2@xmath8 levels of the daughter nucleus are allowed .
if 2@xmath0 decay occurs to an excited level of a nucleus , the electromagnetic deexcitation process follows .
the energy release @xmath14 for double beta processes is taken from the table of the atomic masses @xcite .
for each transition to the ground or an excited level , various modes ( with emission of two neutrinos or majoron , neutrinoless decays due to nonzero neutrino mass or right - handed admixture in the weak interaction , etc . ) and mechanisms ( two - nucleon @xmath15 and @xmath16-isobar n@xmath17 ) of double beta decay are possible . below we give the list of double beta processes which could be simulated with the decay4 : 1 . _ * @xmath18 * _ decay with neutrino mass _ * , @xmath19 * _ transition _ * , @xmath15-*_mechanism ; 2 . _ * @xmath18 * _ decay with right - handed currents , _ * @xmath20 * _ transition _ * , @xmath15-*_mechanism ; 3 . _
* @xmath18 * _ decay with right - handed currents , _ * @xmath20 * _ and _ * @xmath21 * _ transitions _ * , * _ n_*@xmath17-*_mechanism ; 4 .
_ * @xmath22 * _ decay , _ * @xmath20 * _ transition , _ * @xmath15-*_mechanism ; 5 .
_ * @xmath18 * _ decay with emission of majoron , _ * @xmath20 * _ transition _ * , @xmath15-*_mechanism ; 6 . _
* @xmath18 * _ decay with double majoron emission _ * , @xmath20 * _ transition , _ * @xmath15-*_mechanism ; decay with charged _ * @xmath23 * _ majoron or massive vector majoron ; 7 . _
* @xmath18 * _ decay with right - handed currents , _ * @xmath21 * _ transition , _ * @xmath15-*_mechanism ; 8 . _
* @xmath22 * _ decay , _ * @xmath21 * _ transition , _ * @xmath15- * _ and n_*@xmath17-*_mechanisms ; 9 . _
* @xmath24*_@xmath13_*@xmath11 * _ decay ; 10 . _ * @xmath25*_@xmath13_*@xmath11 * _ decay _ * , @xmath20 * _ and _ * @xmath21 * _ transitions ; 11 . _ * @xmath26*_@xmath13 _ * * _ decay ; 12 . _ * @xmath27*_@xmath13 _ * * _ decay , _ * @xmath19 * _ and _ * @xmath21 * _ transitions .
the theoretical formulae for the energy and angular distribution @xmath28 of emitted electrons or positrons are based on works @xcite .
for example , for the first process@xmath29@xmath30 where @xmath31 is the kinetic energy of the @xmath32-th e@xmath7 or e@xmath8 ( in units of the electron mass @xmath33 ) , @xmath34 is the momentum ( in units of @xmath35 ) , @xmath36 is the fermi function , @xmath37 is the atomic number of the daughter nucleus ( @xmath38 for 2@xmath10 and @xmath39 for 2@xmath11 decay ) , @xmath40 is the angle between the particles directions , @xmath41 is the energy available for the particles ( @xmath42 for 2@xmath43 decay and @xmath44 for 2@xmath11 decay , @xmath45 is the energy of the populated level of the daughter nucleus ) , @xmath46
the decay4 describes six decay modes : @xmath10 , @xmath47 @xmath5 and @xmath4 decays , electron capture and @xmath11 decay ( ec ) and isomeric transition ( it ) .
the modes @xmath48 ( @xmath49 , @xmath1 , @xmath5 , @xmath4 , ec , it ) , their probabilities @xmath50 available decay energies @xmath51 and isotopes half - life @xmath52 are taken from the ensdf @xcite or nudat @xcite databases .
the decay mode @xmath48 is sampled according to the probabilities @xmath53 . @xmath54
* decay . *
the endpoint energies in @xmath10 decay @xmath55 are related with the energy release @xmath56 and the level energies @xmath57 of the daughter nucleus by the equation @xmath58 kinetic energy of the beta particle @xmath59 is sampled in accordance with the distribution @xmath60 where @xmath61 is the factor of forbiddenness .
probability of the internal bremsstrahlung in beta decay and energy - angular distribution of bremsstrahlung @xmath9 quanta are calculated as in @xcite .
@xmath62 , @xmath63 * , * @xmath64 * decays . * the energies of particles @xmath65 ( @xmath66 @xmath3 @xmath4 )
are related with the level energies @xmath57 of the daughter nucleus by the equation @xmath67 where @xmath68 @xmath69 is the mass numbers of the parent and daughter nuclei , respectively . * ec ( electron capture and * @xmath70 * decay ) . *
the ensdf database includes the information on the probabilities @xmath71 ( for ec ) and @xmath72 ( for @xmath11 decay ) for @xmath32-th level of the daughter nucleus .
if the level is populated in @xmath11 decay , energy of positron is sampled according to ( 3 ) , where @xmath73 if the @xmath32-th nuclear level was populated in the electron capture , the number of the atomic subshell @xmath74 ( @xmath75 @xmath76 @xmath77 @xmath78 @xmath79 @xmath80 @xmath81 @xmath82 @xmath83 @xmath84 @xmath81 @xmath85 @xmath86 @xmath87 @xmath81 @xmath88 ) , where primary electron vacancy is created , is sampled according to probabilities @xmath89 @xcite : @xmath90,\ ] ] where @xmath91 is the electron capture transition angular momentum , @xmath92 is the relative occupation number for partially filled subshells @xmath74 ( @xmath93 @xmath94 is the number of electrons in the subshell @xmath74 , @xmath95 is the maximal number of electrons in the subshell ) , @xmath96 is the neutrino energy , @xmath97 is the electron binding energy in the parent atom , and @xmath98 is @xmath74 subshell angular momentum .
the squared amplitudes @xmath99 of the bound - state electron radial wave functions and the electron binding energy @xmath100 are taken from @xcite .
the internal bremsstrahlung probability and spectra of bremsstrahlung @xmath9 quanta in allowed electron capture transition from @xmath74 atomic subshell is calculated in accordance with @xcite . * nuclear deexcitation process .
* the nuclear deexcitation process occurs if a daughter nucleus is in the excited @xmath32-th level with energy @xmath57 .
the electromagnetic transition from @xmath32-th to @xmath101-th level is sampled according to probabilities @xmath102 where @xmath103 is the branching ratio of electromagnetic transition from @xmath32-th level to @xmath101-th taken from the nudat or ensdf databases .
there are three possible modes of electromagnetic transition with emission of : ( 1 ) @xmath9 quantum with energy @xmath104 ( 2 ) conversion electron with energy @xmath105 ( the condition @xmath106 should be fulfilled ) , here @xmath97 is the electron binding energy on the @xmath74 subshell@xmath107 ( 3 ) conversion electron - positron pair with total energy @xmath108 ( if @xmath109 ) . to sample the mode , the respective probabilities @xmath110 , @xmath111 , @xmath112 are used , where @xmath113 for all transitions beside e0 , @xmath114 for e0 transition , where @xmath115 , @xmath116 are the coefficients of the internal electron and pair conversion , respectively , and @xmath117 , @xmath118 are the intensities of internal electron and pair conversion , respectively .
total @xmath115 , partial subshell @xmath119 , shell @xmath120 coefficients of electron internal conversion is related by @xmath121 where @xmath122 is the shell index and @xmath123 is the subshell index of the @xmath122 shell @xmath124 .
coefficients @xmath115 and @xmath125 are taken from the nudat or ensdf databases . in case if their values are absent , the coefficients are calculated as @xmath126/(1+\delta _ { ij}^2),\ ] ] where the values of partial coefficients @xmath127 are taken from @xcite . here
@xmath128 is transition multipolarity , and @xmath129 is the mixing ratio of different multipolarities in the @xmath130 transition . for pure @xmath131 transitions ,
electron conversion coefficients are calculated according to formulae from @xcite .
coefficient of internal pair conversion @xmath132 is given by formulae similar to ( 10 ) , and partial coefficients @xmath133 , and electron - positron energy and angular distributions are calculated according to formulae from @xcite . * x rays and auger electrons .
* the vacancies in the atomic shells are created in the electron capture or in the result of internal electron conversion .
ionized atom is deexcited filling the vacancies by electrons from higher atomic shells ; x rays and auger electrons are emitted in this process .
for the sampling of the atomic deexcitation process , the values of the electron binding energies and occupation numbers , radiative and radiationless partial widths are taken from the eadl library @xcite .
the type of process ( radiation or auger electron emission ) is sampled accordingly to the values of radiative and radiationless partial widths . for the vacancy in the @xmath134 subshell
, the energy of x ray for radiative process , in which higher vacancy @xmath135 is created , and the energy of auger electron for nonradiative process ( vacancies @xmath135 and @xmath136 are created ) are given by @xmath137 correction @xmath138 is found from equations in @xcite .
multivacancy corrections for energies and partial widths are included also .
the x ray and auger electron energy spreading is accounted according to lorentzian distribution function @xcite .
* decay chains and time characteristics . * time interval @xmath139 between appearance of the unstable state @xmath140 of the nucleus or atom and its decay with emission of the particle is sampled in accordance with the relation @xmath141 where @xmath142 is a random number uniformly distributed in the range @xmath143 , @xmath144 is a half - life of the unstable state .
the decay4 can also simulate the full decay chain of the parent nuclide together with its daughters .
as an example , the generated spectra of particles , emitted in decay of nuclides from full @xmath145u chain , are shown in fig .
activities of all daughter nuclei are calculated by the decay4 .
* angular correlations between emitted particles . *
the direction ( polar and azimutal angles @xmath146 , @xmath147 of the particle @xmath32 is sampled as isotropical if one of the next conditions is fulfilled : 1 ) the particle is the first in the decay of unoriented parent nuclear state @xmath32 ; 2 ) @xmath148 ( @xmath149 is nuclear spin before emission of the particle @xmath150 3 ) @xmath151 ( @xmath152 is the time given by ( 13 ) , @xmath153 is user defined parameter ( time ) to account the influence of external fields which violate the angular correlation ) ; 4 ) @xmath154 @xmath155 or @xmath156 @xmath157 ( @xmath154 and @xmath156 are angular momenta of particles @xmath32 and @xmath158 if particle @xmath159 exists)@xmath107 5 ) one of the values @xmath160 , @xmath149 , @xmath161 , @xmath154 , @xmath156 is unknown and could not be evaluated . in case , if particle @xmath162 does not satisfy any of the above conditions but particle @xmath32 satisfies one of them , the polar angle @xmath163 of particle @xmath162 in the coordinate system of particle @xmath32 is sampled in according to correlation function @xcite @xmath164 where @xmath165 is the second possible angular momentum of the particle ( for the mixture of multipolarities ) , @xmath166 is type of the particle @xmath32 ( @xmath167 @xmath168 @xmath47 @xmath169 @xmath170 are angular momenta functions , and @xmath140 is even . in this case correlation function is independent from the azimutal angle @xmath171 .
if the next emitted particle @xmath172 exists , which does not satisfy any condition among 1)5 ) , its direction is determined by the more complicated correlation function , which depends on directions of previous particles .
correlation of linear polarisations of @xmath9 quanta is also taken into account @xcite .
the code decay4 was successfully used in several underground experiments for detectors design and optimization , simulation of the backgrounds ( with the help of the geant package ) and for evaluation of the results .
some examples are listed below . 1.1 .
cadmium tungstate crystal scintillators ( enriched in @xmath174cd to 83% ) were used to study @xmath174cd @xcite .
the background of the @xmath174cdwo@xmath175 crystal ( 15.2 @xmath176 ) in the energy region of interest ( q@xmath177=2805 kev ) was equal to @xmath1780.6 counts / y@xmath179kg@xmath180kev . with the 19175 h statistics the half - life limit for neutrinoless 2@xmath0 decay of @xmath174cd
was set : t@xmath181 y ( 90% c.l . ) @xcite .
limits on 0@xmath182 modes with emission of one ( m1 ) or two ( m2 ) majorons were established also : t@xmath183 y and t@xmath184 y ( 90% c.l . ) @xcite . comparing these limits with the theory ,
the restrictions on the neutrino mass @xmath185 ev and majoron - neutrino coupling constant @xmath186 were derived , which are among the most sensitive results for other nuclei @xcite .
the study of the @xmath187 decay of @xmath188gd was carried out with the help of the 95 @xmath176 gd@xmath189sio@xmath190:ce crystal scintillator .
the background was reduced to @xmath1781.0 cpd / kev@xmath179 kg in the vicinity of q@xmath191 energy ( @xmath1781.73 mev ) .
the improved half - life limit was set for @xmath192 decay of @xmath188gd : t@xmath19310@xmath194 y at 68% c.l . @xcite .
two radiopure caf@xmath189:eu crystal scintillators ( 370 g each ) have been used to study @xmath187 decay of @xmath195ca and double electron capture of @xmath196ca as well as for dark matter search .
the highest up - to - date half - life limits were reached for @xmath24 and @xmath25 double electron capture of @xmath197ca : t@xmath19810@xmath199 y and t@xmath20010@xmath199 y ( 68% c.l . ) @xcite .
the study of the @xmath201 decay of @xmath202cd was performed with the help of two low background nai(tl ) crystals and enriched ( to 68% ) @xmath203cd samples ( @xmath178154 g ) .
new t@xmath204 limits for the @xmath205 , @xmath11/ec and ec / ec decay of @xmath202cd have been set in the range of ( 0.3@xmath2064)@xmath17910@xmath207 y at 90% c.l . @xcite .
nemo collaboration experiment performed in the frejus underground laboratory to study 2@xmath0 decay of @xmath208mo @xcite . for background simulation the first version of the decay4 code @xcite was used . the clear two neutrino 2@xmath0 signal
( 1433 events during 6140 h ) was observed , leading to a half - life t@xmath209(stat)@xmath210(syst)@xmath211 y @xcite .
heidelberg - kiev collaboration : the background simulations were performed for the genius project which aim is to increase the present sensitivity for 2@xmath0 decay and dark matter search .
contributions from the cosmogenic activity produced in the ge detectors and from their radioactive impurities as well as from contamination of the liquid nitrogen and other materials were calculated .
external @xmath9 , @xmath212 and neutron backgrounds were considered also .
the results of calculations evidently show feasibility of the genius experiment @xcite .
the event generator decay4 was used successfully in many underground experiments ( kiev , roma - kiev , dama- , nemo- , genius - collaborations ) . +
this work was supported in part by the science and technology center of ukraine ( project # 411 ) . | the computer code decay4 is developed to generate initial energy , time and angular distributions of particles emitted in radioactive decays of nuclides and nuclear ( atomic ) deexcitations .
data for description of nuclear and atomic decay schemes are taken from the ensdf and eadl database libraries .
the examples of use of the decay4 code in several underground experiments are described . -10pt
-10pt 0.4 cm * o.a .
ponkratenko , v.i .
tretyak and yu.g .
zdesenko * _ institute for nuclear research , prospect nauki 47 , msp 03680 kiev , ukraine _ |
low temperature plasmas play a crucial role in materials processing @xcite . particularly capacitive
radio frequency discharges are important .
such plasmas , which were first introduced in the 1960s , have undergone a continuous development keeping up with the increasing requirements of the hitech industry .
one trend is the attempt to increase productivity by increasing the size of the processed wafers or substrates . in microelectronics , wafer sizes of @xmath0 mm
are now state of the art , the transition to @xmath1 mm is envisioned @xcite .
the manufacturing of flat panel displays or photovoltaic solar cells requires discharges in the meter range @xcite .
the size of the processing equipment obviously must meet these requirements .
a second recent trend is the use of higher , often multiple frequencies of up to few hundred mhz believed to ensure better power coupling , higher plasma density , and thus higher productivity as well @xcite .
these developments have implications on the physical models which can be used for realistic simulations of plasma - based manufacturing processes .
one such implication is of particular importance : the simultaneous increase of driving frequency , plasma density , and discharge size puts the plasmas into a regime where the commonly used `` electrostatic approximation '' of maxwell s equations no longer suffices .
that means that @xmath2 ( and thus @xmath3 ) is not justified anymore .
( in this regime the skin depth and/or wavelength of the surface waves propagating in the sheath regions may become comparable or smaller than the reactor dimension ) .
recent experimental studies have revealed that strong plasma and field non - uniformities appear to be connected with electromagnetic effects @xcite .
the theoretical aspects of electromagnetic effects in large - area , high frequency discharges were first systematically studied by lieberman et al .
they considered an axisymmetric slab model with uniform plasma density in the bulk and the matrix sheath ( i.e. , constant ion density and zero electron density ) , treating the plasma as a dielectric and assuming the sheath width to be constant in time .
also , a natural boundary " condition for a ccp discharge with the dielectric sidewall that all the rf current flows through the discharge ( so that the magnetic field at the sidewall is constant in the axial direction and no external source of electromagnetic excitation is involved ) was employed . under these conditions , they treated the problem analytically and gave criteria , based on the estimate of change in the power deposition profile produced by ohmic heating due to these effects , for significance of the standing wave and skin effects .
in essence , a standing wave consists of surface waves , see , e.g. , @xcite , excited in the cavities formed in the powered ( grounded ) sheath , bounded by the electrode , powered ( grounded ) sheath interface and the sidewall ; the surface waves in the grounded and powered regions are weakly coupled with each other through the bulk .
the kind of skin effect meant here manifests itself in tendency of the induced electric field in an rf - driven discharge to reduce current at the axis of the discharge and increase it close to the discharge periphery @xcite , similar to the skin effect in metals conducting rf current ( see , e.g. , @xcite ) .
several detailed studies were conducted by chabert et al .
@xcite who used a self - consistent transmission line model to account for the interaction of rf heating and plasma density . in recent years
the number of contributions to the field has multiplied , for a review see , e.g. , ref . @xcite .
the analytical models mentioned above , being significantly limited by the underlying assumptions , leave many important questions unanswered .
numerical simulations should be used to describe the phenomena in more realistic situations . so far a number of numerical studies based on the fluid models have been conducted for the large scale ccp discharges driven at high frequencies @xcite .
all those approaches were based on the full set of maxwell s equations , which contains electromagnetic radiation .
the underlying equations are hyperbolic and one has to satisfy the courant criterion on numerical stability provided an explicit algorithm , such as the popular finite - difference time - domain ( fdtd ) algorithm , is used for their numerical solution .
the courant criterion forces the time step to be small and the overall simulation time very long .
one can resort to implicit algorithms for the numerical solution of maxwell s equations , however , it must be noted that the existing implicit numerical schemes are rather intricate . in this paper
we propose an alternative numerical approach for studying the large scale , high frequency driven capacitively coupled discharges , which is able to describe the electromagnetic phenomena occurring in such discharges .
the approach is based on the darwin approximation for solving the system of maxwell s equations @xcite .
the darwin approximation reduces the original system of hyperbolic equations to a set of elliptic equations by neglecting the transversal part of the displacement current in ampere s law , which is justified by a scale analysis of the phenomena of interest .
since the resulting elliptic equations do not support electromagnetic radiation in vacuum , the corresponding courant criterion is removed and the simulation time step can be chosen as in electrostatic simulations .
the required numerical scheme is a relatively simple and straightforward modification of the well - established electrostatic algorithms . whereas numerical approaches based on the darwin approximation have proven successfull in unbounded plasmas ( see , e.g. , ref .
@xcite ) , application of the darwin approximation in plasmas having plasma - wall interfaces is more intricate .
it is therefore important that the approach proposed in this article avoids the computationally expensive iterative schemes used in the previous implementations of the darwin approximation - based aglorithms for bounded plasmas ( see refs .
@xcite ) .
we demonstrate that the darwin approximation is in very good agreement with the fully electromagnetic linear treatment of the problem described in @xcite . however , realistic plasma density profiles are far from the model used in the latter work , hence next we study the standing wave and skin effect phenomena employing a nonlinear fluid model based on the darwin approximation with more realistic plasma profiles treated in time domain . unlike some works ( see , e.g. , ref .
@xcite ) , which consider excitation of the electromagnetic fields in the discharge coming from external sources , we assume the
natural " excitation of the electromagnetic fields in the ccp discharge generated by the voltage / current that drives the discharge .
obviously , such kind of excitation is always present in the discharge . also
, our darwin - fluid model accounts for finite electron inertia effects which is necessary in plasmas with low collisions driven at high frequencies .
the paper is organized as follows : in section ii we briefly describe the assumptions of the darwin approximation and its applicability . in section iii
we describe the choice of representation for the electromagnetic fields that is particularly convenient for the numerical implementation and a model geometry used in the examples studied in this paper , the boundary conditions are also given in this section . in section
iv we use a linear fluid model solving for the electromagnetic fields in frequency domain both with the darwin approximation and the full system of maxwell s equations for two example cases with pronounced standing wave and skin effects , respectively , to corroborate that the darwin approximation gives an accurate description of the phenomena in question for the parameters of interest . in section
iv we describe a nonlinear fluid model accounting for the finite electron inertia effects and also using the darwin approximation to solve for the electromagnetic fields .
we demonstrate in a model simulation the standing wave effect for a case with more realistic plasma density profiles and sheath dynamics than in cases investigated in section iv .
finally , we summarize the main results of our study in section vi .
in general , the electrodynamic behavior of any high frequency discharge can be described by the kinetic equation , such as boltzmann s equation , supplemented with the system of maxwell s equations along with appropriate boundary conditions . reduced fluid models can be constructed from this general system by taking moments of boltzmann s equation and postulating closure relations .
the dynamics of the particle distribution function of charged species @xmath4 under influence of electromagnetic forces and collisions is determined by boltzmann s equation , @xmath5 in turn , the electromagnetic field is described by maxwell s equations @xmath6 where @xmath7 and @xmath8 .
direct attempts at numerical solution of - usually require very long simulation time .
this is connected to the fact that this system supports electromagnetic waves in vacuum , which must be resolved in explicit algorithms in order to avoid numerical instabilities .
another approach is to use an implicit algorithm for the numerical solution , where unresolved electromagnetic modes of no interest are damped numerically .
however , such algorithms are cumbersome to implement and often resort to artificial numerical constructions ( see , e.g. , @xcite ) .
a conceptually different alternative is to remove the stiffest time scale in electromagnetic simulation by reduction of the original system of equations using the assumption about the time and spatial scales of the expected phenomena .
namely , we use the fact that for the far most ccp discharges of interest the electrode size is small compared to the wavelength of the electromagnetic wave in vacuum corresponding to the highest frequency driving harmonics . following the logic of @xcite
, one can recast and into a normalized form where @xmath9 with @xmath10 and @xmath11 the typical spatial and time scale of the problem , @xmath12 to the zeroth order in @xmath13 yields @xmath14 , so that to this order the electric field is purely longitudinal ( irrotational ) ; we will denote this part of the electric field @xmath15 .
to the same order the displacement current drops out of . to describe the electromagnetic effects one needs to include the next order in @xmath13 , whereby @xmath16 with @xmath17 the transversal ( solenoidal ) part of the electric field , @xmath18 . and constitute the darwin approximation .
one can see that this model eliminates electromagnetic waves , but keeps an important part of the physics , namely electromagnetic effects in the relatively low frequency phenomena ( when the corresponding wavelength in vacuum is larger than the system size ) .
note that the corresponding frequencies are called ( very ) high frequencies in the standard terminology of the ccp discharges .
it is important to note that by dropping the transversal part of the displacement current the continuity equation remains satisfied .
the relation of the darwin model to the full system of maxwell s equations can also be illustrated by the cold unmagnetized plasma dispersion relation in an unbounded plasma , @xmath19 fig .
1 displays the solutions @xmath20 of this relation for a typical parameter combination ( @xmath21 , @xmath22 ) .
three distinct regimes can be discerned .
slow phenomena ( @xmath23 ) are governed by the collisional skin effect @xmath24 , with @xmath25 the collisionless skin - depth . in the intermediate frequency range ( @xmath26 ) , the collisionless skin effect prevails , @xmath27 finally , for large frequencies ( @xmath28 ) this dispersion relation describes propagation of the electromagnetic waves having @xmath29 . in contrast
, the dispersion relation of the darwin model is @xmath30 the darwin model then covers the middle ground between the electrostatic approximation and a fully electromagnetic treatment .
this example , however , serves only illustrative purpose and is not general in stating that the darwin approximation is inapplicable for @xmath31 as the dispersion relation of waves in bounded plasmas can significanly differ from eq .
( 11 ) ( as is the case in the present study , for example ) .
the decisive conclusion regarding applicability of the darwin approximation for description of a phenomenon in question should be made on the basis of the smallnes of the @xmath13 parameter defined above .
historically , darwin s original derivation @xcite was designed to describe a set of charged particles in free space with velocities small compared to the speed of light .
he started from the lagrangian description of the @xmath32-body problem with the interaction incorporated via the retarded lienard - wichert potentials .
a formal expansion in the smallness parameter @xmath33 , carried up to the second order , resulted in a simplified lagrangian with instantaneous ( not retarded ) potentials . in the same expansion order , relativistic corrections to the equations of motion appeared .
for a recent analysis of that approach , see @xcite .
for the sake of concreteness , we will consider the same problem setup for studying the standing wave and skin effect phenomena in ccp discharges as was suggested by lieberman et al .
@xcite , i.e. , a ccp discharge having cylindrical geometry symmetric in the azimuthal direction bounded by a dielectric sidewall ( see fig .
2 ) . as all the numerical models described in this article will use the same field equations stemming from the darwin approximation of maxwell s equation , we will describe them in this section .
we saw in the previous section that the electric field in the darwin approximation is separated into two parts , a longitudinal part , which is curl - free , and a transversal part , which is divergence - free .
hence , we choose to represent the electric field in the following way , @xmath34 where the first ( second ) term expresses the longitudinal ( transversal ) part of the electric field .
the longitudinal potential @xmath35 is governed by poisson s equation , @xmath36 with @xmath37 the charge density . to find an equation for the transversal vector - potential @xmath38 we first take the curl of ampere s law in the darwin approximation , which results in @xmath39 where we used .
then , one can exploit faraday s law using the chosen representation for the transversal field ,
@xmath40 following @xcite , we consider only tm modes with @xmath41 and @xmath42 , and @xmath43 , so that @xmath44 and @xmath45 hold automatically due to the symmetry of the problem . in all the following models we study the electromagnetic response to the `` natural excitation '' of a ccp discharge , prescribing a harmonically varying constant magnetic field at the sidewall , which is in agreement with the previously made assumption that all the rf current , which drives the discharge , flows through it , and no external excitation source is involved .
correspondingly , boundary conditions for @xmath46 is @xmath47 at the electrodes ( @xmath48 and @xmath49 ) , which is needed in order for the tangential electric field to vanish at the electrodes , and reads @xmath50 at the sidewall ( @xmath51 ) ( provided no rf current leaves the discharge through the dielectric sidewall ) .
boundary conditions for @xmath52 are obtained by taking time derivative of the boundary conditions for @xmath46 .
further , boundary conditions for @xmath53 are @xmath54 at the electrodes , so that the tangential component of the transversal electric field vanishes , and @xmath55 at the sidewall .
the latter boundary condition for the @xmath53 at the sidewall can be obtained as follows . from the condition that the radial component of the transversal electric field vanishes at the dielectric sidewall one can deduce that @xmath56 at @xmath51 . integrating this equation with respect to @xmath57 at @xmath51 , matching the boundary conditions for @xmath53 at the electrodes and setting the common constant to zero
, one obtains the boundary condition sought . as the boundary conditions for
the longitudinal potential will be slightly different in the two models studied in this work , we will specify the boundary conditions for each particular model separately .
in this section we will demonstrate that the darwin approximation is a valid approach for investigation of the electromagnetic phenomena in ccps by comparing the full electromagnetic solution to solution obtained with the darwin approximation for the problem investigated by lieberman et al . @xcite .
the model plasma studied there has uniform ion density and a stepwise density for the electron component , such that electrons have the same density as ions in the bulk and zero density in the sheath , the sheath boundaries are considered to be stationary .
the plasma is modeled as a dielectric , change of the electric field in the sheaths due to motion of the bulk electron plasma component is modeled through the surface charge created at the interfaces between the bulk plasma and the sheath regions .
this approach is valid for small oscillations of the bulk plasma ( see @xcite ) , a more general case of moving sheath boundary is considered in the next section .
then , using the representation of the electric field given in ( which holds in a general case as well ) , one can obtain the following system of equations , @xmath58 provided all quantities are proportional to @xmath59 . in and , @xmath60 in the sheath and @xmath61 in the bulk .
this system of equations describes both fully em case ( if @xmath62 ) and the darwin approximation ( if @xmath63 in the sheath and @xmath64 in the bulk ; this expression can be easily obtained in the usual way by using the equation of motion and ampere s law in the darwin approximation ) .
once one obtains @xmath38 and @xmath65 from solution of and , the transversal electric field can be obtained from @xmath66 and the longitudinal electric field from ampere s law , which provides @xmath67 .
results in figs .
3 and 4 show the power deposition profiles for two selected cases with similar parameters to the corresponding cases in @xcite . in all cases
@xmath68 @xmath69 , and @xmath70 .
the radial profiles shown in figs .
3 and 4 are normalized values of @xmath71 ( with @xmath72 being predominantly capacitively generated electric field ) and @xmath73 @xcite ( with @xmath74 being predominantly induced electric field ) , calculated with the full system of maxwell s equations and the darwin approximation . not only do both models demonstrate same behavior qualitatively , but also quantitative difference lies within a numerical error for the parameters chosen . consequently , from this moment on we will suppose that the darwin approximation is a valid approach for description of the standing wave and skin effects .
although the model problem used @xcite and in the previous section gives a certain insight into the standing wave and skin effects , it is still quite far from realistic ccp plasmas .
the assumptions of stationary sheath boundary and stepwise electron density profile are the main drawbacks of the model .
most of the previous analytical and numerical treatments of the standing wave and skin effects in the literature also assumed stationary sheath boundary with the electrostatic sheath model ( e.g. , @xcite ) , which makes the corresponding numerical algorithm faster , but strips the model of adequate description of possible nonlinear phenomena . in @xcite
the authors studied the problem entirely in the time domain , however , the commonly adopted drift - diffusion approximation also used in that work is clearly not applicable to the usual range of parameters in low - pressure ccp discharges , when the frequency of electron collisional momentum transfer is smaller than the driving frequency .
in particular , the highly collisional regime studied in @xcite renders direct comparison with results of @xcite impossible , as the latter work was made under the low - collisionality ansatz . to fully resolve the nonlinear discharge dynamics
we propose another model , where the sheath boundary is allowed to move , the plasma density profiles have a more realistic shape , and the effects of finite electron mass are retained , which makes the model capable of more general treatment of the electromagnetic phenomena under question in low - pressure ccp discharges than the model considered in the previous section .
the electromagnetic fields in the sheath region are calculated self - consistently with the plasma dynamics and are treated in the same way as the rest of the electromagnetic fields in the entire discharge . to keep the model simple
yet comprising all the effects of interest , we adopt a fluid description for the electron plasma component , whereas ions are assumed to be immobile . to make the model realistic , we adopt the axial ion density profile from a 2d electrostatic pic / mcc simulation ( with the same geometry and plasma parameters as those of the fluid simulation studied here ) , taken at the half - radius radial position ( since the density profiles in the electrostatic case are essentially uniform , we feel that such an approach is justified ) . in this way
we choose to focus on the coupled dynamics of electrons and the electromagnetic fields in the entire discharge , omitting the issue of discharge sustainment .
analogously , we substitute the plasma energy balance equation with the assumption of electron isothermality .
the model geometry is the same as considered in the previous section .
the evolution of electrons is governed by the momentum equation , @xmath75 and the continuity equation , @xmath76 the boundary conditions for is that the radial component of the electron velocity , @xmath77 , vanishes at the sidewall , which comes from the assumption that all the electrons are reflected back into the discharge there , and that the axial component of the electron velocity at the driven ( grounded ) electrodes is @xmath78 , respectively , which corresponds to the kinetically limited flux .
the boundary conditions for are , in accordance with the boundary conditions for , that the radial flux to the sidewall vanishes , and the axial flux to the electrodes is kinetically limited , so that @xmath79 at the driven ( grounded ) electrodes , respectively .
we take the initial ion density profile to be of the gaussian form and the initial electron density profile to be of the same form as the ion density profile .
the numerical discharge simulation gradually develops electron - depleted sheath regions ( see fig.5 ) .
the fields are calculated self - consistently with evolution of the electron plasma component using to ) .
one can distinguish two different sets of boundary conditions depending on the way the discharge is driven , either with a voltage source or a current source .
whereas boundary conditions for are same for both cases ( see the discussion at the end of section iii ) , the boundary conditions for and are different for each case . in this case
the potential at the driven electrode is prescribed , @xmath80 .
this gives dirichlet boundary condition at the driven electrode , the boundary conditions at the sidewall and the grounded electrode being same in both cases ( neumann boundary condition at the sidewall , @xmath81 and at the grounded electrode dirichlet boundary condition @xmath82 .
( see also the discussion in section iii ) .
the time derivative of the total axial current through the discharge , which enters the sidewall boundary condition for , must be calculated under the circumstances by integrating the plasma current density time derivative , @xmath83 using and over the surface of any discharge cross - section where the plasma current is dominant , for example , in the midplane .
the potential at the driven electrode is calculate in this case from kirchhoff s law of current continuity at the driven electrode , where the sum of the electron current and the displacement current in the sheath must match the current supplied by the source .
we seek solution to @xmath35 in the form @xmath84 , where @xmath85 , where @xmath86 is the stationary , and @xmath87 the harmonically changing parts of the potential at the driven electrode , respectively .
the boundary conditions are : @xmath88 at the driven electrode , @xmath89 at the grounded electrode , and @xmath90 at the sidewall ; @xmath91 satisfies @xmath92 with boundary conditions @xmath93 at both electrodes , and @xmath94 at the sidewall . from the current balance at the driven electrode one
has @xmath95 . substituing the chosen form for @xmath35 into this equation ,
one obtains an equation for the potential at the driven electrode @xmath96 the boundary condition for at the sidewall is particularly simple in the case of current - driven discharge , as one can directly use the value of the time derivative of the total current supplied by the source .
finally , we want to briefly discuss some details of numerical implementation of ) , , , , and .
we use a leapfrog - like scheme to integrate and in time , for which we use grids staggered in time for velocity and density , the density and field values being taken at integral time levels and velocity at the half- time levels .
then , the order of the whole numerical scheme is the following : 1 .
is solved treating the nonlinear second term on the left - hand - side semi - implicitly thereby linearizing this term with respect to the density taken at the new time level .
the velocity on the right hand side is time - centralized by taking the average between the new and old time levels , so that it is also semi - implicit .
it is worth mentioning that during the stage of the sheath formation an artificial viscosity term is helpful in getting rid of numerical instabilities arising at the plasma - sheath interface .
once smooth sheaths are formed , the viscosity term can be switched off .
the density on the right - hand side of is treated semi - implicitly and is time - centralized using the same technique as in the previous procedure .
is solved with the boundary conditions depending on the way the discharge is driven according to the techniques described above .
the source term in should be calculated using the expression @xmath97 with @xmath98 and @xmath99 calculated with help of and .
an attempt at direct evaluation of @xmath100 will lead to numerical instability because numerical discretization of a time derivative introduces numerical dispersion and finite field propagation time , whereas conceptually the darwin approximation expects instantaneous field change in the whole domain ( see , e.g. , @xcite ) .
when calculating the source term for , it is also useful to write the term @xmath101 as @xmath102 , where @xmath25 is the collisionless skin depth and we used .
the first term in this expression describes the collisionless skin effect and can be transferred to the left hand side to be treated implicitly for better numerical stability .
finally , is solved using the boundary conditions discussed in section iii .
one can see that because of the chosen representation for the transversal field as @xmath103 the cumbersome and computationally expensive iterative scheme needed for making sure that the calculated @xmath104 is divergence - free used in the previous implementations of the darwin codes ( see , e.g. , @xcite ) is avoided .
this is because @xmath104 calculated by taking the curl of @xmath38 guarantees that @xmath104 is divergence - free automatically . to account for the electromagnetic effects in 2d geometry of the problem we described
, one needs to solve only two additional elliptic equations using the same field solver that is used for solving poisson s equation .
it is a small price for removing the very expensive courant criterion connected with propagation of the light wave , though . in fig .
6 we show the radial profile of relative power deposition due to the ohmic heating by the axial electric field resulting from a current - driven discharge simulation showing a significant standing wave effect for a case with relatively low collisionality ( @xmath105 s@xmath106 ) .
parameters are similar to the parameters for the case shown on the top part of fig .
the inductive power deposition is negligible and is not shown . for comparison
we show results of the linear problem of the section iv with the stepwise electron density profile and equivalent sheath width @xmath4 calculated from the equation @xmath107 ( see @xcite ) with the period averaged densities shown in fig .
the non - uniformity of the radial power deposition profile is slightly more pronounced in the case of the nonlinear fluid model of the present section with the more realistic electron axial density profile , but in general both models are in good agreement .
note that the radial profile of relative power deposition obtained with a purely electrostatic simulation would be uniform .
as one can see , the electromagnetic component is an essential part of the standing wave effect for the parameters in question , despite the fact that surface waves propagating in sheath and generating standing waves may also exist in the electrostatic description ( see , e.g. , @xcite ) .
the evidence of the model nonlinearity can be seen on the fourier spectrum of the time - dependent voltage drop across the current - driven discharge ( fig .
one can clearly see that apart from the fundamental harmonic , the spectrum exhibits response at the third harmonic also .
the amplitudes of other harmonics are noticeably supressed , which might be attributed to the fact that the fluid models tend to over - relax the arsiing density gradients leading to damping of high harmonics .
in this paper we have demonstrated that the darwin approximation is adequate for description of the standing wave and skin effects present in the high frequency ccp discharges by comparing relevant electromagnetic fields calculated using the approximation and solution of the full set of maxwell s equations for two different typical cases exhibiting such effects in a model problem .
we considered the `` natural '' excitation of such effects coming from the current driving the discharge needed for its very existence .
the darwin approximation reduces originally hyperbolic maxwell s equations , by dropping the transversal current in ampere s law , to a set of elliptic equations which do not support light waves propagating in vacuum .
this has very significant consequence for the numerical algorithms based on that approach that the corresponding courant criterion is also removed , enabling same time step size as in electrostatic algorithms .
we have shown that the darwin approximation can be straightforwardly implemented in fluid models , requiring only a slight modification of a possibly existing electrostatic code with the eventual great computational pay - off .
the choice of the representation for the transversal field proposed by us helps to avoid the computationally demanding iterative schemes needed for its divergence cleaning used in other implementations . using a reduced nonlinear fluid code self - consistently evolving plasma and electromagnetic fields in time domain ,
on the example of a case exhibiting standing wave effect we have shown that the electromagnetic effects persist for more realistic profiles and sheath dynamics .
the proper description of low - pressure ccp discharges ultimately requires a kinetic tool as kinetic effects are essential in such discharges .
the use of the darwin approximation is , of course , not limited to fluid models .
the sheath dynamics , the heating mechanisms other than the ohmic heating , the nonlocal transport phenomena all demand kinetic treatment . the darwin numerical algorithm for the solution of the electromagnetic fields designed in this work can be also used in a particle - in - cell code . the reduced fluid model developed in this work
can , in turn , serve as benchmark for such a code .
the authors gratefully acknowledge fruitful discussions with prof .
r. grauer ( ruhr university bochum ) and support by the deutsche forschungsgemeinschaft via the sonderforschungsbereich trr 87 .
m. a. lieberman and a. j. lichtenberg , _ principles of plasma discharges and materials processing _ ( wiley , new york , 2005 ) , 2nd ed.f s. samukawa , m. hori , s. rauf , k. tachibana , p. bruggeman , g. kroesen , j. c. whitehead , a. b. murphy , a. f. gutsol , s. starikovskaia j. phys .
phys . * 45 * , 253001 ( 2012 ) a. e. park , b. u. cho , and j. k. lee .
ieee trans .
plasma sci . * 31 * , 628 , ( 2003 ) j. schmitt .
m. elyaakoubi , and l. sansonnens , plasma sources sci .
technol . * 11 * , 206 ( 2002 ) j. perrin , j. schmitt , c. hollenstein , a. howling , and l. sansonnens plasma , phys .
fusion . * 42 * , 353 ( 2000 ) h. h. goto , h. d. lwe , and t. ohmi , j. vac .
technol . a * 10 * , 3048 , ( 1992 ) n. nakano and t. makabe , j. phys .
d * 28 * , 31 , ( 1995 ) l. sansonnens and j. schmitt , appl .
lett . * 82 * , 182 ( 2003 ) a. a. howling , l. sansonnes , j. ballutaud , and c. hollenstein , j. appl .
phys * 96*. 5429 ( 2004 ) a. perret , p. chabert , j .- p . booth , j. jolly , j. guillon , and p. auvray , appl .
. lett . * 83 * , 243 ( 2003 ) p. chabert , j .- l .
raimbault , j .-
rax , and a. perret , phys .
plasmas * 11 * , 4081 ( 2004 ) g. a. hebner , e. v. barnat , p. a. miller , a. m. paterson , and j. p. holland , plasma sources sci .
technol . * 15 * , 879 ( 2006 ) p. a. miller , e. v. barnat , g. a. hebner , a. m. paterson , and j. p. holland , plasma sources sci . technol . * 15 * , 889 ( 2006 ) m. a. lieberman , j .-
booth , p. chabert , j .-
m . rax , and m. m. turner , plasma sources sci .
technol . * 11 * , 283 ( 2002 ) k. j. bowers , thesis ( phd ) , university of california , berkeley , source dai - b 62/07 , p. 3254
( 2002 ) t. mussenbrock , t. hemke , d. ziegler , r.p .
brinkmann , m. klick , plasma sources sci . technol . *
17 * , 025018 ( 2008 ) s. ramo , j. r. whinnery , and t. van duzer , fields and waves in communication electronics , 3rd ed . , new york , ny : wiley , 1994 p. chabert , j .- l .
raimbault , j .-
m . rax , and m. a. lieberman , phys .
plasmas * 11 * , 1775 ( 2004 ) p. chabert , j. phys .
* 40 * , r63 ( 2007 ) z. chen , s. rauf , k. collins , j. appl
. phys . * 108 * , 073301 ( 2010 ) s. rauf , k. bera , k. collins , 2008 plasma sources sci . technol .
* 17 * 035003 ( 2008 ) y. yang and m.j .
kushner , plasma sources sci
. technol .
* 19 * , 055011 ( 2010 ) i. lee , d.b .
graves , and m.a .
lieberman , plasma sources sci .
* 17 * 015018 ( 2008 ) y .- r .
zhang , x. xu , s .- x .
zhao , a. bogaerts , y .- n .
wang , phys .
plasmas * 17 * 113512 ( 2010 ) c. g. darwin , philos . mag .
, * 39 * , no . 233 , 537 ( 1920 ) h. schmitz and r. grauer , j. of comp .
physics * 214 * , 738 ( 2006 ) d. w. hewett and j. k. boyd , j. comp .
70 * , 166 ( 1987 ) d. w. hewett , comp .
. comm . * 84 * , 243 ( 1994 ) m. r. gibbons and d. w. hewett , j. comp . phys .
* 120 * , 231 ( 1995 ) m. r. gibbons and d. w. hewett , j. comp .
phys . * 130 * , 54 ( 1997 ) p. a. raviart and e. sonnendrcker , numer .
math * 73 * , 329 ( 1996 ) t. b. krause , a. apte , p. j. morrison , phys .
plasmas * 14 * , 102112 ( 2007 ) c. w. nielson and h. r. lewis , methods comput
* 16 * , 367 ( 1976 ) r. p. brinkmann , j. phys .
phys . * 42 * 194009 ( 2009 ) d. j. cooperberg , phys .
plasmas , * 5 * , 853 ( 1998 ) | the darwin approximation is investigated for its possible use in simulation of electromagnetic effects in large size , high frequency capacitively coupled discharges .
the approximation is utilized within the framework of two different fluid models which are applied to typical cases showing pronounced standing wave and skin effects . with the first model it is demonstrated that darwin approximation is valid for treatment of such effects in the range of parameters under consideration .
the second approach , a reduced nonlinear darwin approximation - based model , shows that the electromagnetic phenomena persist in a more realistic setting .
the darwin approximation offers a simple and efficient way of carrying out electromagnetic simulations as it removes the courant condition plaguing explicit electromagnetic algorithms and can be implemented as a straightforward modification of electrostatic algorithms .
the algorithm described here avoids iterative schemes needed for the divergence cleaning and represents a fast and efficient solver , which can be used in fluid and kinetic models for self - consistent description of technical plasmas exhibiting certain electromagnetic activity . |
theoretical stellar evolution models are of huge importance to stellar and galactic astrophysics because they are basically the only way of deriving the age , internal structure and composition of stars and galaxies from simple observational data ( e.g. meynet et al . 2005 ;
nordstrm et al .
the accuracy of the results depends on the predictive power of the models , but this is severely limited by the parameteric treatment of several effects , including convective core overshooting ( @xmath0 ) , the efficiency of convective energy transport ( @xmath1 ) , the effects of rotation , magnetic fields and mass loss , and the need to use theoretically - calculated opacities .
the main offender is @xmath0 , changes in which cause major modifications to the predicted lifetimes , chemical yields and luminosities of massive stars ( e.g. maeder & meynet 1989 ) .
this in turn has a large effect on the predicted properties and formation rates of evolved objects such as core - collapse supernovae ( eldridge & tout 2004 ) and black holes .
also , changes in @xmath1 can affect the derived ages of globular clusters by up to 2gyr ( chaboyer 1995 ) , whilst reasonable variation in @xmath0 can alter the derived ages of open clusters by a factor of two ( bragaglia & tosi 2006 ) .
it is clear that this must be improved .
theoretical models can be calibrated with observational constraints derived using detached eclipsing binaries ( debs ) ( e.g. pols et al .
1997 ; vandenberg et al .
the study of these objects allows us to accurately measure the masses , radii and @xmath2s of two stars of the same age and chemical composition ( andersen 1991 ) .
observational errors in mass and radius can be below 1% ( e.g. , southworth et al . 2004e , 2005b ) .
accurate @xmath2s can also be found , allowing luminosities and distances to be directly measured .
debs have been used to constrain @xmath0 ( andersen et al .
1990 ) and its dependence on mass ( ribas et al .
2000 ) ; @xmath1 was also studied by ludwig & salaris ( 1999 ) .
however , the stength of these results has been compromised by the lack of information available for each object . for a deb ,
the radii and effective temperatures are known for two specific masses .
when comparing to theory , two of these four datapoints ( usually the radii ) are required to deduce the age and metal abundance of the system .
the remaining two are generally not accurate enough to do anything more than indicate the helium abundance or , if you assume a normal helium abundance , perhaps the extent of convective overshooting .
further information is needed to derive accurate constraints on @xmath0 or @xmath1 .
one possibility is to measure the surface metal abundances of the stars from high - dispersion spectroscopy , but this is rarely done .
another possibility is to study debs which display apsidal motion ( rotation of the oritentation of an eccentric orbit due to tidal and relativistic effects ; claret & willems 2002 ) but these are less common and require observations over a long period of time .
.[eboc ] parameters of well - studied debs in open clusters . only systems with masses and radii known to within 5%
are included , sorted by decreasing primary star mass . [ cols="<,<,^,^,^,<",options="header " , ] the study of eclipsing binaries in stellar clusters allows more constraints to be placed on theoretical models ( e.g. thompson et al .
2000 ; lebreton et al .
as the cluster stars , including both components of the deb , were born in the same place at the same time , theoretical predictions must be able to match their observed properties for a single age and chemical composition . as a useful byproduct ,
the age and composition of the cluster can be derived from the theoretical models ( e.g. southworth et al .
2004ad ) and its distance can be measured empirically from the properties of the deb ( e.g. southworth et al .
fundamentally , this method requires a simultaneous fit of the morphology of the cluster in colour - magnitude diagrams and the properties of the deb in mass - radius - temperature diagrams .
we are currently undertaking a research project to study debs which are members of open clusters with a range of ages and metallicities , in order to study how @xmath0 and other theoretical model parameters vary with mass and composition . whilst results
so far have been very encouraging , it has still not been possible to match a well - studied deb to a well - studied cluster ; work to remedy this is in progress .
one concept which has , however , come to light is that studying several debs in open cluster allows you to put four or more stars with the same age and composition but with differing masses into one mass - radius or mass - temperature plot ( southworth et al .
this raises the possibility of obtaining `` standard sequences '' of eclipsing stars in one cluster with accurately known masses , radii , @xmath2s and surface gravities . to this end
we have obtained photometry and spectroscopy of the eclipsing binaries v1481cyg , v2261cyg and v2263cyg in the young open cluster ngc7128 .
preliminary light curves for these are shown in figs .
[ fig1 ] and [ fig2 ] .
preliminary ( no debiassing or flat - fielding ) strmgren @xmath3 light curve of the eclipsing binary v2263cyg in ngc7128 ( @xmath4 is at the top and @xmath5 at the bottom ) .
as it is semi - detached it can not be used to constrain single - star evolutionary theory , but it can provide an accurate distance to the cluster . ] | detached eclipsing binaries are very useful objects for calibrating theoretical stellar models and checking their predictions .
detached eclipsing binaries in open clusters are particularly important because of the additional constraints on their age and chemical composition from their membership of the cluster .
i compile a list containing absolute parameters of well - studied eclipsing binaries in open clusters , and present new observational data on the b - type systems v1481cyg and v2263cyg which are members of the young open cluster ngc7128 . |
more than 30 years ago efimov realized that a three - body system could have a large number of bound states when two or three of the two - body subsystems simultaneously have ( virtual or bound ) @xmath1-states sufficiently close to zero energy @xcite .
this effect has been discussed in a number of subsequent publications @xcite .
external fields can be used to tune the effective two - body interaction aiming at approaching the zero energy condition @xcite .
the picture describing the efimov effect is that one particle effectively has large - distance interaction simultaneously with both the other two particles building up a coherent wave function fully exploiting the interactions .
when only one scattering length is large the effect does not occur since the effective large - distance interactions only involve two of the particles while the third can avoid contributing by being far away without interacting .
the atomic helium trimer system is usually considered to be the most promising candidate for naturally occurring efimov states @xcite .
it has so far escaped direct experimental detection @xcite .
other systems may however be much better candidates possibly with many efimov states .
the purpose of this letter is to pin - point both the class of best suited candidates and the crucial properties optimizing the occurrence conditions .
the hyperspheric adiabatic expansion combined with the faddeev decomposition of the wave function has proven very efficient for investigations of weakly bound and spatially extended three - body systems @xcite .
efimov states can then be computed and without loss of generality even with the simplifying restriction that the potentials act only on @xmath1-waves which are the only contributors at the asymptotic large distances @xcite .
the hyperradius @xmath2 defined by @xmath3 where @xmath4 is the mass , @xmath5 the coordinate of particle @xmath6 and @xmath7 is an arbitray normalization mass . thus @xmath2 is a measure of the average size of the system . after solving the eigenvalue problem for the remaining ( angular )
coordinates the method provides a hyperradial equation with mass @xmath7 and an effective potential @xmath8 , which for intermediate distances of @xmath2 has the simple form @xmath9 where @xmath10 is a positive or negative constant depending on the interactions and the average effective range @xmath11 and average scattering length @xmath12 will be defined later . this potential has the generic form for the efimov states . at distances smaller or comparable with the interaction ranges
@xmath8 is more complicated and generally without divergence as @xmath0 .
this small distance region provides the scale for the energies of the possible efimov states and is otherwise completely unimportant for spatially extended states . at distances larger than the scattering lengths
the potential has the form @xcite @xmath13 where the reduced mass is @xmath14 , the @xmath1-wave scattering length is @xmath15 for system @xmath16 , and the average scattering length is defined as @xmath17 when all three particles are identical the expression for @xmath12 reduces to the common value of @xmath15 .
the efimov states are found at intermediate distances as solutions corresponding to the potential @xmath8 in eq.([e20 ] ) .
the wave function is @xmath18 , where @xmath19 is the modified bessel function with an imaginary index and the binding energy is @xmath20 . by expansion of @xmath19 for small and large values of @xmath2
the radial wave function @xmath21 is @xcite @xmath22 where the zero point for the first oscillation in @xmath2 is assumed to be @xmath23 , which in analogy to eq.([e40 ] ) roughly could be defined as @xmath24 reducing to the common two - body effective radius @xmath25 when all particles are identical .
the form of the wave function in eq.([e50 ] ) is easily found by confirming that @xmath26 is a solution to the schdinger equation with the potential in eq.([e20 ] ) valid at intermediate distances where the energy term can be neglected .
the exponential decrease at large distance occurs for all bound states for distances @xmath27 larger than 1 , where the effective radial potential falls off at least as fast as @xmath28 .
thus @xmath21 oscillates periodically as @xmath29 .
the number of efimov states @xmath30 corresponds then to the number of oscillations between the average interaction range @xmath23 and @xmath31 , i.e. @xmath32 the energies and sizes of these states are related by @xmath33 which reveals the exponential increase of sizes towards infinity and decrease of energies towards zero , respectively .
this behavior originates from the generic @xmath34 potential in eq .
( [ e20 ] ) .
the all decisive parameter @xmath35 must now be determined for systems where efimov states may occur .
at least two of the three scattering lengths must be large .
the formal analysis in @xcite leads to trancendental equations for @xmath35 . for identical bosons we obtain the usual efimov equation rewritten with real quantities , i.e. @xmath36 with the solution @xmath37 .
the result is independent of the mass of the particles . for non - identical bosons , still when all three scattering lengths are large ,
we find instead @xmath38 where @xmath39 and @xmath40 @xmath41 in general the solution
@xmath35 then depends on two ratios of masses , e.g. @xmath42 and @xmath43 . when all masses are equal ( @xmath44 , @xmath45 ) eq .
( [ e90 ] ) reduces to three simpler equations @xmath46 where 2 on the right hand side produce eq .
( [ e80 ] ) while @xmath47 is the only real solution for @xmath48 . when only the two scattering lengths @xmath49 and @xmath15 ( not @xmath50 ) are large the efimov equation for @xmath35 becomes @xmath51 where the solution now only depends on the angle @xmath52 varying between 0 and @xmath53 .
still @xmath52 in eq .
( [ e110 ] ) in turn depends on the above two mass ratios .
the conclusion is that the strength of the potential has to be found by solving eq .
( [ e90 ] ) when all scattering lengths are large and eq .
( [ e120 ] ) when only two scattering lengths are large .
the efimov effect does not occur when only one scattering length is large .
( [ e80 ] ) is the limit of eq .
( [ e90 ] ) when all masses are equal .
the parameters are the three angles @xmath52 , which through eq .
( [ e110 ] ) are functions of two independent mass ratios .
we show in fig .
[ fig1 ] the 3d plot of @xmath35 as function of these independent parameters .
the smallest values slightly above 1 are found in a valley passing the symmetric point of three equal masses and extending towards one very large mass and one moderate mass ratio around unity .
this case with two light masses and one heavy mass corresponds to the smallest @xmath35 where the efimov effect is least pronounced . on the other hand large values of @xmath35 extending to infinity
are obtained for two heavy and one light mass .
this is seen in fig .
[ fig1 ] when both coordinates are large ( @xmath54 , @xmath55 , @xmath56 ) , and along both axes when the other coordinate is very small ( @xmath57 and @xmath58 ) .
thus moderate values of @xmath35 arise for symmetric systems while exceedingly large strengths are possible for asymmetric systems . in all cases
the assumption is that all three two - body subsystems simultaneously have an @xmath1-state close to the threshold of binding .
this is most likely for a symmetric system of identical bosons where only one interaction is involved . for non - indentical particles two tuned subsystems
is the least demanding to exhibit the efimov effect .
if furthermore two of the clusters are identical particles only one independent requirement of an @xmath1-state close to zero is left . in this case where only two subsystems contribute ( two large scattering lengths ) the strength @xmath35 is found from eq .
( [ e120 ] ) as function of the angle @xmath52 corresponding to the subsystem with the small third scattering length .
the result displayed in fig .
[ fig2 ] show the divergence as @xmath59 for @xmath60 and the linear convergence to zero as @xmath61 for @xmath62 . when all masses are equal @xmath63 and eq .
( [ e120 ] ) reduces to eq .
( [ e80 ] ) with the left hand side divided by two .
the resulting solution @xmath64 is roughly two times smaller than 1.0063 obtained when all three subsystems contribute .
the angle @xmath52 is a function of two mass ratios .
when @xmath65 is much smaller than both the other masses @xmath52 approaches zero and @xmath35 is very large .
when @xmath65 is much larger than at least one of the other masses @xmath52 approaches @xmath66 and @xmath35 is very small .
thus extremely large @xmath35 is only possible for two contributing subsystems when the particle @xmath67 related to both these subsystems has a comparatively small mass .
the efimov effect occurs in all cases but is only pronounced for large @xmath35 . in fig .
[ fig3 ] we show solutions to eq .
( [ e90 ] ) corresponding to three very large scattering lengths for a series of different mass ratios .
the striking features are as described in connection with fig .
[ fig1 ] that @xmath35 is very small for one heavy and two light particles , relatively small for similar masses and huge for one light and two heavy particles . comparing figs .
[ fig2 ] and [ fig3 ] we conclude that the mass ratios determine the order of magnitude of the strength @xmath35 whereas the contribution from the third subsystem is marginal .
large values of @xmath35 favor occurrence of ( many ) efimov states , see eq .
( [ e60 ] ) . it should be emphasized that at least two scattering lengths must be sufficiently large to allow the effect in the first place .
an interaction in one two - body subsystem of longer range than the generic @xmath0 potential prohibits occurrence . for nuclei
this leaves only two neutrons combined with a charged ordinary nucleus @xcite .
there might be a chance to produce one efimov state but the second would not appear as illustrated by @xmath68li ( @xmath69li@xmath70 ) where @xmath71 according to eq .
( [ e70 ] ) corresponds to an increase of the radius by a factor of @xmath72 . in atomic and molecular physics
the possibilities are much better . for identical particles
@xmath35 is mass independent .
the molecular prototype is the atomic helium trimer @xmath73he@xmath74 with an excited efimov state of binding energy 0.18@xmath75ev and radius about 50 . correspondingly @xmath76 and the radius increase between neighboring efimov states is a factor of 22 .
equal masses and only two contributing subsystems change this factor to about 542 implying that the radius of the second efimov state then would exceed 1 @xmath77 m . by far
the most favorable case is one light and two heavy identical particles .
one small step in this direction is the asymmetric helium trimer @xmath78he@xmath73he@xmath79 with one bound ( pronounced halo ) state of binding around 1 @xmath77ev and radius about 13 @xcite . substituting one helium atom by an alkali atom reduces the two - body binding energy @xcite and produce spatially extended three - body systems like @xmath80li@xmath73he@xmath81he and @xmath82na@xmath78he@xmath79 with @xmath83 @xcite . in @xmath84li@xmath73he@xmath81he we anticipate an efimov state with a size larger than the 50 expected in @xmath73he@xmath74 .
other combinations with one helium and two alkali atoms are easily conjectured from the results in @xcite .
for example @xmath78he@xmath82na@xmath79 and @xmath78he@xmath85cs@xmath79 give @xmath86 implying that the second efimov state only is 13 , 3.1 times larger than the first .
another type of combinations are even more favorable .
one charge is allowed , since the destructive coulomb - like long - range interaction still is not present .
one electron and two identical atoms ( or molecules ) then maximize the @xmath35-value .
it is sufficient with a large scattering length for the two electron - atom ( molecule ) systems and if also the atom - atom ( molecule - molecule ) contributes the effect could be even larger
. then @xmath35 could be as large as 100 ( electron - atom ( or molecule ) mass ratio @xmath87 ) and the radius increase per state from eq .
( [ e70 ] ) could be less than 5 % .
the number of efimov states within practical reach would increase dramatically while they still remain relatively stable @xcite .
however , this scenario requires a large electron - atom scattering length or equivalently an @xmath1-state energy very close to zero . since the scattering length @xmath88
at least should be larger than 5 - 10 the electron binding @xmath89 should at least be smaller than 0.1 ev .
this is an established fact for the bound negative ions ca , ti and sr and furthermore a number of atoms ( he , be , n , ne , mg , ar , mn , zn , kr , cd , xe , hg , rn ) can not bind an electron although the distance from the threshold in general is unknown @xcite .
the three - body system should now consist of the electron plus a pair of the above atoms chosen as close as possible to the threshold of @xmath1-state binding . here
@xmath73he@xmath90he is the obvious combination but unbound with respect to @xmath73he@xmath79 leaving the possible efimov states as excited states with a large decay probability .
substituting @xmath84li for one of the helium atoms reduces the atomic binding and @xmath73he@xmath91li is then a promising candidate . other examples like
mg@xmath92 mg , mg@xmath92ne , mg@xmath92ar , mg@xmath92kr and mg@xmath92xe were investigated in @xcite .
adding an electron to the weakly bound two - body systems ar - ar , cd - cd , cl - xe , cs - hg , he - hg , hg - hg @xcite , also present efimov candidates , but pairs like h - h , ca - ca , li - li , na - na could be more suitable . instead of atoms we could try to use molecules with very small electron binding , combining them pairwise and adding an electron , see @xcite for candidates .
if this should create efimov states the internal structure of the molecules is not allowed to change within the pair .
this might be possible but requires a separate investigation .
the efimov conditions of two zero energy subsystems can be fulfilled without lower lying bound two - body states , but it is much more probable to encounter systems where the state of zero energy is an excited two - body state .
the resulting efimov states are clearly more unstable and their structure more difficult to study directly .
however , signals should show up in scattering experiments @xcite .
the properties of the efimov states are in any case determined by the scattering lengths and the size of @xmath35 as discussed above .
the strength of the attraction in the effective hyperradial @xmath0 potential for three particles is for large scattering lengths determined by simple equations depending only on two independent mass ratios .
we survey this mass dependence and provide information about favorable mass combinations .
the larger the strength the more pronounced is the efimov effect .
the density of efimov states increases with the square root of the strength which for existing mass combinations can vary by several orders of magnitude .
two contributing subsystems for one light and two heavy particles is much more favorable than three contributing subsystems with similar masses .
careful choices of the three particles can then simultaneously optimize the efimov conditions of large strength and large scattering lengths .
this may allow detection of several or many members in a series of efimov states or alternatively allow detection of the first efimov state in cases when the conditions are less well fulfilled .
we suggest a number of possible combinations of an electron and two neutral atoms or molecules , or alternatively one light atom and two heavy atoms or molecules . | the conditions for occurrence of the efimov effect is briefly described using hyperspherical coordinates .
the strength of the effective hyperradial @xmath0 potential appearing for two or three large scattering lengths is computed and discussed as function of two independent mass ratios of the three constituent particles .
the effect is by far most pronounced for asymmetric systems with three very different masses .
one efimov state may by chance appear in nuclei .
many states could be present for systems with one electron and two neutral atoms or molecules .
estimates of the number of states and their sizes and energies are given . |
finite group theorists have established formulas that enable expression of interesting properties of a group @xmath0 in terms of quantitative statements on sums of values of its characters .
there are many examples @xcite .
we describe a representative one .
consider the commutator map@xmath1 : g\times g\rightarrow g;\text { \ \ } [ x , y]=xyx^{-1}y^{-1 } , \label{cm}\]]and for @xmath2 denote by @xmath3 _ { g}$ ] the set @xmath3 _ { g}=\{(x , y)\in g\times g;$ ] @xmath4=g\}.$ ] in @xcite ore conjectured that for a finite non - commutative simple group @xmath0 the map ( [ cm ] ) is onto , i.e. , @xmath5 _
{ g}\neq 0,$ ] for every @xmath6 the quantity @xmath5 _ { g}$ ] is a class function on @xmath0 and frobenius developed the formula for its expansion as a linear combination of irreducible characters .
frobenius formula is @xmath7_{g}}{\#g}=1+\sum_{1\neq \rho \in irr(g)}\frac{\chi _ { \rho } ( g)}{\dim ( \rho ) } , \label{o - sum}\]]where for @xmath8 in the set @xmath9 of isomorphism classes of irreducible representations aka irreps of @xmath10 we use the symbol @xmath11 to denote its character . estimating the sum in the right - hand side of ( o - sum ) for certain classes of elements in several important finite classical groups was a major technical step in the recent proof @xcite of the ore conjecture . given the ore conjecture thus , the following question naturally arises : * question .
* what is the distribution of the commutator map ( [ cm ] ) ? in @xcite shalev conjectured that for a finite non - commutative simple group @xmath0 the distribution of ( [ cm ] ) is approximately uniform , i.e. , @xmath12 in a well defined quantitative sense ( i.e. , in a family of groups of growing size ) .
this conjecture is wide open @xcite .
it can be proven for the finite symplectic group @xmath13 is a power of an odd prime number @xmath14 . ] invoking its explicit character table , and probably also for @xmath15 @xcite . as was noted by shalev in @xcite ,
one can verify the uniformity conjecture for elements of @xmath0 with small centralizers , using schur s orthogonality relations for characters @xcite .
however , relatively little seems to be known about shalev s conjecture in the case of elements with relatively large centralizers see figure ore for numerical illustration in the case of @xmath16 and the transvection element @xmath17 in @xmath0 which is given by @xmath18to suggest a possible approach for the resolution of the uniformity conjecture , let us reinterpret ( [ c - s ] ) as a statement about extensive cancellation between the terms@xmath19which are called the dimensions of the irreducible representations of a finite group @xmath0 tend to come in certain layers according to order of magnitude .
for example see figure [ dim - irrsp6_5 ] for illustration it is known dl , lu1 that the dimensions of the irreducible representations of @xmath16 are given by some `` universal '' set of polynomials in @xmath20 in this case the degrees of these polynomials give a natural partition of @xmath21 according to order of magnitude of dimensions@xmath22 since the dimension of the representation of a group @xmath0 is what appears in the denominator of ( [ c - r ] ) , it seems reasonable to expect that in ( [ c - s ] ) * ( a ) * _ character ratios of lower dimensional representations tend to contribute larger terms_. * ( b ) * _ the partial sums over low dimensional representations of `` similar '' size exhibit cancellations . _ a significant amount of numerical data collected recently with cannon and goldstein supports assertions ( a ) and ( b ) .
for example , in figure [ cr - t - sp6_5 ] we plot the numerical values of the character ratios of the irreducible representations of @xmath23 evaluated at the transvection @xmath17 ( [ tr ] ) .
more precisely , for each @xmath24 we marked by a circle the point the nearest integer value to a real number @xmath25 @xmath26 and find that the overall picture is in agreement with ( a ) and ( b )
. moreover see figures [ dim - irrsp6_5 ] and cr - t - sp6_5 for illustration the numerics shows that , although the majority of representations are `` large '' , their character ratios tend to be so small that adding all of them contributes very little to the entire ore sum ( [ c - s ] ) .
the above example illustrates that a possible obstacle to getting group theoretical properties by summing over characters , as in formula ( [ o - sum ] ) , is lack of control over the representations with relatively small dimensions .
in particular , it seems that a systematic knowledge on the * `` small '' * representations of finite classical groups could lead to proofs of some important open conjectures , which are currently out of reach .
however , relatively little seems to be known about these small representations @xcite . in this note
we develop a language suggesting that the small representations of the finite classical groups can be systematically described by studying their restrictions to unipotent subgroups , and especially , using the notion of * rank * of a representation @xcite .
in addition , we develop a new method , called the " * * eta correspondence " * * _ _ , _ _ to construct small representations .
we conjecture that our construction is exhaustive .
finally , we use our construction to give a strong estimate on the dimension of the small representations in terms of their rank . for the sake of clarity of exposition we treat in this note only the case of the finite symplectic groups @xmath27 .
* acknowledgement .
* we would like to thank john cannon and steve goldstein for the help in generating the numerical data appearing in this note .
we are grateful to bob guralnick for important conversations .
thanks go also to the following institutions : mpi bonn , texas a&m , uw - madison , weizmann institute , and yale , where the work on this note was carried out .
finally , s.g .
thanks the uw - madison for choosing him as a vilas associate for the period of working on this project .
let us start with the numerical example of the dimensions of the irreducible representations of the group @xmath28 the beginning of the list appears in figure [ numerical - dim - irrsp6_5 ] .
these numbers see also figure [ dim - irrsp6_5]reveal the story of the hierarchy in the world of representations of finite classical groups .
a lot of useful information is available on the `` minimal '' representations of these groups , i.e. , the ones of lowest dimensions @xcite . in the case of @xmath27
these are the @xmath29 components of the oscillator ( aka weil ) representations _ _
_ _ @xcite , @xmath30 of dimension @xmath31 and @xmath30 of dimension @xmath32 which in figure [ numerical - dim - irrsp6_5 ] are the ones of dimensions @xmath33 .
in addition , a lot is known about the `` big '' representations of the finite classical groups , i.e. , those of considerably large dimension ( see @xcite and references therein ) .
we will not attempt to define the `` big '' representations at this stage , but in figure [ numerical - dim - irrsp6_5 ] the ones of dimension @xmath34 and above fall in that category .
however , relatively little seems to be known about the representations of the classical groups which are in the range between `` minimal '' and `` big '' @xcite . in figure numerical - dim - irrsp6_5 those form the layer of @xmath35 representations of dimensions between @xmath36 and @xmath37 in this section we introduce a language that will enable us to define the `` small '' representations of finite classical groups .
this language will extend well beyond the notion of minimal representations and will induce a partition of the set of isomorphism classes of irreducible representations which is closely related to the hierarchy afforded by dimension .
in particular , this language gives an explicit organization of all the representations in figure [ numerical - dim - irrsp6_5 ] , and explains why this list is , in a suitable sense , complete .
the key idea we will use is that of the _ rank _ of a representation .
this notion was developed in the 1980s by howe , in the context of unitary representations of classical groups over local fields @xcite , but it has not been applied to finite groups . for the sake of clarity of exposition , in this note we give the definition of rank only in the symplectic case , leaving the more general treatment to future publication .
we start by discussing necessary ingredients from the structure theory of the symplectic group .
let @xmath38 be a @xmath39-dimensional symplectic vector space over the finite field @xmath40 in order to simplify certain formulas , let us assume that @xmath41where @xmath42 and @xmath43 are vector spaces dual to each other with pairing @xmath44 , and that the symplectic form @xmath45 is the natural one which is defined by that pairing , i.e. , @xmath46 note that @xmath42 and @xmath43 are maximal isotropic aka _ lagrangian_subspaces of @xmath47 consider the symplectic group @xmath48 of elements of @xmath49 which preserve the form @xmath50 denote by @xmath51 the subgroup of all elements in @xmath52 that preserve @xmath42__. _ _ the group @xmath53 is called the _
siegel parabolic _ @xcite and can be described explicitly in terms of the decomposition ( [ ld ] ) @xmath54where @xmath55 is the inverse of the transpose of @xmath56 in particular , @xmath53 has the form _ _ _ _ of a semi - direct product , known also as its levi decomposition @xcite , @xmath57where @xmath58 , called the _ unipotent radical _ of @xmath53 , is the normal subgroup @xmath59the group @xmath60 is abelian and we have a tautological @xmath61-equivariant isomorphism @xmath62where @xmath63 denotes the space of symmetric bilinear forms on @xmath64 and the @xmath61 action on @xmath63 is the standard one .
in addition , if we fix a non - trivial additive character @xmath65 of @xmath66 we obtain a @xmath61-equivariant isomorphism@xmath67where @xmath68 denotes the space of symmetric bilinear forms on @xmath69 the @xmath61 action on @xmath68 is the standard one , the symbol @xmath70 stands for the pontryagin dual ( group of characters ) of @xmath71 and @xmath72 for every @xmath73 where @xmath74 indicates the trace of the composite operator @xmath75 now , take an irreducible representation @xmath8 of @xmath52 and look at the restriction to @xmath60 .
it decomposes @xcite as a sum of characters with certain multiplicities@xmath76the function @xmath77 and its support will be called , respectively , the of @xmath8 , and the of @xmath78 and will be denoted by @xmath79 and @xmath80 respectively .
we would like to organize the decomposition ( [ ft1 ] ) in a more meaningful way .
note that the restriction to @xmath60 of a representation @xmath8 of @xmath52 can be thought of as the restriction to @xmath60 of the restriction of @xmath8 to @xmath53 . using ( [ ngl ] ) , this implies @xcite : the @xmath60-spectrum of a representation @xmath8 of @xmath52 is @xmath61 invariant .
that is , @xmath81 if @xmath82 and @xmath83 define equivalent symmetric bilinear forms on @xmath42 . the first major invariant of a symmetric bilinear form is its rank .
it is well known @xcite that , over finite fields of odd characteristics , there are just two isomorphism classes of symmetric bilinear forms of a given rank @xmath84 they are classified by their discriminant @xcite , which is an element in @xmath85 .
we denote by @xmath86 and @xmath87 , the two classes of symmetric bilinear forms , these whose discriminant is the coset of squares , and these whose coset consists of non - squares , respectively ; or we will denote the pair of them , or whichever one is relevant in a given context as @xmath88 if @xmath82 is a form of rank @xmath89 , we will also say that the associated character @xmath90 has rank @xmath89 .
we may also refer to the character as being of type @xmath91 or type @xmath92 , according to the type of @xmath82 . with this notation
, we can reorganize the expansion ( [ ft1 ] ) of @xmath93 namely , we split the sum according to the ranks of characters , and within each rank we split the sum into two partial sub - sums according to the two isomorphism classes of the associated forms:@xmath94note that , formula ( [ ft2 ] ) implies , by evaluation at the identity of @xmath60 , that the dimension of @xmath8 must [email protected] .
, a weighted sum of the cardinalities @xmath96 of the isomorphism classes of symmetric bilinear forms .
it is easy to write formulas for these cardinalities @xcite .
we have@xmath97where @xmath98 denotes the grassmannian of @xmath89-dimensional subspaces of @xmath99 , the symbol @xmath100 stands for the group of automorphisms of @xmath101 and @xmath102 is the isometry group of a non - degenerate form of type @xmath103 on @xmath101 i.e. , it is @xmath104 in case of a form from @xmath86 and likewise with @xmath91 and @xmath92 switched .
in particular , using standard formulas @xcite for @xmath105 @xmath106 and @xmath107 we obtain @xmath108 from ( [ card2 ] ) we get , in particular , that the smallest non - trivial orbits are those of rank one forms . using ( [ card1 ] ) we see that these have size @xmath109 it follows from this that the smallest possible dimension of a non - trivial irreducible representation @xmath8 of @xmath52 should satisfy@xmath110indeed , we have the following lemma : [ trivial]the only irreducible representation of @xmath52 with @xmath60-spectrum concentrated at zero is the trivial one .
the proof of lemma [ trivial ] is easy , but to avoid interrupting this discussion , we defer it to appendix [ p - of - lemma - trivial ] . a representation whose dimension attaining the lower bound ( [ lowestdim ] ) would contain each rank one character of one type , and nothing else . since @xmath111 is such a small subgroup of @xmath52 , it is unclear whether to expect such a representation to exist . in particular , it would be irreducible already on the siegel parabolic @xmath53 , and it would be the smallest possible faithful representation of @xmath53 .
it turns out , however , that it does exist ; in fact , there are two @xcite .
[ smallest]there are two irreducible representations of @xmath52 of dimension @xmath112 , one containing either one of the two rank one @xmath61 orbits in @xmath113 what is the next largest possible dimension ?
well , one more - the @xmath60-spectrum could include a rank one orbit , and a trivial representation .
it turns out that these also exist @xcite .
[ next]there are two irreducible representations of @xmath52 of dimension @xmath114 one whose @xmath60-spectrum contains one of the rank one orbits in @xmath113 for a proof of propositions [ smallest ] and [ next ] , see section smallest rep .
the existence of the above smallest possible representations , plus considerations of tensor products , tell us that , for any orbit @xmath115 in @xmath70 there will be representations of @xmath52 whose @xmath60-spectrum contains the given orbit , together with orbits of smaller rank . since the size see formulas ( [ card1 ] ) and
( [ card2])of the orbits @xmath115 is increasing rapidly with @xmath116 , representations whose @xmath60-spectrum is concentrated on orbits of smaller rank can be expected to have smaller dimensions .
this motivates us to introduce the following key notion in our approach for small representations .
let @xmath8 be a representation of @xmath117 1 .
we say that @xmath8 is of @xmath118 denoted @xmath119 iff the restriction @xmath120 contains characters of rank @xmath116 , but of no higher rank .
if @xmath8 is of rank @xmath116 and contains characters of type @xmath121 but not of type @xmath122 , then we say that @xmath8 is of @xmath123 ; and likewise with @xmath91 and @xmath92 switched .
let us convey some intuition for this notion using numerical data obtained for the irreducible representations of the group @xmath124see figure [ ranksp6_5 ] .
the computations of the multiplicities and rank in this case reveal a striking compatibility with the families of representations appearing in the list of figure [ numerical - dim - irrsp6_5 ] .
for example , it shows that the trivial representation is the one with rank @xmath125 the @xmath29 components of the two oscillator representations are those of rank @xmath126 and they split into @xmath30 of type @xmath127 and @xmath30 of type @xmath128 the @xmath35 representations of dimensions between @xmath129 and @xmath130 are the ones of rank @xmath131 and they split into @xmath132 of type @xmath133 and @xmath134 of type @xmath135 ; and above that the `` big '' representations are those with rank @xmath136 the main quest now is for a systematic construction of the `` low rank '' irreducible representations . in the next section
we take the first step toward that goal by treating the smallest non - trivial representations of @xmath52 which are of rank @xmath126see propositions [ smallest ] and [ next ] .
where do the smallest representations of @xmath52 come from ?
a conceptual answer to this question was given by weil in @xcite .
they can be found by considering the heisenberg _ _ _ _ group .
the _ heisenberg _ group attached to @xmath38 is a two - step nilpotent group that can be realized by the set@xmath137with the group law@xmath138 in particular , the center @xmath139 of the heisenberg group @xmath140is equal to its commutator subgroup .
moreover , the commutator operation in @xmath141 induces a skew - symmetric bilinear form on @xmath142 that coincides with the original symplectic form .
the group @xmath143 is the analog over a finite field of the lie group associated with the canonical commutation relations ( ccr ) of werner heisenberg , of uncertainty principle fame .
we would like to describe the representation theory , i.e. , the irreducible representations , of the heisenberg group .
this theory is simultaneously simple and deep , with fundamental connections to a wide range of areas in mathematics and its applications .
take an irreducible representation @xmath144 of @xmath143 .
then , by schur s lemma , the center @xmath139 will act by scalars@xmath145where @xmath146 is the the identity operator on the representation space of @xmath144 , and @xmath147 is a character of @xmath148 called the _ central character _ of @xmath144 .
if @xmath149 then @xmath144 factors through @xmath142 , which is abelian , so @xmath144 is itself a character of @xmath47 the case of non - trivial central character is described by the following celebrated theorem @xcite : up to equivalence , there is a unique irreducible representation @xmath150 with given non - trivial central character @xmath65 in @xmath151 we will call the ( isomorphism class of the ) representation @xmath152 the _ heisenberg representation _ associated to the central character @xmath153 there are many ways to realize ( i.e. , to write explicit formulas for ) @xmath154 @xcite . in particular , it can be constructed as induced representation from any character extending @xmath65 to any maximal abelian subgroup of @xmath143 @xcite . to have a concrete one , note that the inverse image in @xmath143 of any lagrangian subspace of @xmath155 will be a maximal abelian subgroup for which we can naturally extend the character @xmath65 .
for example , consider the lagrangian @xmath156 and the associated maximal abelian subgroup @xmath157 with character @xmath158 on it , given by @xmath159then we have the explicit realization of @xmath160 given by the action of @xmath161 by right translations , on the space @xmath162 in particular , we have @xmath163 a compelling property of the heisenberg group is that it has a large automorphism group . in particular , the action of @xmath52 on @xmath155 lifts to an action on @xmath143 by automorphisms leaving the center point - wise fixed .
the precise formula is @xmath164 @xmath165 it follows from the _ stone von neumann
mackey _ theorem , that the induced action of @xmath52 on the set @xmath166 will leave fixed each isomorphism class @xmath160 @xmath167 this means that , if we fix a vector space @xmath168 realizing @xmath160 then for each @xmath169 in @xmath52 there is an operator @xmath170 which acts on space @xmath168 and satisfies the equation @xmath171which is also known as the exact _ egorov identity _
@xcite in the mathematical physics literature .
note that , by schur s lemma , the operator @xmath170 is defined by ( [ egorov ] ) up to scalar multiples .
this implies that for any @xmath172 we have @xmath173 where @xmath174 is an appropriate complex number of absolute value @xmath175 .
it is well known ( see @xcite for explicit formulas ) that over finite fields of odd characteristic this mapping can be lifted to a genuine representation .
[ or]there exists and @xmath176 where still there is a canonical lift @xcite . ] a representation @xmath177that satisfies the egorov identity ( [ egorov ] ) .
we will call @xmath178 the _ oscillator _ representation .
this is a name that was given to this representation in @xcite due to its origin in physics @xcite .
another popular name for @xmath178 is the _
weil _ representation , following the influential paper @xcite .
[ schrodinger model]we would like to have some useful formulas for the representation @xmath178 .
note that the space ( [ ind ] ) is naturally identified with@xmath179on the space ( [ space ] ) we realize the representation @xmath178 .
this realization is sometime called the _ schrdinger model . _ in particular , in that model for every @xmath180 we have @xcite \(a ) @xmath181 ( y)=\psi ( \frac{1}{2}a(y , y))f(y),\medskip $ ] where @xmath182 is symmetric ; \(b ) @xmath183 ( y)=\frac{1}{\gamma ( b,\psi ) } \sum\limits_{y^{\prime } \in y}\psi ( b(y , y^{\prime } ) ) f(y^{\prime } ) , \medskip $ ] where @xmath184 is symmetric , and @xmath185 the quadratic gauss sum ; \(c ) @xmath186 ( y)={\genfrac{(}{)}{}{}{\det ( c)}{q}}f(c^{-1}y),\medskip $ ] where @xmath187 @xmath188 its transpose inverse , and @xmath189 is the legendre symbol the legendre symbol @xmath190 or @xmath191 according to @xmath192 being a square or not , respectively@xmath22 ] .
it turns out that the isomorphism class of @xmath178 does change when varying the central character @xmath65 in @xmath194 however , this dependence is weak .
the following result indicates that there are only two possible oscillator representations . for a character @xmath65 in @xmath195 denote by @xmath196 @xmath197 the character @xmath198 [ oscillators]we
have @xmath199 iff @xmath200 for some @xmath201 for a proof of proposition [ oscillators ] , see appendix p - of - proposition - oscillators .
using formula ( a ) given in remark [ schrodinger model ] , it is easy to determine the rank of the oscillator representation .
each representation @xmath178 is of rank @xmath175 .
one isomorphism class is of type @xmath127 and the other is of type @xmath202 . in addition , the oscillator representations are slightly reducible .
the center @xmath203 acts on the representation @xmath178see remark [ schrodinger model ] for the explicit action of @xmath204 we have the direct sum decomposition @xmath205with @xmath206where @xmath207 is the subspace of `` even vectors '' , i.e. , vectors on which @xmath208 acts trivially , and @xmath209 is the subspace of `` odd vectors '' , i.e. , vectors on which @xmath208 acts via the sign character .
the above discussion also implies the following : the decomposition ( [ even - odd ] ) is the decomposition of @xmath178 into irreducible representations . to conclude , our study of the oscillator representation has established propositions [ smallest ] and [ next ] .
more precisely , the representations ( [ even - odd ] ) have rank one , they are of type @xmath210 , and have the required dimensions .
where do higher rank representations of @xmath52 come from ? * * * * this section will include an answer to this question in the regime of `` small '' representations .
more , precisely we give here a systematic construction of rank @xmath116 irreducible representations of @xmath52 in the so - called `` stable range '' _ _ _ _ @xmath211we will also refer to such representations as `` small '' or `` low rank '' representations .
let @xmath212 be a @xmath116-dimensional vector space over @xmath213 and let @xmath214 be an inner product ( i.e. , a non - degenerate symmetric bilinear form ) on @xmath212 .
the pair @xmath215 is called a _ quadratic space .
_ we denote by @xmath216 the isometry group of the form @xmath217 consider the vector space @xmath218the tensor product of @xmath155 and @xmath212 @xcite .
it has a natural structure of a symplectic space , with the symplectic form given by @xmath219 the groups @xmath48 and @xmath216 act on @xmath218 via their actions on the first and second factors , respectively,@xmath220both actions preserve the form @xmath221 and moreover the action of @xmath52 commutes with that of @xmath216 , and vice versa .
in particular , we have a map @xmath222which embeds the two factors @xmath52 and @xmath216 in @xmath223 and they form a pair of commuting subgroups . in fact , each is the full centralizer of the other inside @xmath224 thus , the pair @xmath225 forms what has been called in @xcite a _ dual pair _ of subgroups of @xmath226 .
we write down a specific model for the oscillator representation on the central character , but we record which symplectic group it belongs to . ]
@xmath227 of @xmath226 which is convenient for us when we consider the restriction of @xmath228 to the subgroups @xmath52 and @xmath216 .
let us identify the lagrangian subspace @xmath229 of @xmath218 with @xmath230 this enables to realize ( see remark schrodinger model ) the representation @xmath228 on the space of functions @xmath231 in this realization , the action of an element @xmath182 of the siegel unipotent @xmath60 of @xmath52 is given by @xmath232where for @xmath233 we denote by @xmath234 the quadratic form @xmath235and we denote by @xmath236 the trace of the composite operator @xmath237 in addition , in this model the action of an element @xmath238 is given by @xmath239 consider the oscillator representation @xmath228 of @xmath226 .
for the rest of this section we make the following choice .
if @xmath65 is the central character we use to define @xmath240 then the character @xmath241 @xmath242 of @xmath66 is the one we use in ( [ symxdual ] ) to identify @xmath68 and @xmath113 with this choice of parameters we can make the following precise statement : [ rank - as - rep - of - sp]assume that @xmath243 as a representation of @xmath244 @xmath228 is of rank @xmath116 and type form @xmath82 on @xmath43 is of type @xmath245 if @xmath246 is isometric to @xmath247 @xmath248 for a proof of proposition [ rank - as - rep - of - sp ] , see appendix p - of - proposition - rank - as - rep - of - sp .
now , consider the restriction , via the map ( [ embedding ] ) , of @xmath227 to the product @xmath249 we decompose this restriction into isotypic components for @xmath216:@xmath250where @xmath251 is a representation of @xmath117 although the factors @xmath251 in ( [ res - dp ] ) will in general not be irreducible , we can say something about how they decompose . let us denote by @xmath252the sets of ( equivalence classes of ) irreducible representations of @xmath52 of rank @xmath118 and of rank @xmath116 and type @xmath253 respectively .
the next theorem
the main result of this note announces that each @xmath254 has a certain largest `` chunk '' , which is in fact what we are searching for .
[ etac]assume that @xmath243 the following hold true : 1 . *
rank * @xmath116 * * piece . * * for each @xmath255 in @xmath256 the representation @xmath251 contains a unique irreducible constituent @xmath257 of rank @xmath116 ; all other constituents have rank less than @xmath116 .
* injection . *
the mapping @xmath258 gives an embedding @xmath259 3 .
[ spectrum]*spectrum . *
the multiplicity of the orbit @xmath245 in @xmath260 is @xmath261 for a proof of theorem [ etac ] , see appendix [ p - of - thm - etac ] .
it also seems that this construction should produce all of the rank @xmath116 representations .
we formulate this as a conjecture .
[ exhaustion]assume that @xmath243 we have@xmath262where @xmath263 and @xmath264 represent the two isomorphism classes of inner products of rank @xmath116 .
note that by ( [ eta ] ) the union in ( [ ex ] ) is indeed disjoint .
conjecture [ exhaustion ] is backed up by theoretical observations and numerical computations
see subsection [ numerics ] for illustration .
we give now several remarks that , in particular , will clarify the novelty of our main result , and will also explain why we decided to call ( [ eta ] ) the .
we would like to comment that ( a ) * eta vs. theta correspondence over local fields * : : _ considering the groups _ @xmath52 _ and _
@xmath216 _ over a local field , one can associate , in a similar fashion as above , to every irreducible representation _ @xmath255 _ of _ @xmath265 _ a representation _ @xmath251 of @xmath117 it will in general not be irreducible and the question is , what component to select from it ?
one option is to take the `` minimal '' piece of @xmath266 indeed , it turns out that @xmath251 has a unique irreducible quotient @xmath267 the assignment @xmath268is the famous theta correspondence , which has been studied by many authors ( * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) for its usefulness in the theory of automorphic forms .
a second option is to take the _
`` maximal '' _ piece of @xmath266 indeed , repeating in the local field case , verbatim , the scheme we proposed above , we find that @xmath251 has a largest chunk in the form of a unique irreducible sub - representation @xmath257 of rank @xmath116 , which will equal @xmath269 exactly when @xmath251 is irreducible .
the assignment @xmath270is our eta correspondence ( [ eta ] ) .
application of the new correspondence to representation theory of classical groups over local fields , will be a subject for future publications .
( b ) * eta correspondence over finite fields * : : as noted by several authors ( see @xcite , and in particular akt where the case of unipotent representations was considered ) the theta correspondence is not defined over finite fields . the eta correspondence comes as the appropriate construction in this case .
this is also the reason we use a related , although different , notation for the correspondence ( [ eta ] ) .
finally , we would like to make the following remark on the generality of our work .
the notion of rank for the group @xmath271 over local fields was described in @xcite .
the theory for general classical groups over local fields was developed by li in @xcite , and it was extended to all semi - simple algebraic groups over local fields by salmasian in @xcite .
the development of the eta correspondence ( [ eta ] ) for all finite classical groups will be discussed in future publications . for expositional purposes , in this note
we describe only the case of the finite symplectic groups .
conjecture [ exhaustion ] is backed up by numerical data collected for the groups @xmath272 @xmath273 ; @xmath274 , @xmath275 and @xmath276 indeed , the magma computations done with cannon and goldstein , for the various sizes of symplectic groups , repeatedly confirm the assertion made in the exhaustion conjecture , i.e. , identity ( [ ex ] ) .
for example , the number of conjugacy classes in the orthogonal groups @xmath277 and @xmath278 together is @xmath279 each contributes @xmath30 classes ; the number of conjugacy classes in the groups @xmath280 and @xmath281 together is @xmath282 one contributes @xmath283 classes and the other @xmath284 classes .
in addition , the number of conjugacy classes in the groups @xmath285 and @xmath286 together is @xmath287 , each contributes @xmath288 classes .
hence , the computations of the multiplicities and rank for the groups @xmath124 and @xmath289 presented in figures [ ranksp6_5 ] and [ ranksp8_3 - 012][ranksp8_3 - 34a][ranksp8_3 - 34b ] , respectively , give the required numerical confirmation of ( [ ex ] ) in these cases .
we would like to clarify the strong relationship between the dimension of a representation of @xmath52 and its rank .
fix @xmath290 and consider a rank @xmath116 irreducible representation @xmath291 let us assume that @xmath8 appears in the image of the eta correspondence ( [ eta ] ) .
namely , there exist @xmath292 such that @xmath293see section [ ec ] . using part spectrum of theorem [ etac ] , and the dimension formula ( [ dimform ] )
, we have @xmath294the point now is see figure [ dim - ranksp6_5 ] for illustration that the term @xmath295 dominates the right hand side of ( [ dim - eta - tau - form ] ) .
indeed , we have [ dim]let @xmath257 be a rank @xmath290 irreducible representation of @xmath52 associated to an irreducible representation @xmath255 of @xmath296 then @xmath297 a proof of theorem [ dim ] will be given in a sequel paper .
the term @xmath298 can be estimated explicitly .
for example we have @xmath299 theorem [ dim ] seems to substantially extend the current knowledge ls , n , sh , tz on the dimensions of representations of the finite symplectic groups ( see lemma 2.3 . in @xcite ) . although dimension tends to increase with rank , because of the factor @xmath300 in ( [ dim - eta - tau - form ] ) , it may happen , see figures ranksp8_3 - 34a[ranksp8_3 - 34b ] , that a representation of rank @xmath116 has larger dimension than one of rank @xmath301 .
however , for a given @xmath116 , if @xmath302 is large enough then the representations of rank @xmath116 will have smaller dimension than those of rank @xmath301 .
for example , it seems that if @xmath303 is sufficiently large , then for @xmath126 , one can take @xmath304 , and for @xmath305 one can take @xmath306see figures [ dim - ranksp6_5 ] and [ ranksp8_3 - 012 ] for illustration . in general , using theorem [ dim ] and the known estimates on the dimensions of the largest irreducible representations of the orthogonal groups , we have the following result : [ compatibility]for sufficiently large @xmath307 in the regime @xmath308the rank @xmath116 representations appearing in the image of the eta correspondence ( [ eta ] ) always have smaller dimension then those of rank @xmath301 .
the exact computation leading to a verification of proposition compatibility will be given in a sequel paper .
if @xmath309 then @xmath120 is trivial .
the lemma now follows from the well known fact that the @xmath60 conjugates generate the group @xmath52 @xcite .
consider the automorphism @xmath310 given by @xmath311 it can be extended to an automorphism of the semi - direct product of @xmath52 with @xmath143 , by letting it act trivially on @xmath52 .
the equivalence of oscillator representations follows .
the fact that for a non - square @xmath312 , the representations @xmath178 and @xmath313 are not isomorphic , can be verified using the realization given in remark [ schrodinger model ] .
this completes the proof of the proposition .
the proposition follows immediately from equation ( [ op - model - n ] ) in section [ schrodinger - vu ] .
we give a proof of theorem [ etac ] * * * * that is an elementary application of the _
double commutant theorem _ @xcite .
then @xmath321 acts semi - simply on @xmath322 and as a representation of @xmath323 we have@xmath324where @xmath325 are all the irreducible representations of @xmath317 , and @xmath326 are all the irreducible representations of @xmath327 in particular , we have a bijection between irreducible representations of @xmath317 and @xmath328 and moreover , every isotypic component for @xmath317 is an irreducible representation of @xmath329 let us start with several preliminary steps .
we work with the schrdinger model of @xmath228 appearing in section schrodinger - vu .
it is realized on the space @xmath330and there , the actions of an element @xmath331 of the siegel unipotent radical @xmath60 @xmath332 and an element @xmath333 are given by formulas ( op - model - n ) and ( [ op - model - o ] ) , respectively . in particular ,
we have for the rest of the section , we fix a transformation @xmath337 which is and consider the character subspace @xmath338we would like to have a better description of the space @xmath339 .
the orthogonal group @xmath216 acts naturally on @xmath340 and we denote by @xmath341 the orbit of @xmath17 under this action . note that , because @xmath17 is onto , the action of @xmath216 on @xmath341 is free . in particular
, we can identify @xmath341 with @xmath265 and the peter
weyl theorem @xcite for the regular representation implies we would like now to describe the commutant of @xmath216 in @xmath346 considering the group @xmath347 of automorphisms of @xmath42 that stabilize the form @xmath348 we obtain two commuting actions@xmath349moreover , we have write @xmath352for various irreducible representations @xmath353 of @xmath52 .
then@xmath354 in addition , @xmath355 is a @xmath356-module , and so is each @xmath357 hence , identity ( [ d ] ) gives a decomposition of @xmath358 into ( not necessarily irreducible ) submodules for @xmath356 . but
proposition [ commutant ] together with the double commutant theorem says that @xmath355 is irreducible as a @xmath356-module .
therefore , exactly one of the @xmath359 will be non - zero , and it defines an irreducible representation of @xmath356 , which has dimension equal to @xmath360 , by equation ( [ reg - of - obeta ] ) . to conclude
, there exists a unique irreducible sub - representation @xmath361 @xmath251 of rank @xmath116 and type @xmath362 the multiplicity of the orbit @xmath245 in @xmath260 is @xmath363 and finally , the double commutant theorem implies that for @xmath364 in @xmath365 we have @xmath366 this completes the proof of theorem [ etac ] . using the delta basis @xmath367 @xmath368 we can verify claim [ regular ] , by showing that if @xmath369 then there exists @xmath238 such that @xmath370 indeed , let @xmath89 be the composition @xmath371where the first and second isomorphisms are these induced by @xmath372 and @xmath373 respectively , and @xmath374 is the radical of @xmath375 this completes the proof of proposition [ regular ] . to verify this assertion , note that we have a short exact sequence@xmath376where @xmath377 , @xmath378 is the orthogonal group of @xmath379 and @xmath380 is the appropriate unipotent group .
the group @xmath378 acts simply transitively on the orbit @xmath341 , as does the group @xmath216 , and these two actions commute with each other .
if we use the map @xmath381 to identify @xmath216 with @xmath341 , then the action of @xmath216 becomes the action of @xmath216 on itself by left translation , and the action of @xmath378 can be identified with the action of @xmath216 on itself by right translation . by the peter - weyl theorem @xcite , we conclude that the groups @xmath378 and @xmath216 generate mutual commutants in the operators on @xmath382 .
the groups @xmath356 and @xmath216 generate mutual commutants on @xmath383 .
this completes the proof of proposition commutant .
aubert a.m. , kraskiewicz w. , and przebinda t. , howe correspondence and springer correspondence for dual pairs over a finite field .
_ lie algebras , lie superalgebras , vertex algebras and related topics , proceedings of symposia in pure mathematics , vol . 92 , ams ( 2016 ) .
_ waldspurger j .-
l . , demonstration dune conjecture de dualite de howe dans le cas @xmath14- adique , @xmath386 .
_ festschrift in honor of i. i. piatetski - shapiro on the occasion of his sixtieth birthday , israel math .
, vol . 2 ( 1990 ) , 267324 . _ | finite group theorists have established many formulas that express interesting properties of a finite group in terms of sums of characters of the group .
an obstacle to applying these formulas is lack of control over the dimensions of representations of the group . in particular , the representations of small dimensions tend to contribute the largest terms to these sums , so a systematic knowledge of these small representations could lead to proofs of important conjectures which are currently out of reach . despite the classification by lusztig of the irreducible representations of finite groups of lie type , it seems that this aspect remains obscure . in this note
we develop a language which seems to be adequate for the description of the * * `` small '' * * representations of finite classical groups and puts in the forefront the notion of * * rank * * _ _ _ _ of a representation .
we describe a method , the * `` eta correspondence '' , * to construct small representations , and we conjecture that our construction is exhaustive . we also give a strong estimate on the dimension of small representations in terms of their rank . for the sake of clarity , in this note
we describe in detail only the case of the finite symplectic groups . |
open quantum systems often exhibit the phenomenon of resonances .
resonances correspond to quasi - stationary states which have a long life - time but eventually decay into the continuum .
( a particle , initially within the system , escapes to infinity . )
one approach to the problem of resonances is based on the study of the analytic properties of the scattering matrix , @xmath0 , in the complex energy plane .
resonances correspond to the poles , @xmath1 , of @xmath0 on the unphysical sheet @xcite .
an alternative approach amounts to solving the stationary schrdinger equation , with the boundary condition of an outgoing wave only .
this condition , which makes the problem non - hermitian , describes a particle ejected from the system .
the schrdinger equation with such boundary condition admits complex eigenvalues @xmath2 , which correspond to the resonances @xcite .
this kind of approach leads in a natural way to a non - hermitian effective hamiltonian , and it has been used for a long time in scattering theory , including scattering in chaotic and disordered systems ( @xcite , and references therein ) .
there are many examples of resonances in atomic and nuclear physics .
more recently , there has been much interest in resonant phenomena in the field of chaotic and disordered systems ( for a recent review see @xcite ) .
there is considerable amount of work concerning the distribution @xmath3 of resonance widths in one - dimensional disordered chains @xcite .
numerical studies presented in that work demonstrate that , in a broad range of @xmath4 , @xmath5 , with the exponent @xmath6 being close to 1 .
( a similar behavior pertains also to two- and three - dimensional systems with localized states @xcite ) .
an analytical calculation @xcite has been performed for a one - dimensional continuous ( white - noise ) potential , in the semiclassical limit , when the localization length is large .
it has been shown in that work that in a broad range of @xmath4 , @xmath3 is well fitted by a function @xmath7 . for sufficiently small @xmath4 ,
however , @xmath8-behavior was obtained , followed by a sharp cutoff at still smaller @xmath4 , due to the finite size of the sample .
a different analytical approach was recently developed in @xcite , for a discrete tight - binding random chain .
the limit of strong disorder ( i.e. , opposite to that of ref .
@xcite ) was considered and the @xmath8-behavior ( for a semi - infinite chain ) was derived . in the present paper
we develop a new approach to the problem of resonances .
the approach is based on counting the number of poles of a resolvent of the corresponding non - hermitian hamiltonian .
we derive a set of equations for a semi - infinite disordered chain , coupled to a ( semi - infinite ) lead .
this set contains , in principal , the full solution of the problem , for an arbitrary coupling strength and disorder .
the equations simplify considerably in the weak coupling regime .
for this case we rigorously derive the asymptotically exact @xmath8-rule and present a simple scaling formula , which contains only the product of the localization length and the density of states , at the relevant energy .
the system is depicted in fig.1 .
black dots , labelled by @xmath9 , designate sites along the semi - infinite disordered chain .
each site of the chain is assigned a site energy , @xmath10 .
different @xmath10 s ( @xmath9 ) are independent random variables chosen from some distribution @xmath11 .
open circles , labeled by @xmath12 represent a perfect semi- infinite lead to which the chain is coupled .
all sites of the lead are assigned @xmath13 .
the lead simulates the free space outside the chain .
all nearest neighbor sites of the chain are coupled to each other by a hopping amplitude @xmath14 , and the same is true for all nearest neighbor sites of the lead . the only exception to
this rule is the pair ( 0,1 ) which provides coupling between the chain and the lead .
the hopping amplitude for this pair is taken to be equal @xmath15 .
this allows us to tune the coupling from @xmath16 ( closed chain ) to @xmath17 ( fully coupled chain ) .
as was mentioned in the introduction , the most direct approach to the problem of resonances is based on solving the stationary schrdinger equation with the boundary condition of an _ outgoing wave _ ( see fig.1 ) .
the schrdinger equation for the entire system ( chain + lead ) is a set of coupled equations : @xmath18 we recall that @xmath19 for @xmath20 ( the lead ) and it is random for @xmath21 ( the chain ) .
eqs.(1 - 4 ) are to be solved subjected to the boundary condition of a outgoing wave in the lead , i.e. @xmath22 , for @xmath23 , with @xmath24 .
the complex wave vector @xmath25 is related to @xmath26 by @xmath27 . using the plane wave shape of the outgoing wave in the lead , it is straightforward to eliminate from eqs.(1 - 4 ) all @xmath28 s with @xmath20 , thus reducing the problem to a system of equations for the amplitudes @xmath28 on the sites of the disordered chain alone ( @xmath9 ) : @xmath29 with the condition @xmath30 . here
@xmath31 for @xmath32 , but not for @xmath33 .
this end site is assigned a complex energy @xmath34 where the parameter @xmath35 describes the coupling strength to the outside world .
thus , the resonances are given by the complex eigenvalues of the non - hermitian effective hamiltonian defined in ( [ nhschrod1d ] ) .
this non - hermitian hamiltonian , @xmath36 , differs from the hermitian hamiltonian , @xmath37 , of the corresponding closed system ( i.e. , with @xmath38 ) only by the complex correction to the energy of the first site ( eq.([epsilon ] ) ) : @xmath39 where @xmath40 is the projection on site 1 .
we set the energy scale by taking @xmath41 and denote by @xmath42 the complex variable @xmath43 .
note that , since @xmath44 is related to @xmath42 via @xmath45 , the effective hamiltonian @xmath36 depends on @xmath42 .
thus , eq . ( [ nhschrod1d ] )
does not constitute a standard eigenvalue problem and the eigenvalues of @xmath36 have to be determined self - consistently .
defining formally the resolvent @xmath46 we can write @xmath47^{-1}= { 1\over 1+\eta e^ { i\tilde k } gp}g , \ ] ] where @xmath48 is the `` unperturbed '' resolvent .
resonances correspond to the singularities of the matrix @xmath49 , in the complex @xmath42-plane , or to the roots of the equation @xmath50 where @xmath51 is the @xmath52-element of @xmath53 in the site representation .
let us write @xmath51 as @xmath54 where @xmath55 is the self - energy for site @xmath56 . eq.([roots ] ) can be written as @xmath57 denoting the zeroes of @xmath58 by @xmath59 , we write the density of resonances ( dor ) , for a given realization of the disorder , as @xmath60 .
this expression refers directly to the semi - infinite chain and it should be understood as the @xmath61 limit of the analogous expression for a finite chain , of size @xmath62 . the limit is well defined for any @xmath63 and no division of the sum by @xmath62 is necessary @xcite , in contrast to the usual case of the density of states ( on the real axis ) for a hermitian problem .
it is convenient to define a @xmath64-function of a complex variable @xmath65 as @xmath66 and to use the relation @xmath67 where the prime indicates a derivative with respect to @xmath42 .
this relation is a generalization of the corresponding relation for a real variable .
it is valid if @xmath58 is analytic in a domain containing the zeroes @xmath59 and @xmath68 .
indeed , under such conditions in the neighborhood of a zero , the equation @xmath69 can be solved by an analytic function @xmath70 and the jacobian of the change of variables from @xmath71 to @xmath72 is @xmath73 due to the cauchy - riemann equations .
we will assume that @xmath74 with probability one .
substituting eq.([f ] ) for @xmath58 , we have @xmath75 where @xmath76 stands for the expression @xmath77 . next we average eq.([rho ] ) over all realizations of the set @xmath78 of the random site energies ( we denote this average by angular brackets ) . since the self - energy @xmath55 does not depend on the energy @xmath79 of the first site , we can single out the variable @xmath79 and average over it explicitly .
this leads to the following expression for the average dor : @xmath80\delta(z-\epsilon - s_1+\eta\lambda)>,\ ] ] the two correlated random variables , @xmath81 and @xmath82 depend on the set @xmath83 of the random site energies .
note that we do not need to know the joint probability distribution for @xmath81 and @xmath82 .
indeed , only three special combinations of these two variables appear in eq.([rho1 ] ) .
it is therefore useful to introduce three functions @xmath84 and write ( [ rho1 ] ) as @xmath85.\end{aligned}\ ] ] we emphasize that @xmath81 refers to a closed ( semi - infinite ) chain , so that one can use the standard relations for various quantities of the hermitian problem @xcite . in particular
, we ll need the relation @xmath86 where @xmath87 is the self energy for site 2 , with site 1 excluded .
it immediately follows that @xmath88 with the help of the definitions ( 15)-(17 ) and the recursion relations ( 19 ) , ( 20 ) it is straightforward to derive integral relations between the functions @xmath89 , @xmath90 and @xmath91 .
for instance , using ( 20 ) , @xmath92 can be rewritten as @xmath93 the variable @xmath94 is now `` isolated '' ( i.e. , all other quantities in ( [ h1 ] ) do not depend on this variable ) and can be averaged over .
this leads to @xmath95.\ ] ] the two other equations can be derived in a similar way , yielding : @xmath96\ ] ] and @xmath97 let us stress that the above defined functions of the complex argument @xmath98 are not complex valued functions in the usual sense but rather shorthand notations for a function of two real variables , @xmath99 and @xmath100 .
for instance , eq .
( [ f ] ) for the function @xmath101 , which is the probability distribution for the real and imaginary part of the self - energy @xmath102 , can be written more explicitly as @xmath103 the three equations , ( 22)-(24 ) , supplemented by the expression ( 18 ) provide the general framework for computing the average dor , @xmath104 , in the complex plane @xmath105 , for any strength of disorder and for an arbitrary coupling @xmath106 .
we do not attempt to solve the problem in its full generality but rather restrict ourselves to the small @xmath106 , i.e. weak coupling case .
from now on we assume that the coupling constant @xmath106 is small and develop a `` linear response theory '' with respect to @xmath106 . in this limit
the width of all resonances becomes proportional to @xmath106 , so that the `` cloud '' of resonances in the @xmath107-plane gets squeezed towards the real axes . in order to define a meaningful @xmath108 limit for the dor
, one must stretch the @xmath109-axis by a factor of @xmath110 .
it is natural to switch from the original variable @xmath105 to a new variable , @xmath111 , where @xmath112 and @xmath113 , where the minus sign accounts for the fact that the resonances are located in the lower half plane of the complex variable @xmath42 .
thus , in the @xmath114-plane resonances appear in the upper half - plane and @xmath115 is the resonance width , in units of @xmath106 .
the limiting ( i.e. , in the @xmath116 limit ) average dor , @xmath117 , in the complex @xmath114-plane is related to the original dor , @xmath118 , as : @xmath119 next , we turn our attention to the functions @xmath120 and @xmath91 in the @xmath108 limit .
the function @xmath121 is the probability distribution for the real and imaginary part of the self - energy @xmath81 .
the shape of this function depends on the point @xmath105 . in the @xmath108 limit ,
as has been just explained , one should set @xmath122 , where @xmath115 does not depend on @xmath106 .
furthermore , since the imaginary part of @xmath81 is proportional to @xmath106 , we set @xmath123 . the distribution , @xmath124 , for the new variables ( @xmath125 is related to the distribution @xmath121 for the `` old '' variables by @xmath126 .
the distribution @xmath124 has a well defined @xmath108 limit and it satisfies an integral equation , obtained from ( [ f1 ] ) by transforming to the new variables and taking the @xmath127 limit : @xmath128 the simplifying feature of the @xmath108 limit is that functions @xmath90 and @xmath91 are easily expressible in terms of @xmath89 .
more precisely , we have to define functions @xmath53 and @xmath37 , of the new variables : @xmath129 these functions , in complete analogy with @xmath124 , have a well defined @xmath108 limit and satisfy integral relations which are derived from ( [ h2 ] ) , ( [ g ] ) by writing them in explicit form ( compare to ( [ f1 ] ) ) , transforming to the new variables and taking the @xmath108 limit .
this yields : @xmath130,\end{aligned}\ ] ] and @xmath131.\end{aligned}\ ] ] one can easily check that ( 29 ) and ( 30 ) are satisfied by @xmath132 now we can derive an expression for @xmath117 @xmath133 , in terms of the function @xmath65 , by taking the @xmath108 limit in eq.(18 ) and using ( [ dor ] ) and ( [ gh ] ) : @xmath134 note that , since @xmath135 , in the @xmath108 limit @xmath136 approaches @xmath137 which is related to @xmath138 by @xmath139
. therefore @xmath140 in ( [ rho3 ] ) is given by @xmath141 it is convenient to define a new variable , @xmath142 , so that @xmath143 is the imaginary part of self - energy ( the resonance width ) in units of @xmath144 .
finally , we denote @xmath145 by @xmath146 ( the real part of the self - energy ) and define the probability distribution @xmath147 , instead of @xmath124 .
@xmath147 satisfies the integral equation @xmath148 where the integration variable @xmath149 was replaced by @xmath150 . the expression ( [ rho3 ] ) for the average dor now reads : @xmath151 while performing integration in equations ( [ f2 ] ) , ( [ rho4 ] ) , one should keep in mind that the function @xmath147 is identically zero for @xmath152 .
this property follows from the basic recursion relation for the self energy @xmath153 .
the integral equation ( [ f2 ] ) , supplemented by the expression ( [ rho4 ] ) , completely defines the problem of resonances in the weak coupling limit .
note that the coupling strength , @xmath106 , does not appear explicitly in these equations .
it only determines the units in which the resonance width is measured ( the widths of all resonances is proportional to @xmath106 ) .
thus , @xmath117 is determined solely by the properties of the closed system and , in this sense , eqs.([f2 ] ) , ( [ rho4 ] ) describe the regime of linear response with respect to the coupling strength @xmath106 .
the function @xmath147 describes the joint probability distribution for the real and imaginary part of the self - energy ( of the closed system and for @xmath42 approaches the real axis ) and the integral equation ( [ f2 ] ) has appeared previously in the study of anderson localization @xcite .
integration of ( [ f2 ] ) over @xmath143 yields the integral equation for the probability distribution , @xmath154 , of the real part of the self - energy : @xmath155 this equation is very useful in the theory of one - dimensional localization @xcite , because the knowledge of @xmath154 allows one to compute the localization length , @xmath156 , according to : @xmath157 we shall use this relation in the next section .
we were not able to obtain the complete analytic solution of the integral equation ( [ f2 ] ) .
it is possible , however , to find the large-@xmath143 asymptotics of the function @xmath147 .
we state the result and give the proof later : to leading order in the small quantity , @xmath158 , @xmath159 where @xmath160 is the usual density of states , on the real energy axis , for the closed chain .
an important feature of this result is that the large-@xmath143 asymptotics does not contain @xmath146 .
therefore , by plugging in ( [ large j ] ) into ( [ rho4 ] ) we immediately obtain the simple result for the small-@xmath115 asymptotic of the average dor : @xmath161 note that the factors containing @xmath162 had cancelled , so that @xmath138 enters only via the density of states and the localization length .
this cancellation is the consequence of the obvious scaling property of the general equation ( [ rho4 ] ) , namely : the dependence on @xmath163 drops out if one transforms to a new variable , @xmath164 , and to a new function , @xmath165 .
thus , the asymptotic ( @xmath166)- behavior is universal , in the sense that it holds for any degree of disorder and for any @xmath167 .
it remains to prove the result stated in ( [ large j ] ) .
it is convenient to define the function @xmath168 and to study its large-@xmath143 asymptotics .
an integral equation for @xmath169 is derived by integrating ( [ f2 ] ) over the second argument of @xmath40 . while performing this integration one should remember that the function @xmath170 is identically zero for @xmath171 .
it follows from this property that for @xmath172 @xmath173 combining both cases , i.e. , @xmath172 and @xmath174 , one obtains the following integral equation : @xmath175 we are interested in the large-@xmath143 behavior of @xmath169 .
it is useful to define the laplace transform @xmath176 and to study its small-@xmath177 behavior .
the integral equation for the transform is obtained directly from ( [ w2 ] ) : @xmath178 where @xmath179 anticipating that @xmath169 is proportional to @xmath158 , we look for the solution ( in the small @xmath177 limit ) of ( [ w4 ] ) in the form @xmath180 this ansatz satisfies eq.([w4 ] ) if @xmath181 and @xmath182 with @xmath183 since the site energy distribution , @xmath11 , is normalized to 1 , it immediately follows from ( [ w7 ] ) that @xmath184 is a constant . we denote this constant by @xmath185 and determine it as follows .
define a function @xmath186 which is the non - trivial solution of the integral equation @xmath187 note that the kernel @xmath188 is not symmetric with respect to @xmath146 and @xmath189 and that equation ( [ w10 ] ) , unlike ( [ w7 ] ) , is not solved by a constant .
the solution of ( [ w10 ] ) is : @xmath190 we now go back to ( [ w8 ] ) , multiply it by @xmath191 , integrate over @xmath146 and , with the help of ( [ w10 ] ) , obtain : @xmath192 which , using ( [ w9 ] ) and the aforementioned result @xmath193 , yields the value of @xmath194 : @xmath195 where the explicit form of @xmath191 ( see ( [ w11 ] ) ) has been used at the last step .
the integral in the numerator is known in localization theory @xcite and is equal to the density of states @xmath160 .
the integral in the denominator is equal to @xmath196)(see ( [ f3 ] ) ) .
thus , in the small-@xmath177 limit , @xmath197 , with @xmath198 . as a consequence , in the large-@xmath143 limit , @xmath199 and @xmath200 which completes the proof of ( [ large j ] ) .
one can also show that ( [ w8 ] ) has a unique solution @xmath201 and the behavior of @xmath202 suggests that @xmath203 and @xmath204
we studied the average density of resonances ( dor ) for a semi - infinite disordered chain , coupled to the outside world by a ( semi - infinite ) perfect lead .
the main result of this work is the expression ( 18 ) for the average dor , supplemented by the three integral equations , ( 22 - 24 ) , for the three functions , @xmath205 and @xmath89 .
this set of equations provide the general framework for calculating the average dor , for an arbitrary disorder and coupling strength .
we applied these general equations to the case of weak coupling and derived an asymptotically exact expression for the average dor , in the limit of small resonance width ( eq.(39 ) ) .
this expression is universal , in the sense that it holds for any degree of disorder and everywhere in the ( unperturbed ) energy band .
it is worthwhile to emphasize the essential difference between the average dor , as defined in this paper , and the probability distribution of the resonance widths , often used in the literature ( e.g.@xcite ) . for a finite size chain , of @xmath62 sites ,
the two quantities differ only by a factor @xmath62 . however , when @xmath62 approaches @xmath206 , the probability distribution shrinks towards a delta - function , while the average dor approaches a perfectly well defined limit .
since we have taken a semi - infinite chain from the start , it was essential for us to work with the average dor , rather than with the probability distribution of the resonant widths .
this enabled us to obtain the @xmath166-behavior in the @xmath207 limit , in contrast to the statement of @xcite that in that limit the @xmath166-behavior has no region of applicability .
the authors acknowledge useful conversations with e. gurevich and o. kenneth .
one of the authors ( b.s . ) acknowledges the hospitality of epfl where most of this work was done . the work was supported in part by the israel science foundation ( isf ) . | we study the average density of resonances ( dor ) for a semi - infinite disordered chain , coupled to the outside world by a ( semi - infinite ) perfect lead .
a set of equations is derived , which provides the general framework for calculating the average dor , for an arbitrary disorder and coupling strength .
these general equations are applied to the case of weak coupling and an asymptotically exact expression for the averaged dor is derived , in the limit of small resonance width .
this expression is universal , in the sense that it holds for any degree of disorder and everywhere in the ( unperturbed ) energy band .
21 cm 16 cm -0.15 cm |
a central set in a euclidean space has symmetry with respect to reflection through a point , called its center .
a closed embedded smooth hypersurface of @xmath3 is called an ovaloid if all its principal curvatures , with respect to the outer unit normal , are positive everywhere .
an ovaloid of dimension one is also called an oval . the compact transverse cross - sections of a cylinder over a central ovaloid in @xmath3 , @xmath4 , with hyperplanes are central ovaloids .
a similar result holds also for quadrics , which are the level sets of quadratic polynomials in @xmath3 , @xmath4 .
their compact transverse cross - sections with hyperplanes are ellipsoids , which are central ovaloids .
following solomon who showed that these two kinds of examples provide the only complete smooth hypersurfaces in @xmath5 , whose ovaloid cross - sections are central , we show that the same conclusion also holds in @xmath3 , @xmath1 .
roughly , we will say that a smooth hypersurface @xmath6 , @xmath1 , has the _ central ovaloid property _ , or _ cop _ ,
if * @xmath2 intersects at least one hyperplane transversally along an ovaloid , and *
every such ovaloid is central .
given this set - theoretic definition of _ cop _ , our main result can be stated as follows : _ a complete , connected smooth hypersurface @xmath7 in @xmath3 , @xmath1 , with _ cop _ is either a cylinder over a central ovaloid , or a quadric .
_ more precisely , the _ central ovaloid property _ and our main theorem can be stated in terms of mappings as follows : [ ds1.1:ccut ] if @xmath7 is a smooth manifold , @xmath8 , @xmath1 , is an immersion , and @xmath9 is a hyperplane transversal to @xmath10 , a cross - cut of @xmath10 relative to @xmath11 is a compact connected component @xmath12 .
[ ds1.1:cop ] let @xmath7 be a smooth manifold , and @xmath13 , be an immersion , @xmath1 , then @xmath10 is said to have the _ central ovaloid property _ , or , _ cop _
, if * there exists at least one cross - cut whose image is an ovaloid , and * every such ovaloid is central .
given this precise terminology , the main theorem can be stated as follows : let @xmath7 be a smooth , connected manifold , and @xmath13 , @xmath1 , be a proper , complete immersion with _ cop _ , then @xmath14 is either a cylinder over a central ovaloid or a quadric .
historically , the first theorem of this sort was proved by w. blaschke .
@xcite suppose every plane transverse , and nearly tangent to , a smooth convex surface @xmath15 intersects @xmath16 along a central loop .
then @xmath16 is a quadric .
b. solomon , in 2009 , removed the convexity assumption and first considered hypersurfaces of revolution as follows : @xcite let @xmath6 , @xmath4 , be a hypersurface of revolution . if @xmath2 intersects every hyperplane nearly perpendicular to its axis of rotation in a central set , then @xmath2 is a quadric . using the rotationally symmetric case ,
solomon then proved the @xmath17 analogue of our main theorem in 2012 .
@xcite let @xmath18 have the _ central oval property _
, i.e. , @xmath2 intersects some plane along an oval , and every such oval is central .
then a complete , connected , smooth surface @xmath2 with _ central oval property _ must either be a cylinder over a central oval or a quadric .
our main result is the generalization of solomon s result @xcite to all dimensions @xmath4 .
the proof is not inductive , and although it is , at certain times , a direct generalization of solomon s method , some essentially new difficulties arose .
moreover , the technical tools required to carry on the argument are more sophisticated and the calculations are more elaborate . as in solomon s paper , we obtain our main result by first working the local version . if an immersed hypersurface @xmath2 intersects some hyperplane along an ovaloid as our definition of _ cop _ requires , then some neighborhood , in @xmath2 , of that ovaloid embeds into @xmath3 as a tube with _ cop_. the proof that these tubes are either cylindrical or quadric constitute the local version of our work . unless noted otherwise , @xmath19 denotes the open interval @xmath20 .
suppose @xmath21 , @xmath4 , is an embedding of the form @xmath22 where @xmath23 and @xmath24 are smooth , and for each fixed @xmath25 , the map @xmath26 parameterizes an ovaloid with center of mass at the origin .
a transversely convex tube is any embedded annulus that , after an affine isomorphism , can be parameterized this way .
a transversely convex tube in _ standard position _ is the image of an embedding @xmath27 of the form . if we discard the central curve @xmath28 of a transversely convex tube @xmath29 in _ standard position _ , we get the rectification @xmath29 , denoted @xmath30 , the image of @xmath31 we say that @xmath30 splits when we can separate variables : @xmath32 for some functions @xmath33 and @xmath34 parameterizing a fixed ovaloid . as in solomon s paper
we get a splitting lemma : * proposition * [ ps5.2:split ] ( splitting lemma)*. * _ if a transversely convex tube @xmath29 in _ standard position _ has _ cop _
, then its rectification @xmath35 splits . _ to obtain this splitting lemma we derive a pair of partial differential equations satisfied by the function @xmath36 which , for each @xmath25 , yields the support function @xmath37 of the ovaloid @xmath38 . using the second pde we cook up a strictly elliptic operator with bounded coefficients and use harnack s inequality to get our splitting lemma .
after we are able to split the function @xmath39 we use the first pde to obtain the following local version of our main result for the rectified tube : * proposition * [ ps5.2:classofrcttube]*. * _ suppose @xmath29 is a transversely convex tube in _ standard position _ with _
cop_. then its rectification @xmath30 is either _ 1 . a cylinder over a central ovaloid , or 2
affinely isomorphic to a hypersurface of revolution .
* proposition * [ ps5.3:axslemm ] ( axis lemma)*. * combining axis lemma and @xcite we can conclude that a transversely convex tube with _ cop _ is either cylindrical or quadric .
this local version of our main result can be stated as * proposition * [ ps5.4:lclvrs ] ( local version)*. * _ a transversely convex tube with _ cop _ is either a cylinder over a central ovaloid or a quadric .
_ the final steps of the proof of our main theorem are as follows : given any immersion with _ cop _ and
a cross - cut , lemma [ ls3.3:exttubnbd ] ( tubular neighborhood ) shows the existence of a neighborhood of this cross - cut foliated by diffeomorphic copies .
then lemma [ ls3.3:exsttrncvxtube ] ( existence of transversely convex tube ) shows that this neighborhood is embedded onto a transversely convex tube with _ cop _ about the image of the cross - cut , when that image is an ovaloid . by proposition [ ps5.4:lclvrs ] ( local version ) ,
the tube is either cylindrical or quadric .
but the boundaries of such a tube , in either case , are again images of cross - cuts and therefore , using lemma [ ls3.3:exttrncvxtube ] ( extension of transversely convex tube ) we can push the boundaries of the annular region in @xmath2 and the tubular region in @xmath3 a little further .
since @xmath2 is connected and complete , the extension process stops only when the annular region fills up @xmath2 and the image is either a complete cylinder or a quadric .
we would like to thank our advisor bruce solomon for giving us this project , and providing guidance .
we would also like to express our gratitude to the indiana university mathematics department for creating an academically challenging and supportive environment , and to the graduate school for the financial support that made our studies possible .
in this section , we provide the necessary basic definitions and auxiliary lemmas from convex geometry .
all definitions and lemmas come from the book of schneider @xcite .
this section ends with a computation of the support function of an ellipsoid . in this subsection
, we provide the definition of convex set , convex body , and strictly convex body .
a set @xmath40 is convex if for every @xmath41 it contains the segment @xmath42 = \big\{\ , ( 1-\lambda ) x + \lambda y \colon 0 \leq \lambda \leq 1 \,\big\}.\ ] ] a set @xmath43 is called a convex body if @xmath44 , convex , and compact . by @xmath45
we denote the set of all convex bodies in @xmath3 .
@xmath40 , the affine hull of @xmath46 , @xmath47 , is the smallest affine subspace of @xmath48 containing @xmath46 .
the relative boundary , @xmath49 , is the boundary of @xmath46 relative to its affine hull .
a convex body @xmath50 is called a strictly convex body if @xmath51 does not contain any line segment . in this subsection
, we introduce the support function and list its important properties that will be relevant to us in later parts .
in particular , we remark how the gradient of a support function determines the boundary points of a convex set .
we also give a rigorous definition of the centrix and we end this subsection with a calculation of the support function of an ellipsoid .
let @xmath43 , @xmath52 , be a convex body .
the support function @xmath53 of @xmath54 is defined on @xmath3 by @xmath55 [ ls2.2:spfnpr ] let @xmath50 , then the support function @xmath56 of @xmath54 has the following properties : 1 .
@xmath56 is sublinear .
[ ls2.2:spfnpr2 ] 2 .
@xmath57 and @xmath58 for all @xmath59 .
[ ls2.2:spfnpr3 ] 3 . given @xmath60 , @xmath61 .
[ ls2.2:spfnpr5 ] 4 .
@xmath62 for all @xmath63 and for all @xmath64 , @xmath65 .
[ ls2.2:spfnpr6 ] the item of lemma [ ls2.2:spfnpr ] shows that a support function is sublinear .
however , the converse is also true .
[ ls2.2:subvcxbd ] @xcite if @xmath66 is a sublinear function , then there exists a unique convex body @xmath67 with support function @xmath68 .
let @xmath50 be a convex body and @xmath63 , then @xmath69 is called the support hyperplane of @xmath54 with outer normal @xmath64 , @xmath70 is called the supporting halfspace of @xmath54 with outer normal @xmath64 , and @xmath71 is called the support set of @xmath54 with outer normal @xmath64 . for @xmath72 , @xmath73 is the signed distance of the support hyperplane @xmath74 to the origin .
note that the support set of convex body does not need to be a singleton .
in particular , the convex body @xmath75 in has a support set , which is a line segment .
however , if the support set is a singleton then we have an important analytical consequence about the support function .
[ ls2.2:diffsptfct ] @xcite let @xmath50 and @xmath76 .
the support function @xmath56 is differentiable at @xmath64 if and only if @xmath77 .
in this case @xmath78 [ ds2.2:centrix ] let @xmath50 and @xmath56 be differentiable in @xmath79 , then the centrix of @xmath54 is defined as @xmath80 note that the function @xmath81 and hence the centrix of @xmath54 are positively homogeneous of degree zero .
we end this subsection with a computation of the support function of an ellipsoid .
this support function will play an important rule in the proof of the main theorem .
[ es2.5:sptfnctellipsoid ] [ support function of ellipsoid ] given @xmath82 positive numbers @xmath83 @xmath84 , the function @xmath85 defined as @xmath86^{1/2}\ ] ] is sublinear and differentiable in @xmath79 . by lemma [ ls2.2:subvcxbd ] there
exists a unique convex body @xmath50 , with @xmath87 and @xmath88 we want to show that the set consisting of the gradients of @xmath56 equals the complete ellipsoid @xmath89 let @xmath90 , then the gradient of @xmath56 at @xmath91 equals @xmath92^{2}}{\lambda_{i}^{2 } } & = \frac{1 } { h_{k}^{2}(x ) } \sum_{i=1}^{n } \lambda_{i}^{2 } x_{i}^{2 } = 1 .
\end{aligned}\ ] ] on the other hand let @xmath93 satisfy @xmath94 so we can conclude that @xmath95 and since @xmath54 is convex the equality must hold , and @xmath54 is the convex body with boundary an origin centered ellipsoid . from now on @xmath87 will also be called the support function of the origin centered ellipsoid @xmath96 given in .
given any ellipsoid @xmath97 in @xmath3 there exist @xmath60 and @xmath65 so that @xmath98 , hence the support function @xmath99 of @xmath100 can be calculated using the properties and of the support function listed in lemma [ ls2.2:spfnpr ] in the following manner : @xmath101 for every @xmath63 .
here , we provide the necessary basic definitions and auxiliary lemmas from multilinear algebra and differential geometry .
the references for this material are the books of federer @xcite , hirsch @xcite .
we end this section by sketching proofs of the existence and extension of a transversely convex tube . in this subsection
, we introduce a principal geometric object of interest , the ovaloid in @xmath3 . using hadamard s theorem @xcite , in the case @xmath4
, we note that every ovaloid has a support parameterization . writing the laplacian in polar
coordinates and using lemma [ ls2.2:diffsptfct ] we give an analytical expression for the support parameterization .
lastly , we recall that ovaloids are preserved under affine isomorphisms of @xmath3 , find a transversely convex tube about the image of a cross - cut when that image is an ovaloid , and determine sufficient conditions for being able to extend it .
an ovaloid @xmath102 , @xmath103 is a closed embedded smooth hypersurface with all principal curvatures positive everywhere with respect to the outer unit normal . in particular , an ovaloid of dimension one , @xmath104 , also called an oval , is an embedded smooth loop in @xmath105 such that , given the outer unit normal vector field @xmath106 and the unit tangent vector field @xmath107 along @xmath108 @xmath109 with @xmath110 , called the geodesic curvature , is positive everywhere .
it is a simple exercise of differential and convex geometry to show that that the gauss map of an oval is a diffemorphism and that an oval bounds a strictly convex body .
the analogous statement holds in higher dimensions by a theorem due to hadamard @xcite , with an alternate proof in @xcite .
hadamard s theorem implies that any ovaloid @xmath102 , @xmath4 is the boundary of a strictly convex body and that its gauss map is a diffeomorphism . using this remark
we can make the following definition .
the support parameterization of the ovaloid @xmath102 , @xmath103 , that bounds a strictly convex body @xmath54 is defined as the inverse of the gauss map @xmath111 since @xmath112 is bijective and for each @xmath72 , the outer unit normal vector @xmath64 to the hyperplane @xmath74 at the points of @xmath113 is also normal to the boundary @xmath114 , we can conclude that @xmath115 must be a singleton . now using this observation and lemma [ ls2.2:diffsptfct ]
we can write @xmath116 so by the sublinearity property of the support function , introduced in item of lemma , and the representation of @xmath117 in polar coordinates we can conclude that the support parameterization is also equal to @xmath118 for all @xmath72 . using the the @xmath119-operator defined in @xcite we get , up to a sign , a unique coordinate representation for the unit normal field defined along a smooth orientable hypersurface .
[ rs3.2:outrunitnorfld ] if @xmath6 is a hypersurface and @xmath120 is a local parameterization of @xmath2 then up to a sign the unique unit normal field along @xmath121 is defined as @xmath122 when @xmath6 is a closed embedded hypersurface we let the mapping in define the unit normal field pointing into the unbounded component of @xmath123 . during the course of our proof of the main theorem we want to use the fact that the hypotheses and the conclusion are invariant under affine isomorphisms of @xmath3 . to achieve this
, we use the fact that ovaloids are preserved under affine isomorphisms of @xmath3 .
[ ls3.2:glnprsvovld ] if @xmath103 , @xmath124 is an affine isomorphism of @xmath3 , and @xmath102 is an ovaloid , then @xmath125 is again an ovaloid . normal curvatures are preserved under isometries and hence ovaloids are preserved under orthogonal maps and translations .
since each invertible linear map can be written as the composition of a symmetric linear map and an orthogonal map , it suffices to show that ovaloids are preserved under invertible symmetric maps .
the last claim can be easily shown by writing the ovaloid and its image locally as a graph . in this subsection ,
our main goal is to construct a transversely convex tube @xmath29 about the image of a cross - cut when that image is an ovaloid . in order to achieve this goal
, we first need lemma [ ls3.3:exttubnbd ] ( tubular neighborhood ) , which provides a neighborhood for a cross - cut foliated by diffeomorphic copies . since the cross - sections of @xmath29 that are close enough to the image of the cross - cut must also be ovaloids we can use lemma [ ls3.3:exsttrncvxtube ] ( existence of transversely convex tube ) to get the required tube .
another important goal is to give sufficient condition for extensibility of a transversely convex tube .
in particular , we point out in lemma [ ls3.3:exttrncvxtube ] ( extension of transversely convex tube ) that whenever the boundary of the tube is the image of a cross - cut then one can extend the tube a little further . given @xmath76 , define the linear function @xmath126 by @xmath127 . for any given @xmath76 and @xmath128 ,
the level hyperplane @xmath129 is defined by @xmath130 [ ls3.3:exttubnbd ] suppose @xmath2 is a smooth manifold , and @xmath131 is a compact connected embedded hypersurface of @xmath2 .
assume that for some @xmath76 and @xmath128 , @xmath10 is transversal to @xmath129 along @xmath131 , then there exist @xmath132 and an embedding @xmath133 \to m$ ] with the following properties : 1 .
@xmath134 for all @xmath135 , 2 .
@xmath136 for all @xmath135 , @xmath137 , 3 .
@xmath138 for all @xmath135 , @xmath137 .
[ rs3.3:ccutfoliate ] according to lemma [ ls3.3:exttubnbd ] ( tubular neighborhood ) , @xmath131 is a cross - cut of @xmath10 relative to @xmath139 and the neighborhood @xmath140 \,\bigr)$ ] of @xmath131 in @xmath2 is foliated by cross - cuts @xmath141 diffeomorphic to @xmath131 , each a level set of @xmath142 .
since the gradient @xmath143 does not vanish on the compact submanifold @xmath131 we can define the vectorfield @xmath144 in an open neighborhood of @xmath131 . using a standard flow box argument for the vectorfield @xmath27 we then get the required embedding @xmath145 . [ ls3.3:exsttrncvxtube ] suppose @xmath1 , the map @xmath8 is an immersion , @xmath146 is a hyperplane , and @xmath147 is a compact connected embedded hypersurface of @xmath2 .
if @xmath148 is an ovaloid and the function @xmath149 has no critical point on @xmath131 then @xmath10 maps some neighborhood of @xmath131 in @xmath2 onto a transversely convex tube .
we can compose the immersion @xmath10 with the embedding @xmath145 , obtained in lemma [ ls3.3:exsttrncvxtube ] , to get an embedding because @xmath10 embeds @xmath131 onto a simply connected ovaloid and the set of embeddings is open in the strong topology @xcite . since the principal curvatures are continuous there exists an open neighborhood of @xmath148 on the image of @xmath150 so that each horizontal cross section is again an ovaloid .
now suppose @xmath10 , @xmath131 are as above and @xmath151 is the transversely convex tube whose existence is guaranteed by lemma [ ls3.3:exsttrncvxtube ] .
note that if @xmath10 has _
cop _ so does @xmath151 .
let @xmath152 \bigr)$ ] be the neighborhood of @xmath131 that maps under @xmath10 onto @xmath151 as described in lemma [ ls3.3:exttubnbd ] and lemma [ ls3.3:exsttrncvxtube ] .
[ ls3.3:exttrncvxtube ] let @xmath8 be an immersion with _
cop _ , and @xmath153 a cross - cut of @xmath10 relative to @xmath154 , then given a transversely convex tube @xmath151 , with _ cop _ , about @xmath148 , where @xmath142 has no critical point in @xmath155 , there exists @xmath156 so that @xmath157 is a transversely convex tube , with _ cop _ , about @xmath148 , where @xmath142 has no critical point in a larger neighborhood @xmath158 of @xmath131 . since each boundary of the transversely convex tube @xmath151 is the image of a cross - cut we can use lemma [ ls3.3:exsttrncvxtube ] to construct a transversely convex tube about each boundary . then it is an easy exercise to glue these two transversely convex tubes about each boundary to the tube @xmath151 to get the required extension .
finally , the _ cop _ of @xmath10 implies that the extended tube has also _
here , we give the proof of the main theorem . we first show that the conclusion holds locally and then use standard differential topology arguments to get the global result .
the two papers that are cited , @xcite and @xcite , were written by solomon . unless stated otherwise , we assume that @xmath4 throughout this section . in this subsection , we classify the support function of an origin centered ellipsoid as a function whose square is the solution of some differential equation .
we define the centrix of an ovaloid and characterize central symmetry in terms of the centrix .
[ ls5.1:orgcntellipsoid ] let @xmath159 , @xmath4 , be a non vanishing function on @xmath160 , then @xmath161(u ) = 0\ ] ] for each @xmath162 if and only if @xmath163 or @xmath164 is the support function of an origin centered ellipsoid .
@xmath165 : define the linear differential operator @xmath166 as @xmath167 , then given @xmath168 the equality @xmath169 implies @xmath170 , for some constant @xmath171 .
the function @xmath172 can be written as @xmath173 , where @xmath174 , @xmath175 is the particular solution given by @xmath176 @xmath177 .
the homogeneous solution @xmath178 , @xmath179 satisfies @xmath180 and using the fact that the spherical harmonics are the eigenspaces of the laplace - beltrami operator , @xmath181 @xcite , we can conclude that @xmath182 is a spherical harmonic of degree @xmath183 in @xmath184 .
therefore , @xmath182 must be of the form @xmath185 , @xmath186 for some trace - free symmetric matrix @xmath46 .
then the function @xmath172 equals @xmath187 is a positive definite symmetric matrix .
the set @xmath188 is therefore an origin centered ellipsoid .
there exists @xmath189 so that @xmath190 where @xmath191 for each @xmath192 and the set @xmath96 equals @xmath193 px = 1\bigr\}\\ & = \bigl\ { x \in \mathbb{r}^{n-1 } \colon ( px)^{\operatorname{\mathrm{t } } } \operatorname{\mathrm{diag}}\bigl[\ , \lambda_{1}^{-2 } , \dotsc , \lambda_{n-1 } ^{-2 } \,\bigr ] px = 1\bigr\}\\ & = p^{-1 } \bigl\ { x \in \mathbb{r}^{n-1 } \colon x^{\operatorname{\mathrm{t } } } \operatorname{\mathrm{diag}}\bigl[\ , \lambda_{1}^{-2 } , \dotsc , \lambda_{n-1 } ^{-2 } \,\bigr ] x = 1\bigr\}\\ & = p^{-1 } \left\ { x \in \mathbb{r}^{n-1 } \colon \sum_{i=1}^{n-1 } \frac{x_{i}^{2}}{\lambda_{i}^{2 } } = 1 \right \}\\ pe & = \left\ { x \in \mathbb{r}^{n-1 } \colon \sum_{i=1}^{n-1 } \frac{x_{i}^{2}}{\lambda_{i}^{2 } } = 1 \right\}. \end{aligned}\ ] ]
the support function of @xmath194 is calculated in example [ es2.5:sptfnctellipsoid ] as @xmath195 and its square satisfies for all @xmath196 @xmath197^{2 } = \sum_{i=1}^{n-1 } \lambda_{i}^{2 } x_{i}^{2 } & = ( x^{\operatorname{\mathrm{t } } } p)b(p^{-1}x ) = ( p^{-1 } x)^{\operatorname{\mathrm{t}}}b(p^{-1}x)\\ \rightarrow \bigl [ ( h_{pe } \circ p)(x ) \bigr]^{2 } & = x^{\operatorname{\mathrm{t}}}bx = h^{2}(x ) . \end{aligned}\ ] ] @xmath198 so we can conclude that @xmath199^{2}$ ] , @xmath200 is the support function of an origin centered ellipsoid and @xmath201 or @xmath202 , depending on whether the nonvanishing function @xmath163 is positive or negative .
@xmath203 : for any given @xmath83 , @xmath192 , let @xmath204 x = 1 \bigr\}\ ] ] be an origin centered ellipsoid with principal axes lying on the coordinate axes and @xmath200 the support function of @xmath96 .
we need to show that @xmath205 on @xmath160 for some constant @xmath171 .
since @xmath206 is positively homogeneous of degree @xmath183 , using the expression of the laplacian in polar coordinates we can compute for any @xmath162 @xmath207 @xmath208 holds for all @xmath162 .
thus we can conclude that @xmath209 \equiv 0 \quad \text{on $ \mathbb{s}^{n-2}$.}\ ] ] a general ellipsoid @xmath97 centered at the origin is of the form @xmath210 , where @xmath211 . according to the item of lemma [ ls2.2:spfnpr ] ,
the support function @xmath99 satisfies @xmath212 for each @xmath162 , by applying the representation of @xmath213 in general coordinates to the particular case @xmath214 we get @xmath215 and hence the spherical laplacian of @xmath216 equals @xmath217 - \left .
\frac{\partial^{2}}{\partial r^{2 } } \right\rvert_{r=1 } h_{e}^{2}\bigl[rp(u)\bigr]\\ & = \delta_{\mathbb{r}^{n-1 } } h_{e}^{2}(pu ) - 2(n-1 ) h_{e}^{2 } ( pu ) = \delta_{\mathbb{s}^{n-2 } } h_{e}^{2 } ( pu ) \end{aligned}\ ] ] therefore , for each @xmath162 we can compute @xmath218 so we can conclude that @xmath219 \equiv 0 $ ] on @xmath160 . the centrix of an ovaloid @xmath102 is defined as the centrix of the strictly convex body @xmath54 that it bounds .
namely , following definition [ ds2.2:centrix ] , if @xmath163 is the support function of @xmath54 , and hence of @xmath108 by definition , then the centrix of @xmath108 is defined as @xmath220 where @xmath221 is the support parameterization of @xmath108 , @xmath222 , and the second the equality follows because @xmath223 is positively homogeneous of degree zero .
[ ls5.1:oddsptparam ] the centrix @xmath225 , when restricted to @xmath226 , coincides with @xmath227 .
the centrix is constant if and only if @xmath108 has central symmetry . in that case
, @xmath228 support parameterizes the origin centered ovaloid @xmath229 , where @xmath230 .
the first claim follows directly from the definition of @xmath28 and @xmath231 . to prove the second claim , assume first that @xmath232 is constant and consider the map @xmath233 which is the reflection through @xmath234 . given @xmath72
@xmath235 = 2 c_{0 } - \gamma(u ) = 2 \frac{\gamma(u ) + \gamma(-u)}{2 } - \gamma(u ) = \gamma(-u).\ ] ] since @xmath72 is arbitrary , we can conclude that @xmath236 and @xmath237 has central symmetry with center @xmath234 . on the other hand , if @xmath108 is central with center @xmath234 , then for every @xmath72 @xmath238}{2 } = \frac{\gamma(u ) + \gamma(-u)}{2 } = c(u)\ ] ] end hence the centrix @xmath28 is constant .
in order to show the last claim assume that @xmath108 has the center of symmetry @xmath234 and @xmath239 according to the item of lemma [ ls2.2:spfnpr ] , for every @xmath72 , the support parameterization of @xmath240 satisfies @xmath241 and hence @xmath242 so we can conclude that the odd part @xmath243 of @xmath244 support parameterizes the origin centered ovaloid @xmath229 .
our goal is to reparameterize the tube @xmath29 with the use of the support map . using this new parameterization
we will construct the height function @xmath245 that lets us parameterize nearly horizontal ovaloid cross - sections as graphs over horizontal ovaloid cross - sections . given any @xmath246 , @xmath247 , the @xmath248-tilted hyperplane is given by @xmath249 where @xmath250 is called the tilt direction , @xmath251 is the @xmath252-intercept , and @xmath248 is the slope of this hyperplane along @xmath250 direction a transversely convex tube @xmath29 in _ standard position _ can be reparameterized as follows : @xmath253 where @xmath254 is a smooth map such that for every @xmath25 , the map @xmath255 support parameterizes the ovaloid @xmath256 , where @xmath257 is the ovaloid with center of mass at @xmath258 . the smooth map @xmath112
is called the support map of @xmath29 . in particular ,
the construction of the map @xmath112 is carried out as follows : @xmath262 = \bar{\nu } \bigl(\ , c(z ) + \alpha(u , z ) , z \,\bigr)\\ & = \left ( \parbox{4.5cm}{\centering{outer unit normal in $ \mathbb{r}^{n-1}$ to the ovaloid $ \mathcal{o}(z ) - ( c(z ) , z)$ at $ ( \alpha(u , z ) , 0)$ } } , \quad z \right)\\ & = \left ( \frac{\star \bigl[\ , \bigwedge \nolimits_{n-2 } \partial_{1 } \alpha(u , z ) \,\bigr ] \bigl ( \frac{\partial}{\partial u_{1 } } \wedge \cdots \wedge \frac{\partial}{\partial u_{n-2 } } \bigr)}{\bigl\lvert \star \bigl[\ , \bigwedge \nolimits_{n-2 } \partial_{1 } \alpha(u , z ) \,\bigr ] \bigl ( \frac{\partial}{\partial u_{1 } } \wedge \cdots \wedge \frac{\partial}{\partial u_{n-2 } } \bigr ) \bigr\rvert } , \quad z \right ) \end{aligned}\ ] ] the map @xmath263 \bigl ( \frac{\partial } { \partial u_{1 } } \wedge \cdots \wedge \frac{\partial}{\partial u_{n-2 } } \bigr)}{\bigl\lvert \star \bigl[\ , \bigwedge \nolimits_{n-2 } \partial_{1 } \alpha(u , z ) \,\bigr ] \bigl ( \frac{\partial}{\partial u_{1 } } \wedge \cdots \wedge \frac{\partial}{\partial u_{n-2 } } \bigr ) \bigr\rvert}\ ] ] is smooth because @xmath119 is an @xmath264-linear map and @xmath265 is smooth .
the map @xmath266 is bijective because for every @xmath25 the ovaloid @xmath267 has a globally defined unit normal field , and the differential has a maximal rank because for every @xmath25 @xmath268 \bigl ( \frac{\partial}{\partial u_{1 } } \wedge \cdots \wedge \frac{\partial}{\partial u_{n-2 } } \bigr)}{\bigl\lvert \star \bigl[\ , \bigwedge \nolimits_{n-2 } \partial_{1 } \alpha(u , z ) \,\bigr ] \bigl ( \frac{\partial}{\partial u_{1 } } \wedge \cdots \wedge \frac{\partial}{\partial u_{n-2 } } \bigr ) \bigr\rvert}\ ] ] is a diffemorphism , which follows from a simple observation in the case @xmath17 ( oval ) , and from hadamard s theorem @xcite in the case @xmath1 . using the inverse function theorem
we can conclude that @xmath269 is a diffeomorphism and we can define the reparameterization @xmath270 of @xmath29 as follows : @xmath271 @xmath272 = x(u , z ) = \bigl ( c(z ) + \alpha(u , z ) , z \bigr)$ ] where @xmath273 \bigl ( \frac{\partial}{\partial u_{1 } } \wedge \cdots \wedge \frac{\partial}{\partial u_{n-2 } } \bigr)}{\bigl\lvert \star \bigl[\ , \bigwedge \nolimits_{n-2 } \partial_{1 } \alpha(u , z ) \,\bigr ] \bigl ( \frac{\partial}{\partial u_{1 } } \wedge \cdots \wedge \frac{\partial}{\partial u_{n-2 } } \bigr ) \bigr\rvert}\ ] ] is the outer unit normal in @xmath274 to the ovaloid @xmath275 at @xmath276 . or equivalently , @xmath277 is the outer unit normal to the ovaloid @xmath256 at the point @xmath278.therefore , if we define the map @xmath279 as @xmath280 we can conclude that @xmath281 support parameterizes the ovaloid @xmath282 for each @xmath25 and @xmath283 for each @xmath284 . because of the transverse convexity of @xmath29 and the proof of lemma [ ls3.3:exsttrncvxtube ] , given any @xmath285 , there exists @xmath289 so that for every @xmath247 , @xmath290 , @xmath291 is again an ovaloid . with this remark ,
our objective now is to show that for any given @xmath285 there exist @xmath292 and a height function @xmath293 so that for every @xmath294 , @xmath295 lets us parameterize the ovaloid @xmath291 by the map @xmath296 we define the smooth map @xmath297 as @xmath298 consider @xmath290 , and observe that @xmath299 \iff y(u , z ) \in \overline{\mathcal{o}}_{\tau}(z_{0 } , \epsilon)\\ & \iff \bigl(\ , \gamma(u , z ) , z \,\bigr ) \in \mathcal{t } \cap p_{\tau , z_{0}}(\epsilon)\\ & \iff \bigl(\ , \gamma(u , z ) , z \,\bigr ) \in p_{\tau , z_{0}}(\epsilon)\\ & \iff z = z_{0 } + \epsilon \bigl [ \tau \cdot \gamma(u , z ) \bigr]\\ & \iff g(\epsilon , u , z ) = 0 .
\end{aligned}\ ] ] fix a local parameterization @xmath302 of @xmath160 and let @xmath303 , then the map @xmath304 satisfies
@xmath305 therefore , the implicit function theorem applied to @xmath304 yields the existence of an open neighborhood of @xmath306 , where @xmath307 is an open interval about @xmath308 , @xmath309 is an open subset of @xmath310 , and a unique smooth function @xmath311 such that @xmath312 & = 0 \quad \text{for all } \epsilon \in v_{1 } \text { and } v \in v_{2}\\ \rightarrow g \circ \phi \bigl [ \epsilon , v , g(\epsilon , v ) \bigr ] & = g ( \epsilon , v ) - z_{0 } - \epsilon \tau \cdot \gamma \bigl ( \eta(v ) , g(\epsilon , v )
\bigr ) = 0\\ g \circ
\phi \bigl [ 0 , v , g(0 , v ) \bigr ] & = g(0 , v ) - z_{0 } = 0\\ g(0 , v ) & = z_{0 } \quad \text{for all $ v \in v_{2}$.}\end{aligned}\ ] ] so there exists a unique smooth function @xmath313 defined as @xmath314 since @xmath160 can be covered by finitely many overlapping open sets @xmath315 , @xmath316 , by taking the intersection @xmath317 , using the uniqueness of the smooth function @xmath318 and by letting @xmath319 we can construct a smooth function @xmath320 that satisfies @xmath321 using the observation in we get for every @xmath294 , @xmath322 & \iff g(\epsilon , u , z ) = 0\\ & \iff z = z(\epsilon , u)\end{aligned}\ ] ] and hence the ovaloid @xmath291 can be parameterized as @xmath323 therefore , the projected ovaloid @xmath324 , for each @xmath247 and @xmath294 , can be parameterized as @xmath325 using the equation we can compute the derivative for each @xmath294 and @xmath162 as @xmath326 [ ps5.2:sptreparam ] given @xmath285 and @xmath247 there exist @xmath332 and a differentiable @xmath333-parameter family of diffeomorphisms @xmath334 such that given the map @xmath335 support parameterizes @xmath324 for each @xmath294 .
the initial map @xmath336 is the identity map on @xmath160 , with the initial @xmath248-derivative given by @xmath337 \tau^{\operatorname{\mathrm{t}}}(u)\\ & = \left ( u \cdot \frac{\partial \gamma}{\partial z}(u , z_{0 } ) \right ) \tau^{\operatorname{\mathrm{t}}}(u ) .
\end{aligned}\ ] ] let @xmath338 be the interval obtained in definition [ es5.2:impfctdom ] of the height function @xmath245 , then for each @xmath339 the map @xmath340 parameterizes @xmath324 , and @xmath341 is smooth . define the smooth map @xmath342 and for every @xmath294 the map @xmath343 which , according to remark [ rs3.2:outrunitnorfld ] , is the outer unit normal field along the closed embedded hypersurface @xmath344 .
the map @xmath345 is a diffeomorphism because @xmath346 parameterizes the ovaloid @xmath324 and the outer unit normal field @xmath347 , in , is a diffeomorphism as a result of a simple observation in the case @xmath17 ( oval ) , and from hadamard s theorem @xcite in the case @xmath1 .
since @xmath348 is bijective and its differential has full rank , by the inverse function theorem , @xmath348 is a diffeomorphism with the smooth inverse @xmath349 where @xmath350 , @xmath294 .
since @xmath351 is smooth , @xmath352 is a @xmath333-parameter family of diffeomorphisms . for each @xmath353 ,
since @xmath354 parameterizes @xmath355 , the map @xmath356 support parameterizes @xmath324 because the unit vector @xmath162 is the outer unit normal to the ovaloid @xmath357 at @xmath358 fix @xmath162 , @xmath364 , and choose a smooth curve @xmath365 so that @xmath366 and @xmath367
. then for every @xmath339 @xmath368 = \left .
\frac{\partial}{\partial t } \right\rvert_{t=0 } u \cdot ( \gamma_{\epsilon } \circ \theta_{\epsilon } ) \bigr [ c(t ) \bigr ] .
\notag\\ \intertext{taking the $ \epsilon$-derivative at $ \epsilon = 0 $ and using the smoothness of $ \gamma$ we obtain } 0 & = \left .
\frac{\partial}{\partial \epsilon } \right\rvert_{\epsilon=0 } \left .
\frac{\partial}{\partial t } \right\rvert_{t=0 } \bigl\ { u \cdot \gamma \bigl(\ , \theta_{\epsilon}(c(t ) ) , z\bigl ( \epsilon , \theta_{\epsilon}(c(t ) ) \bigr ) \,\bigr ) \bigr \ } \notag\\ & = \left .
\frac{\partial}{\partial t } \right\rvert_{t=0 } \left\ { u \cdot \left .
\frac{\partial}{\partial \epsilon } \right \rvert_{\epsilon=0 } \gamma \bigl(\ , \theta_{\epsilon}(c(t ) ) , z\bigl ( \epsilon , \theta_{\epsilon}(c(t ) ) \bigr ) \
, \bigr)\right\}. \label{es5.2:drvcompsptfnc1 } \end{aligned}\ ] ] @xmath369 is a smooth curve on @xmath370 whose @xmath248-derivative at @xmath359 equals @xmath371 since @xmath372 , @xmath373 , using the equation we can compute @xmath374 and hence the tangent vector equals @xmath375 the computation of the derivative in can be continued as @xmath376 & = ( \partial_{1}\gamma ) \bigl ( c(t ) , z_{0 } \bigr ) \left ( \left .
\frac{\partial}{\partial \epsilon } \right\rvert_{\epsilon=0 } \theta_{\epsilon } \bigl [ c(t ) \bigr ] \right)\\ & \phantom{(\partial_{1}\gamma ) \bigl [ c(t ) , z_{0 } \bigr ] + } + ( \partial_{2 } \gamma ) \bigl ( c(t ) , z_{0 } \bigr ) \bigl(\ , \tau \cdot \gamma \bigl ( c(t ) , z_{0 } \bigr ) \,\bigr ) \end{aligned}\ ] ] @xmath377 \right\}\\ & \phantom{\left .
\frac{\partial}{\partial t } \right\rvert_{t=0 } \left\ { u \cdot ( \partial_{1}\gamma ) \bigl ( c(t ) \right\ } } + \left .
\frac{\partial}{\partial t } \right\rvert_{t=0 } \bigl\ { u \cdot ( \partial_{2 } \gamma ) \bigl ( c(t ) , z_{0 } \bigr ) \bigl ( \tau \cdot \gamma \bigl ( c(t ) , z_{0 } \bigr ) \bigr ) \bigr\}\\ & = u \cdot \left
. \frac{\partial}{\partial t } \right\rvert_{t=0 } \left\ { \operatorname{\mathrm{d}\!}\gamma \bigl ( c(t ) , z_{0 } \bigr ) \left ( \left .
\frac{\partial}{\partial \epsilon } \right\rvert_{\epsilon=0 } \bigl ( \theta_{\epsilon } \bigl [ c(t ) \bigr ] , z_{0 }
\bigr ) \right ) \right\}\\ & \phantom{\left .
\frac{\partial}{\partial t } \right\rvert_{t=0 } \left\ { u \cdot ( \partial_{1}\gamma ) \bigl ( c(t ) \right\ } } + u \cdot \left
. \frac{\partial}{\partial t } \right\rvert_{t=0 } \left\ { \left ( \frac{\partial}{\partial z } \right \rvert_{z = z_{0 } } \gamma \bigl ( c(t ) , z \bigr ) \right ) \bigl ( \tau \cdot \gamma \bigl ( c(t ) , z_{0 } \bigr ) \bigr)\\ & = u \cdot \left .
\frac{\partial}{\partial t } \right\rvert_{t=0 } \left .
\frac{\partial}{\partial \epsilon } \right \rvert_{\epsilon=0 } \gamma \bigl(\ , \theta_{\epsilon } \bigl ( c(t ) \bigr ) , z_{0 } \,\bigr)\\ & \phantom{\left .
\frac{\partial}{\partial t } \right\rvert_{t=0 } \left\
{ u \cdot ( \partial_{1}\gamma ) \bigl ( c(t ) \right\ } } + u \cdot \left\ { \left ( \left . \frac{\partial}{\partial t } \right\rvert_{t=0 } \left .
\frac{\partial}{\partial z } \right\rvert_{z = z_{0 } } \gamma \bigl ( c(t ) , z \bigr ) \right ) \bigl(\tau \cdot \gamma ( u , z_{0 } ) \bigr ) \right\}\\ & \phantom{\left .
\frac{\partial}{\partial t } \right\rvert_{t=0 } \left\ { u \cdot ( \partial_{1}\gamma ) \bigl ( c(t ) \right\ } } + u \cdot \left\ { \left ( \left .
\frac{\partial}{\partial z } \right\rvert_{z = z_{0 } } \gamma(u , z ) \right ) \left ( \tau \cdot \left .
\frac{\partial}{\partial t } \right\rvert_{t=0 } \gamma \bigl ( c(t ) , z_{0 } \bigr ) \right ) \right\ } \end{aligned}\ ] ] @xmath378 = \operatorname{\mathrm{d}\!}\gamma_{0}(u)v , \quad \text{we continue as}\ ] ] @xmath379 \bigl ( \tau \cdot \operatorname{\mathrm{d}\!}\gamma_{0}(u)v \bigr ) .
\end{aligned}\ ] ] for each @xmath380 , @xmath281 support parameterizes @xmath381 , and @xmath382 \in \operatorname{\mathrm{t}}_{\gamma(u , z ) } \mathcal{o } ( z ) = \operatorname{\mathrm{t}}_{u } \mathbb{s}^{n-2}\\ & \rightarrow u \cdot \left .
\frac{\partial}{\partial t } \right\rvert_{t=0 } \gamma \bigl ( c(t ) , z\bigr ) = 0 \end{aligned}\ ] ] therefore , we can continue the computation as @xmath383 \bigl ( \tau \cdot \operatorname{\mathrm{d}\!}\gamma_{0}(u)v \bigr)\\ & = u \cdot \left .
\frac{\partial}{\partial \epsilon } \right\rvert_{\epsilon=0 } \operatorname{\mathrm{d}\!}\gamma_{0 } \bigl [ \theta_{\epsilon } ( u ) \bigr ] \bigl ( \operatorname{\mathrm{d}\!}\theta_{\epsilon}(u)v \bigr ) + \left [ u \cdot \frac{\partial \gamma}{\partial z } ( u , z_{0 } ) \right ] \bigl ( \tau \cdot \operatorname{\mathrm{d}\!}\gamma_{0}(u)v \bigr ) .
\end{aligned}\ ] ] in summary , we have the equality @xmath384 \bigl ( \operatorname{\mathrm{d}\!}\theta_{\epsilon}(u)v \bigr ) + \left [ u \cdot \frac{\partial \gamma}{\partial z } ( u , z_{0 } ) \right ] \bigl ( \tau \cdot \operatorname{\mathrm{d}\!}\gamma_{0}(u)v \bigr ) . \end{aligned}\ ] ] note that @xmath385 support parameterizes @xmath386 and so for every @xmath387 @xmath388 \bigl ( \operatorname{\mathrm{d}\!}\theta_{\epsilon}(u)v \bigr ) = 0.\ ] ] when we take the @xmath248-derivative of the equation at @xmath359 we get @xmath389 \bigl ( \operatorname{\mathrm{d}\!}\theta_{\epsilon}(u)v \bigr ) \bigr\}\\ & = \left [ \left .
\frac{\partial}{\partial \epsilon } \right\rvert_{\epsilon=0 } \theta_{\epsilon}(u ) \right ] \cdot \operatorname{\mathrm{d}\!}\gamma_{0}(u)v + u \cdot \left .
\frac{\partial}{\partial \epsilon } \right\rvert_{\epsilon=0 } \operatorname{\mathrm{d}\!}\gamma_{0 } \bigl [ \theta_{\epsilon}(u ) \bigr ] \bigl ( \operatorname{\mathrm{d}\!}\theta_{\epsilon}(u)v \bigr ) \end{aligned}\ ] ] @xmath390 \bigl(\operatorname{\mathrm{d}\!}\theta_{\epsilon}(u)v \bigr ) = - \left [ \left .
\frac{\partial}{\partial \epsilon } \right \rvert_{\epsilon=0 } \theta_{\epsilon}(u ) \right ] \cdot \operatorname{\mathrm{d}\!}\gamma_{0}(u)v.\ ] ] using the equality in the equation , we get @xmath391 \cdot \operatorname{\mathrm{d}\!}\gamma_{0}(u)v + \left [ u \cdot \frac{\partial \gamma}{\partial z}(u , z_{0 } ) \right ] \bigl ( \tau \cdot \operatorname{\mathrm{d}\!}\gamma_{0}(u)v \bigr)\\ & = \left [ \left ( u \cdot \frac{\partial \gamma}{\partial z}(u , z_{0 } ) \right ) \tau - \left .
\frac{\partial}{\partial \epsilon } \right\rvert_{\epsilon=0 } \theta_{\epsilon}(u ) \right ] \cdot \operatorname{\mathrm{d}\!}\gamma_{0}(u)v\\ \intertext{since $ \operatorname{\mathrm{d}\!}\gamma_{0}(u)v \in \operatorname{\mathrm{t}}_{\gamma(u , z_{0 } ) } \mathcal{o}(z_{0 } ) = \operatorname{\mathrm{t}}_{u}\mathbb{s } ^{n-2}$ we can replace $ \tau$ with $ \tau^{\operatorname{\mathrm{t}}}(u)$ } & = \left [ \left ( u \cdot \frac{\partial \gamma}{\partial z}(u , z_{0 } ) \right ) \tau^{\operatorname{\mathrm{t}}}(u ) - \left .
\frac{\partial } { \partial \epsilon } \right\rvert_{\epsilon=0 } \theta_{\epsilon}(u ) \right ] \cdot \operatorname{\mathrm{d}\!}\gamma_{0}(u)v \end{aligned}\ ] ] holds for every @xmath392 . since @xmath393 is an isomorphism we can conclude that @xmath394 holds for every @xmath162 .
our goal is to get an analytic formulation of central symmetry .
since the constancy of the centrix is equivalent to central symmetry , an appropriate derivative of the centrix , which we call the symmetry obstruction , will provide the needed formulation . given any tilt direction @xmath247 , @xmath285 , and @xmath332 as in proposition [ ps5.2:sptreparam ] ,
define the @xmath333-parameter family of smooth maps @xmath395 + \gamma \bigl [ \theta_{\epsilon}(-u ) \bigr]}{2 } \end{aligned}\ ] ] namely , for each @xmath294 , @xmath396 is the centrix of the projected ovaloid @xmath397 .
let @xmath398 denote the odd part of @xmath281 for each @xmath399 and define the maps @xmath400 & \bar{\theta}_{\epsilon } \colon \mathbb{s}^{n-2 } \to \mathbb{s}^{n-2 } & & \\ & \bar{\theta}_{\epsilon } = ( \theta_{\epsilon } \circ \rho)(u ) .
& & \end{aligned}\ ] ] then we can compute for each @xmath162 @xmath401 @xmath402 which can be equivalently written as @xmath403 supose the horizontal cross - section @xmath404 of a transversely convex tube @xmath29 is central about @xmath405 for each @xmath399 .
then for any tilt direction @xmath247 and @xmath285 , we have @xmath406\\ & + \bigl [ \tau \cdot c(z_{0 } ) \bigr ] c'(z_{0 } ) + \bigl [ \tau \cdot \gamma^{-}(u , z_{0})\bigr ] \frac{\partial \gamma^{-}}{\partial z } ( u , z_{0 } ) . \end{aligned}\ ] ] using the notation in the centrix can be written as @xmath407 and the @xmath248-derivative at @xmath359 of the first summand equals @xmath408
\right ) . \label{es5.2:dervcompcnt1 } \end{aligned}\ ] ] since @xmath409 , it follows that @xmath410 and using the equation the derivative in simplifies to @xmath411 + \frac{\partial \gamma}{\partial z}(u , z_{0 } ) \bigl [ \tau \cdot \gamma(u , z_{0 } ) \bigr]\ ] ] for each @xmath25 , @xmath281 support parameterizes the projected ovaloid @xmath404 with center of symmetry at @xmath412 , @xmath413 , and according to lemma [ ls5.1:oddsptparam ] , @xmath414 support parameterizes the ovaloid @xmath415 . using the previous equality and the equation we can conclude that @xmath416 for each @xmath162 and @xmath25 . using these observations
we can compute the @xmath248-derivative of @xmath417 at @xmath359 as follows @xmath418 + \frac{\partial \gamma}{\partial z}(u , z_{0 } ) \bigl [ \tau \cdot \gamma(u , z_{0 } ) \bigr ] \right\}\\ & \quad+ \frac{1}{2 } \left\{(\partial_{1 } \gamma ) ( -u , z_{0 } ) \left [ \left .
\frac{\partial}{\partial \epsilon } \right \rvert_{\epsilon=0 } \bar{\theta}_{\epsilon}(u ) \right ] + \frac{\partial \gamma}{\partial z}(-u , z_{0 } ) \bigl [ \tau \cdot \gamma(-u , z_{0 } ) \bigr ] \right\}\\ & = \frac{1}{2 } \left\ { ( \partial_{1 } \gamma^{- } ) ( u , z_{0 } ) \left [ \left .
\frac{\partial}{\partial \epsilon } \right \rvert_{\epsilon=0 } \theta_{\epsilon}(u ) \right ] \right\}\\ & \quad + \frac{1}{2 } \left\ { \left ( c'(z_{0 } ) + \frac{\partial \gamma^{-}}{\partial z } ( u , z_{0 } ) \right ) \bigl ( \tau \cdot \bigl [ c(z_{0 } ) + \gamma^{-}(u , z_{0 } ) \bigr ] \bigr ) \right\}\\ & \qquad + \frac{1}{2 } \left\ { ( \partial_{1 } \gamma ) ( u , z_{0 } ) \left [ \left .
\frac{\partial}{\partial \epsilon } \right \rvert_{\epsilon=0 } \bar{\theta}_{\epsilon}(u ) \right ] \right\}\\ & \quad \qquad + \frac{1}{2 } \left\ { \left ( c'(z_{0 } ) - \frac{\partial \gamma^{-}}{\partial z } ( u , z_{0 } ) \right ) \bigl ( \tau \cdot \bigl [ c(z_{0 } ) - \gamma^{-}(u , z_{0 } ) \bigr ] \bigr ) \right\ } \end{aligned}\ ] ] @xmath419\right\ } + \bigl [ \tau \cdot c(z_{0 } ) \bigr ] c'(z_{0})\\ & \qquad + \bigl [ \tau \cdot \gamma^{-}(u , z_{0 } ) \bigr ] \frac{\partial \gamma^{-}}{\partial z}(u , z_{0 } ) \end{aligned}\ ] ] the proposition [ ps5.2:sptreparam ] gives the @xmath248-derivatives of @xmath327 and @xmath420 as @xmath421 \tau^{\operatorname{\mathrm{t}}}(u)\ ] ] @xmath422 \tau^{\operatorname{\mathrm{t}}}(-u)\\ & = u \cdot \left [ \frac{\partial \gamma^{-}}{\partial z}(u , z_{0 } ) - c'(z_{0 } ) \right ] \tau^{\operatorname{\mathrm{t}}}(u ) \end{aligned}\ ] ] plugging the derivatives and into the equation we get @xmath423\\ & + \bigl [ \tau \cdot c(z_{0 } ) \bigr ] c'(z_{0 } ) + \bigl [ \tau \cdot \gamma^{-}(u , z_{0})\bigr ] \frac{\partial \gamma^{-}}{\partial z } ( u , z_{0 } ) . \qedhere \end{aligned}\ ] ] [ the symmetry obstruction ] given any tilt direction @xmath247 and @xmath285 , @xmath324 is central if and only if @xmath424 is constant , or equivalently , @xmath425 for every @xmath162 and @xmath364 , where the vector field @xmath426 along @xmath427 is extended radially constant in an open neighborhood of @xmath160 in @xmath184 to a vector field in an open subset of @xmath184 . when @xmath324 has central symmetry for all sufficiently small @xmath428 , which _
cop _ requires , we have @xmath429_{u}\ ] ] for every @xmath162 and @xmath364 .
therefore , the derivative in forms an obstruction to central symmetry and hence to _ cop_. the spherical divergence of the vectorfield @xmath430 equals @xmath431(u ) & = \sum_{j=1}^{n-2 } e_{j}(u ) \cdot \nabla_{e_{j}(u)}^{\mathbb{s}^{n-2 } } \left [ \left .
\frac{\partial } { \partial \epsilon } \right\rvert_{\epsilon=0 } c_{\epsilon}(\cdot , z_{0 } ) \right]_{u}\\ & = \sum_{j=1}^{n-2 } e_{j}(u ) \cdot \overline{\nabla}_{e_{j}(u ) } \left [ \left .
\frac{\partial}{\partial \epsilon } \right\rvert_{\epsilon=0 } c_{\epsilon}(\cdot , z_{0 } ) \right]_{u}\end{aligned}\ ] ] where @xmath432 is the local orthonormal field constructed in definition [ ds3.2:orthnmfield ] .
since the equality above holds for all @xmath433 we can define a function for each tilt direction @xmath247 as @xmath434(u).\end{aligned}\ ] ] the central ovaloid property of the tube @xmath29 implies @xmath435 for every @xmath162 , @xmath25 , and @xmath247 . for each @xmath436
, @xmath414 support parameterizes the horizontal cross - section @xmath437 of the rectified tube @xmath35 .
therefore , there exists a smooth function @xmath438 which , for each @xmath436 , yields the support function of the ovaloid @xmath415 .
we call @xmath163 the transverse support function of @xmath35 .
[ ps5.2:pde ] on a transversely convex tube @xmath35 with central ovaloid property , the transverse support function @xmath163 of @xmath439 satisfies the two partial differential equations : @xmath440 \ , \bigr\rvert_{u}\\ & \
+ \frac{\partial}{\partial z } \bigl[\ , \delta_{\mathbb{s}^{n-2 } } h(u , z ) + ( n-2 ) h(u , z ) \,\bigr ] \nabla_{\mathbb{s}^{n-2 } } h(\cdot , z ) \ , \bigr\rvert_{u}\\ & \ + 2 \left\ { h(u , z ) \nabla_{\mathbb{s}^{n-2 } } \frac{\partial h}{\partial z}(\cdot , z ) \ , \bigr\rvert_{u } + \nabla_{\textstyle \nabla_{\mathbb{s}^{n-2 } } \left .
\frac{\partial h}{\partial z } ( \cdot , z ) \,\right\rvert_{u } } ^{\mathbb{s}^{n-2 } } \nabla_{\mathbb{s}^{n-2 } } h ( \cdot , z ) \ , \bigr\rvert_{u }
\right\ } = 0 \end{aligned}\ ] ] @xmath441\\ & - \bigl [ \delta_{\mathbb{s } ^{n-2 } } h(u , z ) + ( n-2 ) h(u , z ) \bigr ] \frac{\partial h}{\partial z}(u , z ) = 0 \end{aligned}\ ] ] our aim is to show that for an given @xmath162 , @xmath443 , and @xmath247 @xmath444 \bigr\rvert_{u } \right.\\ & \phantom{= f_{\tau}(u , z_{0 } ) = \tau^{\operatorname{\mathrm{t}}}(u ) \cdot \ } + \frac{\partial}{\partial z } \left [ \delta h(u , z_{0 } ) + ( n-2)h(u , z_{0 } ) \right ] \nabla h(\cdot , z_{0 } ) \ ,
\bigr\rvert_{u } \notag\\ & \phantom{= f_{\tau}(u , z_{0 } ) = \tau^{\operatorname{\mathrm{t}}}(u ) \cdot \ } + 2 \left .
\left ( h(u , z_{0 } ) \nabla \frac{\partial h}{\partial z}(\cdot , z_{0 } ) \ , \bigr\rvert_{u } + \nabla_{\textstyle \nabla \left .
\frac{\partial h}{\partial z } ( \cdot , z_{0 } ) \ ,
\right \rvert_{u } } \nabla h ( \cdot , z_{0 } ) \ ,
\bigr\rvert_{u } \right ) \right\ } \notag\\ & \phantom{= f_{\tau}(u , z_{0 } ) } + ( \tau \cdot u ) \left\ { h(u , z_{0 } )
\frac{\partial}{\partial z } \bigl [ \delta h(u , z_{0 } ) + ( n-2 ) h(u , z_{0 } ) \bigr ] \right.\\ & \phantom{= f_{\tau}(u , z_{0 } ) = \tau^{\operatorname{\mathrm{t}}}(u ) \quad}\left .
- \bigl [ \delta h(u , z_{0 } ) + ( n-2 ) h(u , z_{0 } ) \bigr ] \frac{\partial h}{\partial z}(u , z_{0 } ) \right\}\notag \end{aligned}\ ] ] since @xmath247 is arbitrary and the expressions in curly brackets do not depend on @xmath250 we can conclude that each one of the two expressions must vanish and thus we get the two partial differential equations and . throughout the proof assume that each @xmath455 , @xmath456 , @xmath457 , @xmath452 and @xmath106 is extended radially constant in a neighborhood of @xmath160 in @xmath184 to a vectorfield in an open subset of @xmath184 . using the equation
the function @xmath458 can be expanded as @xmath459(u)\\ & = \operatorname{\mathrm{div}}\left [ v \mapsto \bigl ( \tau \cdot \gamma^{-}(v , z_{0 } ) \bigr ) \frac{\partial \gamma^{-}}{\partial z}(v , z_{0 } ) \right](u ) + \operatorname{\mathrm{div}}\bigl [ v \mapsto \bigl ( \tau \cdot c(z_{0 } ) \bigr ) c'(z_{0 } ) \bigr](u)\\ & \quad + \operatorname{\mathrm{div}}\left [ v \mapsto \left ( \frac{\partial \gamma^{-}}{\partial z}(v , z_{0 } ) \cdot v \right ) ( \partial_{1 } \gamma^{-})(v , z_{0 } ) \bigl [ \tau^{\operatorname{\mathrm{t}}}(v ) \bigr ] \right](u ) .
\end{aligned}\ ] ] clearly the second summand vanishes and @xmath460 is equal to @xmath461(u)}_{{\ensuremath{\text{\textcircled{\raisebox{-0.2ex}{\textbf{1}}}}}}}\\ & \quad + \underbrace{\operatorname{\mathrm{div}}\left [ v \mapsto \left ( \frac{\partial \gamma^{-}}{\partial z } ( v , z_{0 } ) \cdot v \right ) ( \partial_{1 } \gamma^{-})(v , z_{0 } ) \bigl [ \tau^{\operatorname{\mathrm{t}}}(v ) \bigr ] \right](u)}_{{\ensuremath{\text{\textcircled{\raisebox{-0.3ex}{\textbf{2 } } } } } } } \end{aligned}\ ] ] the computation of the part @xmath462 of the equation is carried out as follows : @xmath463(u)\\ & = \sum_{j=1}^{n-2 } e_{j}(u ) \cdot \overline{\nabla}_{e_{j}(u ) } \left [ v \mapsto \bigl ( \tau \cdot \gamma^{- } ( v , z_{0})\bigr ) \frac{\partial \gamma^{-}}{\partial z}(v , z_{0 } ) \right]_{u}\\ & = \underbrace{\sum_{j=1}^{n-2 } e_{j}(u ) \cdot \left\ { \underbrace{e_{j}\bigl [ v \mapsto \tau \cdot \gamma^{-}(v , z_{0})\bigr](u)}_{{\ensuremath{\text{\textcircled{\raisebox{-0.2ex}{\textbf{1 } } } } } } } \frac{\partial \gamma^{-}}{\partial z}(u , z_{0 } ) \right\}}_{{\ensuremath{\text{\textcircled{\raisebox{-0.3ex}{\textbf{2}}}}}}}\\ & \quad + \underbrace{\sum_{j=1}^{n-2}\bigl [ \tau \cdot \gamma^{-}(u , z_{0 } ) \bigr ] e_{j}(u ) \cdot
\underbrace{\overline{\nabla}_{e_{j}(u ) } \left [ v \mapsto \frac{\partial \gamma^{-}}{\partial z}(v , z_{0 } ) \right]_{u}}_{{\ensuremath{\text{\textcircled{\raisebox{-0.2ex}{\textbf{3}}}}}}}}_{{\ensuremath{\text{\textcircled{\raisebox{-0.2ex}{\textbf{4 } } } } } } } \end{aligned}\ ] ] the part @xmath462 of the equation can be expanded as @xmath464(u ) = \tau \cdot \overline{\nabla}_{e_{j}(u ) } \bigl [ v \mapsto \gamma^{-}(v , z_{0 } ) \bigr ] \ , \bigr\rvert_{u}\\ & = \tau \cdot \overline{\nabla}_{e_{j}(u ) } \bigl [ v \mapsto \nabla h(\cdot , z_{0 } ) \ , \bigr\rvert_{v } + h(v , z_{0})v \bigr ] \ , \bigr\rvert_{u}\\ & = \tau \cdot \overline{\nabla}_{e_{j}(u ) } \nabla h(\cdot , z_{0 } ) \ , \bigr\rvert_{u }
+ \tau \cdot \overline{\nabla}_{e_{j}(u ) } \bigl[v \mapsto h(v , z_{0})v \bigr ] \ , \bigr\rvert_{u}\\ & = \tau^{\operatorname{\mathrm{t}}}(u ) \cdot \nabla_{e_{j}(u ) } \nabla h(\cdot , z_{0 } ) \
, \bigr\rvert_{u } + \bigl[(\tau \cdot u)u \bigr ] \cdot \overline{\nabla}_{e_{j}(u ) } \nabla h(\cdot , z_{0 } ) \ , \bigr\rvert_{u}\\ & \qquad + \tau \cdot \left\ { e_{j}\bigr[h(\cdot , z_{0})\bigr](u)u + h(u , z_{0 } ) \overline{\nabla}_{e_{j}(u)}n \ , \bigr\rvert_{u } \right\ } \end{aligned}\ ] ] recalling the fact that @xmath465 for each @xmath162 we obtain @xmath466(u)}_{{\ensuremath{\text{\textcircled{\raisebox{-0.2ex}{\textbf{1 } } } } } } } & = \tau^{\operatorname{\mathrm{t}}}(u ) \cdot \underbrace{\nabla_{e_{j}(u ) } \nabla h(\cdot , z_{0 } ) \ , \bigr\rvert_{u}}_{{\ensuremath{\text{\textcircled{\raisebox{-0.3ex}{\textbf{2}}}}}}}\\ & + ( \tau \cdot u ) \underbrace{u \cdot \overline{\nabla}_{e_{j}(u ) } \nabla h(\cdot , z_{0 } ) \ , \bigr\rvert_{u } } _ { { \ensuremath{\text{\textcircled{\raisebox{-0.2ex}{\textbf{3}}}}}}}\\ & + \tau \cdot \bigl\ { e_{j}\bigl[h(\cdot , z_{0})\bigr](u)u + h(u , z_{0})e_{j}(u)\bigr\}. \end{aligned}\ ] ] the part @xmath467 of the equation can be written as @xmath468 e_{k } \right]_{u}\\ & = \sum_{k=1}^{n-2 } \nabla_{e_{j}(u ) } \bigl ( e_{k } \bigl [ h(\cdot , z_{0 } ) \bigr ] e_{k } \bigr ) \bigr\rvert_{u}\\ & = \sum_{k=1}^{n-2 } e_{j}\bigl ( e_{k } \bigl [ h(\cdot , z_{0 } ) \bigr ] \bigr)(u ) e_{k}(u ) + e_{k } \bigl [ h(\cdot , z_{0 } ) \bigr](u ) \underbrace{\nabla_{e_{j}(u)}e_{k } \ , \bigr\rvert_{u}}_{=0}\\ & = \sum_{k=1}^{n-2 } e_{j}\bigl ( e_{k } \bigl [ h(\cdot , z_{0 } ) \bigr ] \bigr)(u ) e_{k}(u ) .
\end{aligned}\ ] ] the part @xmath469 of the equation can be simplified as @xmath470 \ , \big\rvert_{u } - \overline{\nabla}_{e_{j}(u ) }
n \ , \bigr\rvert_{u } \cdot \nabla h(\cdot , z_{0 } ) \ , \bigr\rvert_{u}\\ & = - e_{j}(u ) \cdot
\nabla h(\cdot , z_{0 } ) \ , \bigr\rvert_{u}. \end{aligned}\ ] ] since @xmath471 the part @xmath462 of the equation can be written as @xmath472(u)\\ & = \tau^{\operatorname{\mathrm{t}}}(u ) \cdot \left\ { \sum_{k=1}^{n-2 } e_{j } \bigl ( e_{k } \bigl [ h(\cdot , z_{0 } ) \bigr ] \bigr)(u ) e_{k}(u ) + h(u , z_{0 } ) e_{j}(u ) \right\}\\ & \phantom{\tau^{\operatorname{\mathrm{t}}}(u ) \cdot } + ( \tau \cdot u ) \bigl\
{ e_{j } \bigl [ h(\cdot , z_{0 } ) \bigr](u ) - e_{j}(u ) \cdot
\nabla h(\cdot , z_{0 } ) \ , \bigr\rvert_{u}\bigr\}. \end{aligned}\ ] ] in order to compute the part @xmath467 of the equation we first observe that @xmath473 \notag\\ & = e_{j}(u ) \cdot \left [ \nabla \frac{\partial h}{\partial z}(\cdot , z_{0 } ) \ ,
\bigr\rvert_{u } + \frac{\partial h } { \partial z}(u , z_{0})u \right ] = e_{j}(u ) \cdot \nabla \frac{\partial h}{\partial z}(\cdot , z_{0 } ) \ ,
\bigr\rvert_{u } \end{aligned}\ ] ] and then by using the equation we can write @xmath474(u ) \left [ e_{j}(u ) \cdot
\nabla \frac{\partial h}{\partial z}(\cdot , z_{0 } ) \ ,
\bigr\rvert_{u } \right]\\ & = \sum_{j=1}^{n-2 } \left\ { \tau^{\operatorname{\mathrm{t}}}(u ) \cdot \left ( \sum_{k=1}^{n-2 } e_{j } \bigl ( e_{k } \bigl [ h(\cdot , z_{0 } ) \bigr ] \bigr)(u ) e_{k}(u ) + h(u , z_{0 } ) e_{j}(u ) \right ) \right.\\\ & \hspace{1 cm } + ( \tau \cdot u ) \bigl ( e_{j } \bigl [ h(\cdot , z_{0 } ) \bigr](u ) - e_{j}(u ) \cdot
\nabla h(\cdot , z_{0 } ) \ , \bigr\rvert_{u}\bigr ) \biggr\ } \left [ e_{j}(u ) \cdot
\nabla \frac{\partial h}{\partial z}(\cdot , z_{0 } ) \ , \bigr\rvert_{u } \right ] \end{aligned}\ ] ] @xmath475 e_{j } \bigl ( e_{k } \bigl [ h(\cdot , z_{0 } ) \bigr ] \bigr)(u ) \right ) e_{k}(u)\\ & + \tau^{\operatorname{\mathrm{t}}}(u ) \cdot \sum_{j=1}^{n-2 } \left [ e_{j}(u ) \cdot \nabla \frac{\partial h}{\partial z}(\cdot , z_{0 } ) \ , \bigr\rvert_{u } \right ] h(u , z_{0 } ) e_{j}(u)\\ & + ( \tau \cdot u ) \left ( \sum_{j=1}^{n-2 } \left [ e_{j}(u ) \cdot
\nabla \frac{\partial h}{\partial z}(\cdot , z_{0 } ) \ , \bigr \rvert_{u } \right ] e_{j } \bigl [ h(\cdot , z_{0 } ) \bigr](u ) - \nabla h(\cdot , z_{0 } ) \
, \bigr\rvert_{u } \cdot \nabla \frac{\partial h}{\partial z}(\cdot , z_{0 } ) \
, \bigr\rvert_{u } \right ) \end{aligned}\ ] ] where the last dot product is obtained as @xmath476 e_{j}(u ) \cdot \nabla h(\cdot , z_{0 } ) \ , \bigr\rvert_{u}\\ & = \nabla h(\cdot , z_{0 } ) \ , \bigr\rvert_{u } \cdot \sum_{j=1}^{n-2 } \left [ e_{j}(u ) \cdot
\nabla \frac{\partial h } { \partial z}(\cdot , z_{0 } ) \ , \bigr\rvert_{u } \right ] e_{j}(u)\\ & = \nabla h(\cdot , z_{0 } ) \ , \bigr\rvert_{u } \cdot \nabla \frac{\partial h}{\partial z } ( \cdot , z_{0 } ) \ , \bigr\rvert_{u } \end{aligned}\ ] ]
so we can conclude that the part @xmath467 of the equation equals @xmath477(u ) e_{j}(u ) \cdot \frac{\partial \gamma^{-}}{\partial z}(u , z_{0 } ) \\ & = \tau^{\operatorname{\mathrm{t}}}(u ) \cdot \biggl\ { \sum_{k=1}^{n-2 } \biggl ( \sum_{j=1}^{n-2 } \biggl [ e_{j}(u ) \cdot \nabla \frac{\partial h}{\partial z}(\cdot , z_{0 } ) \ , \bigr\rvert_{u } \biggr ]
e_{j } \bigl ( e_{k } \bigl [ h(\cdot , z_{0 } ) \bigr ] \bigr)(u ) \biggr ) e_{k}(u ) \\ & \hspace{2.5 cm } + h(u , z_{0 } ) \nabla \frac{\partial h}{\partial z}(\cdot , z_{0 } ) \ , \bigr\rvert_{u } \biggr\ } \end{aligned}\ ] ] note that @xmath478 terms vanishes and is obtained because @xmath479 e_{j}\bigl [ h(\cdot , z_{0 } ) \bigr](u ) = \nabla h(\cdot , z_{0 } ) \
, \bigr\rvert_{u } \cdot \nabla \frac{\partial h}{\partial z } ( \cdot , z_{0 } ) \ , \bigr\rvert_{u}.\ ] ] in oder to compute the part @xmath480 of the equation , which is @xmath481 \sum_{j=1}^{n-2 } e_{j}(u ) \cdot \overline{\nabla}_{e_{j}(u)}\left [ v \mapsto \frac{\partial \gamma^{-}}{\partial z}(v , z_{0 } ) \right]_{u}\ ] ] we first expand the part @xmath469 of the same equation as @xmath482_{u } = \overline{\nabla}_{e_{j}(u ) } \left [ v \mapsto \nabla \frac{\partial h}{\partial z}(v , z_{0 } ) + \frac{\partial h}{\partial z}(v , z_{0})v \right]_{u}\\ & = \overline{\nabla}_{e_{j}(u ) } \left [ v \mapsto \nabla \frac{\partial h}{\partial z}(v , z_{0 } ) \right]_{u } + \overline{\nabla}_{e_{j}(u ) } \left [ v \mapsto \frac{\partial h}{\partial z}(v , z_{0 } ) n_{v } \right]_{u}\\ & = \overline{\nabla}_{e_{j}(u ) } \left [ \nabla \frac{\partial h}{\partial z}(\cdot , z_{0 } ) \right]_{u } + e_{j } \left [ \frac{\partial h}{\partial z}(\cdot , z_{0 } ) \right ] ( u ) \ , u + \frac{\partial h}{\partial z } ( u , z_{0 } ) \overline{\nabla}_{e_{j}(u ) } n \ , \bigr\rvert_{u } \end{aligned}\ ] ] and when we sum over @xmath483 the dot product with @xmath484 we get @xmath485_{u}\\ & = \sum_{j-1}^{n-2 } e_{j}(u ) \cdot \overline{\nabla}_{e_{j}(u ) } \left [ \nabla \frac{\partial h}{\partial z } ( \cdot , z_{0 } ) \right]_{u } + \sum_{j=1}^{n-2 } \frac{\partial h}{\partial z}(u , z_{0})\\ & = \operatorname{\mathrm{div}}\left [ \nabla \frac{\partial h}{\partial z}(\cdot , z_{0 } ) \right](u ) + ( n - 2 ) \frac{\partial h}{\partial z}(u , z_{0 } ) \\ & = \delta \frac{\partial h}{\partial z}(\cdot , z_{0 } ) \ , \bigr\rvert_{u } + ( n - 2 ) \frac{\partial h}{\partial z}(u , z_{0 } ) = \left .
\frac{\partial}{\partial z } \right\rvert_{z = z_{0 } } \bigl [ \delta h(\cdot , z )
\ , \bigr\rvert_{u } + ( n - 2)h(u , z ) \bigr ] .
\end{aligned}\ ] ] meanwhile , @xmath486\\ & = \tau^{\operatorname{\mathrm{t}}}(u ) \cdot \nabla h(\cdot , z_{0 } ) \ , \bigr\rvert_{u } + ( \tau \cdot u ) h(u , z_{0 } ) .
\end{aligned}\ ] ] therefore , the part @xmath480 of the equation equals @xmath487 \sum_{j=1}^{n-2 } e_{j}(u ) \cdot \overline{\nabla}_{e_{j}(u ) } \left [ v \mapsto \frac{\partial \gamma^{-}}{\partial z}(v , z_{0 } ) \right]_{u}\\ & = \tau^{\operatorname{\mathrm{t}}}(u ) \cdot \left\ { \left .
\frac{\partial}{\partial z } \right\rvert_{z = z_{0 } } \bigl [ \delta h(\cdot , z ) \ , \bigr\rvert_{u } + ( n - 2)h(u , z ) \bigr ] \nabla h ( \cdot , z_{0 } ) \ , \bigr\rvert_{u}\right\}\\ & \qquad + ( \tau \cdot u ) h(u , z_{0 } ) \left .
\frac{\partial}{\partial z } \right\rvert_{z = z_{0 } } \bigl [ \delta h(\cdot , z )
\ , \bigr\rvert_{u } + ( n - 2)h(u , z ) \bigr ] \end{aligned}\ ] ] then adding the equations and we can conclude that the part @xmath462 of the equation equals @xmath488(u)\\ & = \tau^{\operatorname{\mathrm{t}}}(u ) \cdot \biggl\ { \left .
\frac{\partial}{\partial z } \right\rvert_{z = z_{0 } } \bigl [ \delta h(\cdot , z ) \
, \bigr\rvert_{u } + ( n - 2 ) h(u , z ) \bigr ] \nabla h(\cdot , z_{0 } ) \ , \bigr\rvert_{u}\\ & \hspace{0.8cm}+ h(u , z_{0 } ) \nabla \frac{\partial h}{\partial z } ( \cdot , z_{0 } ) \ , \bigr\rvert_{u}\\ & \hspace{0.8cm}+ \sum_{k=1}^{n-2 } \left(\sum_{j=1}^{n-2 } \left [ e_{j}(u ) \cdot \nabla \frac{\partial h } { \partial z } ( \cdot , z_{0 } ) \ , \bigr\rvert_{u } \right ] e_{j}\bigl ( e_{k}\bigl [ h ( \cdot , z_{0 } ) \bigr ] \bigr)(u)\right ) e_{k}(u ) \biggr\}\\ & + ( \tau \cdot u ) \biggl\ { h(u , z_{0 } ) \left .
\frac{\partial}{\partial z } \right\rvert_{z = z_{0 } } \bigl [ \delta h(\cdot , z ) \ , \bigr\rvert_{u } + ( n - 2 ) h(u , z ) \bigr ] \biggr\ } \end{aligned}\ ] ] now we want to compute the part @xmath467 of the equation @xmath489 \right](u).\ ] ] note that @xmath490 is a vector field on @xmath160 and @xmath491 support parameterizes the ovaloid @xmath492 , where for every @xmath493 @xmath494 thus we can conclude that @xmath495 $ ] is a vector field on @xmath496 . @xmath497 & = \overline{\nabla}_{\tau^{\operatorname{\mathrm{t}}}(v ) } \gamma^{- } ( \cdot , z_{0 } ) \ , \bigr\rvert_{v }
= \overline{\nabla}_{\tau^{\operatorname{\mathrm{t}}}(v ) } \bigl [ \nabla h(\cdot , z_{0 } ) + h(\cdot , z_{0 } ) n \bigr ] \ , \bigr\rvert_{v}\\ & = \overline{\nabla}_{\tau^{\operatorname{\mathrm{t}}}(v ) } \nabla h(\cdot , z_{0 } ) \ , \bigr\rvert_{v } + \overline{\nabla}_{\tau^{\operatorname{\mathrm{t}}}(v ) } h(\cdot , z_{0})n \ , \bigr\rvert_{v}\\ & = \overline{\nabla}_{\tau^{\operatorname{\mathrm{t}}}(v ) } \nabla h(\cdot , z_{0 } ) \ , \bigr\rvert_{v } + \bigl [ \tau^{\operatorname{\mathrm{t}}}h(\cdot , z_{0 } ) \bigr](v ) v + h(v , z_{0 } )
\overline{\nabla}_{\tau^{\operatorname{\mathrm{t}}}(v ) } n \ , \bigr\rvert_{v}\\ & = \overline{\nabla}_{\tau^{\operatorname{\mathrm{t}}}(v ) } \nabla h(\cdot , z_{0 } ) \ , \bigr\rvert_{v } + \bigl [ \tau^{\operatorname{\mathrm{t}}}h(\cdot , z_{0 } ) \bigr](v ) v + h(v , z_{0 } ) \tau^{\operatorname{\mathrm{t}}}(v ) .
\end{aligned}\ ] ] since @xmath498 \in \operatorname{\mathrm{t}}_{v } \mathbb{s}^{n-2}$ ] , @xmath499 decomposes as @xmath500(v)v\ ] ] and @xmath501 & = \bigl ( \nabla_{\tau^{\operatorname{\mathrm{t}}}(v ) } \nabla h(\cdot , z_{0 } ) \ , \bigr\rvert_{v } - \bigl [ \tau^{\operatorname{\mathrm{t}}}h(\cdot , z_{0 } ) \bigr](v)v \bigr)\\ & \quad + \bigl [ \tau^{\operatorname{\mathrm{t}}}h(\cdot , z_{0 } ) \bigr](v ) v + h(v , z_{0 } ) \tau^{\operatorname{\mathrm{t}}}(v)\\ & = \nabla_{\tau^{\operatorname{\mathrm{t}}}(v ) } \nabla h(\cdot , z_{0 } ) \ , \bigr\rvert_{v } + h(v , z_{0 } ) \tau^{\operatorname{\mathrm{t}}}(v ) .
\end{aligned}\ ] ] therefore , the divergence in can be expanded as @xmath502_{u}\\ & = \underbrace{\sum_{j=1}^{n-2 } e_{j}(u ) \cdot \biggl\ { e_{j } \left [ v \mapsto \frac{\partial \gamma^{-}}{\partial z } ( v , z_{0 } ) \cdot v \right](u ) \bigl ( \nabla_{\tau^{\operatorname{\mathrm{t}}}(u ) } \nabla h(\cdot , z_{0 } ) \ ,
\bigr\rvert_{u}+ h(u , z_{0 } ) \tau^{\operatorname{\mathrm{t}}}(u ) \bigr ) \biggr\}}_{\textstyle { \ensuremath{\text{\textcircled{\raisebox{-0.2ex}{\textbf{6}}}}}}}\\ & + \underbrace{\sum_{j=1}^{n-2 } e_{j}(u ) \cdot \biggl\ { \left ( \frac{\partial \gamma^{-}}{\partial z}(u , z_{0 } ) \cdot u \right ) \underbrace{\nabla_{e_{j}(u ) } \nabla_{\tau^{\operatorname{\mathrm{t } } } } \nabla h(\cdot , z_{0 } ) \ , \bigr \rvert_{u}}_{\textstyle { \ensuremath{\text{\textcircled{\raisebox{-0.2ex}{\textbf{1 } } } } } } } \biggr\}}_{\textstyle { \ensuremath{\text{\textcircled{\raisebox{-0.3ex}{\textbf{2}}}}}}+ { \ensuremath{\text{\textcircled{\raisebox{-0.2ex}{\textbf{3}}}}}}+ { \ensuremath{\text{\textcircled{\raisebox{-0.2ex}{\textbf{4}}}}}}+ { \ensuremath{\text{\textcircled{\raisebox{-0.2ex}{\textbf{5}}}}}}}\\ & + \underbrace{\sum_{j=1}^{n-2 } e_{j}(u ) \cdot \biggl\ { \left ( \frac{\partial \gamma^{-}}{\partial z}(u , z_{0 } ) \cdot u\right ) \underbrace{\overline{\nabla}_{e_{j}(u ) } \bigl
[ v \mapsto h(v , z_{0 } ) \tau^{\operatorname{\mathrm{t}}}(v ) \bigr ] \ , \bigr\rvert_{u}}_{\textstyle { \ensuremath{\text{\textcircled{\raisebox{-0.2ex}{\textbf{7 } } } } } } } \biggr\}}_{\textstyle { \ensuremath{\text{\textcircled{\raisebox{-0.2ex}{\textbf{8 } } } } } } } \end{aligned}\ ] ] using the fact @xmath503 and gauss equations we get @xmath504 } z & = - ( s_{n}x \cdot z ) s_{n}y + ( s_{n}y \cdot z ) s_{n}x\\ & = - ( x \cdot z)y + ( y \cdot z)x \end{aligned}\ ] ] for all @xmath27 , @xmath270 , @xmath505 . so if we let @xmath506 , @xmath507 , and @xmath508 then the part @xmath462 of the equation can be written as @xmath509}_{\textstyle { \ensuremath{\text{\textcircled{\raisebox{-0.2ex}{\textbf{1 } } } } } } } } \nabla h(\cdot , z_{0})}_{\textstyle { \ensuremath{\text{\textcircled{\raisebox{-0.3ex}{\textbf{2}}}}}}}\\ & \underbrace{- \bigl [ s_{n}e_{j } \cdot \nabla h(\cdot , z_{0 } ) \bigr ] s_{n}\tau^{\operatorname{\mathrm{t}}}}_{\textstyle { \ensuremath{\text{\textcircled{\raisebox{-0.2ex}{\textbf{4 } } } } } } } { } + \underbrace{\bigl[s_{n}\tau^{\operatorname{\mathrm{t } } } \cdot \nabla h(\cdot , z_{0})\bigr ] s_{n}e_{j}}_{\textstyle { \ensuremath{\text{\textcircled{\raisebox{-0.2ex}{\textbf{5 } } } } } } } \end{split}\ ] ] the parts @xmath467 , @xmath469 , @xmath480 , and @xmath510 of the equation yield the corresponding parts of the equation .
we will first compute the parts of the equation .
the part @xmath462 of the equation can be calculated as @xmath511(u ) = \nabla_{e_{j}(u ) } \tau^{\operatorname{\mathrm{t } } } \ , \bigr\rvert_{u } - \nabla_{\tau^{\operatorname{\mathrm{t}}}(u ) } e_{j } \
, \bigr \rvert_{u } = \overline{\nabla}_{e_{j}(u ) } \tau^{\operatorname{\mathrm{t } } } \ , \bigr\rvert_{u } - \overline{\nabla}_{\tau^{\operatorname{\mathrm{t}}}(u ) } e_{j } \bigr\rvert_{u}\\ & = \overline{\nabla}_{e_{j}(u ) } \bigl [ v \mapsto \tau - ( \tau \cdot v)v \bigr ] \ , \bigr\rvert_{u } - \bigl [ \underbrace{\nabla_{\tau^{\operatorname{\mathrm{t}}}(u ) } e_{j } \ , \bigr\rvert_{u}}_{= 0 } { } + \bigl ( \overline{\nabla}_{\tau^{\operatorname{\mathrm{t } } } ( u ) } e_{j } \ , \bigr\rvert_{u } \cdot u\bigr)u \bigr ] \end{aligned}\ ] ] @xmath512u + ( \tau \cdot u ) \overline{\nabla}_{e_{j}(u ) } n \ ,
\bigr\rvert_{u } \bigr)\\ & \hspace{5.5 cm } - \bigl ( \underbrace{\overline{\nabla}_{\tau^{\operatorname{\mathrm{t}}}(u ) } e_{j } \cdot n \ , \bigr\rvert_{u}}_{=0 } { } - e_{j}(u ) \cdot \overline{\nabla}_{\tau^{\operatorname{\mathrm{t}}}(u ) } n \ , \bigr\rvert_{u } \bigr)u\\ & = - \bigl ( \bigl [ \tau \cdot e_{j}(u ) \bigr ] u + ( \tau \cdot u ) e_{j}(u ) \bigr ) + \bigl [ e_{j}(u ) \cdot \tau^{\operatorname{\mathrm{t}}}(u ) \bigr]u = - ( \tau \cdot u ) e_{j}(u ) .
\end{aligned}\ ] ] the parts @xmath480 and @xmath510 of the equation can be written as @xmath513 \cdot \nabla h(\cdot , z_{0 } ) \ , \bigr\rvert_{u }
\bigr ) s_{n } \bigl [ \tau^{\operatorname{\mathrm{t}}}(u ) \bigr ] & = - \bigl ( e_{j}(u ) \cdot
\nabla h(\cdot , z_{0 } ) \ , \bigr\rvert_{u } \bigr ) \tau^{\operatorname{\mathrm{t}}}(u)\\ \big ( s_{n } \bigl [ \tau^{\operatorname{\mathrm{t}}}(u ) \bigr ] \cdot \nabla h(\cdot , z_{0 } ) \ , \bigr\rvert_{u }
\bigr ) s_{n } \bigl [ e_{j}(u ) \bigr ] & = \bigl ( \tau^{\operatorname{\mathrm{t}}}(u ) \cdot \nabla h(\cdot , z_{0 } ) \ , \bigr\rvert_{u }
\bigr ) e_{j}(u ) .
\end{aligned}\ ] ] using the part @xmath462 , the part @xmath467 of the equation can be written as @xmath514 } \nabla h(\cdot , z_{0 } ) \
, \bigr\rvert_{u } & = - ( \tau \cdot u ) \nabla_{e_{j}(u ) } \nabla h(\cdot , z_{0 } ) \ , \bigr\rvert_{u}\\ & = - ( \tau \cdot u ) \nabla_{e_{j}(u ) }
\left ( \sum_{k=1}^{n-2 } e_{k } \bigl [ h(\cdot , z_{0 } ) \bigr ] e_{k } \right)_{u}. \end{aligned}\ ] ] recall that @xmath515 is the support parameterization of the ovaloid @xmath516 given by @xmath517 and hence the part @xmath467 of the equation equals @xmath518(u ) } \nabla h(\cdot , z_{0 } ) \ , \bigr\rvert_{u } \right ] \notag\\ & = \frac{\partial h}{\partial z}(u , z_{0 } ) \sum_{j=1}^{n-2 } e_{j}(u ) \cdot \nabla_{-(\tau \cdot u)e_{j}(u ) } \nabla h(\cdot , z_{0 } ) \
, \bigr\rvert_{u } \notag\\ & = - ( \tau \cdot u ) \frac{\partial h}{\partial z}(u , z_{0 } ) \sum_{j=1}^{n-2 } e_{j}(u ) \cdot \nabla_{e_{j}(u ) } \nabla h(\cdot , z_{0 } ) \
, \bigr\rvert_{u } \notag\\ & = ( \tau \cdot u ) \left [ - \delta h(\cdot , z_{0 } ) \ ,
\bigr\rvert_{u } \frac{\partial h}{\partial z}(u , z_{0})\right ] .
\label{es5.2:pde2;1p2 } \end{aligned}\ ] ] using the equation we can easily compute @xmath519(u ) & = e_{j } \left [ \frac{\partial h}{\partial z}(\cdot , z_{0 } ) \right](u ) .
\end{aligned}\ ] ] the part @xmath520 of the equation is equal to @xmath521 \ , \bigr\rvert_{u}\\ & = e_{j } \bigl [ h(\cdot , z_{0 } ) \bigr](u ) \tau^{\operatorname{\mathrm{t}}}(u ) + h(u , z_{0 } ) \overline{\nabla}_{e_{j}(u ) } \bigl[v \mapsto \tau - ( \tau \cdot v)v \bigr ] \ , \bigr\rvert_{u}\notag\\ & = e_{j } \bigl [ h(\cdot , z_{0 } ) \bigr](u )
\tau^{\operatorname{\mathrm{t}}}(u ) + h(u , z_{0 } ) \bigl [ - \overline{\nabla}_{e_{j}(u ) } ( \tau \cdot n)n \ , \bigr\rvert_{u } \bigr ] \notag\\ & = e_{j } \bigl [ h(\cdot , z_{0 } ) \bigr](u ) \tau^{\operatorname{\mathrm{t}}}(u ) - h(u , z_{0 } ) \bigl [ \bigl ( \tau \cdot \overline{\nabla}_{e_{j } ( u)}n \ , \bigr\rvert_{u } \bigr)u + ( \tau \cdot u ) \overline{\nabla}_{e_{j}(u ) } n \ , \bigr\rvert_{u } \bigr ] \notag\\ & = e_{j } \bigl [ h(\cdot , z_{0 } ) \bigr](u ) \tau^{\operatorname{\mathrm{t}}}(u ) - h(u , z_{0 } ) \bigl [ \tau \cdot e_{j}(u ) \bigr]u - h(u , z_{0 } ) ( \tau \cdot u ) e_{j}(u ) .
\end{aligned}\ ] ] using the part @xmath520 , part @xmath522 of the equation can be computed as @xmath523 \ , \bigr\rvert_{u } \right\ } \notag\\ & = \frac{\partial h}{\partial z}(u , z_{0 } ) \left\ { \sum_{j=1}^{n-2 } e_{j } \bigl [ h(\cdot , z_{0 } ) \bigr](u ) \bigl ( \tau^{\operatorname{\mathrm{t}}}(u ) \cdot e_{j}(u ) \bigr ) - ( n - 2)(\tau \cdot u ) h(u , z_{0 } ) \right\ } \notag\\ & = \tau^{\operatorname{\mathrm{t}}}(u ) \cdot \left\ { \sum_{j=1}^{n-2 } e_{j } \bigl [ h(\cdot , z_{0 } ) \bigr](u ) \frac{\partial h}{\partial z}(u , z_{0 } ) e_{j}(u ) \right\ } \notag\\ & \hspace{6 cm } + ( \tau \cdot u ) \left [ - ( n - 2 ) h(u , z_{0 } ) \frac{\partial h}{\partial z}(u , z_{0 } ) \right ] \notag\\ & = \tau^{\operatorname{\mathrm{t}}}(u ) \left\ { \frac{\partial h}{\partial z}(u , z_{0 } ) \nabla h(\cdot , z_{0 } ) \ , \bigr\rvert_{u } \right\ } + ( \tau \cdot u ) \left [ - ( n - 2 ) h(u , z_{0 } ) \frac{\partial h}{\partial z}(u , z_{0 } ) \right ] .
\label{es5.2:pde2;1p3 } \end{aligned}\ ] ] the part @xmath469 of the equation equals @xmath524 \ , \bigr\rvert_{u } \notag\\ & = \frac{\partial h}{\partial z}(u , z_{0 } ) \nabla_{\tau^{\operatorname{\mathrm{t}}}(u ) } \left ( \sum_{j=1}^{n-2 } e_{j } \cdot \overline{\nabla}_{e_{j } } \nabla h(\cdot , z_{0 } ) \right)_{u } \notag\\ & = \frac{\partial h}{\partial z}(u , z_{0 } ) \nabla_{\tau^{\operatorname{\mathrm{t}}}(u ) } \delta h(\cdot , z_{0 } ) \ , \bigr\rvert_{u } \notag\\ & = \frac{\partial h}{\partial z}(u , z_{0 } ) \nabla_{\tau^{\operatorname{\mathrm{t}}}(u ) } \bigr [ v \mapsto \delta h(v , z_{0 } ) \bigr ] \ , \bigr \rvert_{u } \notag\\ & = \tau^{\operatorname{\mathrm{t}}}(u ) \cdot \left\ { \frac{\partial h}{\partial z}(u , z_{0 } ) \nabla \bigl [ \delta h(\cdot , z_{0 } ) \bigr ] \ , \bigr \rvert_{u } \right\}. \label{es5.2:pde2;1p3 } \end{aligned}\ ] ] the part @xmath480 of the equation can be written as @xmath525 \tau^{\operatorname{\mathrm{t}}}(u ) \right\ } \notag\\ & = - \frac{\partial h}{\partial z}(u , z_{0 } ) \sum_{j=1}^{n-2 } \bigl [ e_{j}(u ) \cdot \nabla h(\cdot , z_{0 } ) \ , \bigr \rvert_{u } \bigr ] e_{j}(u ) \cdot \tau^{\operatorname{\mathrm{t}}}(u ) \notag\\ & = \tau^{\operatorname{\mathrm{t}}}(u ) \cdot \left\ { - \frac{\partial h}{\partial z}(u , z_{0 } ) \nabla h(\cdot , z_{0 } ) \ , \bigr\rvert_{u } \right\}. \label{es5.2:pde2;1p4 } \end{aligned}\ ] ] the part @xmath510 of the equation is equal to @xmath526 e_{j}(u)\right\ } \notag\\ & = \frac{\partial h}{\partial z}(u , z_{0 } ) \bigl [ \tau^{\operatorname{\mathrm{t}}}(u ) \cdot \nabla h(\cdot , z_{0 } ) \ , \bigr\rvert_{u } \bigr ] \sum_{j=1}^{n-2 } e_{j}(u ) \cdot e_{j}(u ) \notag\\ & = \tau^{\operatorname{\mathrm{t}}}(u ) \cdot \left\{(n - 2 ) \frac{\partial h}{\partial z}(u , z_{0 } ) \nabla h(\cdot , z_{0 } ) \ , \bigr\rvert_{u } \right\}. \label{es5.2:pde2;1p5 } \end{aligned}\ ] ] by summing the equations , , , and we get @xmath527 \ , \bigr\rvert_{u } - \frac{\partial h}{\partial z}(u , z_{0 } ) \nabla h(\cdot , z_{0 } ) \ , \bigr\rvert_{u}\\ & \phantom{\tau^{\operatorname{\mathrm{t}}}(u ) \cdot \bigl\ { \frac{\partial h}{\partial z}(u , z_{0 } ) \nabla \bigl [ \delta h(\cdot , z_{0 } ) \bigr ] \ , \bigr\rvert_{u } - { } } + ( n -2 ) \frac{\partial h}{\partial z}(u , z_{0 } ) \nabla h(\cdot , z_{0 } ) \
, \bigr \rvert_{u } \bigr\}\\ & \qquad + ( \tau \cdot u ) \left\ { - \frac{\partial h}{\partial z}(u , z_{0 } ) \delta h(\cdot , z_{0 } ) \ , \bigr\rvert_{u } \right\ } \end{aligned}\ ] ] the part @xmath528 of the equation can be computed as @xmath529(u ) \bigl ( \nabla_{\tau^{\operatorname{\mathrm{t}}}(u ) } \nabla h(\cdot , z_{0 } ) \ , \bigr\rvert_{u}+ h(u , z_{0 } ) \tau^{\operatorname{\mathrm{t}}}(u ) \bigr ) \biggr\ } \notag\\ & = \sum_{j=1}^{n-2 } e_{j}(u ) \cdot \left\ { e_{j}\left [ \frac{\partial h}{\partial z}(\cdot , z_{0 } ) \right](u ) \
\nabla_{\tau^{\operatorname{\mathrm{t}}}(u ) } \nabla h(\cdot , z_{0 } ) \ , \bigr\rvert_{u}\right\}\notag\\ & \hspace{4 cm } + \sum_{j=1}^{n-2 } e_{j}(u ) \cdot \left\ { e_{j } \left [ \frac{\partial h}{\partial z}(\cdot , z_{0 } ) \right](u ) \ h(u , z_{0 } )
\tau^{\operatorname{\mathrm{t}}}(u ) \right\}\notag \end{aligned}\ ] ] @xmath530(u ) \ \bigl\ { e_{j}(u )
\cdot \nabla_{\tau^{\operatorname{\mathrm{t}}}(u ) } \nabla h(\cdot , z_{0 } ) \ , \bigr\rvert_{u } \bigr\}\notag\\ & \hspace{4 cm } + \tau^{\operatorname{\mathrm{t}}}(u ) \cdot \sum_{j=1}^{n-2 } h(u , z_{0 } ) e_{j } \left [ \frac{\partial h}{\partial z}(\cdot , z_{0 } ) \right](u ) \
e_{j}(u ) \notag\\ & = \sum_{j=1}^{n-2 } e_{j } \left [ \frac{\partial h}{\partial z}(\cdot , z_{0 } ) \right](u ) \ \nabla_{\tau^{\operatorname{\mathrm{t}}}(u ) } \bigl [ e_{j } \cdot \nabla h(\cdot , z_{0 } ) \bigr ] \
, \bigr\rvert_{u } + \tau^{\operatorname{\mathrm{t}}}(u ) \cdot h(u , z_{0 } ) \nabla \frac{\partial h}{\partial z}(\cdot , z_{0 } ) \ , \bigr\rvert_{u } \notag\\ & = \sum_{j=1}^{n-2 } e_{j } \left [ \frac{\partial h}{\partial z}(\cdot , z_{0 } )
\right](u ) \ \bigl\ { \tau^{\operatorname{\mathrm{t}}}(u ) \cdot \nabla \bigl [ e_{j } \cdot \nabla h(\cdot , z_{0 } ) \bigr ] \ , \bigr\rvert_{u } \bigr\ } \notag\\ & \hspace{4 cm } + \tau^{\operatorname{\mathrm{t}}}(u ) \cdot h(u , z_{0 } ) \nabla \frac{\partial h}{\partial z}(\cdot , z_{0 } ) \ , \bigr\rvert_{u } \notag\\ & = \tau^{\operatorname{\mathrm{t}}}(u ) \cdot \left\ { \sum_{j=1}^{n-2 } e_{j } \left [ \frac{\partial h}{\partial z}(\cdot , z_{0 } ) \right](u ) \ \nabla \bigl [ e_{j } \cdot \nabla h(\cdot , z_{0 } ) \bigr ] \ , \bigr\rvert_{u } + h(u , z_{0 } ) \nabla \frac{\partial h}{\partial z } ( \cdot , z_{0 } ) \ , \bigr\rvert_{u } \right\}. \label{es5.2:pde2;1p1 } \end{aligned}\ ] ] by summing the equations , and we find that the divergence at @xmath162 @xmath531 \right ] ( u)\ ] ] equals @xmath532 \ , \bigr\rvert_{u } + h(u , z_{0 } ) \nabla \frac{\partial h}{\partial z}(\cdot , z_{0 } ) \ , \bigr \rvert_{u}\\ & \phantom{\tau^{\operatorname{\mathrm{t}}}(u ) \cdot \biggl\ { \frac{\partial h}{\partial z}(u , z_{0 } ) \nabla \bigl [ \delta h(\cdot , z_{0})}+ \sum_{j=1}^{n-2 } e_{j } \left [ \frac{\partial h}{\partial z}(\cdot , z_{0 } ) \right](u ) \
\bigl [ e_{j } \cdot \nabla h(\cdot , z_{0 } ) \bigr ] \ , \bigr\rvert_{u } \biggr\}\\ & \quad + ( \tau \cdot u ) \left\ { - \bigl [ \delta h(\cdot , z_{0 } )
\ , \bigr\rvert_{u } + ( n -2)h(u , z_{0 } ) \bigr ] \frac{\partial h}{\partial z}(u , z_{0})\right\ } .
\end{aligned}\ ] ] finally the value of the function @xmath458 at @xmath533 can be found by adding the equations and @xmath534(u)\\ & = \operatorname{\mathrm{div}}\left [ v \mapsto \bigl ( \tau \cdot \gamma^{-}(v , z_{0 } ) \bigr ) \frac{\partial \gamma^{-}}{\partial z}(v , z_{0 } ) \right](u)\\ & \hspace{4 cm } + \operatorname{\mathrm{div}}\left [ v \mapsto \left ( \frac{\partial \gamma^{-}}{\partial z}(v , z_{0 } ) \cdot v \right ) ( \partial_{1}\gamma^{-})(v , z_{0 } ) \bigl [ \tau^{\operatorname{\mathrm{t}}}(u ) \bigr ] \right ] \end{aligned}\ ] ] @xmath535 \nabla h(\cdot , z_{0 } ) \ , \bigr\rvert_{u}\\ & \hspace{1.8 cm } + 2 h(u , z_{0 } ) \nabla \frac{\partial h}{\partial z}(u , z_{0 } ) \ , \bigr\rvert_{u } + \underbrace{\sum_{j=1}^{n-2 } e_{j } \left [ \frac{\partial h}{\partial z}(\cdot , z_{0 } )
\right](u ) \ \nabla
\bigl [ e_{j } \cdot \nabla h(\cdot , z_{0 } ) \bigr](u)}_{\textstyle { \ensuremath{\text{\textcircled{\raisebox{-0.2ex}{\textbf{1 } } } } } } } \\ & \hspace{3.8 cm } + \underbrace{\sum_{k=1}^{n-2 } \left ( \sum_{j=1}^{n-2 } e_{j } \left [ \frac{\partial h } { \partial z}(\cdot , z_{0})\right](u ) \ e_{j } \bigl [ e_{k}h(\cdot , z_{0 } ) \bigr](u ) \right)e_{k}(u)}_{\textstyle { \ensuremath{\text{\textcircled{\raisebox{-0.3ex}{\textbf{2}}}}}}}\\ & \hspace{3.8 cm } + \frac{\partial h}{\partial z}(u , z_{0 } ) \nabla \bigl [ \delta h(\cdot , z_{0 } ) +
( n -2 ) h(\cdot , z_{0 } ) \bigr ] \ , \bigr\rvert_{u }
\biggr\ } \end{aligned}\ ] ] @xmath536\\ - \bigl [ \delta h(\cdot , z_{0 } ) \ , \bigr\rvert_{u } + ( n - 2 ) h(u , z_{0 } ) \bigr ] \frac{\partial h}{\partial z}(u , z_{0 } ) \biggr\ } \end{gathered}\ ] ] @xmath537(u ) \
\bigl [ e_{j } \cdot \nabla h(\cdot , z_{0 } ) \bigr](u)\\ & = \sum_{j=1}^{n-2 } e_{j } \left [ \frac{\partial h}{\partial z}(\cdot , z_{0 } ) \right](u ) \sum_{k=1}^{n-2 } e_{k } \bigl [ e_{j } \cdot \nabla h(\cdot , z_{0 } ) \bigr](u ) e_{k}(u)\\ & = \sum_{j , k=1}^{n-2 } e_{j } \left [ \frac{\partial h}{\partial z}(\cdot , z_{0 } ) \right](u ) \bigl\ { \underbrace{\nabla_{e_{k}(u ) } e_{j } \ , \bigr\rvert_{u}}_{=0 } \cdot \nabla h(\cdot , z_{0})\\ & \hspace{5 cm } + e_{j}(u ) \cdot \nabla_{e_{k}(u ) } \nabla h(\cdot , z_{0 } ) \ , \bigr\rvert_{u}\bigr\ } e_{k}(u)\\ & = \sum_{j , k=1}^{n-2 } e_{j } \left [ \frac{\partial h}{\partial z}(\cdot , z_{0 } ) \right](u ) \left\ { e_{j}(u ) \cdot
\sum_{l=1}^{n-2 } \nabla_{e_{k}(u ) } \bigl ( e_{l } \bigl [ h(\cdot , z_{0 } ) \bigr ] e_{l } \bigr)_{u } \right\ } e_{k}(u)\\ & = \sum_{j , k , l=1}^{n-2 } e_{j } \left [ \frac{\partial h}{\partial z}(\cdot , z_{0})\right](u ) \biggl\ { e_{j}(u ) \cdot
\bigl [ e_{k } \bigl [ e_{l } ( h ( \cdot , z_{0 } ) ) \bigr](u ) \
e_{l}(u)\\ & \hspace{5 cm } + e_{l } \bigl[h(\cdot , z_{0})\bigr](u ) \underbrace{\nabla_{e_{k}(u)}e_{l } \ , \bigr\rvert_{u } } _ { = 0}\bigr ] \biggr\ } e_{k}(u)\\ & = \sum_{j , k=1}^{n-2 } e_{j } \left [ \frac{\partial h}{\partial z}(\cdot , z_{0 } ) \right](u ) \left\ { \sum_{l=1}^{n-2 } e_{j}(u ) \cdot \bigl [ e_{k } \bigl [ e_{l}(h ( \cdot , z_{0 } ) ) \bigr](u ) \bigr ] e_{l}(u)\right\ } e_{k}(u ) \end{aligned}\ ] ] @xmath538(u ) \
e_{k } \bigl [ e_{j}(h(\cdot , z_{0 } ) ) \bigr](u ) \ , e_{k}(u)\\ & \mspace{-58mu } = \sum_{k=1}^{n-2 } \left\ { \sum_{j=1}^{n-2 } e_{j } \left [ \frac{\partial h}{\partial z } ( \cdot , z_{0 } ) \right](u ) \ e_{k } \bigl [ e_{j}(h(\cdot , z_{0 } ) ) \bigr](u ) \right\ } e_{k}(u)\\ & \mspace{-58mu } = \sum_{k=1}^{n-2 } \left\ { \sum_{j=1}^{n-2 } e_{j } \left [ \frac{\partial h}{\partial z } ( \cdot , z_{0 } ) \right](u ) \ e_{j } \bigl [ e_{k}(h(\cdot , z_{0 } ) ) \bigr](u ) \right\ } e_{k}(u ) .
\end{aligned}\ ] ] the change of order of differentiation at the last step is permissible because for every smooth function @xmath539 , @xmath162 and for every @xmath540 , @xmath541 @xmath542 f \bigr)(u ) & = \bigl [ \nabla_{e_{j}(u ) } e_{k } \ , \bigr\rvert_{u } - \nabla_{e_{k}(u ) } e_{j } \ , \bigr\rvert_{u } \bigr ] f = 0 \\ e_{j}(e_{k}f)(u ) & - e_{k}(e_{j}f)(u ) = 0 \end{aligned}\ ] ] so we can conclude the parts @xmath462 and @xmath467 of the equation are equal and now we want to show that they are also equal to @xmath543 @xmath544 \nabla_{e_{j}(u ) } \bigl [ \nabla h(\cdot , z_{0 } ) \bigr ] \ , \bigr\rvert_{u}\\ & = \sum_{j=1}^{n-2}e_{j } \left [ \frac{\partial h}{\partial z}(\cdot , z_{0 } ) \right ] ( u ) \sum_{k=1}^{n-2 } \nabla_{e_{j}(u ) } \bigl ( v \mapsto e_{k } \bigl [ h(\cdot , z_{0 } ) \bigr](v ) \ , e_{k}(v ) \bigr)_{u } \end{aligned}\ ] ] @xmath545(u ) \bigl\ { e_{j } \bigl [ e_{k } \bigl [ h(\cdot , z_{0 } ) \bigr ] \big](u ) \ , e_{k}(u ) + e_{k } \bigl [ h(\cdot , z_{0 } ) \bigr](u ) \underbrace{\nabla_{e_{j } ( u ) } e_{k } \ , \bigr\rvert_{u}}_{=0}\bigr\}\\ & = \sum_{j , k=1}^{n-2 } e_{j } \left [ \frac{\partial h}{\partial z}(\cdot , z_{0 } ) \right](u ) \
e_{j } \bigl [ e_{k } \bigl [ h(\cdot , z_{0 } ) \bigr ] \big](u ) \ , e_{k}(u)\\ & = \sum_{k=1}^{n-2 } \left\ { \sum_{j=1}^{n-2 } e_{j } \left [ \frac{\partial h}{\partial z}(\cdot , z_{0 } ) \right ] ( u ) \
e_{j } \bigl [ e_{k } \bigl [ h(\cdot , z_{0 } ) \bigr ] \bigr ] ( u ) \right\ } e_{k}(u ) .
\end{aligned}\ ] ] @xmath546 \ , \bigr\rvert_{u}\\ & \hspace{3.3cm}+ \left .
\frac{\partial}{\partial z } \right\rvert_{z = z_{0 } } \bigl [ \delta h(\cdot , z_{0 } ) + ( n -2)h(u , z ) \bigr ] \nabla h(\cdot , z_{0 } ) \ , \bigr\rvert_{u}\\ & \hspace{3.3cm}+2 \left ( h(u , z_{0 } ) \nabla \frac{\partial h}{\partial z}(\cdot , z_{0 } ) \ ,
\bigr\rvert_{u } + \nabla_{\textstyle \nabla \left .
\frac{\partial h}{\partial z}(\cdot , z_{0 } ) \ , \right\rvert_{u } } \nabla h ( \cdot , z_{0 } ) \ ,
\bigr\rvert_{u}\right ) \bigg\}\\ & \phantom{f_{\tau}(u , z_{0 } ) = } + ( \tau \cdot u ) \ , \biggl\{h(u , z_{0 } ) \left .
\frac{\partial}{\partial z } \right \rvert_{z = z_{0 } } \bigl [ \delta h(\cdot , z_{0 } )
\ , \bigr\rvert_{u } + ( n -2 ) h(u , z ) \bigr]\\ & \phantom{f_{\tau}(u , z_{0 } ) = } \hspace{3.8 cm } - \bigl [ \delta h(\cdot , z_{0 } ) \ , \bigr\rvert_{u } + ( n - 2 ) h(u , z_{0 } ) \bigr ] \frac{\partial h}{\partial z}(u , z_{0 } ) \biggr\}. \end{aligned}\ ] ] as we mentioned in the beginning the expressions in the curly brackets do not depend on @xmath247 , @xmath547 for all @xmath247 and
hence these expressions must individually vanish for each @xmath162 and @xmath285 .
let @xmath548 be the transverse support function of the rectified tube @xmath549 and @xmath550 be the support map of @xmath551 given by @xmath552 if we can show that @xmath553 is the product of @xmath554 and the support function @xmath555 of a fixed ovaloid , then @xmath556 + \bigl [ r(z)h_{0}(u ) \bigr]u\\ & = r(z ) \bigl [ \nabla_{\mathbb{s}^{n-2 } } h_{0}(u ) + h_{0}(u)u \bigr ] = r(z ) \gamma^{\circ}(u ) \end{aligned}\ ] ] where @xmath557 is the support parameterization of a fixed ovaloid . according to proposition [ ps5.2:pde ] , the transverse support function @xmath163 satisfies the partial differential equation , which is equivalent to @xmath558 at height @xmath559 , using @xcite , we have @xmath560 for every @xmath162 , where @xmath561 is the @xmath562 principal curvature of the ovaloid @xmath563 , @xmath564 .
therefore , the equation is equivalent to @xmath565 for every @xmath162 and @xmath436 . and the condition is equivalent to @xmath566 define the operator @xmath567 , then the transverse support function h satisfies @xmath568 let @xmath569 be the support function of the ovaloid @xmath570 .
our aim is to show that @xmath571 for some positive function @xmath554 of height @xmath245 . by construction @xmath572 for each @xmath162 and
assume that there exists @xmath573 satisfying @xmath574 and @xmath575 for each @xmath162 .
since @xmath576 is a strictly positive function , @xmath576 and @xmath577 are continuous on a compact set , we can fix @xmath578 so that @xmath579 on @xmath160 .
define the positive constants @xmath580 and @xmath581 as follows : @xmath582 the subset @xmath583 defined as @xmath584 is nonempty because there exists @xmath162 with @xmath585 the subset @xmath586 is also closed because @xmath587 is a continuous function .
now we want to show that @xmath586 is also open and then conclude that @xmath588 .
let @xmath589 , and consider a local coordinate system @xmath590 of @xmath160 about @xmath64 @xmath591 consider the function @xmath592 , which is smooth and satisfies @xmath593 because @xmath594 for every @xmath493 .
the operator @xmath595 , under the local coordinate system @xmath590 , is strictly elliptic in the domain @xmath596 for some @xmath597 .
therefore , the operator @xmath166 and the smooth function @xmath598 in @xmath596 satisfy the hypotheses of harnack s inequality @xcite and hence exists a positive constant @xmath599 depending only on @xmath600 and @xmath166 so that for some closed ball @xmath601 @xmath602 and the last equality holds because @xmath603 and @xmath604 .
so we can conclude that @xmath605 } = 0.\ ] ] for the open ball @xmath606 of radius @xmath607 and centered at the origin , @xmath608 $ ] is an open neighborhood of @xmath589 in which @xmath587 vanishes identically and hence @xmath609 \subseteq u.\ ] ] so we can conclude that @xmath610 on @xmath160 .
we show that when @xmath29 is a transversely convex tube in _ standard position _ , and @xmath35 is not a cylinder , then there exists a linear isomorphism that fixes the @xmath252-axis while making each horizontal cross - section @xmath615 of @xmath35 simultaneously spherical .
according to lemma [ ps5.2:split ] ( splitting lemma ) , the transverse support function @xmath163 can be written as @xmath616 for each @xmath436 and @xmath162 .
the partial differential equation can therefore be written as @xmath617 ( u ) + ( n -2 ) \nabla
h_{0}(u)\bigr\}\\ & \quad + r(z)r'(z ) \bigl\ { \bigl [ \delta h_{0}(u ) \bigr ] \nabla h_{0}(u ) + ( n -2 ) h_{0}(u ) \nabla h_{0}(u )
\bigr \}\\ & \quad + 2 \left\ { r(z)r'(z ) h_{0}(u ) \nabla h_{0 } ( u)+ \nabla_{\textstyle \nabla r'(z ) h_{0}(u ) } \nabla r(z)h_{0}(u)\right\}\\ & = r(z)r'(z ) \bigl\ { h_{0}(u ) \nabla \bigl [ \delta h_{0}\bigr](u ) + 2(n -1)h_{0}(u ) \nabla h_{0}(u)\\ & \qquad \phantom{r(z)r'(z ) \bigl\ { } + \bigl
[ \delta h_{0}(u ) \bigr ] \nabla h_{0}(u ) + 2 \nabla_{\textstyle \nabla h_{0}(u ) } \nabla h_{0}(u ) \bigr\}. \end{aligned}\ ] ] if @xmath35 is not cylindrical , then there exists @xmath443 , @xmath618 and since @xmath619 we must have @xmath620 \nabla h_{0}(u ) + h_{0}(u ) \nabla \bigl [ \delta h_{0 } \bigr](u ) + 2(n-1 ) h_{0 } ( u ) \nabla h_{0}(u ) \\ & \phantom{{}= \bigl [ \delta h_{0}(u ) \bigr ] \nabla h_{0}(u ) + h_{0}(u ) \nabla \bigl [ \delta h_{0 } \bigr](u ) } + \underbrace{2 \nabla_{\textstyle \nabla h_{0}(u ) } \nabla h_{0}(u)}_{\textstyle { \ensuremath{\text{\textcircled{\raisebox{-0.2ex}{\textbf{1 } } } } } } } \end{aligned}\ ] ] for each @xmath162 .
let s consider the smooth function @xmath621 on @xmath160 and calculate its spherical gradient @xmath622(u ) = \nabla \bigl [ \delta h_{0}^{2 } \bigr](u ) + 2 ( n -1 ) \nabla h_{0}^{2}(u)\\ & = \nabla \bigl [ 2 h_{0 } \delta h_{0 } + 2 \lvert \nabla h_{0 } \rvert^{2 } \bigr](u ) + 4(n - 1 ) h_{0}(u ) \nabla h_{0}(u ) \\ & = 2 \bigl [ \delta h_{0}(u ) \bigr ] \nabla h_{0}(u ) + 2 h_{0}(u ) \nabla \bigl [ \delta h_{0}\bigr](u ) + 4 ( n -1)h_{0}(u ) \nabla h_{0}(u)\\ & \phantom{{}= 2 \bigl [ \delta h_{0}(u ) \bigr ] \nabla h_{0}(u ) + 2 h_{0}(u ) \nabla \bigl [ \delta h_{0}\bigr](u)}+ \underbrace{2 \nabla \bigl [ \nabla h_{0 } \cdot \nabla h_{0 } \bigr]}_{\textstyle { \ensuremath{\text{\textcircled{\raisebox{-0.2ex}{\textbf{1}}}}}}}. \end{aligned}\ ] ] the part @xmath462 of the equation can be expanded using the @xmath623 as follows : @xmath624(u ) \notag\\ & = 2 \sum_{j , k=1}^{n-2 } \bigl [ ( e_{k}h_{0})(u ) \
e_{k } ( e_{j}h_{0})(u ) \bigr ] e_{j}(u ) \notag\\ \phantom{2 \nabla_{\textstyle \nabla h_{0}(u ) } \nabla h_{0}(u ) } & = 2 \sum_{j=1}^{n-2 } \left [ \sum_{k=1}^{n-2 } ( e_{k}h_{0 } ) ( u ) \
e_{k}(e_{j } h_{0})(u ) \right ] e_{j}(u ) \notag\\ & = 2 \sum_{j=1}^{n-2 } \left [ \sum_{k=1}^{n-2 } ( e_{k}h_{0 } ) ( u ) \
e_{j}(e_{k } h_{0})(u ) \right ] e_{j}(u ) \end{aligned}\ ] ] where the change of the order of differentiation at the last equality is possible because of the observation .
again by using the @xmath625 , the part @xmath462 of the equation can expanded as follows : @xmath626(u ) = 2 \sum_{j=1}^{n-2 } e_{j } \bigl [ \nabla h_{0 } \cdot \nabla h_{0 } \bigr](u ) e_{j}(u)\\ & = 2 \sum_{j=1}^{n-2 } 2 \bigl [ \nabla_{e_{j}(u ) } \nabla h_{0}(u ) \cdot \nabla h_{0}(u ) \bigr ] e_{j}(u ) \\ & = 4 \sum_{j=1}^{n-2 }
\left\ { \nabla_{e_{j}(u ) } \left [ \sum_{k=1}^{n-2 } ( e_{k}h_{0 } ) \ , e_{k}\right](u ) \cdot \nabla h_{0}(u ) \right\ } e_{j}(u)\\ \end{aligned}\ ] ] @xmath627 \cdot \nabla h_{0}(u ) \bigr\ } e_{j}(u ) \notag\\ & = 4 \sum_{j , k=1}^{n-2 } e_{j}(e_{k}h_{0})(u ) \left\ { e_{k}(u ) \cdot \sum_{l=1}^{n-2 } ( e_{l}h_{0})(u ) \
, e_{l } ( u ) \right\ } e_{j}(u ) \notag\\ & = 4 \sum_{j=1}^{n-2 } \left [ \sum_{k=1}^{n-2 } ( e_{k}h_{0})(u ) \
e_{j}(e_{k}h_{0})(u ) \right ] e_{j}(u ) .
\end{aligned}\ ] ] therefore we can conclude that @xmath628(u ) = 2 \nabla_{\textstyle \nabla h_{0}(u ) } \nabla h_{0}(u)\ ] ] and hence @xmath629(u ) & = \bigl [ \delta h_{0}(u ) \bigr ] \nabla h_{0}(u ) + h_{0}(u ) \nabla \bigl [ \delta h_{0 } \bigr](u)\\ & \qquad + 2(n-1 ) h_{0}(u ) \nabla h_{0}(u ) + 2 \nabla_{\textstyle \nabla h_{0}(u ) } \nabla h_{0}(u)\\ & = 0 .
\end{aligned}\ ] ] for each @xmath162 .
using lemma [ ls5.1:orgcntellipsoid ] we can conclude that the positive function @xmath576 on @xmath160 is the support function of an origin centered ellipsoid .
since the support function @xmath163 satisfies @xmath630 , the item of lemma [ ls2.2:spfnpr ] implies that every horizontal cross section @xmath631 of @xmath35 is a homothetic copy of @xmath632 .
therefore , there exists an affine isomorphism of @xmath633 mapping @xmath632 onto the unit sphere @xmath634 .
extend this affine isomorphism to @xmath3 by fixing the last coordinate and hence @xmath35 is affinely isomorphic to a hypersurface of revolution in @xmath635 .
[ linearity criterion ] [ ls5.3:affcrv ] a curve @xmath636 defined on an open interval is affine if and only if it is locally odd , i.e. , for each @xmath637 there exists @xmath638 so that @xmath639 for each @xmath640 . without loss of generality
we can assume that @xmath29 lies in _ standard position _ because a general transversely convex tube , by definition , is affinely isomorphic to the one in _ standard position _ and the central curve is preserved under affine isomorphisms .
using proposition [ ps5.2:classofrcttube ] and the invariance of the central curve under affine isomorphisms we can further assume that @xmath35 is either a cylinder or a hypersurface of revolution with @xmath641 as its axis and show that the _ cop _ assumption forces the central curve of @xmath29 to be a straight line .
* cylindrical case : * assume that @xmath35 is a cylinder over a central ovaloid @xmath570 at height @xmath308 .
each horizontal cross - section @xmath404 at height @xmath642 translates to @xmath570 .
let @xmath112 be the support parameterization of @xmath570 , then a parameterization of @xmath29 can be obtained as @xmath643 fix @xmath644 , @xmath162 , consider the @xmath64-diameter of @xmath645 , which is the line segment joining the unique two points at which @xmath645 has normal line @xmath646
. since @xmath112 support parameterizes @xmath570 , the endpoints of the @xmath64-diamater are @xmath647 and @xmath648 .
the tangent hyperplanes of @xmath29 at @xmath647 and @xmath648 are @xmath649 , \ , \partial_{2}x(u , b)(1 ) \bigr\}\\ & = u^{\perp } \oplus \mathbb{r}(c'(b ) , 1)\\ & = ( -u)^{\perp } \oplus \mathbb{r}(c'(b ) , 1)\\ & = \mathrm{span } \bigl\ { \partial_{1}x(-u , b ) \bigl [ ( -u)^{\perp } \bigr ] , \ , \partial_{2}x(-u , b)(1 ) \bigr\}\\ & = \operatorname{\mathrm{t}}_{x(-u , b ) } \mathcal{t}. \end{aligned}\ ] ] fix the @xmath64-diameter of @xmath645 as axis and tilt the horizontal hyperplane @xmath650 in a direction perpendicular to the axis with some small slope @xmath651 to get a new hyperplane @xmath652 . for sufficiently smal @xmath651
, the intersection @xmath653 will remain a central ovaloid because @xmath29 has _
cop_. since @xmath652 contains the @xmath64-diameter of @xmath645 , the endpoints @xmath647 and @xmath648 of that diameter remain on @xmath654 independently on @xmath651 . since the tangent hyperplanes to @xmath655 at these points are parallel and their intersections with @xmath652 form codimension one planes tangent to @xmath656 at @xmath647 and @xmath648 , those planes are also parallel .
we can conclude that the @xmath64-diameter of @xmath645 remains a diameter of @xmath656 independently on @xmath651 , and hence @xmath657 remains the center of symmetry of @xmath656 for each @xmath162 and for all sufficiently small @xmath651 .
the center of @xmath656 remains fixed as we vary @xmath651 .
every point sufficiently close to @xmath645 on @xmath29 belongs to @xmath656 for some @xmath658 and small enough @xmath651 , so that by _ cop _
, its reflection through @xmath657 also lies on @xmath29 .
it follows that an entire neighborhood @xmath586 of @xmath645 in @xmath29 has reflection symmetry through @xmath657 . in this neighborhood , pick two points @xmath659 , @xmath660 , which are the endpoints of a diameter of horizontal cross - section and denote the center of symmetry @xmath657 as @xmath661 .
we can define the reflection through @xmath661 as @xmath662 which maps @xmath659 and @xmath660 respectively to @xmath663 and @xmath664 , which belong to a horizontal cross - section in the neighborhood @xmath586 .
since the differentials satisfy @xmath665 and @xmath666 and @xmath667 we can conclude that @xmath668 .
this implies that the segment @xmath669 $ ] is a diameter and the midpoint of @xmath670 $ ] is mapped to the midpoint of @xmath669 $ ] .
since this holds in the neighborhood @xmath586 of @xmath645 we can choose @xmath638 so that @xmath671\\ & = c(b ) - c(b - t ) \end{aligned}\ ] ] for each @xmath672 . since @xmath644 is arbitrary then by using lemma [ ls5.3:affcrv ] ( linearity criterion ) we can conclude that @xmath28 is affine .
* horizontally spherical case : * each horizontal plane @xmath650 cuts the tube @xmath29 in an @xmath264- sphere centered at @xmath657 , @xmath644 .
let @xmath673 denote the square of the radius of this sphere and @xmath674 for each @xmath644 .
fix @xmath683 , @xmath684 , let @xmath685 denote the orthogonal complemet of @xmath250 in @xmath686 with basis @xmath687 , and @xmath688 . since @xmath29 has _ cop _ , and lies in _ standard position _ we can find a small slope @xmath689 , and an @xmath690-intercept @xmath691 such that the plane @xmath692 given in the new coordinates by @xmath693 intersects @xmath29 in a central ovaloid @xmath108 with center having @xmath694 coordinate @xmath680 . from now on we will identify the curve @xmath695 , the plane @xmath692 , the tube @xmath29 , and the ovaloid @xmath108 with their coordinate representations under the basis @xmath696 because such an identification constitutes an orthogonal isomorphism .
namely , we make the following identifications : for the curve @xmath23 given by @xmath697 we let @xmath698 , where @xmath699 . the plane @xmath692 becomes @xmath700 the tube is replaced with @xmath701^{2 } + \sum_{i=1}^{n-2 } \bigl [ y_{i } - \bar{c}_{i+1}(z ) \bigr]^{2 } = f(z ) \right\}\ ] ] and lastly the ovaloid @xmath702 is replaced with @xmath703^{2 } + \sum_{i=1}^{n-2 } \bigl [ y_{i } - \bar{c}_{i+1}(z ) \bigr]^{2 } = f(z ) \right\ } \label{es5.3:ovaloidyz - crd}\\ & = \bigl\ { ( y , z ) \in \mathbb{r}^{n-2 } \times \mathbb{r } \colon \lvert z \rvert < 1 , \ , \lvert y - c^{\perp}(z ) \rvert^{2 } = \lambda(z ) \bigr\ } \notag \end{aligned}\ ] ] let @xmath707 , @xmath708 , @xmath709 , @xmath710 , @xmath711 , then the cross - section @xmath712 of the ovaloid is an @xmath713-sphere centered at @xmath714 with radius @xmath715 .
when @xmath716 and @xmath717 , choose two antipodal points @xmath718 pick the point @xmath719 and let @xmath244 be the support parameterization of @xmath108 , then there exists a unique @xmath720 with @xmath721 .
since @xmath108 has the center of symmetry whose @xmath694-coordinate is @xmath680 , @xmath722 must also lie on @xmath723 .
the ovaloid @xmath108 has the same tangent hyperplane @xmath724 at the two endpoints of the diameter connecting @xmath719 to @xmath725 .
this hyperplane , when intersected with @xmath726 , becomes the tangent plane to @xmath727 at @xmath719 and @xmath728 .
therefore , we can conclude that @xmath729 , and the chord joining @xmath719 and @xmath730 passes through the center of @xmath108 which has coordinates @xmath731 in the @xmath732 coordinate system . by using the expression of the ovaloid @xmath108 in the @xmath732
coordinate system we get @xmath736^{2 } + \bigl\lvert ( c^{\perp}(\beta ) + y ) - c^{\perp}(\beta + t ) \bigr\rvert^{2 } \\ & = \left [ \frac{\bar{\beta } + t}{m } - \mu_{\beta}^{+}\bar{c}_{1}(t ) - \mu_{\beta}^{-}\bar{c}_{1}(t ) \right]^{2 } + \bigl\lvert c^{\perp}(\beta ) + y - \mu_{\beta}^{+}c^{\perp}(t ) - \mu_{\beta}^{-}c^{\perp}(t ) \bigr
\rvert^{2}. \end{aligned}\ ] ] @xmath737^{2 } + \bigl\lvert ( c^{\perp}(\beta ) - y ) - c^{\perp } ( \beta - t ) \bigr\rvert^{2}\\ & = \left [ \frac{\bar{\beta } - t}{m } - \mu_{\beta}^{+}\bar{c}_{1}(t ) + \mu_{\beta}^{-}\bar{c}_{1}(t ) \right]^{2 } + \bigl\lvert c^{\perp}(\beta ) - y - \mu_{\beta}^{+}c^{\perp}(t ) + \mu_{\beta}^{-}c^{\perp}(t ) \bigr \rvert^{2}. \end{aligned}\ ] ] when we subtract the equation from the equation we get @xmath738^{2}\!+\ ! \bigl\lvert ( c^{\perp}(\beta ) - \mu_{\beta}^{+}c^{\perp}(t ) ) + ( y - \mu_{\beta}^{- } c^{\perp}(t ) ) \bigr\rvert^{2}\\ & - \left [ \left ( \frac{\bar{\beta}}{m } - \mu_{\beta}^{+}\bar{c}_{1}(t ) \right ) - \left ( \frac{t}{m } - \mu_{\beta } ^{- } \bar{c}_{1}(t ) \right ) \right]^{2}\ ! - \ !
\bigl\lvert ( c^{\perp}(\beta ) - \mu_{\beta}^{+}c^{\perp}(t ) ) - ( y - \mu_{\beta}^{-}c^{\perp}(t ) ) \bigr\rvert^{2}\\ & = 4 \left ( \frac{\bar{\beta}}{m } - \mu_{\beta}^{+}\bar{c}_{1}(t ) \right)\left ( \frac{t}{m } - \mu_{\beta}^{- } \bar{c}_{1}(t ) \right ) + 4 \bigl(c^{\perp}(\beta ) - \mu_{\beta}^{+}c^{\perp}(t)\bigr ) \cdot \bigl(y - \mu_{\beta } ^{-}c^{\perp}(t)\bigr ) \end{aligned}\ ] ] recall that @xmath739^{2 } > 0 \end{aligned}\ ] ] for sufficiently small @xmath740 because the ovaloid @xmath108 can not contain its center of symmetry and hence @xmath741 .
so we can conclude that @xmath712 is a nondegenerate sphere @xmath742 \times \{\beta + t\}\ ] ] for small enough @xmath740 .
pick any two distinct points @xmath743 , @xmath744 , @xmath745 , \beta + t \bigr)\\ \bar{y } ' = ( y ' , \beta + t ) & = ( c^{\perp}(\beta + t ) + \xi ' , \beta + t ) \in \mathcal{o}\\ & = \bigl ( c^{\perp}(\beta ) + \bigl [ c^{\perp}(\beta + t ) - c^{\perp}(\beta ) + \xi ' \bigr ] , \beta + t \bigr ) \end{aligned}\ ] ] for two distinct points @xmath746 , @xmath747 .
the central symmetry condition of the ovaloid @xmath108 and the equations , allow us to write the following two equalities @xmath748 - \mu_{\beta}^{-}c^{\perp}(t)\bigr ) \end{aligned}\ ] ] using the coordinates of @xmath743 and @xmath749 - \mu_{\beta}^{-}c^{\perp}(t)\bigr ) \end{aligned}\ ] ] using the coordinates of @xmath750 .
subtracting the equation from the equation we get @xmath751 since the equation holds for every @xmath746 , @xmath747 we can conclude that @xmath752 \end{aligned}\ ] ] holds for small enough @xmath753 depending on @xmath754 and hence using lemma [ ls5.3:affcrv ] ( linearity criterion ) we can conclude that @xmath755 is affine .
the curve @xmath23 has coordinates @xmath756 with respect the the orthonormal basis @xmath757 and since @xmath755 is affine it must be of the form @xmath758 therefore , the curve @xmath28 must be of the form @xmath759 with respect to the basis @xmath760 . using @xmath761 as the tilt direction and the argument above we get @xmath762 with respect to the basis @xmath760 and
hence we can conclude that the curve @xmath763 is affine .
let @xmath29 be a transversely convex tube with _
cop_. by definition @xmath29 is affinely isomorphic to a transversely convex tube @xmath764 in _
standard position_. as central ovaloids are mapped to central ovaloids by affine isomorphisms , @xmath764 also has _
cop_. the central curve of @xmath765 is affine by lemma [ ps5.3:axslemm ] ( axis lemma ) and hence @xmath764 is affinely isomorphic to a rectified transversely convex tube @xmath766 with _ cop _ in _ standard position_. using proposition [ ps5.2:classofrcttube ] we can conclude that @xmath766 is either a cylinder over a central ovaloid or affinely isomorphic to a hypersurface of revolution with _
cop_. as @xmath29 is affinely isomorphic to @xmath767 we can conclude that @xmath29 is either a cylinder over a central ovaloid or affinely isomorphic to a hypersurface of revolution with _
cop_. in the case when @xmath29 is affinely isomorphic to a hypersurface of revolution with _ cop _ we can use @xcite to conclude that @xmath29 is affinely isomorphic to a quadric and hence it is itself a quadric . for every affine isomorphism @xmath768 , @xmath769 is again a proper , complete , immersion with cop if @xmath10 has these properties and @xmath770 is again a cylinder or a quadric depending on whether @xmath14 is a cylinder or a quadric . according to the _ cop _ definition [ ds1.1:cop ] , there exists a cross - cut @xmath12 so that @xmath771 is an ovaloid .
by of the affine invariance of the problem we can assume that @xmath772 and use lemma [ ls3.3:exsttrncvxtube ] ( existence of transversely convex tube ) to construct a transversely convex tube with _ cop _
about @xmath771 with the following properties : * case * @xmath778 in this case , @xmath774 has no critical point in @xmath779 , @xmath780 is an embedding and @xmath10 embeds @xmath781 into @xmath782 as a transversely convex tube @xmath783 with _
cop_. therefore , using proposition [ ps5.4:lclvrs ] ( local version ) , we can conclude that @xmath783 is either a cylinder over a central ovaloid , or a quadric , and because of the affine invariance of the problem we can further assume that @xmath641 is the axis of @xmath783 . assume first that @xmath783 is a cylinder .
our aim is to show that this is not possible by producing a contradiction to the choice of maximal and minimal heights .
since @xmath780 is an embedding and @xmath784 = \dim m$ ] , using the inverse function theorem we can conclude that it is a diffeomorphism onto its image , which is an open subset of @xmath2 .
define the set @xmath785 as @xmath786 we eventually want to show that @xmath787 is a submanifold .
but we first argue that the set @xmath787 exists , is compact and belongs to the boundary of @xmath788 .
the limit @xmath789 exists : fix a cauchy sequence @xmath790 with @xmath791 , @xmath651 , then there exist @xmath792 , @xmath793 smooth curve connecting @xmath794 to @xmath795 so that for each @xmath449 and @xmath796 @xmath797 then the distance between the points @xmath798 and @xmath799 has a bound @xmath800 \leq \mathrm{length } \ , \beta_{ij } = \mathrm{length } \ , \alpha_{ij } <
\epsilon\ ] ] whenever @xmath449 , @xmath796 because @xmath2 has the pullback metric .
therefore , @xmath801 is a cauchy sequence and the limit @xmath789 exists because @xmath802 is a complete metric space , which follows because @xmath10 is a complete immersion .
@xmath803 \cap f^{-1 } \bigl [ \mathcal{o}(b ) \bigr]$ ] and it is compact . take @xmath804 , then @xmath805 and hence @xmath806 $ ] , @xmath807 , @xmath808 = b$ ] .
this implies that @xmath809 and we get one inclusion . on the other hand ,
if we take @xmath810 , @xmath811 with @xmath812 , @xmath813 , @xmath814 , then @xmath815 is a cauchy sequence , @xmath816 and @xmath817 .
therefore , the equality holds and the set @xmath787 is compact because @xmath10 is a proper map .
moreover the definition of @xmath787 implies that @xmath818 .
a similar argument shows that @xmath819 \cap f^{-1 } \bigl [ \mathcal{o}(a ) \bigr]\ ] ] is compact and we can conclude that @xmath820 \setminus \psi_{a , b}(\gamma \times ( a , b))\\ & = \gamma_{a } \cup \gamma_{b}. \end{aligned}\ ] ] now we want to show that @xmath787 is a submanifold of @xmath2 of dimension @xmath821 and conclude that @xmath787 is a cross - cut of @xmath10 relative to @xmath822 .
this conclusion , together with lemma [ ls3.3:exttrncvxtube ] ( extension of transversely convex tube ) will contradict the maximality of @xmath777 .
because of the affine invariance of the problem we can assume that @xmath823 , @xmath824 and denote the map @xmath825 as @xmath145 . note that because @xmath826 is a cylinder , for every @xmath827 @xmath828 hence @xmath829 .
therefore , we can conclude that @xmath830 does not vanish on the compact set @xmath831 and we can define the vectorfield @xmath832 in an open neighborhood @xmath833 of @xmath834 . using the existence and uniqueness of flow box for @xmath270 we obtain a triple @xmath835 , where 1 .
@xmath836 is an open neighborhood of @xmath834 in @xmath2 + 2 .
@xmath837 is smooth + 3 . for each @xmath838
, @xmath839 is the integral curve of @xmath270 and @xmath840 + 4 . for each @xmath841
, @xmath842 is a diffeomorphism onto its image . for each @xmath838
the curve @xmath843 satisfies @xmath844 y(c_{q}(t ) ) = 1\ ] ] for each @xmath841 . therefore
, for each @xmath838 the height function @xmath845 has no local maximum or minimum in @xmath846 .
choose @xmath171 so that @xmath847 , @xmath848 , and @xmath849 .
define the submanifold @xmath850 of @xmath2 , which is diffeomorphic to @xmath851 as @xmath852^{-1 } \mathcal{o}(c).\ ] ] its @xmath853-distance to @xmath834 satisfies @xmath854 = \lvert c \rvert < \operatorname{\mathrm{dist}}_{m}(\gamma_{0 } , m \setminus v_{1})\ ] ] and hence we can conclude that @xmath855 .
similarly for every @xmath856 the submanifold @xmath857 satisfies @xmath858 .
@xmath864^{-1 } \mathcal{o } ( c + s)\\ & \iff \phi_{s}(x ) \in \psi(\gamma \times ( a , 0 ) ) \text { and } ( f \circ \phi_{s})(x ) \in \mathcal{o}(c + s)\\ & \iff \phi_{s}(x ) \in \psi ( \gamma \times ( a , 0 ) ) . \end{aligned}\ ] ] the last equivalence is true because @xmath865 implies that its @xmath10 image @xmath866 and @xmath867 yields @xmath868 . by definition @xmath869 \bigr)$ ] , and if this image is not a subset of @xmath870 then there must exist @xmath871 so that @xmath872 , which implies the existence of a local extremum in @xmath873 .
this contradicts the observation in and hence we can conclude that @xmath874 .
assume that there exists @xmath875 , then there is an open neighborhood @xmath876 of @xmath660 with @xmath877 > 0 $ ] .
choose @xmath878 so that for each @xmath879 @xmath880 > 0 $ ] .
then for each @xmath881 @xmath882 \cap \phi_{-c+s}(\gamma_{c } ) & = \emptyset\\ \bigl[\ , w \cap \psi ( \gamma \times ( a , 0 ) \,\bigr ] \cap \gamma_{s } & = \emptyset \end{aligned}\ ] ] and @xmath883 \cap f(\gamma_{s } ) = \emptyset$ ] because @xmath884 is a bijection onto its image .
therefore , we must have for each @xmath881 @xmath885 \cap \mathcal{o}(s ) = \emptyset\ ] ] which is not possible .
this implies that @xmath886 .
let @xmath863 be arbitrary and consider the limit @xmath887 for some sequence @xmath888 in @xmath112 and hence we have @xmath889 \cap f^{-1 } \bigl [ \mathcal{o}(0 ) \bigr ] = \gamma_{0}.\ ] ] so we get @xmath890 and since @xmath891 is a diffeomorphism and @xmath850 is an @xmath821 dimensional compact connected submanifold of @xmath2 we can conclude that @xmath834 is also an @xmath821 dimensional compact connected submanifold of @xmath2 .
therefore , using lemma [ ls3.3:exttubnbd ] ( tubular neighborhood ) we can conclude that @xmath834 is actually a cross - cut and lemma [ ls3.3:exsttrncvxtube ] ( existence of transversely convex tube ) shows that there exists a transversely convex tube about @xmath892 .
finally , lemma [ ls3.3:exttrncvxtube ] ( extension of transversely convex tube ) implies that tube about @xmath893 extends @xmath894 , contradicting the maximality of @xmath895 .
it follows that the transversely convex tube @xmath896 is a quadric hypersurface of revolution about the @xmath897-axis .
therefore there exist a vertical dilation @xmath898 and a vertical translation @xmath899 such that the affine isomorphism @xmath900 satisfies @xmath901 for some @xmath171 . since @xmath783 intersects some horizontal hyperplane along a compact set so does @xmath902 .
after we remove the cylindrical case we can conclude that @xmath902 is one of the following quadrics : therefore , @xmath783 is affinely isomorphic to one of the quadrics above . discarding the cone , on all these hypersurfaces ,
horizontal cross - sections are spheres and hence the maximality and minimality of @xmath777 and @xmath46 , respectively , must be dictated by the condition alone .
this implies that @xmath774 must have critical points on both boundaries of the open set @xmath788 .
namely , there exist @xmath903 , @xmath904 so that @xmath905 but among the quadrics listed above , @xmath906 has multiple critical points only on the sphere , where it attains both a maximum and a minimum .
so we can conclude that @xmath907 $ ] equals the complete ellipsoid @xmath908 .
since @xmath2 is connected we must have @xmath909 .
* case * @xmath910 in this case @xmath911 is an embedding and since @xmath912 the map @xmath145 must be a diffeomorphism onto its image , which is both open and closed in @xmath2 . since @xmath2 is connected @xmath913 .
@xmath914 & = \bigcup_{r > 0 } f \bigl [ \psi(\gamma \times ( -r , r ) ) \bigr]\\ & = \bigcup_{r > 0 } f(u_{-r , r } ) \end{aligned}\ ] ] where @xmath915 for each @xmath916 . for each @xmath916 , @xmath917 is a transversely convex tube with _
cop_. using proposition [ ps5.4:lclvrs ] ( local version )
we can conclude that @xmath917 is either a cylinder over a cental ovaloid with axis @xmath918 or a quadric hypersurface of revolution about @xmath641 .
therefore , @xmath917 must either be a cylinder over a central ovaloid or a tube hyperboloid for each @xmath916 .
assume that there exists @xmath919 and @xmath920 is a cylinder over a central ovaloid , then for every @xmath921 @xmath922 and hence @xmath917 is a cylinder over a central ovaloid .
so we can conclude that @xmath923 is a cylinder over a central ovaloid .
similarly , if there exists @xmath919 so that @xmath920 is a tube hyperboloid , then @xmath14 is a again a tube hyperboloid . *
case * @xmath924 since the reflection about @xmath925 is an affine isomorphism , these two cases are equivalent .
so we assume @xmath926 . for each @xmath927 @xmath928 \text {
is an open subset of $ m$ and}\\ & f(u_{a , b } ) \text { is a transversely convex tube with \emph{cop}. } \end{aligned}\ ] ] @xmath929 can not be a cylinder over a central ovaloid because we can get , as in * case 1 * , a contradiction to the minimality of @xmath46 .
so we can conclude that for each @xmath927 , @xmath929 is either a spherical paraboloid or a convex hyperboloid .
assume that there exists @xmath930 so that @xmath931 is a spherical paraboloid , then for each @xmath932 , @xmath929 is again a spherical paraboloid and @xmath933 is a complete paraboloid .
since @xmath2 is connected , @xmath934 and @xmath14 is a complete paraboloid .
similarly , if there exists @xmath930 so that @xmath931 is a convex hyperboloid , then @xmath935 is a convex hyperboloid .
99 wilhelm blaschke , _ affine differentialgeometrie .
differentialgeometrie der kreis - und kugelgruppen_. gesammelte werke , band 4 , edited by burau , s.s .
chern , k. leichtwei , h.r .
mller , l.a .
santal , simon and k.strubecker , thales - verlag , essen , 1985 . | a hypersurface @xmath0 , @xmath1 , has the central ovaloid property , or _ cop _ , if * @xmath2 meets some hyperplane transversally along an ovaloid , and *
every such ovaloid on @xmath2 has central symmetry .
generalizing work of b. solomon to higher dimensions , we show that a complete , connected , smooth hypersurface with _ cop _ must either be a cylinder over a central ovaloid , or else quadric . |
as theoretically predicted in @xcite and experimentally discovered in @xcite , two - band solid state systems in 3d can host topologically protected `` weyl semimetallic phases '' in which the quasiparticle excitations at the band crossings ( `` weyl points '' ) share some features with weyl fermions from relativistic quantum mechanics .
the experimental signature , namely `` fermi arcs '' of surface states which connect the projected weyl points in the surface brillouin zone , is just as remarkable .
based on this initial success , much effort has been put into the general study of topological semimetallic phases in the hope of predicting and eventually realising new exotic fermionic quasiparticles in condensed matter systems .
most proposals have focused on _ local _ aspects in the sense of finding new types of topological obstructions to locally opening up gaps in semimetal band crossings .
staying in the two - band case , there are generalisations of the basic weyl semimetal phase to `` type - ii '' ones @xcite , as well as `` quadratic '' ones @xcite .
one can consider model hamiltonians with more than two bands and in a different number of spatial dimensions @xcite .
following the example set by topological insulator theory , one may also introduce antiunitary symmetry constraints such as time - reversal and charge - conjugation @xcite . to circumvent certain difficulties in achieving non - trivial obstructions , point symmetries
were also introduced into the game , and a host of possibilities arise @xcite .
our results in this paper have a different focus , concentrating on ( 1 ) isolating in a conceptually simple but rigorous way the general mathematical mechanism allowing for local semimetallic charges , ( 2 ) providing the full _ global _ topological characterisation of semimetal band structures , fermi arcs , and the relation to topological insulator invariants , and ( 3 ) introducing a new family of semimetals whose topological invariants have a very different character to those commonly used in the literature .
figure [ fig : flowchart ] summarises the physical and mathematical concepts which we introduce in this paper . ) which are dual to each other.,scaledwidth=100.0% ] with regards to ( 1 ) , we see , with much hindsight , that the local topological charges protecting basic types of semimetal crossings are completely inherited from the singular vector field which specifies the hamiltonian @xcite . the poincar hopf theorem @xcite then shows how the brillouin zone topology forces a global charge cancellation condition , and the prediction of fermi arcs becomes a corollary of the bulk - boundary correspondence .
this brings us to point ( 2 ) .
while the 3d weyl semimetal may be intuitively understood using a stokes theorem argument to predict `` jumps '' in 2d chern numbers across the weyl points , there are several subtleties involved , not the least of which is the fact that the local charges ( equal to the jumps ) fail to completely characterise the semimetal , and can not predict , a priori , the fermi arc topology @xcite .
it is here that the finer topology of the brillouin torus comes into play : the fact that @xmath0 has non - trivial cycles means that `` the ( fermi ) arc joining two weyl points '' is an ambiguous notion , even up to deformations of the arc ( contrast with the simply - connected spheres ) .
the resolution of this ambiguity requires additional global data which captures the full topology of semimetal band structures , and is closely related to the concept of euler structures introduced by turaev to resolve ambiguities in torsions of reidemeister type @xcite ; both hinge on the fact that the first homology of the underlying manifold may be non - zero . by considering a homological version of _ euler structures _
@xcite , we find that the language of _ euler chains _ very concisely represents all the _ topological _ features of a semimetal .
the euler chain representation directly predicts the _ topology _ of the fermi arcs , and the latter are the experimental signature of a topological semimetal .
furthermore , the notion of topological equivalence between semimetals becomes clear in this language .
a semimetal hamiltonian determines a fermi arc , e.g. through a transfer matrix formalism @xcite , and this determination descends to the level of topological invariants in the sense that equivalent hamiltonians give rise to equivalent fermi arcs .
having isolated the basic mathematical principles underlying the geometry and topology of semimetals , we move on to ( 3 ) : the generalisation of our theory for 2-frames or 2 vector fields along the lines of the atiyah dupont theorem @xcite which is the analogue of the poincar
hopf theorem in this context , where the _ kervaire semicharacteristic _ plays the role of the euler characteristic .
it is a topological invariant of a very different nature to those that had been previously utilised in the literature on topological phases ( e.g. chern classes and related characteristic classes , @xmath1-theory , winding numbers / degrees ) .
this leads us to define some new mathematical notions _ kervaire structures _ and _ kervaire chains _ which are two mutually dual objects characterising a new class of @xmath2-topological semimetals introduced for the first time in this paper , experimentally predicting torsion fermi arcs . in section [ sec : physback ] , we give a quick overview of the use of cohomological characteristic classes to study topological insulating phases , and then connect it to semimetal phases via the mayer vietoris technique of `` patching together '' the non - singular portion ( avoiding the weyl points ) and topologically trivial portions ( small neighbourhoods of the weyl points ) .
the basic `` semimetal mayer
vietoris sequence '' is given a geometric interpretation as an extension problem , parallel to the physical problem of deforming a semimetal band structure into a globally insulating one , in section [ sec : extension ] . in section [ sec : euler ] , we introduce the concept of euler structures and the euler chain representation of a semimetal .
we illustrate with diagrams the basic intuitive ideas , and apply the euler chain concept to fermi arc topology and their rewirings . in section [ sec : higherdimension ] , we study dirac - type hamiltonians parametrised by singular vector fields , emphasising the merits of an abstract coordinate - free analysis with regards to symmetry considerations . in particular , we justify the generality of dirac - type hamiltonians through natural symmetry constraints , in order to avoid introducing `` spurious topology '' through ad - hoc parametrisations of toy model hamiltonians . in section [ sec : semimetalgerbe ] , a geometric connection to gerbes is made in the 4d case , and in section [ sec:5dsemimetal ] , the connection to quaternionic valence bundles in 5d is made , and the geometric interpretation as an extension problem is in section [ sec : extension ] . in section [ sec : torsionsemimetal ] , we introduce a new class of semimetals modelled on hamiltonians which are bilinear in gamma matrices .
the topology of such semimetals is characterised by the notion of kervaire structures and chains , and we predict an interesting phenomenon of @xmath2 surface fermi arcs .
finally we suggest some directions for future work in section [ sec : outlook ] .
* conventions . * in this paper
, @xmath3 will always be a compact connected smooth @xmath4-dimensional manifold without boundary , playing the role of a general parameter manifold for a family of hamiltonians . in specific instances
, it may have an orientation , metric , spin@xmath5 structure etc .
, and may be the brillouin zone for a family of bloch hamiltonians , as additionally specified .
all ( co)-homology groups have integer coefficients unless otherwise indicated .
the real clifford algebra @xmath6 has @xmath7 generators squaring to @xmath8 and @xmath9 generators squaring to @xmath10 , corresponding to an orthonormal basis @xmath11 in @xmath12 for the bilinear form @xmath13 ; @xmath14 refers to @xmath15 , and the complex clifford algebra @xmath16 has @xmath17 anticommuting generators squaring to @xmath10 .
2d chern insulators are classified by a first chern number , equal to the integrated berry curvature 2-form of the valence bands over the brillouin zone @xmath18 .
when the fermi energy @xmath19 lies in a spectral gap , the fermi projection onto the valence bundle @xmath20 may be defined . over @xmath21 the @xmath20
are classified by their rank and first chern class , with the latter generally considered to be the interesting invariant . since the first chern class of @xmath20 is that of the determinant line bundle , the classification problem of 2d chern insulators is formally equivalent to that of u(1 ) line bundles over the brillouin zone for which the _ integral cohomology group _
@xmath22 provides a complete answer . in terms of differential forms
, @xmath23 measures the failure of the ( berry ) curvature 2-form @xmath24 of @xmath20 to be globally exact ( the berry connection 1-form @xmath25 only exists locally ) .
the integral of @xmath24 over @xmath21 yields an integer - valued first chern number , which may be interpreted as a topological obstruction to globally defining a basis of valence bloch eigenstates for @xmath20 .
a typical way to construct a valence line bundle with first chern class @xmath26 is to consider @xmath27 traceless . ]
bloch hamiltonians @xmath28 , parametrised by a nowhere - vanishing smooth vector field @xmath29 on @xmath21 . here
@xmath30 are the pauli matrices .
the spectrum is @xmath31 , and the spectrally - flattened hamiltonian @xmath32 is parametrised by the unit vector map @xmath33 .
note that @xmath34 and @xmath32 determine the same valence line subbundle @xmath20 of the bloch bundle @xmath35 .
the map @xmath36 has an integer degree given by the formula @xmath37)=\text{deg}(\hat{{\bm{{h}}}})[s^2]$ ] , where @xmath38 , [ s^2]$ ] denote the fundamental classes .
the degree is a homotopy invariant , and all degrees do occur . by identifying @xmath39 with @xmath40 ( the bloch sphere construction ) , we see that @xmath20 is given by the pullback of the tautological ( hopf ) line bundle over @xmath40 under @xmath36 . due to the low - dimensionality of @xmath21 , all maps into @xmath41
are approximated by maps into @xmath40 , and so the first chern class of @xmath20 is precisely the degree of @xmath36 . in higher dimensions , and also in the presence of additional symmetries ( e.g. time - reversal or point group symmetries ) constraining the form of the hamiltonians , topological band insulators are classified by more complicated invariants . in the mathematical physics literature ,
these have included higher degree cohomology invariants playing the role of characteristic classes for the valence bundles , e.g. higher chern / instanton numbers , symplectic bundle invariants @xcite , fu kane
mele / fkmm - type invariants @xcite , or generalised cohomology invariants ( @xmath1-theory ) @xcite .
we are primarily interested in @xmath42 or @xmath43 spatial dimensions , and in those dimensions , the relevant `` berry curvature form '' @xmath24 can be a higher - degree form @xcite , and so a higher degree cohomology group @xmath44 comes into play .
cohomology groups with other coefficients also arise quite generally in obstruction theory , which can be used to study semimetal - insulator transitions .
all these ( generalised ) cohomological invariants are also homotopy invariants , and are closely related to degree theory ( as we saw from the 2d chern insulator described above ) . although the prototypical topological semimetal is the weyl semimetal in @xmath45 , the general theory presented in this paper works in higher dimensions .
this is not merely of theoretical interest . in the case of topological insulators , there are concrete proposals and experiments in which topological phases of physical systems in three or fewer dimensions formally realise those in @xmath46 .
some examples are quasicrystalline systems @xcite , time - periodic ( floquet ) insulators @xcite and their photonic counterparts @xcite , and general `` virtual '' topological insulators @xcite .
it is in this spirit that we embark on the mathematical study of semimetallic phases for general @xmath4 . in a semimetal
, the fermi level does not lie in a spectral gap , but instead passes through band crossings at some weyl submanifold @xmath47 of the momentum space manifold @xmath3 .
such crossings may arise as `` accidental degeneracies '' @xcite and , for `` dirac - type hamiltonians '' in @xmath4-dimensions occur generically at points . a quick way to see
this is to consider dirac - type bloch hamiltonians of the form @xmath48 here each @xmath49 is a smooth assignment of @xmath4-component vectors , and @xmath50 is the vector of traceless hermitian @xmath51-matrices representing irreducibly and self - adjointly the @xmath4-generators of a clifford algebra ( for @xmath45 , these are the usual @xmath27 pauli matrices ) . by the clifford algebra relations @xmath52 , the square of @xmath53 is the scalar @xmath54 , so the eigenvalues of @xmath53 are @xmath55 , each with degeneracy equal to half the dimension of the @xmath51-matrices .
this means that a band crossing at @xmath56 requires @xmath4 conditions @xmath57 , and we see that the locus @xmath47 of band crossings generically has codimension @xmath4 .
when @xmath3 is a @xmath4-manifold , @xmath47 is generically a collection of isolated `` weyl points '' , the ( rather misleading ) terminology arising from the @xmath45 case in which the low - energy excitations around a band crossing are essentially described by the weyl equation @xcite .
furthermore , each weyl point @xmath58 serves as a generalised _ monopole _ , and may be assigned a _ local topological charge _ via the unit - vector map restricted to a small enclosing @xmath59-sphere @xmath60 , @xmath61 and defining the local charge @xmath62 to be the degree of @xmath63 . the local charge information of all the crossings for the hamiltonian @xmath64 is concisely summarised by the 0-chain @xmath65 .
the unit vector map @xmath36 on @xmath66 , and its restrictions to @xmath60 , can be used to pullback the hopf line bundle with connection when @xmath45 , the basic gerbe with connection when @xmath67 and the quaternionic hopf line bundle with connection when @xmath68 ( see appendix [ appendix ] for their definitions ) . as explained in section [ sec : extension ] , these geometrical structures arise naturally when analysing dirac - type hamiltonians , their band crossings , and the topology of their valence bundles .
interesting phenomena arise when we study the band crossings _
globally_. for instance , if @xmath29 can be identified as a tangent vector field over @xmath3 , as is often the case in physical models , then the classical poincar
hopf theorem guarantees that @xmath69 where @xmath70 is the euler characteristic of @xmath3 , which is zero for @xmath71 or @xmath4 odd .
this global `` charge - cancellation '' condition is well - known in lattice gauge theory ( where @xmath71 ) as the nielsen ninomiya theorem @xcite and discussed in the semimetal context in @xcite . together with the bulk - boundary correspondence
, it predicts the appearance of surface fermi arcs connecting the projected weyl points on a surface brillouin zone .
these arcs are an experimental signature of topological non - triviality in a semimetal band structure , and have been discovered recently @xcite . more generally , @xmath29 could be a section of some oriented rank-@xmath4 vector bundle @xmath72 over a compact oriented @xmath4-manifold @xmath3 .
the local topological charges at the zeros of @xmath29 are defined as before , as is the euler characteristic of @xmath72 ( evaluating the euler class on the fundamental class ) . when the euler characteristic @xmath73 vanishes ( and when @xmath74 following @xcite ) , a notion of _ euler structures _ can be defined , and they provides links between topological semimetal invariants , seiberg
witten invariants and reidemeister torsions .
more pertinently , euler structures are a torsor over @xmath75 , and they provide an elegant way to understand some global aspects of fermi arc topology , including certain ambiguities first studied in @xcite , as well as the `` rewiring '' of fermi arcs @xcite .
this is already interesting when @xmath71 , for which there are non - trivial euler structures .
we take the following as a working definition : [ defn : abstracthamiltonian ] let @xmath3 be a compact , oriented @xmath4-dimensional riemannian spin@xmath76 manifold with @xmath77 .
abstract dirac - type hamiltonian _ is a smooth vector field @xmath29 over @xmath3 , which is _ insulating _
if @xmath29 is non - singular ( nowhere zero ) , and _ semimetallic _ if @xmath29 has a finite set @xmath47 of singularities .
this definition is motivated by the construction of concrete bloch hamiltonians from such vector fields , generalising , given in section [ sec : higherdimension ] . in @xcite , we carried out an analysis of the global structure of semimetal band structures and fermi arcs .
the mayer
vietoris principle provided the key to understanding the connection between topological semimetals and insulators ( fig .
[ fig : metal - or - insulator ] ) .
in particular , we showed that the local charge information at weyl points needs to be supplemented with some global data to fully characterise semimetal band structure and fermi arcs .
we summarise the key constructions in @xcite , then provide several different ways to understand them .
* definition of @xmath78 .
* let @xmath3 be a compact oriented @xmath4-manifold ( usually the brillouin torus @xmath0 in concrete models ) with @xmath77 , and consider a smooth family of ( concrete ) @xmath79 bloch hamiltonians @xmath80 .
the spectrum comprises @xmath17 bands , and we assume that band crossings at the fermi level ( normalized to @xmath81 ) occur on a finite set @xmath82 of isolated `` weyl points '' . on the complement @xmath66 ,
the bloch hamiltonians are gapped , and there is no difficulty in defining the valence bundle @xmath20 over @xmath66 and its characteristic classes in @xmath83 . for each @xmath58 , choose ( mutually disjoint ) open balls @xmath84 containing @xmath85 , and @xmath86-spheres @xmath87 surrounding @xmath85 .
let @xmath88 and @xmath89 be the respective disjoint unions . thus @xmath3 is covered by @xmath90 and @xmath66 , and
the intersection of the covering sets is @xmath91 which is homotopy equivalent to @xmath92 ( figure [ fig : mv - subspaces ] illustrates the setting for two weyl points ) . used in the semimetal mv - sequence , for two weyl points @xmath93.,scaledwidth=80.0% ] the mv sequence links the cohomology groups of @xmath3 with those of the covering subspaces and the intersection @xmath92 , by an exact sequence @xmath94 the maps which increase cohomological degree are the mv connecting maps , while the others are ( differences of ) restriction maps .
note that @xmath90 is contractible , so the @xmath95 term vanishes for @xmath96 .
when a semimetallic bloch hamiltonian is specified abstractly by a vector field @xmath29 as in definition [ defn : abstracthamiltonian ] , the zero set of @xmath29 will correspond to the weyl points @xmath47 , and we have the subsets @xmath97 exactly as above .
then @xmath29 defines a topological invariant for the semimetal which lives in @xmath98 ( section [ sec : euler ] ) .
the most interesting part of the mv sequence is at @xmath99 , which we call the cohomological _ semimetal mv sequence _ , @xmath100 here @xmath101 is the direct sum of the local charge groups for each @xmath85 , and the last mv connecting map @xmath102 is the `` total charge operator '' which adds up the local charges @xcite . for integer coefficients , @xmath103 and
the charges are @xmath104-valued , but we will also consider @xmath2 coefficients in section [ sec : torsionsemimetal ] .
let us be more explicit in the basic 3d situation , with @xmath27 bloch hamiltonians parametrised by @xmath105 and @xmath106 .
the mv - sequence reads @xmath107 consider local charges @xmath108 which cancel , then there exists ( by exactness ) a line bundle @xmath20 over @xmath109 whose restriction to @xmath110 has chern class given by the local charge at @xmath111 . away from @xmath112 ( with charges @xmath113 )
, @xmath114 has a first chern number @xmath115 on the 2d subtorus in the @xmath116-@xmath117 direction ( blue ) . consider the region @xmath118 between the two blue 2-tori , with a small ball around @xmath119 removed .
this is a 3-manifold whose boundary @xmath120 consists of the two blue 2-tori ( oppositely oriented ) together with a small 2-sphere around @xmath121 .
note that the top and bottom faces are identified , as are the front and back faces so they do not contribute to @xmath120 . everywhere on @xmath118 ,
the valence line bundle is well - defined , and its berry curvature 2-form @xmath24 is closed . by stokes theorem ,
@xmath122 , so the difference between the first chern numbers for the left 2-torus and the right 2-torus is precisely the first chern number on the small 2-sphere , which is @xmath123 .
this argument is easily generalised to higher dimensions and higher curvature forms ( remark [ remark : higherjumps]).,scaledwidth=60.0% ] * intuitive idea behind relation local charges and first chern numbers of a semimetal .
* suppose the vector field @xmath29 on @xmath124 vanishes at @xmath93 with local indices @xmath125 respectively .
for simplicity , we first assume that @xmath111 have the same @xmath126-coordinates , as in figure [ fig : chern - jump ] .
take a trivial bloch bundle @xmath127 , the pauli matrices @xmath128 with respect to a trivialisation , and let @xmath129 .
then on @xmath109 , the valence line bundle @xmath20 is well - defined . at each @xmath130 away from the weyl points , @xmath20 has first chern numbers @xmath131 obtained by pairing @xmath132 with the fundamental class of the 2-torus at @xmath130 ( integrating the berry curvature over the 2-torus ) . the integer @xmath133 is sometimes called a `` weak invariant '' for `` weak '' topological insulators supported on lower - dimensional subtori of the brillouin torus . as a function of @xmath130 , @xmath133 remains constant unless a weyl point is traversed , whence it jumps by an amount equal to the local charge there .
a possible function @xmath133 is given at the bottom of figure [ fig : chern - jump ] , and different choices of @xmath29 can give rise to different @xmath134 .
note that even with @xmath134 and the local charges @xmath135 still do not completely specify the class of @xmath20 there are two other independent first chern numbers @xmath136 which are constant functions of @xmath137 and @xmath138 respectively ( since both weyl points are traversed simultaneously as either @xmath137 or @xmath138 is varied ) . in general , the @xmath111 need not have the same @xmath126-coordinates , and so there are similar jumps in the functions @xmath139 and @xmath140 , see figures [ fig : intersection]-[fig : intersection2 ] .
there is a simple way to concisely and invariantly ( i.e. coordinate - free ) capture all the jumps in the various first chern numbers , through a poincar dual picture in terms of euler chains ( see section [ sec : semimetaleulerchain ] and remark [ remark : higherjumps ] ) . *
all first chern numbers can appear in a weyl semimetal .
* let us also consider @xmath141 which arise as a subbundle of a @xmath142 which is not necessarily trivial .
such @xmath141 are easily obtained by tensoring @xmath143 with a line bundle @xmath144 with @xmath145 .
the `` twisted '' hamiltonian @xmath146 acts on the `` twisted '' bloch bundle @xmath147 , and has fermi line bundle @xmath148 ( defined over @xmath109 ) .
therefore @xmath149 and , in particular , the first chern numbers @xmath150 increases by @xmath151 and similarly for @xmath152 .
in this way , all elements of @xmath153 can be realised as valence line bundles of some semimetallic bloch hamiltonian . in @xmath45 ,
the exact sequence has a direct interpretation as a solution to a geometrical extension problem , which we can interpret in terms of insulator - semimetal transitions .
let us recall that @xmath154 classifies line bundles over @xmath3 ( e.g. determinant valence bundles ) .
so for @xmath155 , the exactness of tells us the following : 1 .
a collection of line bundles @xmath156 with chern classes @xmath157 may be extended to a line bundle over @xmath66 ( comes from a semimetal ) iff @xmath158 ( local charges cancel ) .
a line bundle @xmath159 extends to all of @xmath3 iff each local charge vanishes .
thus a semimetal ( whose valence bundle represents a class in @xmath160 ) can be gapped into an insulator iff all the local charges are zero .
a _ topological semimetal _ with a topologically protected crossing at @xmath85 has non - zero local charge there .
the insulator invariants in @xmath161 ( `` weak '' chern insulators ) appear faithfully in @xmath160 .
thus two semimetals with the same local charge ( same image under @xmath162 ) can be _ globally inequivalent _ , differing by some element in ( the image under @xmath163 of ) @xmath161 .
an important consequence of point 3 is that the local charge information is not sufficient to determine the global band structure of the semimetal .
as explained in @xcite , this ambiguity has crucial consequences for the determination of the resulting surface fermi arcs .
we will have more to say about this in section [ sec : bbc ] , see also figure [ fig : bulk2boundary ] .
more generally , in @xmath77 , a dirac - type bloch hamiltonian in the sense of definition [ defn : abstracthamiltonian ] determines by restriction smooth maps @xmath164 , and the semimetal mv sequence can be interpreted at this level . in appendix
[ appendix ] we describe the relevant geometric structures on @xmath165 , @xmath166 , and these can be pulled back to @xmath92 via all the @xmath63 . 1 . in @xmath45 ,
let @xmath144 be the hopf line bundle over @xmath39 .
consider a collection of pullback line bundles @xmath167 over @xmath60 , such that @xmath168 , where @xmath169 denotes the degree of the map @xmath63 .
although they individually may not extend to @xmath170 , together @xmath171 over @xmath92 does extend to @xmath66 by exactness of the mv sequence .
see the appendix for more information .
2 . in @xmath67 , the @xmath172 part of the semimetal mayer vietoris sequence @xmath173 has the following analogous geometric interpretation .
a collection of pullback basic gerbes @xmath174 with dixmier - douady numbers @xmath175 may be extended to a gerbe over @xmath66 ( comes from a gerbe semimetal ) iff @xmath176 ( local dd charges cancel ) .
the pullback basic gerbe @xmath177 extends to all of @xmath3 iff each local dd charge vanishes .
thus a gerbe semimetal ( with topological invariant in @xmath178 ) can be gapped into an insulator iff all the local dd charges are zero .
a _ topological gerbe semimetal _ with a topologically protected crossing at @xmath85 has non - zero local dd charge there .
the insulator invariants in @xmath179 ( `` weak '' gerbe insulators ) appear faithfully in @xmath178 .
thus two gerbe semimetals with the same local dd charge ( same image under @xmath162 ) can be _ globally inequivalent _ , differing by some element in ( the image under @xmath163 of ) @xmath179 .
conversely , the pullback gerbes @xmath180 over @xmath60 , where @xmath181 is the basic gerbe over @xmath182 , such that @xmath183 .
although the @xmath180 may not individually extend to @xmath170 , the gerbe @xmath184 does extend to @xmath66 by exactness of the mv sequence .
see section [ sec : semimetalgerbe ] and the appendix for more details .
3 . in @xmath68 ,
the analysis is analogous to the case of line bundles done above .
one replaces the line bundle @xmath144 over @xmath40 by the quaternionic line bundle @xmath185 over @xmath186 and one uses the mv sequence in , where the first chern class is replaced by the 2nd chern class .
see section [ sec:5dsemimetal ] and the appendix for more details .
although there is a mv sequence analysis in the case of kervaire semimetals ( section [ sec : kervairestructures ] ) , we are unable to formulate a geometric extension problem in this context ( see section [ sec : outlook ] ) . in section [ sec:2dberry ]
, we briefly alluded to the relation between chern numbers and berry connections associated to valence subbundles for a family of hamiltonians over @xmath3 . in general , a rank-@xmath17 eigen - subbundle has a @xmath187 adiabatic ( non - abelian berry ) connection , whose curvature gives the real ( de rham ) chern classes of the subbundle @xcite . for @xmath188 ,
the curvature 2-form is involved in computing first chern numbers . in the presence of time - reversal symmetry ,
the subbundle is complex even - dimensional ( @xmath189 ) and in fact a quaternionic symplectic bundle with quaternionic dimension @xmath134 ; the berry connection becomes a @xmath190 one , see @xcite and section [ sec : quaternionbundle ] . for @xmath191 , the 4-form obtained from squaring the curvature enters in the computation of second chern numbers . in a similar vein ,
the additional imposition of a chiral symmetry allows a `` semimetal gerbe '' to be associated to the family of hamiltonians , see section [ sec : semimetalgerbe ] .
there is a `` berry connection 2-form '' with 3-form curvature , which computes the dixmier
douady invariants of the gerbe .
fermi arcs are more naturally analyzed in homology , and for this reason , we poincar - dualise @xmath192 with the physical meanings of the first two groups explained in section [ sec : euler ] .
roughly speaking , the poincar dual of a semimetal invariant in @xmath160 is an euler chain in the _ relative _ homology group @xmath193 , whose boundary is the 0-chain of local charge data
. equation can also be exhibited directly as a mayer
vietoris sequence for _ borel moore _ homology @xmath194 @xcite . the groups @xmath195 may be defined for locally compact spaces @xmath118 using _ locally finite _ chains . for compact spaces , @xmath194 and the usual ( singular ) homology @xmath196 coincide . however , @xmath197 whereas @xmath198 , for @xmath84 an open ball containing @xmath85 .
there are isomorphisms @xmath199 , @xmath200 , poincar duality @xmath201 , and restriction maps to @xmath194 of open subsets , see chapter 2.6 of @xcite .
the poincar dual of written in terms of @xmath194 is @xmath202 which is precisely the mv - sequence for @xmath194 ( ix.2.3 of @xcite ) , with respect to the open cover @xmath203 of @xmath3 ; c.f . the `` localization '' long exact sequence , ix.2.1 of @xcite and 2.6.10 of @xcite . the sequences and are the same on account of @xmath204 . the language of borel
moore homology has certain advantages , but for this paper we stay in the more elementary setting of relative homology so as to avoid having to introduce too much technical machinery . in @xcite
, we explained how fermi arcs represent _ relative homology _ classes in @xmath205 , where @xmath206 projects out one torus direction , so @xmath207 is the surface brillouin zone .
the euler chain representing the topological data of a semimetal is pushed forward to a surface fermi arc , and this is the poincar dual picture of the bulk - boundary correspondence , see section [ sec : bbc ] .
since we parametrise dirac - type bloch hamiltonians by vector fields @xmath29 on @xmath3 , we would also like to analyse the semimetal mv - sequence in terms @xmath29 and its homotopies . in the insulating case
, we might want to maintain the gap condition and so consider homotopies through non - singular ( i.e. nowhere vanishing ) vector fields . in the semimetal case ,
we want to at least maintain the gap condition everywhere except perhaps at a finite set of isolated weyl points .
we may require the set of weyl points and their charges to be kept fixed , or we may allow them to move around and be created / annihilated in pairs .
* smooth euler structures . * let @xmath3 be a compact oriented @xmath4-manifold with @xmath208 , and suppose @xmath3 has euler characteristic @xmath209 ( this is automatic in @xmath45 ) so that it has a global non - vanishing vector field .
[ defn : smootheuler ] two non - singular smooth vector fields @xmath210 on @xmath3 are said to be _ homotopic _ ( also called _ _ homologous _ _ through non - singular vector fields can also be analysed @xcite , but we do not consider this in this paper .
] in @xcite ) if for some open ball @xmath211 , the fields @xmath210 are homotopic on @xmath212 in the class of non - singular vector fields on @xmath212 .
the set @xmath213 of homotopy classes of non - singular vector fields is called the set of _ smooth euler structures _ on @xmath3 . as explained in section 5 of @xcite , the first obstruction to such a homotopy between @xmath214
is an element ( written as @xmath215 ) of @xmath216 , canonically isomorphic to @xmath75 by poincar duality .
it turns out that there is a natural free and transitive action of the group of obstructions @xmath217 on @xmath213 .
if we pick a reference nonsingular @xmath218 as the zero , then we can identify @xmath213 with @xmath219 . for @xmath71 ,
a natural choice for @xmath218 is a constant length vector field pointing along a torus cycle ( this gives a trivial insulating phase ) ; reference to such a @xmath218 will be implicit in this case .
we remark that the space of spin@xmath5 structures for @xmath3 is a @xmath154-torsor , and that when @xmath45 , a non - singular @xmath218 determines a spin@xmath5 structure through the unit vector field @xmath220 @xcite , and vice - versa . an equivalent definition of euler structures , which is useful for our generalisation to kervaire structures in section [ sec : kervairestructures ] ,
was given in @xcite as follows .
[ defn : coeuler ] let @xmath3 be a compact oriented @xmath4-manifold with @xmath209 , and let @xmath221 be the sphere bundle for the tangent bundle @xmath222 ( with a riemannian metric ) .
define the set of _ euler structures _ to be the subset @xmath223 for which the restriction to each ( oriented ) fibre @xmath224 generates @xmath225 .
then there is a free and transitive action of @xmath226 on @xmath227 under pullback to @xmath228 and addition in @xmath229 . to pass from the first picture of smooth euler structures as @xmath213 to this second picture , notice that the unit vector field @xmath36 for a non - singular @xmath29 is a map @xmath230 and so a @xmath4-cycle in @xmath228 .
orient @xmath228 by requiring the intersection number of every @xmath231 with this cycle to be @xmath10 .
then the @xmath4-cycle @xmath36 poincar dualises to a @xmath86-cocycle which we write as @xmath232 , and the latter represents an euler structure in @xmath233 in the second picture . to emphasize the cohomology definition of euler structure
, we will sometimes use the term _ co - euler structure _ , to distinguish it from a homological definition to be given in section [ sec : eulerchain ] . as with @xmath213 , there is an identification of the affine space @xmath234 with @xmath219 upon fixing a reference non - singular @xmath218 .
writing @xmath235 for the projection , we note that @xmath236 in the gysin sequence @xmath237 with @xmath238 the pushforward , or integration over the @xmath165 fibers .
the subset @xmath233 comprises @xmath239 $ ] and all its translates by @xmath240 , so @xmath241 gives a bijection from @xmath242 taking @xmath239 $ ] to the identity element .
there is a dual picture of euler structures involving vector fields with finite singularities , which is closely related to the intuition behind fermi arcs .
we sketch the construction here for the case of non - degenerate zeroes , see @xcite for the general case . for a vector field @xmath29 on @xmath3 with finite singularity set @xmath243 ,
let @xmath244 denote the singular 0-chain @xmath245 which encodes the local charge information .
we can also think of @xmath244 as an element of @xmath246 whose weights sum to zero .
let @xmath3 be a compact @xmath4-manifold with @xmath208 and @xmath209 .
euler chain _ for a 0-chain @xmath247 in @xmath3 whose weights sum to zero , is a 1-chain @xmath248 such that @xmath249 is @xmath247 .
an euler chain @xmath250 defines a relative homology class in @xmath251 where @xmath47 is the set of points that @xmath247 is defined on is allowed to be zero on points in @xmath47 . ] .
let @xmath252 denote the subset of relative homology classes of 1-chains in @xmath3 whose boundary is @xmath247 . from the exact sequence @xmath253 we see that @xmath254 is a coset of @xmath75 in @xmath251 and so an affine space for @xmath75 .
in particular , the difference of @xmath255,[l']\in{\mathfrak{eul}}(t,{\mathcal{w}})$ ] is an element of @xmath75 .
note that any 0-chain @xmath247 with total weight zero can be realised as the 0-chain of charges @xmath244 for some singular vector field @xmath29 , by the ( converse of the ) poincar hopf theorem .
euler chains can be thought of as extra global data encoding how the local charge configuration for a singular vector field is `` connected '' . as explained in section [ sec : semimetaleulerchain ] , euler chains are in a precise sense poincar dual to cohomological semimetal invariants .
* non - degenerate homotopies .
* let @xmath210 be two smooth vector fields , then there is a canonical identification of @xmath256 and @xmath257 as affine spaces , using a notion of _ non - degenerate homotopy _ defined as follows .
let @xmath258 be the pullback of the tangent bundle @xmath222 under the projection @xmath259\times t\rightarrow t$ ] .
non - degenerate homotopy _ between @xmath29 and @xmath260
is a section @xmath261 of @xmath258 transverse to the zero section , which restricts to @xmath29 at @xmath262 and @xmath260 at @xmath263 .
such a homotopy exists by perturbing , for instance , a linear homotopy , and allows for the movement @xmath264 of the local charges @xmath265 including the creation and annihilation of pairs of zeros with equal and opposite charges ( fig .
[ fig : euler - chain ] ) . for generic @xmath266 , the intermediate vector field @xmath267
intersects the zero vector field transversally ( e.g. transversality of @xmath267 fails when a pair of weyl points are created or annihilated ) .
the zero set @xmath268 of @xmath261 is a canonically oriented 1-submanifold with boundary @xmath269 , where the superscript indicates whether the 0-chain lies on @xmath270 or on @xmath271 .
let @xmath255\in { \mathfrak{eul}}(t,{\mathcal{w}}_{{\bm{{h}}}})$ ] , then @xmath272 is a 1-cycle on @xmath273\times
t$ ] relative to @xmath271 , since @xmath274 . because @xmath275\times t , \{1\}\times t)=0 $ ] , there is a 2-chain @xmath276 on @xmath273\times t$ ] whose boundary is @xmath272 relative to @xmath271 .
taking @xmath277 , we see that @xmath278 is a 1-chain on @xmath271 with @xmath279 , which defines a class @xmath280\in { \mathfrak{eul}}(t,{\mathcal{w}}_{{\bm{{h}}}'})$ ] .
intuitively , we can think a homotopy @xmath261 as providing the data of how an initial @xmath250 is to be `` carried along @xmath268 '' onto a final @xmath278 .
$ ] goes horizontally , while all coordinates for the manifold @xmath3 are suppressed except one cyclic @xmath281-direction .
the height function at each @xmath266 ( yellow curve ) indicates the length of a tangent vector on @xmath282 in the aforementioned direction .
black arrows represent one component of the initial ( left ) vector field @xmath29 and final ( right ) one @xmath260 , which are taken to be equal in this example .
the zero - section of the tangent bundle of @xmath273\times t$ ] is the blue shaded area .
a non - degenerate homotopy @xmath261 is shown , with the intermediate vector fields @xmath267 ( indicated by yellow dashed curves ) generically having isolated zeroes with charges @xmath283 .
the orange oriented submanifold @xmath268 comprises the singularities of @xmath261 .
the euler chains for @xmath267 ( red lines joining @xmath284 to @xmath121 ) are `` carried along '' @xmath268 , disappear when the weyl points coalesce , then reappear when a pair of weyl points are created .
the homology class of the euler chain for @xmath29 is unchanged by this homotopy , consistent with what happens if the obvious constant homotopy is chosen instead.,scaledwidth=70.0% ] which translates the initial vector field @xmath29 by half a @xmath281-cycle , showing how the euler chain is carried along the zero set ( orange ) of @xmath261 .
rotation by a full cycle will return the euler chain to its original position , with the same homology class.,scaledwidth=70.0% ] which changes the homology class of an euler chain .
note that there is no open ball of @xmath3 within which the singular set of @xmath267 stays for all @xmath285 $ ] .
rather , such an `` enveloping open set '' contains an @xmath281 cycle.,scaledwidth=70.0% ] one shows @xcite that the above assignment @xmath255\mapsto[l']$ ] is independent of the choices of homotopy and of @xmath102 , yielding an affine map @xmath286 .
furthermore , one has the properties @xmath287 , and @xmath288 , so each @xmath256 is canonically isomorphic to a single affine space ( which for oriented @xmath3 can be taken to be the space of smooth euler structures @xmath213 in definition [ defn : smootheuler ] ) over @xmath75 @xcite .
in particular , when a non - singular @xmath218 in @xmath213 has been chosen , a possibly singular @xmath29 has an associated euler chain @xmath250 .
if there is a @xmath261 such that the local charges @xmath289 are simply moved around smoothly and disjointly without creation / annihilation , i.e. there is a smooth 1-parameter family of diffeomorphisms @xmath290 of @xmath3 taking @xmath291 to @xmath289 , @xmath285 $ ] , then the zero set @xmath292 of @xmath261 is simply a set of disjoint lines representing the trajectories of the local charges .
an initial euler chain @xmath250 is `` carried along @xmath292 '' to the final euler chain @xmath278 ( fig . [
fig : euler - chain2 ] ) .
suppose @xmath210 have the same 0-chains of local charges , @xmath293
. then the endomorphism @xmath294 can be thought of as the homological change in the euler chains , @xmath255\mapsto [ l']$ ] , under a non - degenerate homotopy @xmath261 taking @xmath29 to @xmath260 .
specifically , @xmath295 can be pushed forward under the projection @xmath296 to a 1-chain on @xmath3 , also written @xmath297 , which is actually a cycle .
then @xmath280-[l]\in h_1(t,{{\mathbb z}})$ ] is a `` homological difference '' between @xmath29 and @xmath260 ; this difference was called a chern
simons class @xmath298 in @xcite . for @xmath299 having different local charges ,
@xmath298 is defined similarly , as a 1-chain modulo boundaries .
suppose there is a non - degenerate homotopy @xmath261 which stays non - singular outside some open ball in @xmath3 ( necessarily containing @xmath300 )
. then an initial euler chain @xmath250 does not pick up any nonzero element of @xmath75 , so @xmath280=[l]$ ] ( fig . [ fig : euler - chain ] , [ fig : euler - chain3 ] ) .
the upshot of introducing euler chains is that we can use them to represent globally the topological data of a semimetal band structure , whenever vector fields are used to parameterise semimetal hamiltonians .
this is certainly the case for @xmath27 hamiltonians in 3d , and more generally for dirac - type hamiltonians in higher dimensions .
the singular points @xmath243 of @xmath29 are the band crossings of @xmath34 , and the local charge information is contained in the 0-chain @xmath244 .
each @xmath29 can be associated with an euler chain class in @xmath301 by the prescription of the previous subsection , and vector fields with the same local charges but homologically different euler chains can not he homotoped whilst keeping the singular set within an open ball ( in particular , while keeping the singular set fixed ) .
when we generalise to kervaire chains for bilinear hamiltonians later , it will be convenient to use an alternative prescription to associate an euler ( or kervaire ) chain to a semimetal hamiltonian .
this alternative prescription uses a definition of euler structures which follows that given in section [ sec : coeuler ] , but which is defined in terms of singular vector fields .
let @xmath228 be the sphere bundle of the tangent bundle of @xmath3 , and @xmath302 be the restrictions of @xmath228 to the subspaces of @xmath3 appearing in our mv - sequence , and let @xmath296 denote the various bundle projections .
all these sphere bundles have vanishing euler class .
there is also an mv - sequence for the cover @xmath303 of @xmath228 , which has intersection @xmath304 , and we write @xmath305 for its connecting maps .
let @xmath218 be a non - singular vector field on @xmath3 , then @xmath236 .
consider the semimetal mv - sequence and its dual homology sequence , where we have suppressed the integer coefficients , @xmath306&h^{d-1}(t)\ar[d]_{\text{pd}}\ar[rr]^{\iota^ * } & & h^{d-1}(t\setminus w)\ar[d]_{\text{pd}}\ar[rr]^{\beta } & & h^{d-1}(s_w)\ar[d]_{\text{pd}}\ar[rr]^{\sigma } & & h^d(t)\ar[d]_{\text{pd}}\ar[r ] & 0\\ 0\ar[r ] & h_1(t)\ar[rr ] & & h_1(t , w)\ar[rr]^{\partial } & & h_0(w ) \ar[rr]^{\sigma } & & h_0(t)\ar[r ] & 0 } .\ ] ] the local charges can be thought of as a 0-chain @xmath247 on @xmath82 , or dually as an element @xmath307 in @xmath308 ; in either case they are required to have total charge zero . for a 0-chain @xmath247 of local charges on @xmath3 , with total charge 0 and defined on @xmath82 , we write @xmath309 for the image of @xmath254 in @xmath98 under poincar duality ; equivalently this is the inverse image of @xmath307 under @xmath162 .
we combine the cohomology mayer vietoris ( horizontal ) and gysin sequences ( vertical ) for @xmath3 and @xmath228 , @xmath310 & 0\ar[d ] & 0\ar[d ] & \\ 0\ar[r ] & h^{d-1}(t)\ar[d]_{p^*}\ar[r]^{\iota^ * } & h^{d-1}(t\setminus w)\ar[d]_{p^*}\ar[r]^{\beta } & h^{d-1}(s_w)\ar[d]_{p^*}\ar[r]^{\quad\sigma } & \cdots\\ 0\ar[r ] & h^{d-1}({\mathcal{s } } ) \ar[r]\ar[d]_{p _ * } & h^{d-1}({\mathcal{s}}|_{t\setminus w})\oplus h^{d-1}({\mathcal{s}}|_{d_w } ) \ar[r]\ar[d]_{p _ * } & h^{d-1}({\mathcal{s}}|_{s_w } ) \ar[r]^{\quad\widetilde{\sigma}}\ar[d]_{p _ * } & \cdots\\ 0\ar[r ] & h^0(t)\ar[r]\ar[d ] & h^0(t\setminus w)\oplus h^0(d_w)\ar[r]\ar[d ] & h^0(s_w ) \ar[r]\ar[d ] & \cdots\\ & 0 & 0 & 0 & } , \label{mvgysin}\ ] ] where we have augmented the middle gysin sequence by a @xmath311 piece needed to make the bottom two horizontal mv sequences exact .
the horizontal maps are restriction maps ( or differences of ) and so it is easy to see that the diagram commutes .
the middle mv - sequence is , by the knneth theorem , @xmath312 and so the restriction map for the @xmath313 factor has a `` local charge '' component @xmath314 and a second component which is less important for us .
furthermore the restriction map for the @xmath315 factor only lands in the @xmath316 factor .
the relevant part of is then the commuting diagram of short exact sequences @xmath310 & 0\ar[d ] & 0\ar[d ] & \\ 0\ar[r ] & h^{d-1}(t)\ar[d]_{p^*}\ar[r]^{\iota^*\;\ ; } & h^{d-1}(t\setminus w)\ar[d]_{p^*}\ar[r]^{\beta\;\ ; } & \text{ker}_\sigma(h^{d-1}(s_w))\ar[d]_{p^*}\ar[r ] & 0\\ 0\ar[r ] & h^{d-1}({\mathcal{s } } ) \ar[r]^{\iota^*\;\;}\ar[d]_{p_*}\ar[u]_{\hat{{\bm{{h}}}}_{\text{ref}}^ * } & h^{d-1}({\mathcal{s}}|_{t\setminus w } ) \ar[r]^{\tilde{\beta}\;\;}\ar[d]_{p_*}\ar[u]_{\hat{{\bm{{h}}}}_{\text{ref}}^ * } & \text{ker}_{\widetilde{\sigma}_1}(h^{d-1}(s_w ) ) \ar[r]\ar[d ] & 0\\ 0\ar[r ] & h^0(t)\ar[r]^{\iota^*\;\;}\ar[d ] & h^0(t\setminus w)\ar[r]\ar[d ] & 0 & \\ & 0 & 0 & & } .\label{mvgysin2}\ ] ] let @xmath247 be a 0-chain of local charges defined on the finite subset @xmath82 with total weight zero , and @xmath307 its poincar dual .
we define @xmath317 to be the set of @xmath318 such that @xmath319 and @xmath320 .
there is an identification of @xmath317 and @xmath309 as affine spaces for @xmath219 . by chasing through the diagram ,
we see that an element @xmath321 is of the form @xmath322 for some @xmath323 with @xmath324 and some @xmath325 with @xmath326 , i.e. @xmath327 and @xmath328 .
clearly , the action of @xmath329 on @xmath317 by pullback @xmath330 and addition is free , and is also transitive since the difference of @xmath331 is the sum of @xmath332 and @xmath333 , which is something in @xmath334 . given a reference non - singular @xmath218 , it is easy to check that the surjective map @xmath335 respects the local charges @xmath336 , and restricts to a @xmath329-equivariant bijection between @xmath317 and @xmath309 .
the following diagram summarises the various affine spaces for @xmath337 : @xmath338 \ar[d ] & & { \mathfrak{eul}}(t,{\mathcal{w}})\ar[ll ] \ar[d ] & \subset \quad\quad\;\ ; h_1(t , w ) \\ & h^{d-1}(t)\ar[rr]\ar[d ] \ar[u ] & & { \mathfrak{coeul}}(t,{\mathcal{w}})\ar[ll ] \ar[d]\ar[u ] & \subset \quad h^{d-1}(t\setminus w ) \\ h^{d-1}({\mathcal{s}})\quad\supset & \widetilde{{\mathfrak{coeul}}}(t)\ar[rr ] \ar[u ] & & \widetilde{{\mathfrak{coeul}}}(t,{\mathcal{w}})\ar[ll ] \ar[u ] & \subset \quad h^{d-1}({\mathcal{s}}|_{t\setminus w } ) } .\nonumber\ ] ] if @xmath247 is the zero 0-chain , then @xmath339 coincides under @xmath163 with @xmath234 .
let @xmath29 be a vector field with local charges @xmath247 , so @xmath36 defines a @xmath4-cycle on @xmath340 relative to the boundary @xmath304 .
its poincar dual is a @xmath86-cocyle on @xmath340 representing an element of @xmath317 .
pulling back under @xmath220 gives an element @xmath341 $ ] in @xmath309 , whose poincar dual is an euler chain @xmath342 $ ] in @xmath254 . in this way , a semimetal hamiltonian specified by a vector field with finite singularities can be represented by an euler chain for @xmath247 .
[ defn : semimetaltopinv ] suppose @xmath34 is a dirac - type hamiltonian specified by a smooth vector field @xmath29 on @xmath3 with finite singularities @xmath47 .
it has a cohomological topological invariant @xmath341\in h^{d-1}(t\setminus w,{{\mathbb z}})$ ] as in the above paragraph , and its _ euler chain representation _ is the poincar dual @xmath342\in h_1(t , w,{{\mathbb z}})$ ] .
the euler chain representation of a semimetal is extremely useful because it invariantly characterises and generalises the idea of `` jumps in chern numbers '' as a weyl point is traversed , even for complicated local charge configurations .
more precisely , the natural pairing @xmath343 is well - defined for any semimetal with invariant @xmath344 $ ] , and any homology @xmath86-cycle @xmath345 $ ] avoiding @xmath47 .
for example , we can take @xmath345 $ ] to be the fundamental classes of @xmath86-spheres surrounding weyl points , or ( when @xmath71 ) of @xmath86 subtori at a fixed coordinate .
the pairing poincar dualises to the _ _ intersection pairing _ _ , as in 2.6.17 of @xcite . here
@xmath346 is borel moore homology , mentioned in section [ sec : dualmv ] , which is isomorphic to @xmath193 and poincar dual to @xmath98 . ] between @xmath342 $ ] and @xmath345 $ ] , i.e. the ( signed ) intersection between the corresponding euler chain @xmath347 and the @xmath86-cycle @xmath276 in question . as a concrete example in @xmath45 , for those @xmath345 $ ] which are represented by @xmath348-submanifolds @xmath276
, the intersection number computes the integral over @xmath276 of the berry curvature 2-form ( which is defined on @xmath66 ) this yields the familiar first chern numbers ( `` weak '' chern invariants ) and their discontinuities across the weyl points ( fig .
[ fig : intersection]-[fig : intersection2 ] ) . for @xmath71 ,
a projection @xmath349 induces a homology projection @xmath350 , which we can think of as being poincar dual to the bulk - boundary correspondence @xcite . under @xmath351 ,
an euler chain becomes a surface _ fermi arc _ a 1-chain on @xmath207 whose boundary is the projected 0-chain of local charges .
figure [ fig : bulk2boundary ] provides an intuitive picture in @xmath45 explaining how surface fermi arcs are determined to @xmath284 , so the fermi arcs there are _ oppositely oriented _ to those in this paper . ] from bulk euler chains by @xmath351 .
the euler chain description is coordinate - free , and determines the fermi arc connectivity for the projection to _ any _ surface .
are euler chains encoding , in particular , how the first chern numbers @xmath352 vary with @xmath353 . only the values of @xmath133
are indicated .
gapless surface states appear at points in the surface brillouin torus @xmath21 where a first chern number is non - zero , forming a fermi arc connecting the projected weyl points .
the fermi arc is determined from the euler chain homologically by projection @xmath351,title="fig : " ] are euler chains encoding , in particular , how the first chern numbers @xmath352 vary with @xmath353 .
only the values of @xmath133 are indicated .
gapless surface states appear at points in the surface brillouin torus @xmath21 where a first chern number is non - zero , forming a fermi arc connecting the projected weyl points .
the fermi arc is determined from the euler chain homologically by projection @xmath351,title="fig : " ] if we are able to access the complete specification of a semimetal hamiltonian as an operator , then its surface fermi arc is uniquely determined in principle . in practice , we have access to the fermi arcs , and different fermi arcs ( as 1-chains ) come from different hamiltonians ( as operators ) .
the passage from hamiltonian to fermi arc can be modelled , for example , by transfer matrices @xcite . at the _ topological _ level ,
a coarser but more appropriate question is how fermi arcs may be used to distinguish _ topologically distinct _
semimetal hamiltonians . for this latter question ,
we note that _ any _ euler chain representative for a semimetal hamiltonian @xmath34 maps under @xmath351 to a 1-chain which is topologically equivalent to the actual fermi arc for @xmath34 .
thus fermi arcs which are topologically distinct must come from topologically distinct hamiltonians ; our slight abuse of language in calling the 1-chain a `` fermi arc '' does not matter at the level of topological invariants .
certain scenarios involving tuning semimetal hamiltonians and `` rewiring '' their surface fermi arcs were considered in @xcite .
these essentially involve homotopies @xmath261 which fix the local charges @xmath293 ( at least up to a diffeomorphism moving the weyl points ) , and so induce the identity map on the class of euler chains in @xmath354 . under @xmath351 , the surface fermi arcs for
@xmath29 and @xmath260 may become `` rewired '' , but in such a way that their class in @xmath205 remains unchanged .
rewirings which involve a change in @xmath205 necessarily require the prefiguring euler chain in @xmath254 to change homology class , and this can not be achieved by homotopies @xmath261 which fix the local charges @xmath247 .
figures [ fig : intersection]-[fig : intersection2 ] illustrate some examples of rewirings . with one coordinate suppressed ; we can also interpret it as the projected surface brillouin torus @xmath21 .
two homologous euler chains ( the left and right pairs of red directed lines ) having the same local charge configuration are shown , and we can also interpret them as the corresponding surface fermi arcs .
first chern numbers for 2d subtori in the @xmath355 and @xmath356 directions are shown , and may be computed by counting the signed intersection between the euler chain and the subtorus .
the `` straightness '' of the euler chain does not matter , and chern numbers can also be calculated for 2-cycles other than the standard ones with a fixed coordinate.,scaledwidth=90.0% ] representing more semimetal invariants in @xmath357.,scaledwidth=90.0% ] [ remark : higherjumps ] in @xmath46 , the intuition afforded by figures [ fig : chern - jump ] and [ fig : bulk2boundary]-[fig : intersection2 ] carries over in much the same way .
for example , in @xmath67 , the blue surfaces represent hyperplane slices ( 3-tori ) on which the semimetal gerbe of section [ sec : semimetalgerbe ] restricts and has a curvature 3-form .
there are four independent slice directions . as a slice
is moved transversely across a semimetal band crossing with local charge @xmath358 , the dixmier
douady invariant of the gerbe on the slice jumps by @xmath358 . in @xmath68 ,
the slices are 4-tori and and it is the second chern number of the @xmath359-invariant semimetal ( section [ sec:5dsemimetal ] ) which jumps by @xmath358 . [
remark : nonzeroeuler ] it is also possible to define euler structures for @xmath3 even if @xmath360 by introducing a basepoint @xmath361 which `` contains @xmath70 '' , see @xcite .
an euler chain for a 0-chain @xmath247 is then a 1-chain @xmath248 such that @xmath362 .
every vector field @xmath29 on @xmath3 with isolated zeroes @xmath47 admits an euler chain for its 0-chain @xmath244 of local charges in this sense : take @xmath363 , then @xmath364 by the poincar
hopf theorem .
euler structures for @xmath3 with basepoint @xmath365 are then defined as classes of pairs @xmath366 which are considered equivalent if @xmath367 up to a boundary .
there is an action of @xmath75 on these classes by addition to @xmath250 , which is furthermore free and transitive .
so far , we have implicitly assumed that the hamiltonians act on a _ trivial _ bloch bundle @xmath368 and that a trivialisation ( choice of basis ) has been given , so that the dirac - type hamiltonians of the form are well - defined globally .
although this default setting suffices to illustrate many interesting features of semimetals , it is worth mentioning that non - trivial @xmath142 can arise physically and can still be handled mathematically . in general
one has , after a bloch floquet transform , a _ hilbert bundle _
@xmath369 in which the fibre @xmath370 over @xmath56 comprises the quasi - periodic bloch wavefunctions with quasimomentum @xmath371 @xcite .
one typically restricts attention to a low - energy subbundle @xmath142 defined by a finite number of energy bands separated from the rest by spectral gaps .
subsequently , a _ bloch bundle _
@xmath142 will refer to such a truncated finite - rank hermitian vector bundle over a manifold @xmath3 of ( quasi)-momenta , on which the _ bloch hamiltonians _
act fiberwise . in the theory of topological band insulators , it is crucial that @xmath142 or its subbundles can be non - trivializable , e.g. 2d chern insulators are essentially specified by the first chern class of a non - trivializable valence line bundle .
expressions such as should be understood as _
local _ expressions , which continue to make sense on a bloch bundle @xmath142 as long as @xmath142 has the structure of a clifford module bundle ( or spinor bundle ) so that the @xmath79 operators @xmath372 are well - defined globally .
in particular , the analysis of the spectrum of @xmath373 remains the same for _ any _ @xmath142 namely , the @xmath55 eigenspaces degenerate ( i.e. there is a band crossing ) precisely at the zeroes of @xmath29 . on the other hand , the topology of the valence subbundle over @xmath66 , which is defined by the fermi projection @xmath374 , does depend on @xmath142 through the operators @xmath372 .
when @xmath375 , the most general @xmath17-band hamiltonian is _ not _ of dirac - type , and well - defined topological invariants should respect the @xmath187 gauge invariance of the rank-@xmath17 bloch bundle @xmath142 .
nevertheless , dirac - type hamiltonians are generic when certain time - reversal / particle - hole symmetries are imposed .
such additional symmetries restrict the gauge group , and the topological invariants which we will define for insulators / semimetals described by dirac - type hamiltonians , are gauge invariant in this restricted sense .
some examples of such gauge restrictions were studied in @xcite , and we also analyse them in section [ sec : higherdimension ] . the properties of dirac - type hamiltonians which are relevant for defining insulator / semimetal invariants can be abstracted as follows .
an abstract dirac - type hamiltonian is a section of the vector part of some clifford algebra bundle over a momentum space manifold @xmath3 , i.e. the quantization of a vector field @xmath29 specifying the abstract hamiltonian as in definition [ defn : abstracthamiltonian ] .
pointwise in @xmath3 , the spectrum of such a hamiltonian ( as a clifford algebra element ) can be found quite easily ; it depends only on the vector field and has the crucial feature that it degenerates exactly at the zero set @xmath47 of the vector field .
when such hamiltonians are represented on some clifford module bundle @xmath142 ( physically the bloch bundle ) , they become concrete families of hermitian operators with fermi projections etc .
the local topological invariant at a zero @xmath58 is intrinsic to @xmath29 rather than the particular choice of bloch bundle @xmath142 , and exists even when @xmath142 is not a trivial bundle .
furthermore , the symmetries and gauge structure for the hamiltonians become more transparent at this abstract level , and this is especially so in higher dimensions .
* hamiltonians as quantized vector fields . * for @xmath77 ,
let @xmath376 be an oriented rank-@xmath4 real vector bundle over a compact oriented manifold @xmath3 with fibre metric @xmath377 , along with a spin@xmath5 structure ( e.g. take @xmath3 a spin@xmath5 manifold and @xmath376 its tangent bundle , then there is a @xmath154 worth of spin@xmath5 structures to choose from ) .
the clifford algebra bundle @xmath378 is a `` quantization '' of the exterior algebra bundle @xmath379 to allow for multiplication of vectors , both bundles having structure group @xmath380 @xcite . with the spin@xmath5 structure
, we can construct the ( irreducible ) spinor bundle @xmath142 ( the `` bloch bundle '' ) of complex rank @xmath381 , with @xmath382-invariant hermitian inner product , and clifford multiplication gives a _ real _ bundle homomorphism @xmath383 .
thus there is a map @xmath384 in particular , @xmath385 is identified with a subbundle of @xmath378 , and the map @xmath386 takes a section @xmath387 to a concrete dirac - type hamiltonian .
more explicitly , an orthonormal frame @xmath388 for @xmath376 becomes @xmath389 in @xmath378 and satisfies the clifford relations @xmath390 .
the @xmath391 are represented on @xmath142 as traceless self - adjoint endomorphisms @xmath392 which inherit the familiar relations @xmath393 .
thus , a section @xmath394 defines an _ abstract dirac - type hamiltonian _ @xmath395 , which is then represented by a _
concrete dirac - type hamiltonian _
@xmath396 acting on @xmath142 .
the concrete _ dirac - type bloch hamiltonian _ above a point @xmath56 is the operator @xmath397 acting on the fiber @xmath398 . in the clifford algebra bundle , it is already ( @xmath380-invariantly ) true that @xmath399 .
if follows that the spectrum of a bloch hamiltonian is @xmath400 with each eigenvalue @xmath401-fold degenerate , independently of @xmath142 , and this is @xmath382-invariant . precisely at the zeroes of @xmath29 ,
the two @xmath401-fold degenerate energy bands cross , and such a crossing is precisely protected by the topological index of @xmath29 there . more generally , we only require @xmath142 to be a clifford module bundle for @xmath378 , for which @xmath395 acts self - adjointly .
such clifford module bundles are in fact twisted versions of spinor bundles , obtained by tensoring with some vector bundle @xmath402 , so we shall mainly consider only irreducible spinor bundles . in section [ sec : torsionsemimetal ] we will also consider @xmath403 acting skew - adjointly on @xmath142 , giving rise to `` @xmath51-bilinear hamiltonians '' .
physically , one might start off with a given bloch bundle @xmath142 , which is a hermitian @xmath187 vector bundle over @xmath3 obtained by fourier / bloch floquet transform .
in order to have a notion of clifford multiplication and dirac - type hamiltonians , we need some assumptions on @xmath142 allowing it to have the structure of a ( irreducible ) clifford module bundle for some @xmath376 . in the first place , @xmath142 should have the correct ( complex ) rank . in most model hamiltonians considered in the physics literature , @xmath71 and the bloch bundle @xmath142
is assumed to be trivial . in this case , we take the trivial bundle @xmath72 , identified with the tangent bundle of @xmath0 , and @xmath387 is a tangent vector field . then the corresponding dirac - type bloch hamiltonian @xmath404 has gamma matrices which can be taken to be constant over @xmath0 .
formally , a family of @xmath79 bloch hamiltonians over @xmath3 can also be regarded as a family of hamiltonians on an @xmath17-level quantum mechanical system , with adiabatic phases / holonomies , etc .
there is , however , a fundamental difference between @xmath3 as an adiabatic parameter space and as a quasi - momentum space , when antiunitary symmetries are introduced .
the brillouin torus @xmath0 is topologically the space of unitary characters for the translation group @xmath406 of a lattice , so the complex conjugation involved in an antiunitary symmetry operator induces an involution @xmath407 ( a @xmath2-action ) on @xmath0 .
explicitly , when we parameterise @xmath0 by angles @xmath408 , the character @xmath409 labelled by @xmath371 conjugates to the character labelled by @xmath410 . in the context of bloch hamiltonians ,
a fermionic time - reversal symmetry @xmath359 is a lift of @xmath407 to a map on the bloch bundle @xmath142 which is antiunitary on fibres , and satisfies @xmath411 .
compatibility of @xmath359 with @xmath412 is then the condition @xmath413 .
such bundles equipped with a `` quaternionic '' structure @xmath359 were studied in @xcite , and have more recently been investigated in the context of topological insulators @xcite . for bosonic time - reversal , @xmath359 squares to the identity instead , and is sometimes called a `` real structure '' . similarly , an antiunitary particle - hole symmetry @xmath405 _ anticommutes _ with the bloch hamiltonians in a way that respects @xmath407 , namely , @xmath414 .
on the other hand , for a family of hamiltonians parametrised by @xmath3 , we simply have @xmath415 and @xmath416 . in either case , the squares @xmath417 may be @xmath418 , and when both @xmath359 and @xmath405 are present , they may be assumed to commute @xcite .
the same involution @xmath407 is induced on @xmath0 by a spatial inversion symmetry @xmath419 , since a translation by @xmath420 becomes a translation by @xmath421 .
thus we can consider a time - reversal symmetry which concurrently effects spatial inversion and @xmath422 to _ separately _ be symmetries , and indeed , the latter situation requires additional data specifying whether @xmath423 commute or anticommute @xcite , leading in each case to different wigner classes . ] ; we write @xmath424 for such a symmetry operator
. then we have @xmath425 where the effect of @xmath407 cancels ; similarly for @xmath426 . thus the operator @xmath424 is an ordinary quaternionic structure , and the bloch bundle @xmath142 has invariants as a _ symplectic bundle_. * convention . * for this paper , we will simply write @xmath427 with the understanding that @xmath428 is meant whenever @xmath3 is a brillouin torus . thus we always have @xmath415 and @xmath429 .
let us explain the above abstractions in more detail for two - band hamiltonians in @xmath45 , in which @xmath430 .
this is the setting for the basic weyl semimetal . since @xmath431
, any rank-2 hermitian vector bundle @xmath142 can arise as a spinor bundle : take the traceless hermitian endomorphisms @xmath432 of @xmath142 with the hilbert
schmidt real inner product @xmath433 .
then @xmath432 becomes a real orientable rank-3 bundle with structure group @xmath434 liftable to @xmath435 .
an orthonormal frame for @xmath432 gives a set of pauli operators @xmath436 , which is positively oriented if @xmath437 .
such a positively - oriented frame has the form of the standard pauli matrices in some local trivialization of @xmath142 .
the bloch hamiltonian is given locally by @xmath129 .
the @xmath438 has a clifford algebra bundle and a choice of spin@xmath439 structure yields a spinor bundle .
@xmath142 is some twisted version of this spinor bundle , and the twisting can be thought of as changing the spin@xmath439 structure under the canonical action of @xmath154 @xcite .
thus @xmath436 may be thought of as @xmath440 for an orthonormal frame @xmath441 in @xmath442 . from this point of view , a section @xmath29 is quantized to an _ abstract hamiltonian _ @xmath443 , which acts concretely as @xmath444 on @xmath142 .
note that , in particular , the spectrum @xmath445 is invariant under @xmath446 gauge transformations of @xmath142 .
[ rem : gauge3d ] a gauge transformation of the bloch bundle @xmath142 implemented by local @xmath27 unitaries @xmath447 turns @xmath448 into the conjugate @xmath449 .
the result of this conjugation is known to be equal to @xmath450 where @xmath451 is rotated from @xmath452 by the @xmath453 element @xmath454 in @xmath455 similarly , @xmath456 in case the structure group of @xmath142 can be reduced to @xmath457 .
two band - hamiltonians can more generally include a trace term , i.e. @xmath458 , where @xmath459 is not constant and so not eliminable by shifting the overall energy level . indeed , hamiltonians with such @xmath459 terms ( and constant @xmath460 ) can give rise to new phenomena such as `` type - ii weyl semimetals '' @xcite .
it is possible to distinguish the two situations , i.e. eliminate the @xmath459 term , by imposing a natural symmetry constraint .
* quaternionic structure and @xmath405-symmetry .
* let us recall the construction of the spinor representation @xmath461 of @xmath462 .
let @xmath463 be orthonormal vectors in @xmath464 , @xmath465 , and @xmath466 .
the even part @xmath467 is isomorphic to @xmath468 , and for the complexified algebras , @xmath469 .
the spinor representation on @xmath470 is the ( unique complex ) irreducible representation of @xmath471 , and is extended to @xmath472 by setting a chirality condition @xmath473 .
@xmath474 lies in @xmath475 as the unit quaternions , and @xmath461 is an irreducible representation of @xmath476 .
this is just the fundamental representation of @xmath477 which is well - known to be quaternionic .
the quantized bivectors @xmath478 generate @xmath479 . on @xmath461 , they are represented by skew - hermitian operators @xmath480 ( which generate a quaternion algebra ) , and with the chirality element @xmath481 the identity on @xmath461 , we recover the pauli matrices @xmath482 .
since @xmath461 is a quaternionic representation for @xmath476 , there is an antiunitary operator @xmath483 on @xmath461 commuting with @xmath484 such that @xmath485 , from which we deduce that @xmath483 _ anticommutes _ with @xmath486 .
for instance , we can take @xmath487 where @xmath488 is complex conjugation in the basis where @xmath489 are the standard pauli matrices ( this operator is often taken to be the fermionic time - reversal operator on spin-@xmath490 systems ) . if @xmath72 has a spin ( not just spin@xmath5 ) structure , there is also a notion of spinor bundles @xmath142 associated to the spin structure @xcite , and these have structure group @xmath491 . globalising the above constructions , these spinor bundles come with a quaternionic structure @xmath405 anticommuting with @xmath492 .
* reduction to @xmath493 . *
suppose we have a reduction of the structure group of the bloch bundle from @xmath494 to @xmath495 .
this is possible exactly when @xmath496 is trivial , i.e. @xmath497 ( this is a general result about reducing from @xmath187 to @xmath498 ) . alternatively , since @xmath499 , the reduction from @xmath494 to @xmath457 requires a quaternionic structure on @xmath142 , which provides an isomorphism between the line bundles @xmath500 and @xmath501 for an orthogonal splitting @xmath502
. then @xmath503 . in this case
, we can identify @xmath432 with the tangent bundle @xmath504 , as ( trivial ) @xmath453 bundles .
therefore , in the two - band case in 3d , the imposition of a @xmath405-symmetry constraint is one possible way to justify ( 1 ) the restriction from arbitrary @xmath27 bloch hamiltonians to the _ traceless _ ones , and ( 2 ) the triviality of the bloch bundle @xmath142 .
we now generalize the two - band crossings of the previous section to four - band crossings .
these should involve crossings of _ pairs _ of bands , in analogy to the dispersion of massless dirac spinors .
again , the adjective `` dirac '' is loaded and possibly misleading , and our constructions below are more analogous to weyl fermions in 1 + 5 d ( which is still misleading ) . for this subsection
, we take @xmath3 to be a compact oriented 5-manifold , and @xmath505 a trivial hermitian vector bundle . the bundle @xmath432 of traceless hermitian endomorphisms of @xmath142 is now a trivial rank-15 real vector bundle whose sections are bloch hamiltonians @xmath506 with inner product @xmath507 . for notational ease
, we will sometimes suppress the dependence on @xmath371 when dealing with a single operator .
physicists often restrict attention to `` dirac - type '' hamiltonians @xmath404 where @xmath508 is an orthonormal set of @xmath509 hermitian operators on @xmath461 satisfying the clifford algebra relation @xmath52 , and @xmath29 is a 5-component vector field .
the pragmatic justification is ease of manipulation , since the spectrum is again easily found to be @xmath510 , with doubly - degenerate eigenvalues that become four - fold degenerate precisely at the zeroes of @xmath29 ( fig .
[ fig : symmetric spectrum ] ) .
however , we should note that arbitrary gauge transformations will not preserve this particular form of @xmath509 hermitian matrices .
it is therefore useful to find a symmetry which reduces the allowed gauge transformations so that the dirac - form of @xmath34 is automatically preserved .
we do this by invoking a @xmath359-symmetry constraint ( a quaternionic structure ) on the hamiltonians , which will pick out the dirac - type hamiltonians as a rank-5 subbundle of @xmath432 .
in fact , the observation that the proper gauge groups and berry connections / phases for time - reversal invariant hamiltonians are _ quaternionic _ , had already been made in @xcite , and in the @xmath509 case such hamiltonians can be constructed by considering @xmath511 as a spin-@xmath512 representation ( so - called `` quadrupole hamiltonians '' ) .
we give a more direct account , emphasizing the clifford algebra and spin groups responsible for this structure .
consider first a fibre @xmath513 and define the antiunitary operator @xmath514 where @xmath488 is complex conjugation , which satisfies @xmath515 .
we verify that ( real ) linear combinations of the following mutually anticommuting @xmath516 , _ commute _ with @xmath517 : @xmath518 a more invariant way to see this is to notice that an anticommuting set of @xmath509 hermitian operators @xmath519 satisfying @xmath520 , generate ( over the reals ) a copy of @xmath521 inside @xmath522 which thereby commutes with a @xmath523 action generated by @xmath524 ( two anticommuting square roots of @xmath8 ) .
the product of two or three different @xmath525 is not hermitian , but @xmath526 is hermitian and also commutes with @xmath517 .
the remaining ten - dimensional subspace of traceless hermitian operators is spanned by operators @xmath527 and they _ anticommute _ with @xmath517 instead we will interpret this latter subspace of @xmath509 hamiltonians as those compatible with a particle - hole symmetry @xmath405 , and these form an important class to be analysed in section [ sec : z2invariant ] . therefore , traceless hermitian operators of the form @xmath373 are precisely those which are compatible with a quaternionic structure @xmath517 , which we interpret as a time - reversal operator @xmath359 .
conversely , a quaternionic structure @xmath517 identifies @xmath511 with @xmath528 , and @xmath529 is the real subalgebra of @xmath530 which commutes with @xmath517 . globalising the above arguments , we deduce : a quaternionic structure @xmath517 on a rank-@xmath531 bloch bundle picks out a real orientable rank-5 subbundle @xmath532 of commuting dirac - type hamiltonians . the chirality condition @xmath526 for an orthonormal frame @xmath533 for @xmath534 defines the positive orientation .
a quaternionic structure always exists on a trivial bloch bundle , and in this case , we identify @xmath534 with the tangent bundle @xmath376 as @xmath535-bundles .
thus @xmath536 corresponds to a positively - oriented orthonormal frame @xmath537 for @xmath72 . in terms of clifford modules
, the clifford algebra @xmath538 has @xmath539 as the spinor representation in which the central chirality element @xmath540 is the identity operator .
a vector field @xmath29 is quantized to an abstract dirac - type hamiltonian @xmath541 , which is represented concretely on @xmath142 as the hamiltonian @xmath542 isometrically : @xmath543 .
the observation that the space @xmath534 has structure group @xmath535 was made in section 5 of @xcite ( for trivial bundles @xmath142 ) .
generalizing remark [ rem : gauge3d ] , @xmath461 is the spinor representation for @xmath544 , and conjugating @xmath404 by a @xmath545 unitary matrix @xmath447 takes it to @xmath546 where @xmath547 is rotated from @xmath452 by the @xmath535 matrix @xmath548 covered by @xmath447 .
the reduction of the gauge group is important : to stay within the class of dirac - type hamiltonians , we can not allow all @xmath549 gauge transformations of the bloch bundle @xmath142 , but only those in @xmath550
. for a 5-component vector field with finite isolated zeroes @xmath82 , the local index at @xmath58 is again given by the degree of the unit vector map @xmath551 restricted to a small sphere around @xmath85 .
this @xmath104-valued local index measures the obstruction to `` gapping '' out a four - band crossing _ within the family of dirac - type hamiltonians _ ( or @xmath359-invariant hamiltonians ) .
away from @xmath47 , @xmath53 can be spectrally - flattened to @xmath552 , and the fermi projection @xmath553 projects onto a complex rank-2 subspace of @xmath539 .
this subspace is actually a _ quaternionic line _ due to the existence of @xmath517 commuting with @xmath53 , i.e. an element of @xmath186 .
just as a unit 3-vector defines a point in @xmath554 via @xmath555 @xmath556-subspace of @xmath557 , we also have @xmath558 via \{unit 5-vector @xmath29 } @xmath559 \{@xmath556-subspace of @xmath64}. the valence subbundle @xmath20 of a @xmath359-invariant dirac - type hamiltonian in 5d is thus a quaternionic line bundle over @xmath66 .
these quaternionic line bundles are pulled back from the tautological one over @xmath560 , and may have non - trivial _
pontryagin class / second chern class in @xmath561 @xcite .
the _ symplectic _ pontryagin classes @xmath562 for @xmath563 bundles @xcite , should be distinguished from the real pontryagin classes @xmath564 for @xmath565 bundles .
the former comes from @xmath566 and equals @xmath567 of the underlying complex bundle of a quaternionic bundle , while the latter comes from @xmath568 and equals @xmath567 of the complexification of the underlying real bundle of a complex bundle ( e.g. underlying a quaternionic one ) . for the tautological quaternionic line bundle @xmath569 , the symplectic pontryagin class /
second chern class is a generator of @xmath570 . the @xmath571 part of the semimetal mayer
vietoris sequence @xmath572 is slightly harder to interpret in terms of characteristic classes of @xmath20 : unlike @xmath573 and complex line bundles , it is not generally true that @xmath574 classifies all quaternionic line bundles ( equivalently @xmath575-bundles ) over a given space .
for example , @xmath576 is a non - trivial @xmath457 bundle over @xmath577 but @xmath578 so @xmath579 does nt detect this bundle .
nevertheless , @xmath580 with the generators coming from pulling back @xmath185 along the projection @xmath581 for each of the five independent choices of 4-subtori .
thus all elements of @xmath582 do arise as the second chern class of some valence quaternionic line bundle @xmath583 ( weak insulator invariants ) , and their restrictions to @xmath584 form the kernel of the local charge map @xmath162 in .
the other elements @xmath585 have some non - zero local charges under @xmath162 . from a singular vector field
@xmath29 with those local charges @xmath586 , the valence bundle for @xmath64 has @xmath587
. thus for @xmath588 , the @xmath574 groups in can be interpreted as the @xmath579 invariants for the valence bundles @xmath20 of the dirac - type hamiltonians as constructed above .
quaternionic line bundles are already stable for dim@xmath589 @xcite , so we can classify the @xmath20 by ( reduced ) quaternionic @xmath1-theory @xmath590 .
therefore , the semimetal mv - sequence in @xmath590 can be interpreted in terms of obstructions to extending a semimetal valence bundle ( over @xmath591 ) to an insulator one ( over @xmath592 ) .
when we introduced @xmath359 to @xmath509 hamiltonians , we found that they had the form @xmath404 for some 5-component vector @xmath452 .
now introduce a further ( unitary ) chiral symmetry @xmath593 , with @xmath594 which a compatible @xmath34 must _ anticommute _ with . without loss , we can take @xmath595 .
equivalently , this is an additional ( antiunitary ) @xmath405 symmetry , with @xmath596 and @xmath597 : take @xmath598 . the additional anticommutation requirement @xmath599 ( or @xmath600 ) forces @xmath53 to be a linear combination of only the first _ four _ gamma matrices in .
furthermore , the traceless condition on @xmath34 is now automatic .
thus in 4d , we may identify the @xmath601-compatible ( or @xmath602-compatible ) @xmath509 hamiltonians with the tangent bundle of @xmath603 , via @xmath404 , where @xmath604 now have four components .
similarly , @xmath539 is the spinor representation for @xmath605 which has a quaternionic structure @xmath606 compatible with @xmath607 .
the spectrum is again @xmath400 , vanishing at the zeroes @xmath608 of @xmath29 .
the topological invariant for a four - band crossing at a zero @xmath58 is the degree of the the unit vector map @xmath609 .
there is also an analysis based on supercommutants .
the requirement of having symmetry operators @xmath610 is equivalent to having a _ @xmath2-graded representation _ of the clifford algebra @xmath611 @xmath612 are the three anticommuting clifford generators , acting irreducibly as odd operators on the superspace @xmath613 .
the supercommutant , by the super - schur lemma @xcite , is a real superdivision algebra @xmath462 .
more explicitly , we can take @xmath614 the @xmath615-compatible hamiltonians are spanned by the above @xmath462 generators and the grading operator .
having identified the @xmath359 and @xmath593-symmetric @xmath509 hamiltonians as those which are linear combinations of @xmath616 , we now explain how a gerbe can be associated to such hamiltonians .
first , it has become usual to associate a gapped chiral symmetric hamiltonian @xmath34 with a unitary in half the number of dimensions as follows . in a basis
in which @xmath593 is @xmath617 , a chiral symmetric hamiltonian @xmath34 is off - diagonal as is @xmath618 .
thus @xmath619 where @xmath620 is unitary since @xmath621 .
note that this association @xmath622 is basis - dependant ( under @xmath623 transformations which respect @xmath593 ) , and for families of hamiltonians @xmath506 , the topological invariants of the map @xmath624 , which is defined wherever @xmath625 is defined ( i.e. away from @xmath47 ) , can change under `` large '' gauge transformations .
we require a further @xmath359 symmetry such that @xmath359 commutes with @xmath593 , so it is of the form @xmath626 for some @xmath27 unitaries @xmath627 . since @xmath411 , up to a unitary transformation preserving @xmath593 , we can take @xmath628 as in .
having done so , a basis for the @xmath359 and @xmath593 compatible hamiltonians is @xmath616 given in .
note that the unitary transformations which preserve both @xmath593 and @xmath359 form a @xmath629 subgroup of @xmath549 , and the matrices @xmath616 can be conjugated by such unitaries . in the basis
where @xmath630 as in , the spectrally - flattened hamiltonian is @xmath631 with @xmath632 we recognise as a parametrisation of @xmath457 .
thus @xmath624 is a map from @xmath633 , from which we can associate a gerbe and analyse the gerbe extension problem to @xmath3 , using the semimetal mv - sequence as in section [ sec : extension ] .
gauge transformations live in @xmath634 , and they effect @xmath635 for @xmath636 . the roles of @xmath593 and @xmath618 can be exchanged , in the sense that we can write @xmath637 and @xmath638 instead . from this point of view , @xmath618 is the difference of the spectral projections @xmath639 onto the positive and negative eigenbundles @xmath640 ( over @xmath66 ) . since @xmath641
, @xmath620 gives an isomorphism between @xmath642 and @xmath643 , c.f .
section iv of @xcite .
there is another interesting class of @xmath509 bloch hamiltonians in 5d : those _ quadratic _ , or _ bilinear _ , in @xmath51 . as explained in section [ sec : dirac5d ] , a quaternionic structure @xmath517 singles out @xmath534 in @xmath432 as the @xmath517-commuting ( traceless ) hamiltonians , and we identify @xmath534 with the tangent bundle @xmath72 of @xmath3 .
furthermore , the @xmath517-_anticommuting _ hamiltonians form a complementary rank-10 subbundle @xmath644 to @xmath534 in @xmath432 , also with structure group @xmath535 . in terms of the @xmath51-matrices , an orthonormal basis for @xmath644 is given by @xmath645 where @xmath646 is a multi - index and e.g. @xmath647 $ ] .
we should think of @xmath648 as coming from @xmath649 , just as @xmath372 came from an orthonormal frame @xmath650 . in accordance with terminology in the physics literature
, elements of @xmath644 are said to be particle - hole symmetric , with @xmath651 .
thus @xmath405-symmetric hamiltonians are exactly the `` @xmath51-bilinear '' ones they are ( @xmath652 times of ) quantizations of @xmath653 , and we will refer to such hamiltonians simply as _ bilinear hamiltonians_. * hamiltonians parametrised by a pair of vector fields . * for example ,
a _ pair _ of vector fields @xmath654 defines an abstract hamiltonian @xmath655 $ ] in @xmath656 .
this is represented on @xmath142 as the hamiltonian @xmath657={\text{i}}({\bm{{a}}}\wedge{\bm{{b}}})_i\gamma_i$ ] .
note that @xmath658 is geometrically just an oriented area element , and is invariant under rotations in the plane of @xmath659 .
furthermore , @xmath660 iff @xmath659 are linearly dependent .
following @xcite , a pair of vector fields @xmath654 over @xmath3 which are linearly independent everywhere is called a _ tangent @xmath348-field _ ; if @xmath654 is defined only on the complement of a finite number of points @xmath47 , it is called a _
tangent @xmath348-field over @xmath3 with finite singularities _ , or a _ tangent @xmath348-field over @xmath66_. [ rmk : twoterm ] the general element of @xmath661 is not a simple tensor @xmath658 but a sum of two such tensors @xmath662 , which can be taken to be orthogonal to each other this is a linear algebraic fact about canonical forms of antisymmetric matrices @xmath663 thus we can more generally consider @xmath664 for mutually orthogonal pairs of vector fields @xmath654 and @xmath665 .
interestingly , the spectrum of @xmath51-bilinear hamiltonians @xmath664 can be found in much the same way as the linear ones , utilising only the clifford algebra relations .
let @xmath664 be a bilinear hamiltonian as in remark [ rmk : twoterm ] , and let @xmath666 .
then @xmath667 , and all four signs occur .
let @xmath668 .
at the clifford algebra level , @xmath669 thus , the square of @xmath670 is @xmath671 .
if @xmath672 or @xmath673 is zero , we are done .
otherwise , @xmath674 for some positively - oriented orthonormal frame @xmath675 , where we have suppressed the wedge product notation and used the chirality condition @xmath676 .
since @xmath677 has spectrum @xmath418 with each doubly - degenerate , it follows that @xmath678 and @xmath679 , where we have suppressed the dependence on @xmath56 .
the zeroes of @xmath34 are easily read off , and there are two types : ( 1 ) when @xmath680 , and ( 2 ) when @xmath681 .
the former involves a four - fold degenerate zero , so all four bands cross , whereas the latter involves crossing between only a pair of bands ( fig .
[ fig : four - band - crossing ] ) . -quadratic hamiltonians .
if @xmath682 , the spectrum degenerates to the dotted blue lines , which are each 2-fold degenerate.,scaledwidth=60.0% ] when @xmath673 is zero , the four - band spectrum becomes a pair of doubly - degenerate bands , and these degenerate pairs cross exactly at the points where @xmath683 , much like the spectrum of the dirac - type hamiltonians which are linear in @xmath51 .
we are interested in such `` bilinear hamiltonians '' , and they are precisely the pure tensor ones @xmath684={\text{i}}\sum_i ( { \bm{{a}}}(k)\wedge{\bm{{b}}}(k))_i\gamma_i$ ] , specified by a _ single _ pair of vector fields @xmath685 , with @xmath686
. we will _ assume _ that the singularities ( points where @xmath658 vanishes ) form a finite set of isolated points @xmath47 .
in other words , @xmath654 defines a tangent 2-field over @xmath66 . at a singularity @xmath58
, there may be topological obstructions to extending @xmath654 across @xmath85 , in which case , we can not open up a gap at the four - band crossing at @xmath85 by simply deforming @xmath672 into a strictly positive function . despite the spectrum resembling that of a ( linear ) dirac - type hamiltonian , this type of four - band crossing can be topologically protected by a very different and much more subtle mechanism , see section [ sec : atinvariant ] .
we may also consider bilinear hamiltonians @xmath687 specified by a pair of vector fields @xmath654 over a @xmath4-manifold with @xmath688 .
however , it is no longer the case that a @xmath405-symmetry singles out such hamiltonians , and one needs to motivate such model hamiltonians in a different way .
they may nevertheless be studied as a toy model .
a detailed study of tangent 2-fields with finite singularities was carried out in @xcite , where analogues of the poincar hopf theorem for vector fields were obtained .
for this , we need the notion of the local index at a singularity of a tangent 2-field @xmath654 on @xmath66 .
let @xmath689 be a small oriented 4-sphere surrounding @xmath58 , then @xmath654 gives a map @xmath690 to the non - compact stiefel manifold @xmath691 of 2-frames in @xmath692 .
there is a homotopy equivalence between @xmath693 and the ( compact ) stiefel manifold @xmath694 of orthonormal 2-frames in @xmath692 .
the _ local index of @xmath654 at @xmath85 _ is defined to be the homotopy class of @xmath695 in @xmath696 , and is the local obstruction to deforming @xmath654 to remove the singularity at @xmath85 .
recall from the previous subsection the construction of a bilinear dirac hamiltonian @xmath687 from a tangent 2-field @xmath654 over @xmath3 with finite singularities @xmath47 , and that four - band crossings occur precisely at @xmath47
. the local index at @xmath85 of @xmath654 has the physical interpretation as a @xmath2 local topological charge protecting the band crossing @xmath58 . to understand @xmath697 better , recall the fibration @xmath698 where the @xmath699 base parametrises the choice of a first normalised vector @xmath700 , and the fiber parametrises ( locally ) the choice of @xmath701 orthonormal to @xmath700 .
from the homotopy exact sequence , @xmath702 we see that @xmath703 comes from the inclusion of the fiber @xmath704 which has the famous @xmath705 .
thus , we can construct an explicit example of a @xmath2-protected four - band crossing as follows .
choose some local euclidean coordinates @xmath371 for a neighbourhood @xmath706 of @xmath85 centred at 0 , and a trivialization @xmath707 of the tangent bundle above @xmath706 .
take @xmath708 , and for @xmath709 , we choose @xmath701 orthonormal to @xmath700 by letting @xmath710 be given by a generator of @xmath711 .
extend to orthogonal vectors @xmath712 to all of @xmath706 by scaling by @xmath713 .
thus @xmath714 on @xmath715 and @xmath716 .
then @xmath85 is a four - band crossing with non - trivial @xmath2-charge . by taking @xmath717
we obtain a _ _
2-plane field__-distributions _ , see @xcite for a guide to @xmath371-distributions on manifolds . ] defining for each @xmath718 a point in the oriented grassmannian @xmath719 of 2-planes in @xmath692 . for a 2-plane field over @xmath66
, there is similarly a map @xmath720 , and a local index can be defined at @xmath58 .
note that @xmath721 , and at the level of @xmath722 the exact sequence @xmath723 says that the index is in @xmath724 .
although we can consider bilinear hamiltonians specified by a field of oriented area elements , i.e. a 2-plane field along with areas ( magnitudes ) , which will also have @xmath2 local indices , we only consider those specified by a tangent 2-field ( with finite singularities ) in this paper . for tangent 2-fields in @xmath77-dimensions , the local topological index at a point singularity
is defined similarly , as an element of @xmath725 .
these homotopy groups are known to be @xmath2 when @xmath4 is odd and @xmath726 when @xmath4 is even @xcite . unlike the case of vector fields , the global constraint for these indices depends on @xmath4 ( mod 4 ) @xcite and in the following subsection , we will restrict to @xmath727 ( mod 4 ) . for this section ,
we assume that @xmath3 is a compact oriented @xmath4-manifold @xmath3 with @xmath727 ( mod 4 ) and @xmath728 .
most of the constructions of section [ sec : euler ] can be carried out for @xmath2 coefficients , and there is also a generalisation of the poincar
hopf theorem for tangent 2-fields and 2-plane fields with finite singularities @xcite .
this suggests that hamiltonians parametrised by a tangent 2-field , such as the 5d @xmath2-dirac semimetals introduced in section [ sec : z2invariant ] , can have topologically protected crossings which are furthermore constrained by the global atiyah dupont condition .
if @xmath729 , we expect there to be `` @xmath2-charge cancellation '' and also a notion of `` @xmath2-fermi arcs '' .
let us simplify notation and write @xmath730 for the tangent 2-field @xmath654 over @xmath66 , and @xmath731 for its local @xmath2 charge at @xmath58 .
the sum @xmath732 is a global obstruction to deforming @xmath733 to a tangent 2-field on all of @xmath3 . in @xcite , atiyah
dupont showed that @xmath734 is equal to the _ _ real kervaire semicharacteristic__. ] @xmath735 , defined as @xmath736 thus @xmath737 for any @xmath730 .
this is not a vacuous result since conditions on @xmath3 , under which tangent 2-fields with finite singularities exist , are known , see @xcite . in particular , it is sufficient that the stiefel whitney class @xmath738 vanishes ( e.g. @xmath71 ) .
although we have assumed the bilinear hamiltonians to be specified by tangent 2-fields , we could also work more generally with 2-fields for other vector bundles using a generalisation of the atiyah dupont theory in @xcite .
let @xmath3 be a compact oriented @xmath4-manifold @xmath3 with @xmath727 ( mod 4 ) and @xmath729 .
for a 0-chain @xmath739 defined on @xmath47 and with total weight zero , a _ kervaire chain _ for @xmath247 is a 1-chain @xmath740 such that @xmath741 , thus @xmath255\in h_1(t , w,{{\mathbb z}}_2)$ ] .
we write @xmath742 for the subset of homology classes of kervaire chains for @xmath247 .
the @xmath2-versions of the semimetal mv sequence and the dual homology sequence are @xmath306&h^{d-1}(t,{{\mathbb z}}_2)\ar[d]_{\text{pd}}\ar[r ] & h^{d-1}(t\setminus w,{{\mathbb z}}_2)\ar[d]_{\text{pd}}\ar[rr]^{\beta } & & h^{d-1}(s_w,{{\mathbb z}}_2)\ar[d]_{\text{pd}}\ar[rr]^{\sigma } & & h^d(t,{{\mathbb z}}_2)\ar[d]_{\text{pd}}\ar[r ] & 0\\ 0\ar[r ] & h_1(t,{{\mathbb z}}_2)\ar[r ] & h_1(t , w,{{\mathbb z}}_2)\ar[rr]^{\partial } & & h_0(w,{{\mathbb z}}_2 ) \ar[rr]^{\sigma } & & h_0(t,{{\mathbb z}}_2)\ar[r ] & 0 } .\nonumber\ ] ] as before , we define @xmath743 to be the poincar dual image of @xmath742 .
we will then explain how a tangent 2-field with finite singularity set @xmath47 has a topological invariant in @xmath744 and therefore a kervaire chain representation .
let @xmath3 be a compact oriented @xmath4-manifold @xmath3 with @xmath727 ( mod 4 ) and @xmath729 , and @xmath228 be the sphere bundle for its tangent bundle .
a _ kervaire structure _ for @xmath3 is a class in @xmath745 whose restriction to each fibre @xmath746 generates @xmath747 .
define @xmath748 to be the set of kervaire structures for @xmath3 . from the gysin sequence with @xmath2-coefficients ,
we see that @xmath749 acts on @xmath748 freely and transitively by pullback and addition .
let @xmath750 be the bundle of stiefel manifolds @xmath751 associated to the orthonormal frame bundle of @xmath3 ( endowed with some riemannian metric ) . corresponding to the fibration @xmath752 , there is a fibration @xmath753 a ( non - singular ) tangent 2-field @xmath730 on @xmath3 gives rise to a kervaire structure as follows .
let @xmath754 denote the section of @xmath755 determined by @xmath730 , which is a @xmath4-cycle on @xmath755 .
the poincar dual is a @xmath756-cocycle on @xmath755 , and pushforward under @xmath757 gives a @xmath86-cocycle in @xmath228 .
furthermore , the construction of the cocycle is such that its restriction to @xmath231 generates @xmath747 for each @xmath56 .
now consider tangent 2-fields on @xmath3 with finite singularities @xmath47 .
as in section [ sec : semimetaleulerchain ] , we can define a local charge operator @xmath758 .
then for each 0-chain @xmath247 with total charge zero , we define @xmath759 to be the subset of @xmath760 whose restriction to each fibre over @xmath718 generates @xmath747 , and whose local charges is @xmath307 .
each @xmath759 is an affine space for @xmath749 .
suppose @xmath730 is a tangent 2-field on @xmath66 with local charge 0-chain @xmath247 , then it determines a @xmath4-cycle in @xmath761 relative to the boundary @xmath762 , which poincar dualises to a @xmath756-cocycle on @xmath761 .
pushforward under @xmath757 gives a @xmath86-cocycle on @xmath340 representing a class in @xmath759 .
fix a reference non - singular vector field @xmath218 , then @xmath763 identifies @xmath759 with @xmath744 .
the following diagram summarises the various affines spaces for @xmath764 : @xmath765 \ar[d ] & & { \mathfrak{kerv}}(t,{\mathcal{w}})\ar[ll ] \ar[d ] & \subset \quad\quad\;\ ; h_1(t , w,{{\mathbb z}}_2 ) \\ & h^{d-1}(t,{{\mathbb z}}_2)\ar[rr]\ar[d ] \ar[u ] & & { \mathfrak{cokerv}}(t,{\mathcal{w}})\ar[ll ] \ar[d]\ar[u ] & \subset \quad h^{d-1}(t\setminus w,{{\mathbb z}}_2 ) \\
z}}_2)\quad\supset & \widetilde{{\mathfrak{cokerv}}}(t)\ar[rr ] \ar[u ] & & \widetilde{{\mathfrak{cokerv}}}(t,{\mathcal{w}})\ar[ll ] \ar[u ] & \subset \quad h^{d-1}({\mathcal{s}}|_{t\setminus w},{{\mathbb z}}_2 ) } .\nonumber\ ] ] let @xmath34 be a bilinear hamiltonian specified by a tangent 2-field @xmath730 over a compact oriented 5-manifold @xmath3 with @xmath729 , with finite singularities @xmath47 .
it has a cohomological topological invariant @xmath766\in h^4(t\setminus w,{{\mathbb z}}_2)$ ] defined as in the previous paragraph , and its _ kervaire chain representation _ is the poincar dual @xmath767\in h_1(t , w,{{\mathbb z}}_2)$ ] .
the bulk - boundary correspondence proceeds exactly as before , taking a kervaire chain for @xmath247 to a `` @xmath2-fermi arc '' with topological invariant in @xmath768 . unlike the usual @xmath104-fermi arcs , the @xmath2-fermi arcs cancel in pairs so that there is at most one arc between two projected band crossings , and the arc never fully winds around a 1-cycle in @xmath769 .
the topological semimetal - topological insulator relationship encoded in the semimetal mv sequence can , in principle , be refined using an mv sequence in _ differential cohomology _ @xcite .
another refinement comes from additional spatial symmetries .
for example , a spatial inversion symmetry @xmath419 induces an action on the brillouin torus @xmath3 under which @xmath47 should be invariant .
the same action arises if @xmath359 rather than @xmath770 symmetry is present .
the vector field @xmath29 needs to be equivariant , and the local index at @xmath58 is an equivariant degree .
finally , the @xmath1-theory mv sequence can be used ( with @xmath66 replaced by @xmath771 ) to analyse stable topological semimetal invariants .
this paves the way for the development of a new type of t - duality transformation @xmath772 , which would provide a t - dual picture of semimetals along the lines of @xcite in the case of insulators .
we end by posing the question : associated to the mv sequence analysis in the case of kervaire semimetals ( section [ sec : kervairestructures ] ) , is there a geometric extension problem ? _
acknowledgements_. we thank keith hannabuss for some helpful discussions , and siye wu for his useful criticisms .
this work was supported by the australian research council via arc discovery project grant dp150100008 and arc decra grant de170100149 .
to be self - contained , we recall here some universal constructions of geometric objects with connections that are used in the main text . + * the hopf line bundle over @xmath40 . * ( sections [ sec : physback ] and [ sec : extension ] ) recall that the 2d - sphere @xmath40 is defined as the space of complex lines in @xmath773 .
the hopf line bundle @xmath144 over @xmath40 is defined as the sub - bundle of the trivial bundle @xmath774 given by @xmath775 as a sub - bundle of the trivial bundle @xmath774 ( which has the obvious hermitian metric and unitary connection ) @xmath144 comes equipped with a hermitian metric and unitary connection @xmath776 .
explicitly , a natural projection defining @xmath144 is given by @xmath777 , where @xmath778 is the unit vector in @xmath464 and @xmath779 are the pauli matrices , @xmath780 for @xmath781 .
then @xmath782 and the curvature is @xmath783 * the basic gerbe over @xmath785 . * ( sections [ sec : semimetalgerbe ] and [ sec : extension ] ) we identify @xmath182 and consider @xmath786 defined by @xmath787 then the projection to the 2nd factor is a submersion @xmath788 and the fibred product @xmath789 } = \{(z_1 , z_2 , g ) \in \text{u}(1)^2 \times \text{su}(2 ) \big| z_1 , z_2 \ne 1 , z_1 , z_2 \notin { \rm spec}(g)\}.\ ] ] for @xmath790 , let @xmath791 denote the @xmath672 eigenspace of @xmath377 and @xmath792 where @xmath672 is an eigenvalue lying in the open arc component in @xmath793 joining @xmath794 that does not contain @xmath795 .
let @xmath796 denote the determinant line , which is the highest exterior power .
then @xmath797 is a bundle gerbe description of the basic gerbe over @xmath182 , see @xcite , @xcite .
it comes with gerbe connection whose 3-curvature is the wess
witten ( wzw ) integrand @xmath798 and whose dixmier - douady number classifying the gerbe up to equivalence is , @xmath799 + * the quaternionic hopf line bundle over @xmath186 . *
( sections [ sec:5dsemimetal ] and [ sec : extension ] ) recall that the 4d - sphere @xmath186 is defined as the space of quaternionic lines in @xmath528 and that the quaternionic hopf line bundle @xmath185 over @xmath186 is defined as the sub - bundle of the trivial bundle @xmath800 given by @xmath801 as a sub - bundle of the trivial bundle @xmath802 ( which has the obvious hermitian metric and unitary connection ) @xmath185 comes equipped with a hermitian metric and unitary connection @xmath803 .
explicitly , a natural projection defining @xmath185 is given by @xmath804 , where @xmath805 is the unit vector in @xmath692 and @xmath806 are gamma matrices , with @xmath807 for @xmath808 generating the clifford algebra and @xmath526 .
then @xmath809 and the curvature is @xmath810 the 2nd chern number classifying @xmath185 up to isomorphism is , @xmath811 + * the stiefel manifold @xmath812 . * ( section [ sec : torsionsemimetal ] ) it can be defined as the space of all orthonormal 2-frames in @xmath692 . as a homogeneous space , @xmath813 , where we notice that it is a compact , connected , oriented @xmath814-dimensional manifold , which can be identified with the unit sphere @xmath815 subbundle of the tangent bundle of the sphere @xmath699 .
so @xmath812 is a non - principal @xmath816-bundle over @xmath699 with euler class equal to @xmath348 .
the pontryagin number of the bundle @xmath817 is just the pontryagin number of @xmath699 , which is zero since the sphere @xmath699 is the oriented boundary of the @xmath43 dimensional ball . by @xcite ,
the euler and pontryagin numbers together determine the bundle up to isomorphism . via the long exact sequence for the fibration @xmath817
, we see that there is a natural isomorphism , ( this is related to @xcite ) @xmath818 the generator of @xmath819 determines the non - trivial @xmath785-bundle over @xmath577 via the clutching construction , @xmath820 by the pontryagin thom construction , @xmath821 comes from @xmath822 and so relates to the t hooft
polyakov @xmath453 monopole @xcite .
bradlyn , b. , cano , j. , wang , z. , vergniory , m.g . ,
felser , c. , cava , r.j . ,
bernevig , b.a . : beyond dirac and weyl fermions : unconventional quasiparticles in conventional crystals .
science * 353*(6299 ) ( 2016 ) korba , j. : distributions , vector distributions , and immersions of manifolds in euclidean spaces . in : handbook of global analysis ( edited by d. krupka and d. saunders ) , elsevier science , amsterdam , pp .
665724 ( 2008 ) turner , a.m. , vishwanath , a. : beyond band insulators : topology of semimetals and interacting phases . in : m. franz and l. molenkamp ( eds . ) , contemp .
concepts cond . mat .
* 6 * , _ topological insulators _ , pp . 293324 , elsevier , amsterdam ( 2013 ) wigner , e.p . :
unitary representations of the inhomogeneous lorentz group including reflections . in group
theoretical concepts in elementary particle physics , ( ed .
f. grsey ) , vol .
3780 , gordon and breach , new york , ( 1964 ) | the subtle interplay between local and global charges for topological semimetals exactly parallels that for singular vector fields .
part of this story is the relationship between cohomological semimetal invariants , euler structures , and ambiguities in the torsion of manifolds .
dually , semimetal invariants can be represented by euler chains from which the surface fermi arc connectivity can be deduced .
these dual pictures , and the link to insulators , are organised using geometric exact sequences .
we go beyond dirac - type hamiltonians and introduce new classes of semimetals whose local charges are subtle atiyah
dupont
thomas invariants globally constrained by the kervaire semicharacteristic , leading to the prediction of torsion fermi arcs . |
as progress is made towards a new , precision era of cosmology , _ redundancy _ will play an increasingly important role .
as cosmology is an _
observational _ science , it will be crucial to avail ourselves of multiple , independent tests of , and constraints on , competing cosmological models and their parameters .
furthermore , such redundancy may provide the only window on systematic errors which can impede our progress or send us off in unprofitable directions . to illustrate the efficacy of such an approach in modern cosmology , i ll track the baryon density of the universe as revealed early on ( first few minutes ) by big bang nucleosynthesis ( bbn ) , later ( few hundred thousand years ) as coded in the fluctuation spectrum of the cosmic microwave background ( cmb ) radiation , and up to the present , approximately 10 gyr after the expansion began .
as theory suggests and terrestrial experiments confirm , baryon number should be preserved throughout these epochs in the evolution of the universe , so that the number of baryons ( @xmath0 nucleons ) in a comoving volume _ should _ be unchanged from bbn to today . as a surrogate for identifying a comoving volume , we may compare the baryon / nucleon density to the density of cmb relic photons .
except for the additional photons produced when e@xmath1 pairs annihilate , the number of photons in our comoving volume is also preserved .
as a result , the baryon density may be tracked through the evolution of the universe utilizing the nucleon - to - photon ratio @xmath2 where , n_n / n_. since the temperature of the cmb fixes the present number density of relic photons , the fraction of the critical mass / energy density in baryons today ( @xmath3 ) is directly related to @xmath2 , _ 10 10 ^ 10= 274_bh^2 , where the hubble parameter , measuring the present universal expansion rate , is h@xmath4 kms@xmath5mpc@xmath5 . according to the hst key project , @xmath6 ( ( * ? ? ?
* freedman 2001 ) ) . for several decades
now the best constraints on @xmath2 have come from the comparison of the predictions of bbn with the observational data relevant to inferring the primordial abundances of the relic nuclides d , he , he , and li . for recent reviews ,
the interested reader is referred to @xcite , @xcite , @xcite , and references therein .
however , the time is rapidly approaching when new data of similar accuracy will be available to probe the baryon density at later epochs .
indeed , recent cmb data has very nearly achieved this goal .
this same level of precision is currently lacking for the present universe estimates but , such as they are , they do permit us to compare and contrast independent estimates of @xmath2 ( or @xmath7 ) at three widely separated eras in the evolution of the universe . in the next section
the bbn bounds on the nucleon density are presented and their implications for the dark baryon and dark matter problems described .
next , we turn to the cmb estimates of the baryon density , comparing them to those from bbn .
last ( but not least ) , we turn to an estimate of the present universe baryon density utilizing data whose interpretation avoids the need to adopt a relation between mass and light in the universe . in recent years cosmological research has been data - driven and we have become used to looking forward with great anticipation to the latest observational results , and greedily wishing for more . be careful of what you wish for !
not all the new data has led us unerringly along the right path .
indeed , within a few weeks of this meeting new data became available on the cmb angular fluctuation spectrum ( halverson 2001 ; netterfield 2001 ; lee 2001 ) and on the deuterium abundance ( ( * ? ? ?
* levshakov , dessauges - zavadsky , dodorico , & molaro 2001 ) ; ( * ? ? ?
* pettini & bowen 2001 ) ) which change dramatically some of the results i presented in my talk at this symposium . in the interest of preserving some of the historical record
, i will comment on the older data and their implications , while reserving the most recent observations for my conclusions regarding the baryon density of the universe .
during the first few minutes in the evolution of the universe the density and temperature were sufficiently high for nuclear reactions to occur in the time available . as the universe expanded , cooled , and became more dilute , the rolling blackout " became permanent and the universal nuclear reactor went offline .
the abundances of the light nuclei formed during this limited epoch of primordial alchemy were determined , for the most part , by the competition between the time available ( the universal expansion rate ) and the density of the reactants : nucleons ( neutrons and protons ) .
the abundances predicted for nucleosynthesis in the standard cosmological model ( sbbn ) are shown ( as a function of @xmath2 ) in figure [ fig : bbn ] .
the abundances of d , he , and li are rate limited " , being determined by the competition between the nuclear production / destruction rates and the universal expansion rate . as such , they are sensitive to the nucleon density and have the potential to serve as baryometers " . in contrast , the nuclear reactions which build he are so rapid and , because its destruction is inhibited ( due to its strong binding and the lack of a stable nuclide at mass-5 ) , the helium-4 mass fraction , y , is insensitive to @xmath2 . since virtually all the neutrons available at bbn are incorporated into he , y is determined mainly by the neutron - to - proton ratio which is controlled by the competition between the weak interaction rate ( @xmath8-decay " ) and the expansion rate . as a result , since the neutron lifetime is known very accurately , the he abundance is a sensitive probe of the early universe expansion rate .
= 3.5truein of the three relic nuclides whose primordial abundances may be probes of the baryon density ( d , he , li ) , deuterium is the baryometer of choice .
first and foremost , the predicted primordial abundance has a sigificant dependence on the nucleon density ( d / h @xmath9 ) . as a result
, if the primordial abundance is known to , say , 10% , then the baryon density ( @xmath2 ) can be determined to @xmath10 ; truly precision cosmology ! equally important ,
as @xcite showed long ago , bbn is the only astrophysical site where an interesting " abundance of deuterium may be produced ( d / h @xmath11 ) ; the relic abundance is not enhanced by post - bbn production .
furthermore , as primordial gas is cycled through stars , deuterium is completely destroyed ( because of the small binding energy of the deuteron , destruction occurs during the pre - main sequence evolution , when the stars are fully mixed ) . as a result , the abundance of deuterium has only decreased ( or , remained close to its primordial value ) since bbn .
( d / h)_now ( d / h)_bbn , [ d / h1 ] where now " refers to the true " deuterium abundance in any system observed at any time ( the solar system ; the interstellar medium ; high redshift , low metallicity qso absorbers , etc . ) .
there is , however , one potentially serious problem associated with using deuterium as a baryometer .
the atomic spectra of deuterium and hydrogen are identical , save for the isotope shift associated with the deuteron - proton mass difference . as a result , a small amount of hydrogen at the wrong " velocity ( an interloper " ) can masquerade as deuterium , so that ( d / h)_now ( d / h)_obs .
[ d / h2 ] a comparison of equations [ d / h1 ] and [ d / h2 ] reveals the problem : how to relate ( d / h)@xmath12 to ( d / h)@xmath13 ?
one approach is to concentrate on observing deuterium in those high redshift ( high - z ) , low metallicity ( low - z ) systems where not much gas will have been cycled through stars and the deuterium ( now " ) should be essentially primordial ( bbn ) .
if so , there should be no variation ( outside of the statistical errors ) with metallicity . to test for interlopers the best approach ( the favorite of theorists ! ) is to have lots of data .
observers are making great progress towards this goal , but the road has not always been straight . after some false starts , there were four high - z , low - z qso absorption line systems where deuterium had been detected as of the time of this symposium in april 2001 ( burles & tytler 1998a , b ; omeara 2001 ; ( * ? ? ?
* dodorico , dessauges - zavadsky , & molaro 2001 ) ) . a fifth system ( ( * ? ? ?
* webb 1997 ) ) , with possibly a much higher d / h , is widely agreed to have insufficient velocity data to rule out contamination by a hydrogen interloper ( ( * ? ? ?
* kirkman 2001 ) ) .
however , these data on d / h , displayed as a function of metallicity in figure 2 , appear to challenge our expectations ( of a low - metallicity , deuterium plateau " ) in that they suggest an anticorrelation between d / h and metallicity .
has deuterium actually been destroyed since bbn , and is the bbn abundance at least as high as the highest data point ?
this would be most surprising in such low metallicity systems ( see , jedamzik & fuller 1997 ) .
= 3.5truein perhaps one or more of the high " points is too high " because of an interloper . to test for this possibility ,
let s display the same data as a function of the hydrogen column density ( omeara 2001 ) ; see figure 3 .
it might be expected that the lower hydrogen column density absorbers could be more easily contaminated .
two of the systems have relatively low h - column densities ( @xmath14@xmath15 lyman limit systems ( lls ) ) and two have considerably higher n@xmath16 ( @xmath17@xmath15 damped lyman alpha ( dla ) absorbers ) .
= 3.5truein indeed ,
the hint from figure 3 is that interlopers may be playing a role in either or both of the lls .
if so , perhaps we should identify only the dlas with the primordial deuterium abundance .
however , given the small number of data points , this would be premature . at the symposium i adopted a weighted average ( @xmath18d / h@xmath19 ) and called for more data .
observers were nt long in responding .
but , i had nt anticipated what they would find . further observations by @xcite of the @xcite absorber ( the low d / h point at the highest n@xmath16 in fig .
3 ) revealed a more complex velocity structure and led to an upward revision in the derived d / h ( which , by the way , is by more than the previous statistical error estimate , reinforcing the potential for systematic errors to wreak havoc ) .
so far , so good , now that the four d / h determinations are more consistent with one another , alleviating the need to invoke interloper contamination . however , in the meanwhile @xcite weighed in with a new deuterium detection in a dla .
the pettini - bowen abundance is smaller than all the other determinations , indeed smaller than ( although within the errors of ) the presolar nebula abundance of @xcite . the current data ( as of july 2001 )
are shown in figure 4 .
= 3.5truein what to do ?
the dispersion among these four determinations hints that one or more may be wrong . until this puzzle is resolved by more data ( the last refuge of the theorist )
, i will adopt as a default estimate the one derived by @xcite : ( d / h)@xmath20 .
the consequences for the bbn derived baryon density are shown in figure 5 where the overlap between the predicted and observed " primordial abundances are used to constrain @xmath2 .
= 3.5truein from a careful comparison between the bbn predicted abundance ( including errors which are subdominant ) and the adopted primordial value , the baryon density when the universe is less than a half hour old may be constrained ( see figure 6 ) .
for cosmology , this is truly a precision " determination .
whether it is accurate , only time will tell .
_ 10(bbn ) = 5.6 0.5 ( _ bh^2 = 0.020 0.002 ) [ etabbn ] = 3.0truein this range for the baryon density poses some interesting challenges to our view of the universe .
these challenges may be seen by comparing the bbn estimate of with those found by adding up all the baryons associated with the luminous " material observed in the present / recent universe at z @xmath21 ( persic & salucci 1992 ) , and with estimates of the total mass density at present .
these comparisons are shown in figure [ fig : obvsh ] where the various density parameter estimates / ranges are plotted as a function of the hubble parameter h@xmath22 ( recall that , according to @xcite , h@xmath23 ) .
= 3.5truein the gap between the upper limit to luminous baryons and the bbn band is the dark baryon problem " : not all the baryons expected from bbn have been seen in the present universe .
perhaps it is hubris to expect that all baryons will choose to radiate ( or absorb ) in those parts of the spectrum we can see , or which our instruments can record . indeed , from the absorption observed in the lyman - alpha forest at redshifts of order 2 3 ( ( * ? ?
* weinberg , miralda - escud , hernquist , & katz 1997 ) ) , it seems clear that the density of baryons in the present universe is much larger than the upper bound to luminous baryons shown in figure [ fig : obvsh ] .
we return to a different estimate of the baryon density in the present / recent universe in [ sec : sn1a ] .
the gap between the bbn band and the band labelled @xmath24 is one aspect of the classical dark matter problem : the mass density inferred from the motions of galaxies exceeds the baryon density derived from bbn . aside from those solutions which resort to modifying gravity ( sanders ( 2001 ) this volume )
, the standard assumption is that this gap provides evidence for the dominance at present of non - baryonic dark matter .
much of this symposium was devoted to discussions of possible dark matter candidates and the means for their detection .
the early universe is radiation dominated .
as the universe expands and cools , the density in non - relativistic matter becomes relatively more important , eventually dominating after a few hundred thousand years . at this stage perturbations can begin to grow under the influence of gravity and , on scales determined by the relative density of baryons , oscillations in the baryon - photon fluid will develop .
when , at a redshift @xmath25 , the electron - proton plasma combines to form neutral hydrogen , the cmb photons are freed to propagate thoughout the universe .
these cmb photons preserve the record of the baryon - photon oscillations through very small temperature fluctuations in the cmb spectrum which have been detected by the newest generation of experiments , beginning with cobe ( ) and continuing with the exciting boomerang ( ( * ? ? ?
* de bernardis 2000 ) ; ( * ? ? ?
* lange 2000 ) ) and maxima-1 ( ( * ? ? ?
* hanany 2000 ) ) results which appeared somewhat more than a year ago . in figure
[ fig : cmb ] we illustrate the status quo ante with the dramatic and challenging results from those earlier data .
= 3.83truein the relative heights of the odd and even peaks in the cmb angular fluctuation spectrum depend on the baryon density and the early boomerang and maxima-1 data favored a high " baryon density ( compare the bbn case " , @xmath26 , in the upper panel with that for a baryon density some 50% higher shown in the lower panel ) . at the time of this symposium , these were the extant data and they posed a challenge to the consistency of the standard model of cosmology . at the symposium
we were told that new data would be forthcoming shortly and , the observers did nt disappoint . in less than a month new ( and some revised ) data appeared ( halverson 2001 ; netterfield 2001 ; lee 2001 ) which have eliminated the challenge posed by the older data .
although the extraction of cosmological parameters from the cmb data can be very dependent on the priors adopted in the analyses ( see ( * ? ? ?
* kneller 2001 ) ) , the inferred baryon density is robust .
@xcite find , _
10(cmb ) = 6.0 0.6 ( _ bh^2 = 0.022 0.002 ) . in figure
[ fig : lik2 ] is shown the comparison of the likelihoods for the nucleon - to - photon ratio from bbn , when the universe was some tens of minutes old , and from the cmb , some few hundred thousand years later .
the excellent agreement between the two , independent estimates is a spectacular success for the standard model of cosmology and an illustration of the great potential for precision tests of cosmology .
as already discussed in [ sec : bbn ] ( see figure [ fig : obvsh ] ) , the amount of baryons visible " in the present universe is small compared to that expected on the basis of bbn ( and , as seen in [ sec : cmb ] , to that revealed by the cmb data ) .
if most baryons in the present universe are dark , how can their density be constrained ?
there are a variety of approaches .
many depend on assumptions concerning the relation between mass and light , or require the adoption of a specific model for the growth of structure . in the approach utilized here we attempt to avoid such model - dependent assumptions .
instead , we use the data from the snia magnitude - redshift surveys ( ( * ? ? ? * perlmutter 1997 ) ; ( * ? ? ? * schmidt 1998 ) ; ( * ? ? ? * perlmutter 1999 ) ) , along with the _ assumption _ of a flat universe ( which receives strong support from the newest cmb data ( halverson 2001 ; netterfield 2001 ; lee 2001 ) ) to pin down the total matter density ( @xmath24 ) , which will be combined with an estimate of the universal baryon fraction ( @xmath27 ) derived from studying the x - ray emission from clusters of galaxies . for more details on this approach , see @xcite and @xcite .
= 3.0truein in figure [ fig : sn1a ] are shown the snia - constrained 68% and 95% contours in the @xmath28 plane .
the expansion of the universe is currently accelerating for those models which lie above the ( dashed ) q@xmath29 line .
the @xmath30 line is for a flat " ( zero 3-space curvature ) universe . as shown in @xcite , adopting the assumption of flatness and assuming the validity of the snia data , leads to a reasonably accurate ( @xmath31 ) estimate of the present matter density . _
m(snia;flat ) = 0.28 ^+0.08 _ -0.07 . as the largest collapsed objects , rich clusters of galaxies provide an ideal probe of the baryon _ fraction _ in the present universe .
observations of the x - ray emission from clusters of galaxies permit constraints on the hot gas content of such clusters which , when corrected for baryons in stars ( but , not for any dark baryons ! ) , may be used to estimate @xmath32 . from observations of the sunyaev - zeldovich effect in x - ray clusters ,
@xcite constrain the hot gas fraction which @xcite have used to estimate @xmath32 and to derive a present - universe ( @xmath33 gyr ; @xmath34 ) baryon density .
_ 10(snia;flat ) = 5.1 ^+1.8 _ -1.4 ( _ bh^2 = 0.019 ^+0.007 _
-0.005 ) . in figure
[ fig : lik3 ] the corresponding likelihood distribution for the present universe baryon density is shown ( snia " ) along with those derived earlier from deuterium and bbn , and from the cmb fluctuation spectra . although the uncertainties are largest for this present - universe value , it is in excellent agreement with the other , independent estimates .
the abundances of the relic nuclides produced during bbn encode the baryon density during the first few minutes in the evolution of the universe .
of these relics from the early universe , deuterium is the baryometer of choice . the deuterium abundance in relatively unprocessed material , such as the high - z
, low - z qso absorption line systems should be very nearly primordial . at present
there are data for five such systems . although the statistical accuracies of these data are high , the dispersion among them in the derived d / h ratio is surprisingly large , suggesting that systematic errors ( interlopers ?
, complex velocity structure ? ) may effect one or more of these determinations .
nonetheless , these data seem consistent with a primordial abundance ( d / h)@xmath35 ( ( * ? ? ? * omeara 2001 ) ) which was adopted in [ sec : bbn ] to derive @xmath36 ( @xmath37 ) .
several hundred thousand years later , when the universe became transparent to the cmb radiation , the baryon density was imprinted on the temperature fluctuation spectrum which has been observed by the cobe ( ( * ? ? ?
* bennett 1996 ) ) , boomerang ( ( * ? ? ?
* netterfield 2001 ) ) , maxima ( ( * ? ? ?
* lee 2001 ) ) , and dasi ( ( * ? ? ?
* halverson 2001 ) ) experiments . for determining the baryon density ,
these current cmb data have a precision approaching that of bbn : @xmath38 ( @xmath39 ) .
the excellent agreement between the bbn and cmb values ( see fig . [
fig : lik2 ] ) provides strong support for the standard model of cosmology . in the present universe
most baryons are dark ( @xmath40 ) , so that estimates of the baryon density some 10 billion years after the expansion began are more uncertain and , often model - dependent . in [ sec : sn1a ] we combined an estimate of the total matter density ( @xmath24 ) derived from the snia magnitude - redshift data ( ( * ? ? ?
* perlmutter 1997 ) ; ( * ? ? ?
* schmidt 1998 ) ; ( * ? ? ?
* perlmutter 1999 ) ) and the assumption of a flat universe ( @xmath41 ) , with the universal baryon fraction inferred from x - ray observations of clusters of galaxies ( ( * ? ? ?
* grego 2001 ) ) to derive @xmath42 ( @xmath43 ) ( ( * ? ? ?
* steigman , kneller , & zentner 2000 ) ) .
although of much lower statistical accuracy , this estimate of the present universe baryon density is in complete agreement with those from bbn and the cmb ( see fig . [
fig : lik3 ] ) .
i note in passing that if the mass of dark baryons in clusters is similar to the stellar mass , this present - universe baryon density estimate would increase by @xmath44 , bringing it into essentially perfect overlap with the bbn and cmb values .
= 3.5truein the concordance of the standard , hot , big bang cosmological model is revealed clearly by the overlapping likelihood distributions for the universal density of baryons shown in fig .
[ fig : lik3 ] . as satisfying as this agreement
might be , it should impell us to action , not complacency .
how may we test further the standard model ?
one way is to return to the relic nuclides which , so far , have been set aside : he and li . = 3.5truein consistency among the three , independent baryon density estimates permits us to identify a best " value : @xmath45 ( @xmath46 ) . for this baryon - to - photon ratio
the bbn - predicted primordial abundances are : ( d / h)@xmath47 , y@xmath48 , and ( li / h)@xmath49 . these best " estimates are shown by the stars " in figures [ fig : hevsd ] and [ fig : livsd ] . in those figures the bands reflect the bbn - predicted he vs. d and li vs d relations , including the nuclear and weak interaction physics uncertainties .
it is clear that the deuterium abundance is in excellent agreement with current data .
what of he and li ? at present there are two estimates for the primordial abundance of he based on large ( nearly ) independent data sets and analyses of low - metallicity , extragalactic regions ; see fig .
[ fig : hevsd ] .
it " ( ( * ? ? ?
* izotov , thuan , & lipovetsky 1994 ) , ( * ? ? ? * izotov & thuan 1998 ) ) estimate of y@xmath50 is only @xmath51 away from our best fit value of 0.248 , while the os " determination ( ( * ? ? ?
* olive & steigman 1995 ) , ( * ? ? ?
* olive , skillman , & steigman 1997 ) , ( * ? ? ?
* fields & olive 1998 ) ) of y@xmath52 is nearly @xmath53 lower . although it may be tempting to dismiss the os estimate , recent high quality observations of a relatively metal - rich region in the smc by @xcite reveal an abundance ( y@xmath54 ) which is _ lower _ than the it primordial value .
when this abundance is extrapolated to zero metallicity , @xcite find y@xmath55 , in excellent albeit accidental ! agreement with the os value .
the comparisons among different observations and between theory and observations suggest that unaccounted systematic errors ( underlying stellar absorption weakening the helium emission lines ? ) may have contaminated at least some of the data .
the comparison between theory and data for li poses much more of a challenge ; see fig . [
fig : livsd ] . the data point plotted in fig .
[ fig : livsd ] , from @xcite : ( li / h)@xmath56 , is lower by a factor of three ( @xmath57 dex ) than the standard model prediction ( the star " in fig .
[ fig : livsd ] ) .
there are , however , good reasons to believe that the low value of the lithium abundance derived by @xcite from observations of metal - poor , _ old _ stars ( ( * ? ? ?
* ryan , norris & beers 1999 ) ) is not representative of the lithium abundance in the gas out of which these stars formed ( ( * ? ? ?
* pinsonneault 1999 ) ; ( * ? ? ?
* salaris & weiss 2001 ) ; ( * ? ? ?
* pinsonneault 2001 ) ; ( * ? ? ?
* theado & vauclair 2001 ) ) . using a model for lithium depletion via rotational mixing over the long lifetimes of these stars
, @xcite reanalyzed the @xcite data and concluded that the primordial lithium abundance should be higher than the @xcite value by @xmath58 dex , bringing the abundance inferred from the earliest generation of stars in the galaxy in closer agreement with that expected for the best " estimate of the baryon density in the standard cosmological model .
increasingly precise observational data permit us to track the baryon density from the big bang to the present . at widely separated epochs from the first few minutes , through the first few hundred thousand years , to the present universe ,
a consistent value emerges , accurate to better than 10% : @xmath59 ( @xmath60 ) .
such a low baryon density ( @xmath61 ) reinforces the need for non - baryonic dark matter ( @xmath62 ) , which itself appears to be subdominant at present to an unknown form of dark energy .
precision cosmology has led us to an extreme form of the copernican principle ! while this baryon density is fully consistent with present estimates of the primordial deuterium abundance , it challenges the data and astrophysics " used to determine the primordial abundances of he and li . if unresolved by new data , or better astrophysics , these conflicts
might be providing a peek at new physics beyond the standard models of cosmology and/or particle physics .
these are surely exciting times . for valuable advice , assistance , and tutorials
i wish to thank gus evrard , jim felten , jim kneller , jeff linsky , joe mohr , paulo molaro , keith olive , manuel peimbert , max pettini , marc pinsonneault , bob scherrer , evan skillman , david tytler , sueli viegas , terry walker , and andrew zentner .
praise is due mario livio and the efficient staff of the stsci for the smooth organization of an exciting meeting .
my research is supported by doe grant de - fg02 - 91er-40690 . | the primordial abundances of deuterium , helium , and lithium probe the baryon density of the universe only a few minutes after the big bang . of these relics from the early universe , deuterium is the baryometer of choice . after reviewing the current observational status ( a moving target ! ) ,
the bbn baryon density is derived and compared to independent estimates of the baryon density several hundred thousand years after the big bang ( as inferred from cmb observations ) and at present , more than 10 billion years later .
the excellent agreement among these values represents an impressive confirmation of the standard model of cosmology , justifying indeed , demanding more detailed quantitative scrutiny . to this end ,
the corresponding bbn - predicted abundances of helium and lithium are compared with observations to further test and constrain the standard , hot , big bang cosmological model . |
ladder systems have been the focus of much theoretical and experimental work over more than two decades .
one theoretical motivation was to investigate the dimensional crossover between the well - known one - dimensional ( 1d ) physics and the two - dimensional ( 2d ) case , in the search of 2d non - fermi liquid physics .
a second reason stems from experiments and the study of ladder compounds , such as the famous telephone number one sr@xmath1ca@xmath2cu@xmath3o@xmath4 which has a superconducting phase at high pression and for a small hole density .
@xcite the simplest ladder model is that of a two - leg ladder , made of two coupled fermionic chains . in stark contrast to the single chain case , the two - leg ladder system displays a superconducting phase with @xmath5-wave superconductivity for repulsive interactions which stems from the doping of a spin - gapped mott insulating phase at half - filling .
@xcitethis gave the belief that the two - leg ladder problem already contains seeds of the rich physics of the cuprates .
two - leg ladders have thus become over the years a fundamental system for the study of low - dimensional strongly correlated fermions .
various exotic quantum phases have been predicted theoretically depending on the form of the coupling between the two chains and the filling.@xcite experimental realizations of two - leg ladder systems are clearly called for , in particular , to investigate the rich physics of the weak - coupling regime . in this respect ,
ultracold fermionic gases are a promising way to study two - leg ladder problems thanks to the high level of control on interchain hopping and interactions .
@xcite the ladder geometry might be created by considering double - well optical lattices for instance.@xcite a second possible way is to load a two - component fermi gas in a optical lattice and consider higher - lattice orbitals , typically the @xmath0-band , to simulate a fermionic two - leg ladder system .
more precisely , we consider , in this paper , a two - component fermi gas which is loaded in a 1d optical lattice ( running along the @xmath6-direction ) with moderate strength of ( harmonic ) confining potential @xmath7 in the direction ( i.e. , @xmath8 ) perpendicular to the chain .
@xcite it is assumed that all the @xmath9-level of the oscillator are fully occupied while the @xmath0-level , i.e. , @xmath10 , are partially filled .
the resulting lattice fermionic model has been derived in refs . within the tight - binding approximation and takes the following form : @xmath11 , \label{pbandmodel}\end{aligned}\ ] ] where @xmath12 is the orbital index and @xmath13 is the spin index or internal components of the underlying cold atoms . in eq .
( [ pbandmodel ] ) , @xmath14 describes the density operator at site @xmath15 and a pseudo - spin operator for the orbital degrees of freedom has been defined : @xmath16 where @xmath17 are the pauli matrices and a summation over repeated indices is implied .
model ( [ pbandmodel ] ) can be viewed as a two - chain fermionic system without interchain hopping but with density - density interchain interactions and pair - hopping processes between the two orbitals ( see fig . [
fig : p - band-2leg ] ) . -band model ( [ pbandmodel ] ) .
the orbital degrees of freedom plays the role of the legs of the ladder . besides the kinetic term with amplitude @xmath18 for each orbital , there are local interactions at each site @xmath15 : density - density intra - orbital ( amplitude @xmath19 , in red ) and inter - orbital ( amplitude @xmath20 , in green ) , orbital exchange ( amplitude @xmath20 , in blue ) and inter - orbital pair - hopping that breaks u(1)@xmath21 generically ( amplitude @xmath20 , in orange ) .
[ fig : p - band-2leg ] ] the continuous symmetry of model ( [ pbandmodel ] ) for general @xmath22 is : @xmath23 , where u(1)@xmath24 denotes the u(1 ) ( charge ) symmetry related to the conservation of the total number of atoms , while @xmath25 is the internal global su(2 ) ( spin ) symmetry of the two - component fermi gas .
we note that the u(1)@xmath26 continuous symmetry for the orbital degrees of freedom is explicitly broken in eq .
( [ pbandmodel ] ) .
however , the use of harmonic potential in the @xmath8 direction implies a constraint of the two coupling constants @xmath22 and the investigation of model ( [ pbandmodel ] ) along the harmonic line @xmath27 enjoys an u(1)@xmath26 symmetry corresponding to rotation along the @xmath28-axis in the orbital subspace . the u(1)@xmath21 symmetry is trivially realized when @xmath29 where the model reduces to two decoupled su(2 ) hubbard chains .
finally , the @xmath30 case displays also an u(1)@xmath26 symmetry since it is directly related to @xmath29 after a redefinition of the orbital pseudo - spin operator .
we consider , in this paper , the most general case where @xmath22 are not fine - tuned for two main reasons . on the one hand ,
departure from the harmonic line @xmath27 can be investigated by breaking the axial symmetry of the 2d harmonic trap with the introduction of quartic potentials .
@xcite on the other hand , the harmonic line has been wrongly identified in refs . and
we want to make contact with their results obtained when @xmath31 . at half - filling
, it has been shown that model ( [ pbandmodel ] ) along the harmonic line with repulsive interaction describes a haldane phase @xcite of the spin-1 heisenberg chain .
@xcite the latter is known to be the paradigmatic 1d topological phase protected by symmetry.@xcite for an attractive interaction , a haldane phase for the charge degrees of freedom @xcite has also been predicted in ref . by means of a strong - coupling approach and density - matrix renormalization group ( dmrg ) calculations .
@xcite in ref .
, it has been shown that there is adiabatic continuity between weak and strong coupling regime and these two haldane phases are related by a spin - charge interchange symmetry .
the half - filled @xmath0-band model ( [ pbandmodel ] ) along the harmonic line thus paves the way to realize experimentally non - trivial haldane phases in the context of cold fermions . in this paper , by means of a low - energy approach and dmrg calculations , we map out the zero - temperature phase diagram of the @xmath0-band model ( [ pbandmodel ] ) for incommensurate filling and at quarter filling which best avoids three - body looses . in stark contrast to the conclusion of ref .
for attractive interaction , we show that the @xmath0-band model ( [ pbandmodel ] ) away from half - filling does not behave as an attractive 1d hubbard chain with the formation of a 2@xmath32 cdw but instead is a representative of the physics of two - leg fermionic ladders .
in particular , we find that the hallmark of its phase diagram is the emergence of various competing superconducting phases for attractive and repulsive interactions as well as a mott transition with the formation of fully gapped bond - ordering waves ( bow ) .
the rest of the paper is organized as follows . in sec .
[ sec : lowenergy ] , we present our low - energy approach based on a mapping onto majorana fermions and a one - loop renormalization group ( rg ) analysis . a zero - temperature phase diagram for model ( [ pbandmodel ] ) is then deduced in the weakly - interacting regime for incommensurate filling and at quarter filling with fermi vector @xmath33 ( @xmath34 being the lattice spacing ) . in sec .
[ sec : dmrg ] , dmrg calculations are carried out to investigate the intermediate and large - coupling regimes to fully determine the phase diagram of the @xmath0-band model .
we present then our concluding remarks in sec .
[ sec : conclusion ] .
in this section , we perform a weak - coupling approach to model ( [ pbandmodel ] ) away from half - filling .
we focus on incommensurate filling and a special commensurate filling where @xmath33 .
the starting point of this approach is the introduction of left - right moving dirac fermions from the continuum limit of the non - interacting hamiltonian ( [ pbandmodel ] ) with @xmath35 : @xmath36 with @xmath37 , @xmath38 , and @xmath39 . in stark contrast to the half - filled case of ref . , there is a spin - charge separation which strongly simplifies
the analysis and the full hamiltonian of the @xmath0-band model ( [ pbandmodel ] ) decomposes into two commuting pieces : @xmath40 with @xmath41 = 0 $ ] , where @xmath42 governs the physical properties of the charge degrees of freedom and @xmath43 refers for all remaining non - abelian degrees of freedom , i.e. , spin , orbital and spin - orbital degrees of freedom .
let us first consider the simplest part in eq .
( [ spinchargesep ] ) , i.e. , the charge degrees of freedom . for incommensurate filling , on general grounds , the charge part takes the form of a tomonaga - luttinger hamiltonian : @xcite @xmath44 , \label{luttbis}\ ] ] where @xmath45 and @xmath46 are the luttinger parameters . in this low - energy approach ,
the charge excitations are described by the bosonic field @xmath47 and its dual field @xmath48 .
the explicit form of the luttinger parameters for the @xmath0-band model can be extracted from the continuum limit ( [ contlimitdirac ] ) after standard calculations .
we find : @xmath49 where @xmath50 is the fermi velocity and @xmath51 .
for incommensurate filling , no umklapp term appears and the charge degrees of freedom display metallic properties in the luttinger liquid universality class .
@xcite however , at quarter filling , umklapp processes might be generated leading to an additional term in eq .
( [ luttbis ] ) .
such perturbation can be found by means of higher orders of the rg calculations or from lattice symmetries .
@xcite in this respect , the one - step translation symmetry expresses as follows in terms of the charge bosonic field : @xmath52 .
the charge perturbation with the smallest scaling dimension which is compatible with the translation symmetry is then : @xmath53 .
the full charge hamiltonian density for the commensurate filling @xmath33 becomes then equivalent to the quantum sine - gordon model at @xmath54 : @xmath44 -g_u \cos(\sqrt{16 \pi } \phi_c ) .
\label{chargeham}\ ] ] this model is exactly solvable and the development of the strong - coupling regime with a charge gap @xmath55 corresponds to @xmath56 , while for @xmath57 the charge degrees of freedom remain gapless in the luttinger universality class . using the estimate ( [ luttingerparameters ] ) , we find the position of the critical line for a mott transition within the bosonization approach : @xmath58 such that @xmath59 .
one has to be very careful with this estimate since the full expression of @xmath46 as a function of @xmath22 ( [ luttingerparameters ] ) is strictly valid only in the weak - coupling regime .
one needs a complementary approach , i.e. , numerical calculations to conclude on the existence or not of a mott transition for the lattice model ( [ pbandmodel ] ) .
note that the mechanism of the mott transition is very similar to the one which occurs in the repulsive u(4 ) hubbard chain @xcite or in the problem of spin-3/2 cold fermions .
@xcite for @xmath56 ,
the charge degrees of freedom are fully gapped and the strong - coupling regime of the sine - gordon model ( [ chargeham ] ) is described by the pinning of the charged field on the minima : @xmath60 @xmath0 being an integer .
the sign of the umklapp coupling constant @xmath61 is difficult to fix but we expect @xmath62 from an higher - order perturbative expansion in the weak - coupling regime so that:@xmath63 . @xcite the dmrg calculations of sec . [
sec : dmrg ] will shed light on the position of the pinning of the charge field from the determination of the mott - insulating phases of the @xmath0-band model .
the low - lying excitations are massive kinks and antikinks which interpolate between the ground states of the quantum sine - gordon model ( [ chargeham ] ) .
the charges associated to these charge excitations are @xmath64 in units of the electron charge .
we have thus massive holon as low - lying excitations in the charge sector . the low - energy physics of the remaining degrees of freedom , included in @xmath65 of eq .
( [ spinchargesep ] ) , can be inferred from a mapping onto six majorana fermions as in refs .
a simple way to obtain this correspondence is through the introduction of four chiral bosonic fields @xmath66 from the bosonization of dirac fermions:@xcite @xmath67 where the bosonic fields satisfy the following commutation relation : @xmath68 = \frac{i}{4}\delta_{l l'}\delta_{\sigma \sigma'}. \label{commutator}\ ] ] the presence of the klein factors @xmath69 ensures the correct anticommutation of the fermionic operators .
the klein factors satisfy the anticommutation rule @xmath70 and they are constrained so that @xmath71 , with @xmath72 .
hereafter , we will work within the @xmath73 sector .
the next step of the approach is to introduce a bosonic basis which singles out the different degrees of freedom , i.e. , charge , spin , orbital , and spin - orbital degrees of freedom:@xcite @xmath74 from these new bosonic fields , one can now consider a refermionization procedure by introducing six left and right moving majorana fermions through : @xmath75 where @xmath76 are again klein factors which ensure the adequate anticommutation rules for the fermions . using this correspondence , the interacting part of @xmath65 can be expressed in terms of these majorana fermions : @xmath77 with the identification : @xmath78 using standard calculations , the one - loop rg equations for model ( [ majomodel ] ) can de derived : @xmath79 as often in 1d , the rg equations enjoy some hidden discrete symmetries : @xmath80 which correspond to the existence of duality symmetries on the majorana fermions @xcite @xmath81 while the right - moving majorana fermions remain invariant .
these chiral transformations for each majorana fermion correspond , in the continuum limit , to the well - known kramers - wannier duality symmetry of the underlying 1d ising model in a transverse field .
the latter transformation maps the ordered phase to the disorder one which is ordered in terms of the disorder operator .
@xcite starting from the ordered phase of the six underlying ising models in a transverse field , the transformations ( [ dualitiespband ] ) lead then to four possible different gapful phases . a numerical rg analysis of eqs .
( [ rgn=2 ] ) is necessary to determine which phases is indeed reached starting from the initial conditions ( [ couplingsn=2 ] ) . ) .
bottom : for each phase denoted i - iv , we give the dominant superconducting correlations ( see text ) as well as the duality symmetries which relate the phases between them .
we display the central charge @xmath82 or @xmath83 which characterizes the different quantum phase transitions of the model.,title="fig : " ] 0.3 cm ) .
bottom : for each phase denoted i - iv , we give the dominant superconducting correlations ( see text ) as well as the duality symmetries which relate the phases between them .
we display the central charge @xmath82 or @xmath83 which characterizes the different quantum phase transitions of the model.,title="fig : " ] we did this numerical analysis and , as depicted in fig .
( [ fig : phasediag_n=2 ] ) , it reveals that the rg flow goes in the strong - coupling regime in the far infrared ( ir ) along four asymptotic lines : @xmath84 along these rays , the interacting part of the effective hamiltonian ( [ majomodel ] ) simplifies as follows : @xmath85 with @xmath86 . using the duality symmetries ( [ dualitiespband ] ) , all these models reduces to the same model which takes the form of the so(6 ) gross - neveu ( gn ) model with interaction : @xcite @xmath87 this is an example of a dynamical symmetry enlargement ( dse ) where an so(6 ) symmetry , i.e. a higher symmetry than the initial symmetry of the @xmath0-band lattice model , emerges at low - energy .
this phenomenon occurs in a large variety of models with marginal interactions in the scaling limit @xcite as , for instance , in the half - filled two - leg hubbard model @xcite and in the su(4 ) hubbard chain at half - filling @xcite , where an so(8 ) symmetry occurs at low energy .
the emergence of an so(6 ) symmetry has also been obtained within the one - loop rg analysis of various models of doped two - leg ladders . @xcite the main interest of this so(6 ) dse scenario stems from the fact that the isotropic rg ray is described by a massive integrable field theory ( [ gnso6 ] ) when @xmath88 .
@xcite the development of the strong - coupling regime in the so(6 ) model leads to generation of a non - perturbative fermionic mass , i.e. the emergence of a spin - gap phase where all spin , orbital and spin - orbital excitations are fully gapped . the nature of the underlying electronic phases of the @xmath0-band model
can then be inferred by a straightforward semiclassical approach of the so(6 ) model and the application of the duality symmetries ( [ dualitiespband ] ) .
alternatively , one can perform a direct bosonic semiclassical approach of the different models in eq .
( [ so6raysapp ] ) using the identification ( [ refer ] ) . in the following ,
we proceed to this analysis by identifying the phases of fig .
[ fig : phasediag_n=2 ] . in the first phase ,
when @xmath90 and @xmath91 of fig .
( [ fig : phasediag_n=2 ] ) , the numerical analysis of the one - loop rg equations reveals that the rg flow is attracted in the far ir along a special ray ( i ) of eq .
( [ asymptotrg ] ) .
this phase is described by the @xmath89 duality and the resulting physical properties of that phase are governed by the interacting hamiltonian @xmath92 of eq .
( [ so6raysapp ] ) . a straightforward semiclassical approach of the latter model leads to the following pinning of the bosonic fields : @xmath93 the electronic properties of phase i , which includes the harmonic line @xmath94 , depend also on the charge degrees of freedom that are decoupled .
let us first consider the @xmath95 case .
one has then a gapless charge excitation whereas all remaining spin , orbital , and spin - orbital excitations are fully gapped .
all @xmath96 densities are short - ranged due to the pinning of the dual spin - orbital field @xmath97 .
interestingly enough , the leading electronic instability in this phase turns out to be a bond - centered superconducting instability with order parameter : @xmath98 which is odd under the @xmath99 orbital exchange symmetry ( @xmath100 ) .
a similar superconducting instability has been introduced in the study of doped two - leg electronic ladder in ref . .
in the continuum limit , we have : @xmath101 one can then bosonize this operator by means of the identification ( [ bosofer ] ) and the physical basis ( [ basebosons ] ) : @xmath102 . \nonumber
\label{bcssboson}\end{aligned}\ ] ] taking account of the vacuum expectation values ( [ pinningi ] ) , we get @xmath103 so that the equal - time correlation function of the pairing operator has a power - law decay : @xmath104 when @xmath105 , the bond - centered bcs superconducting instability is strongly enhanced with respect to the non - interacting case .
_ we have thus a dominant superconducting instability for repulsive interaction_. while the @xmath96 densities are short - ranged in phase i , @xmath106 densities might compete with the superconducting instability ( [ bcss ] ) .
many @xmath106 density terms can be written in the continuum description like @xmath107 for instance .
the latter gives in the bosonized language : @xmath108 so that in the spin - gapped phase ( [ pinningi ] ) , one has @xmath109 , and therefore a power - law decay for its correlation function : @xmath110 since @xmath111 , one observes that the @xmath106 density correlation function ( [ 4kfcorrel ] ) decays much faster than the superconducting one ( [ bcsscorrel ] ) . in summary , the leading instability of phase
i with @xmath95 is the bond - centered bcs superconducting one . in the mott - insulating phase with @xmath112
, all excitations are fully gapped . the bond - centered bcs superconducting instability ( [ bcss ] )
is now short - ranged since the charge field @xmath47 is pinned ( [ pinnedcharge ] ) and thus its dual field @xmath48 is a strongly fluctuating field . in contrast , the @xmath106 density of eq .
( [ rho4kf1 ] ) can have now a non - zero expectation value , breaking spontaneously the translation symmetry .
we thus expect a mott phase with a two - fold degenerate ground state at quarter filling @xmath113 .
the nature of this phase depends on the sign of the coupling constant @xmath61 of the umklapp perturbation in model ( [ chargeham ] ) . in this respect , we introduce the following order parameters as in the study of the extended quarter - filled two - leg hubbard ladder : @xcite @xmath114 with @xmath115 .
one can obtain a continuum and bosonized representation for these order parameters .
the @xmath106 contribution of the latter plays a crucial role and can be derived using the results of refs .
and we get for the leading contribution : @xmath116 with @xmath117 , @xmath118 . from these results , we find the bosonized descriptions of the order parameters ( [ orderparamottn=2 ] ) : @xmath119 taking into account that in phase i , we have @xmath120 , we have either a uniform @xmath106 bow or uniform @xmath106 cdw depending on the sign of the umklapp term in eq .
( [ chargeham ] ) : @xmath121 we expect in the weak - coupling regime that @xmath122 .
a uniform @xmath106 bow is thus stabilized which is two - fold degenerate and breaks the translation symmetry .
the latter phase is similar to the bow phase obtained in the quarter - filled spin-3/2 so(5 ) chain model .
@xcite in the second phase , when @xmath124 and @xmath125 , the one - loop rg flow is now attracted in the far ir by the asymptote ( ii ) of eq .
( [ asymptotrg ] ) .
this phase is described by the @xmath123 duality . a straightforward semiclassical approach of model @xmath126 in eq .
( [ so6raysapp ] ) leads to the following pinning of the bosonic fields in this phase : @xmath127 in this phase , as we will see in sec .
[ sec : dmrg ] , numerical results find that @xmath95 .
all spin , orbital , and spin - orbital excitations are fully gapped in this phase and the charge degrees of freedom are gapless since @xmath95 . due to the presence of attractive interactions ,
one may call this gapless phase a luther - emery phase as in the spin-1/2 hubbard chain with @xmath128 .
@xcite its physical nature is very different from that of phase i since the expectation values of the bosonic fields ( [ pinningii ] ) are different . as it can readily be seen from eq .
( [ pinningii ] ) , we have again the condensation of a dual field , here the orbital one @xmath129 , which implies that the @xmath96-cdw operator is a strongly fluctuating order . as in phase
i , the only possible cdw quasi - long range order in this phase is a @xmath106 cdw .
however , the @xmath130 cdw of eq .
( [ rho4kf1 ] ) becomes now short - range since the orbital dual field @xmath131 condenses in phase ii . in this respect
, we have to consider another @xmath106 cdw instability : @xmath132 .
the latter can be directly expressed in terms of the bosonic field : @xmath133 using the expectation values ( [ pinningii ] ) , we deduce the leading asymptotics of the equal - time correlation function of the @xmath106 cdw operator : @xmath134 as it will be seen in the next section , we have @xmath135 and , one does not expect , on general ground , that this @xmath106-cdw will be the dominant instability in this phase . as in phase
i , a superconducting instability turns out to be the leading one .
the latter is defined by the following order parameter : @xmath136 which is odd under the @xmath99 orbital symmetry ( @xmath100 ) and antisymmetric with respect to the spin degrees of freedom ( @xmath137 ) .
the order parameter ( [ bcs- ] ) can be expressed directly in terms of the bosonic fields : @xmath138 .
\nonumber \label{bcs - boson}\end{aligned}\ ] ] using the expectation values ( [ pinningii ] ) , we get @xmath139 , so that the equal - time correlation function of the pairing operator ( [ bcs- ] ) has a power - law decay : @xmath140 which dominates the @xmath106 cdw ones ( [ 4kfcorrelbis ] ) when @xmath95 . the next phase , defined by @xmath142 and @xmath125 , is described by the asymptote ( iii ) of the one - loop rg flow .
the harmonic line of the @xmath0-band model with attractive interaction belongs to this phase which is described by the @xmath141 duality with interacting hamiltonian @xmath143 of eq .
( [ so6raysapp ] ) .
the bosonic fields of the bosonization approach are now pinned to the values : @xmath144 a spin - gap is formed and a gapless @xmath145 phase emerges in this attractive regime with @xmath146 since the umklapp term can not gap out the charge degrees of freedom when @xmath147 . in close parallel to the previous cases , one can determine the nature of the leading electronic instability of this luther - emery phase by means of the bosonization approach combined with the pinning ( [ pinningiii ] ) .
the dominant @xmath106 cdw is the one of phase ii , given by eq .
( [ rho4kf2 ] ) with the power - law behavior ( [ 4kfcorrelbis ] ) .
the relevant superconducting instability for phase iii is defined by : @xmath148 which is even under the @xmath99 orbital symmetry ( @xmath100 ) and antisymmetric with respect to the spin degrees of freedom ( @xmath137 ) . in terms of the bosonic fields , it reads as follows : @xmath149 .
\nonumber \label{bcstboson}\end{aligned}\ ] ] using the vacuum expectation values ( [ pinningiii ] ) , we immediately get @xmath150 , so that the equal - time correlation function of this pairing operator has a power - law decay : @xmath151 since @xmath147 , we conclude that the superconducting instability ( [ bcs+ ] ) dominates the @xmath106-cdw ordering ( [ 4kfcorrelbis ] ) .
_ the luther - emery phase is thus governed by a superconducting instability ( [ bcs+ ] ) in stark contrast to the @xmath96-cdw phase predicted by the dmrg study of ref . _ .
the physics of the @xmath0-band model with attractive interaction is thus not similar to that of 1d attractive hubbard model as emphasized in ref . .
the last phase of the @xmath0-band model corresponds to the region where @xmath142 and @xmath91 .
the numerical analysis of the rg flow shows that , here , the one - loop rg flow is attracted in the far ir by the special line ( iv ) of eq .
( [ asymptotrg ] ) .
the resulting phase is described by the @xmath152 duality with interacting hamiltonian @xmath153 of eq .
( [ so6raysapp ] ) which leads to the following pinning for the bosonic fields of the bosonization approach : @xmath154 as before , this pinning leads to the formation of a gapless @xmath145 phase when @xmath95 where the charge degrees are the only critical modes of the problem .
we now consider the standard @xmath5-wave superconducting instability of the two - leg electronic ladder to determine the nature of phase iv : @xmath155 which is even under the @xmath99 orbital symmetry ( @xmath100 ) .
the bosonized expression of this superconducting instability reads as follows : @xmath156 .
\nonumber \label{bcsdboson}\end{aligned}\ ] ] from this expression , we observe that this operator is a fluctuating order , i.e. , has short - ranged correlation , in the previous phases , while , in phase iv , one has from the pinning ( [ pinningiv ] ) : @xmath157 .
the @xmath5-wave superconducting instability ( [ bcsd ] ) becomes dominant in phase iv with the power - law behavior : @xmath158 from the pinning ( [ pinningiv ] ) , we find that the @xmath96-cdw operator is short - ranged while the @xmath106 cdw of phase ii , given by eq .
( [ rho4kf2 ] ) , has a power - law behavior in phase iv : @xmath159 with subleading exponent when @xmath95 with respect to the superconducting instability ( [ bcsdcorrel ] ) . from the duality symmetries ( [ dualitiespband ] )
, we can , as well , discuss the different quantum phase transitions that occur in the @xmath0-band model by investigating self - dual manifolds
. however , fig .
[ fig : phasediag_n=2 ] reveals that the transitions belong to special lines of the lattice model : @xmath160 and @xmath29 . from eq .
( [ pbandmodel ] ) , the @xmath29 describes two decoupled quarter - filled spin-1/2 hubbard chains with @xmath19 coupling constant . when @xmath161 , a spin - gap is formed and therefore the phase ii / phase iii transition is critical with @xmath162 gapless charge modes . when @xmath90 , all degrees of freedom are gapless since the quarter - filled spin-1/2 hubbard chains is known not to exhibit a mott transition .
@xcite the phase i / phase iv transition is thus critical with central charge @xmath83 .
the transition between phase i and phase ii is located along the @xmath163 line of the @xmath0-band model . from the definition of the @xmath0-band model ( [ pbandmodel ] ) , one can show that the latter line corresponds to two decoupled repulsive quarter - filled spin-1/2 hubbard chains which does not have relevant umklapp process . a @xmath83 behavior should occur along this quantum phase transition . finally , the last transition between phase iii and phase iv belongs to the @xmath164 line which takes the form of two decoupled attractive quarter - filled spin-1/2 hubbard chains with a @xmath82 quantum critical behavior .
all these results can be derived by investigating self - dual manifolds of the rg eqs .
( [ rgn=2 ] ) . since the transition lines are located on special high symmetry lines
, one would expect the location of these transitions to remain universal beyond the weak coupling regime where this analysis is valid .
this will be indeed confirmed numerically in sec .
[ subsec : qpt_num ] .
a summary of the phases and quantum phase transitions , obtained from the low - energy approach , can be found in the bottom of fig .
[ fig : phasediag_n=2 ] .
we will now determine the phase diagram of the @xmath165 @xmath0-band model at quarter filling ( i.e. 1 particle per site ) using numerical simulations with the dmrg algorithm .
this will allow to go beyond the one - loop rg analysis done in fig .
[ fig : phasediag_n=2 ] and check the analytical predictions for realistic intermediate or strong couplings @xmath166 ( we fix @xmath167 as the unit of energy ) .
moreover , numerical data are needed to determine the numerical value of the luttinger parameter @xmath46 which allows to compute the dominant correlations .
typically , we have used open boundary conditions ( obc ) and lengths @xmath168 and @xmath169 , keeping up to 4000 states when computing correlations in order to keep a discarded weight below @xmath170 in most regions , although simulations where @xmath171 were found to be more difficult to converge ( discarded weight around @xmath172 , see discussion below ) . for practical purpose , we have mapped the @xmath0-band model onto an _ equivalent _ ( pseudo)spin-1/2 ( where the pseudo spin corresponds to the orbital ) fermionic models on a @xmath173-leg ladder , and we have implemented the abelian u(1 ) symmetry corresponding to the conservation of particles spin . our main result is presented in fig . [
fig : phasediag_pband_n2_n1 ] where we plot the phase diagram vs @xmath174 obtained by computing various correlation functions on @xmath169 systems .
we will present the numerical data below , but we can already discuss the different phases .
first of all , the ( quantum phase ) transition lines are found to be @xmath29 and @xmath175 , which is expected by symmetry since they correspond to special lines of the model , see sec .
[ subsec : qpt ] . a detailed analysis will be given in sec .
[ subsec : qpt_num ] .
second , it is remarkable that phases ii , iii and iv found in the weak - coupling analysis are confirmed to exist in a wide range of parameters .
last , we have computed the luttinger parameter @xmath46 using the dominant correlation , which is always of some bcs type , in each of these phases .
overall , we observe a rather good agreement between the numerical phase diagrams obtained by solving rg equations ( fig .
[ fig : phasediag_n=2 ] ) or the full microscopic model using dmrg ( fig .
[ fig : phasediag_pband_n2_n1 ] ) , except for the large mott phase at this commensurate filling . for incommensurate
filling , we predict that the rg phase diagram would be identical to the numerical one . .
numerical values correspond to the luttinger parameter @xmath46 in each region . ]
we will now present some numerical data that were used to compute this numerical phase diagram .
we have relied mostly on computing superconducting correlation functions ( in various channels ) as well as density ones . to avoid spurious effects due to obc , we have chosen to compute correlations as @xmath176 so that we can determine if they decay algebraically or are short - ranged . in figs .
[ fig : corr_u1_m6_u2_6]-[fig : corr_u1_m6_u2_m2]-[fig : corr_u1_6_u2_m6 ] , such correlations are plotted and correspond respectively to phases ii , iii and iv . in all cases ,
we emphasize that dominant correlations are found to be the superconducting ones , in different channels ( see below ) . indeed
, the luttinger parameter @xmath46 is always larger than 1/2 in the critical phases , otherwise a mott phase is stabilized .
some values of @xmath46 are given on the phase diagram in fig .
[ fig : phasediag_pband_n2_n1 ] . in particular , for phases ii , iii and iv
, we observe an adiabatic continuity from weak to strong coupling , while phase i on the contrary is much reduced and replaced instead by mott phase at intermediate coupling and polarization at strong coupling . in this region of the phase diagram , our numerical results , shown in fig .
[ fig : corr_u1_m6_u2_6 ] for instance when @xmath177 and @xmath178 , are in perfect agreement with the rg predictions and dominant bcs@xmath179 predictions .
however , numerics is needed to compute the luttinger parameter @xmath46 and determine whether pairing correlations dominate over density ones . using eqs .
( [ 4kfcorrelbis])-([bcs - correl ] ) , our numerical fits for both bcs@xmath179 and cdw correlations are compatible with a single @xmath180 , for instance 0.65 for the parameters set chosen on the plot .
all the other correlations are short - ranged as expected .
the leading instability in this phase is therefore the bcs@xmath179 superconducting one .
, @xmath178 ) and @xmath169 obtained with dmrg .
this corresponds to our phase ii .
fitting the dominant bcs@xmath179 correlation with eq .
( [ bcs - correl ] ) leads to @xmath181 , which is also consistent with the @xmath106 cdw correlation function ( see eq .
( [ 4kfcorrel ] ) ) . ]
the phase iii region is particularly interesting since it could be achieved experimentally using the harmonic trapping scheme ( i.e. @xmath182 ) with attractive interactions . as seen in fig .
[ fig : corr_u1_m6_u2_m2 ] , we have found that the dominant correlations are of the pairing type , in the bcs@xmath183 channel . in all this region , they can be fitted using eq .
( [ bcs+correl ] ) to get @xmath184 ( values are given in the phase diagram ) .
in such a case , the dominant density correlations are _ uniform _ and decay as @xmath185 as expected . , @xmath186 ) for @xmath169 obtained with dmrg .
this corresponds to our phase iii .
fitting the dominant bcs@xmath183 correlation with eq .
( [ bcs+correl ] ) leads to @xmath187 . ]
note that this dominant superconducting correlation function was not computed in ref .
, where only cdw signal was discussed . as a result ,
the dominant instability in this luther - emery phase is a superconducting one , in stark contrast to the @xmath96-cdw phase predicted earlier @xcite . in the region corresponding to phase iv , our numerical results in fig .
[ fig : corr_u1_6_u2_m6 ] indicate that dominant correlations are of the pairing type again , but in a different bcs@xmath188 channel , as expected from the low - energy analysis . in the whole region
, we can fit these correlations using eq .
( [ bcsdcorrel ] ) to extract @xmath46 , or equivalently we could fit the subleading cdw correlations with eq .
( [ 4kfcorrelfin ] ) , although it would be less precise .
some values of @xmath46 are given in fig .
[ fig : phasediag_pband_n2_n1 ] and are always larger than 1/2 so that no mott transition is found . , @xmath189 ) and @xmath169 obtained with dmrg .
this corresponds to our phase iv .
fitting the dominant bcs@xmath188 correlation with eq .
( [ bcsdcorrel ] ) leads to @xmath181 , which is also consistent with the @xmath106 cdw correlation function ( see eq .
( [ 4kfcorrel ] ) ) . ] the region @xmath190 , which should correspond to phase i ( see fig .
[ fig : phasediag_n=2 ] ) according to rg solution is more involved to analyze .
we have plotted on fig .
[ fig : bow ] the finite - size scaling of the @xmath191 bow , i.e. the kinetic energy difference measured in the middle of the chain .
data are given along the harmonic line @xmath182 .
extrapolations are compatible with a finite , albeit small , value for intermediate interactions ( for instance @xmath192 , @xmath193 ) , while it seems to vanish in the weak and strong coupling regimes .
it is known that the weak coupling is difficult numerically since the non - interacting starting point correspond to two decoupled fermionic chains with a large total central charge @xmath83 . for instance , the mott transition in the su(@xmath194 ) hubbard model at filling @xmath195 has been discussed quite extensively to occur for a _
critical @xmath196 when @xmath197 based on bosonization and quantum monte - carlo results @xcite as well as dmrg ones @xcite , while some older dmrg simulations had indicated a _ vanishing _
@xmath198 @xcite .
since the charge gap is expected to open in an exponentially way , this is clearly a difficulty for any numerical technique .
however , this regime is perfectly suited for bosonization and weak - coupling rg : indeed , since the luttinger parameter @xmath199 in the non - interacting case , umklapp processes are irrelevant so that finite interactions are necessary to enter the mott phase .
thus , our interpretation is the following : ( i ) for small interaction parameters , rg analysis should be valid and phase i with dominant bcs@xmath200 correlations is expected with @xmath180 so that umklapp processes are irrelevant ; ( ii ) for intermediate interactions , a mott phase occurs with @xmath191 bow and exponentially decaying bcs correlations at large distance ; ( iii ) for very large interactions ( i.e. @xmath201 and @xmath202 ) , we have noticed that the ground - state is ferromagnetically polarized ( hence degenerate ) : the polarized ground - state can be simply understood as two decoupled spinless fermionic chains ( one for each orbital ) , which energy is independent of both @xmath19 and @xmath20 .
intuitively , such a state could be stabilized at large @xmath203 s since other competing states will have higher energies .
still , our numerics is clear on this point as can be checked by comparing ground - state energies in different sectors ( varying the number of particles per spin ) , or measuring @xmath199 from charge correlations ( data not shown ) .
bow order parameter along the harmonic line @xmath182 . ] a simple way to obtain information on the phases , or on their phase transitions , is through the measurements of block von neumann entanglement entropy which is known to scale as @xcite : @xmath204 where @xmath205 is the central charge and @xmath206 the conformal distance . on a finite system with obc , due to friedel oscillations , the fitting can be more involved and one can use for instance the knowledge of the local kinetic bond energies to get more reliable results ( see fig .
11 of ref . for instance ) .
our numerical results should be compared to the analytical predictions made in sec .
[ subsec : qpt ] . in fig .
[ fig : entropies_ll](a ) , we have plotted typical data in phases ii , iii and iv . by removing oscillations using the bond kinetic energy as an additional fitting parameter
, we can obtain smooth functions that perfectly agree with the expected behavior ( [ eq : svn ] ) with @xmath145 .
obtained for @xmath169 and various @xmath166 .
oscillations have been removed using the bond kinetic energies @xmath207 and fitting parameter @xmath208 ( equal respectively to @xmath173 , @xmath209 and @xmath210 ( see text ) . in all cases , we can extract a central charge @xmath211 as expected .
( b ) similar data along the quantum phase transitions which can be fitted with @xmath212 ( using @xmath213 ) in the attractive case and with @xmath214 ( using @xmath215 and @xmath216 , plus a shift for clarity ) in the repulsive case using smaller @xmath168 system.,title="fig : " ] obtained for @xmath169 and various @xmath166 .
oscillations have been removed using the bond kinetic energies @xmath207 and fitting parameter @xmath208 ( equal respectively to @xmath173 , @xmath209 and @xmath210 ( see text ) . in all cases , we can extract a central charge @xmath211 as expected .
( b ) similar data along the quantum phase transitions which can be fitted with @xmath212 ( using @xmath213 ) in the attractive case and with @xmath214 ( using @xmath215 and @xmath216 , plus a shift for clarity ) in the repulsive case using smaller @xmath168 system.,title="fig : " ] now , considering the expected quantum phase transition along the lines @xmath175 and @xmath29 , we can clearly identify different behaviors for attractive vs repulsive interactions ( as expected since there will be a finite spin gap or not respectively ) . in fig .
[ fig : entropies_ll](b ) for attractive interactions , we can fit our entanglement entropies data perfectly with @xmath82 using @xmath169 . in the repulsive case ,
our numerical data are not converged on the same size keeping up to @xmath217 states , so that we plot instead data obtained with @xmath168 and @xmath217 . in this case
, we can not remove entirely the oscillations in the data , but we can get a very good fit using the expected @xmath83 central charge . overall , we get an excellent agreement with the theoretical predictions .
we have presented a comprehensive study of the most general model relevant for one - dimensional @xmath0-band two - component fermionic cold gases with local interactions only .
we have concentrated here on incommensurate filling and quarter - filling . using a state - of - the - art low - energy approach , supplemented with a one - loop rg numerical analysis , we have found that generically , the charge sector decouples from the spin - orbital one which is gapped . as a consequence ,
most of the phase diagram is occupied by standard luttinger liquid phases with a single gapless charge mode ( hence a central charge @xmath145 ) .
nevertheless , we have clarified the nature of the dominant instability and have found that it is always of some bcs superconducting kind , in one of the following channels : bcs@xmath200 , bcs@xmath179 , bcs@xmath183 and bcs@xmath188 , see eqs .
( [ bcss])-([bcs-])-([bcs+])-([bcsd ] ) respectively .
in particular , an interesting bond - centered superconducting instability emerges along the harmonic line for the repulsive interaction .
the nature of the phase transitions between these four superconducting phases is also elucidated and found to behave with central charges @xmath82 or @xmath83 .
our numerical simulations do confirm that phases with dominant bcs@xmath179 , bcs@xmath183 and bcs@xmath188 extend from weak to strong coupling and occupy large regions in the phase diagram .
in particular , for attractive interactions and harmonic trapping , bcs@xmath183 correlations are the dominant ones , different from the @xmath96-cdw phase predicted earlier @xcite . in the last region where the bond - centered superconducting instability bcs@xmath200 is expected at weak coupling , our dmrg data at
quarter - filling ( i.e. , one particle per site ) indicate that a mott phase intervene with fully gapped bond - ordering waves at intermediate coupling , and spontaneous polarization at strong coupling .
we have also numerically confirmed the nature of all the quantum phase transitions present in this model .
the @xmath0-band two - component fermion mode , studied in this paper , is thus an interesting model to explore superconducting instabilities of doped two - leg fermionic ladder systems as well as the occurence of a mott transition .
given the recent progress in realizing su(@xmath194)-symmetric fermi gases @xcite , we hope that some of the phases discussed here could be realized experimentally and probed using local spectroscopy techniques or by measuring short - range correlations .
the authors would like to thank k. totsuka for a related collaboration and useful discussions .
numerical simulations have been performed using hpc resources from genci tgcc , genci idris ( grant x2015050225 ) and calmip ( grant 2015-p0677 ) .
the authors would like to thank cnrs for financial support ( pics grant ) . | the zero - temperature phase diagram of @xmath0-orbital two - component fermionic system loaded into a one - dimensional optical lattice is mapped out by means of analytical and numerical techniques .
it is shown that the @xmath0-band model away from half - filling hosts various competing superconducting phases for attractive and repulsive interactions . at quarter filling , we analyze the possible formation of incompressible mott phases and in particular for repulsive interactions , we find the occurrence of a mott transition with the formation of fully gapped bond - ordering waves . |
event - by - event fluctuations of kinematic characteristics and particle yields are believed to be one of the important probes to study the dynamics of heavy - ion collisions .
transverse momentum and multiplicity fluctuations are expected to be modified when the system approaches the phase boundary between hadron gas and quark - gluon plasma ( qgp ) .
it has been also argued that significant transverse momentum and multiplicity fluctuations should appear for systems hadronizing near the second - order critical qcd end - point @xcite .
the qcd phase diagram - ( @xmath1 ) , where @xmath2 is the temperature and @xmath3 bariochemical potential - can be scanned both by varying the system size / centrality and energy and therefore a possible nonmonotonic evolution of event - by - event ( @xmath4 and multiplicity ) fluctuations with beam energy , system size , or centrality may be used as an indication of the phase transition and the qcd critical point @xcite .
the na49 experiment at the cern sps studied both the system size dependence and the energy dependence of transverse momentum fluctuations . in the analysis of the na49 data
the @xmath5 fluctuation measure - proposed in @xcite - was used to quantify transverse momentum fluctuations on event - by - event basis ( the definition and some properties of the @xmath5 measure are presented in the next section and in @xcite ) .
the na49 results exhibit a significant nonmonotonic structure of transverse momentum fluctuations with the system size at the top sps energy @xcite .
qualitatively the same structure was found for all possible charge selections i.e. for all charged , negatively charged , and positively charged particles
. however , the preliminary na49 results on the energy dependence of transverse momentum fluctuations over the whole sps energy range manifests an unexpected effect : the @xmath0 measure shows a different behavior for different charge selections ( see @xcite and fig .
[ fipt_urqmd_onlyone ] ) .
the @xmath0 is independent of energy and it is consistent with zero for negatively charged particles , but it significantly increases for lower sps energies for both all charged and positively charged particles . in this work
it will be shown that the effect observed by the na49 experiment is connected with protons only , and it can be explained by event - by - event impact parameter fluctuations or - more precisely - by a correlation between the number of protons in the forward hemisphere and the number of protons that are closer to the production region .
the final conclusions from this analysis are important for all experiments measuring event - by - event fluctuations , especially for those that are using all charged particles in the analysis .
it is also important for detectors with acceptances extending to the beam / target spectator domain . due to a specific choice of the rapidity region ( midrapidity only ) the relativistic heavy ion collider ( rhic )
experiments appear to be rather insensitive to the effects of impact parameter fluctuations - even if the analysis is performed by use of all charged particles registered in the detectors .
the analysis presented in this article is based on preliminary na49 results @xcite and on events generated within the urqmd approach @xcite .
following the authors of @xcite one defines a single - particle variable @xmath6 with the overbar denoting averaging over a single - particle inclusive distribution .
further , one introduces the event variable @xmath7 , which is a multiparticle analog of @xmath8 , defined as @xmath9 where the summation runs over particles in a given event .
note , that @xmath10 , where @xmath11 represents averaging over events . finally , the @xmath0 measure is defined as @xmath12 where @xmath13 is the average multiplicity of the considered particles .
there are two most important properties of the @xmath0 measure .
if the system consists of particles that are emitted independently from each other ( no interparticle correlations ) @xmath0 equals zero . on the other hand ,
if a+a is an incoherent superposition of many independent n+n interactions ( superposition model ) , then @xmath0 is independent of centrality and it has the same value for a+a and n+n collisions . several effects may lead to nonzero value of @xmath0 . among them
are those that occur on an event - by - event basis ( event - by - event fluctuations of the inverse slope parameter , existence of different event classes , i.e. , `` plasma '' and `` normal '' events ) , but there are also interparticle correlations due to bose - einstein statistics , coulomb effects , resonance decays , flow , jet production , etc .
to study the energy dependence of transverse momentum fluctuations the na49 experiment used samples of central pb+pb collisions at 20@xmath14 , 30@xmath14 , 40@xmath14 , 80@xmath14 and 158@xmath14 gev energy , corresponding to @xmath15 = 6.27 , 7.62 , 8.73 , 12.3 , and 17.3 gev , respectively .
the fraction of the total inelastic cross section of nucleus+nucleus collisions ( @xmath16 ) was set as equal to 7.2% . in the analysis only tracks with @xmath17 gev / c were used .
for all five sps energies the forward - rapidity region was selected as @xmath18 , where @xmath19 is the particle rapidity calculated in the center - of - mass reference system ( particles were not identified in this analysis and their rapidities were calculated assuming the pion mass for all particles ) . within the studied rapidity
range the azimuthal angle acceptance of the na49 detector - ( @xmath20 ) - is not uniform and it is described by the analytical curves given in @xcite . additionally , it was demonstrated @xcite that at lower sps energies , the na49 tpc acceptance extends to the projectile spectator domain and therefore the sample of particles can be contaminated by beam particles . for energies above 80@xmath14 gev the beam region and the region of particles - accepted by the na49 for @xmath4 fluctuations analysis - do not overlap .
the preliminary na49 results on the energy dependence of transverse momentum fluctuations are presented in fig .
[ fipt_urqmd_onlyone ] ( points ) .
separate panels correspond to all charged particles , negatively charged and positively charged particles used in the analysis .
a significant decrease of transverse momentum fluctuations with increasing energy is observed for positively charged particles , whereas for the negatively charged ones @xmath0 values are independent of energy .
the @xmath0 values measured by the na49 experiment are compared to the predictions of the urqmd model @xcite .
the urqmd generator is a microscopic transport model producing hadrons via formation , decay , and rescattering of resonances and strings .
the urqmd approach simulates multiple interactions of both target / beam nucleons and newly produced particles , excitation , and fragmentation of color strings and the formation and decay of hadronic resonances . in this
analysis default parameters of the model have been used ( meson - meson and meson - baryon scattering included ) . for each energy , from the sample of minimum bias pb+pb events , the most central 7.2% interactions have been selected , similarly as in the real na49 events .
such selection corresponds to the impact parameter range @xmath21 fm ( in the generated histogram of impact parameter values the number of entries @xmath22 was set as proportional to @xmath23 [ fm ] ) . the na49 experiment used all charged particles , originating from the main vertex , to determine the @xmath0 measure .
it means practically that only main vertex pions , protons , and kaons and their antiparticles were used in the analysis , because particles coming from the decays of @xmath24 , @xmath25 , @xmath26 and @xmath27 are believed to be rejected by a set of track selection criteria .
therefore the analysis of the urqmd events has been carried out also by use of all charged pions , protons , and kaons and their antiparticles . in the analysis of the urqmd events the same kinematic and acceptance restrictions
have been applied as in the case of the na49 data . in principle , the selected @xmath4 - azimuthal angle acceptance ( @xmath28 ) is common for all five sps energies and it corresponds to the restrictions used in the na49 analysis ( for details see @xcite ) .
[ fipt_urqmd_onlyone ] presents the comparison of the @xmath0 versus energy for the preliminary na49 data ( points ) and for the urqmd model ( lines ) .
both the preliminary na49 data and the urqmd model show qualitatively the same behavior .
the urqmd model confirms a significant decrease of transverse momentum fluctuations with increasing energy for positively charged particles . on the other hand , there is no energy dependence of @xmath0 for negatively charged particles . in the previous article of the na49
@xcite the system size dependence of transverse momentum fluctuations was studied and a nonmonotonic structure of @xmath0 appeared , but the shape of this dependence was very similar for all three charge selections .
therefore , the fact that the energy dependence of @xmath0 is qualitatively different for various charge selections seems to be quite surprising . as the @xmath0 value for all charged particles
can be treated as a nontrivial combination of @xmath0 for negatively and positively charged particles ( additional sources of fluctuations , however , are not excluded ) , a main emphasis should be placed on understanding the qualitative difference between positively and negatively charged particles .
it should be also stressed that , - although the same kinematic and acceptance restrictions are used for the preliminary na49 data and for the urqmd events , - the @xmath0 values should not be directly compared in these two cases .
one of the reasons is a fact that the urqmd model does not include effects of short - range correlations ( bose - einstein and coulomb ) .
moreover , it has been already shown by the ceres experiment that the measured values of mean @xmath4 fluctuations , calculated by use of the urqmd model with default parameters , can be underestimated @xcite and that switching off secondary scattering ( in particular meson - baryon rescattering ) results in the values of @xmath4 fluctuations similar to those observed in the data .
it has been also checked that this underestimation ( when rescattering processes are included in the urqmd model ) can be even stronger for lower energies .
the above effect , observed by the ceres experiment , might help in a qualitative explanation of the underestimation of the urqmd @xmath0 values in fig .
[ fipt_urqmd_onlyone ] .
the similar structure of the energy dependence of the @xmath0 measure for the na49 data and for the urqmd model encourages us to search for an explanation of the origin of the increased @xmath0 values for positively charged particles at lower sps energies by analyzing effects incorporated in the model simulations .
although the track - by - track identification is not always possible in the data it is instructive to check - by use of the urqmd model - what is the particle content in the studied forward - rapidity region . the na49 sample of negatively charged particles is composed mainly of negative pions ( 94.4% of all negatives for 20@xmath14 gev and 89.4% of all negatives for 158@xmath14 gev ) and negatively charged kaons ( 5.6% for 20@xmath14 gev and 9.2% for 158@xmath14 gev ) , whereas the number of antiprotons can be treated as negligible ( below 0.1% for 20@xmath14 gev and 1.6% for 158@xmath14 gev ) .
the sample of positively charged particles is less homogeneous , as it contains positive pions ( 42.8% for 20@xmath14 gev and up to 69% for 158@xmath14 gev ) , positive kaons ( from 8.5% for 20@xmath14 gev to 10.4% for 158@xmath14 gev ) and protons ( 48.7% for 20@xmath14 gev and 20.6% for 158@xmath14 gev ) . within the na49 kinematic and acceptance region
the total multiplicity of the final state protons remains nearly constant for all five energies ( 11.1 - 13.2 ) , in contrary to the multiplicities of newly produced ( generated ) particles such as @xmath29 or @xmath30 .
therefore , the fraction of protons in a sample of positively charged particles is significantly higher for lower energies and thus one can expect that the increased @xmath0 values for lower energies might be indeed due to protons . the qualitative verification of this hypothesis is presented in fig . [ fipt_urqmd_pi_p_k ] where the @xmath0 values are presented separately for different types of final state particles .
the right panel of fig .
[ fipt_urqmd_pi_p_k ] shows that the @xmath0 values for protons only ( solid , thick curve ) are very close to those for all positively charged particles ( solid , thin curve ) thus confirming that protons indeed are responsible for the increased @xmath0 values at lower energies .
the @xmath0 measure obtained for newly produced particles such as pions , kaons ( positively and negatively charged ) or antiprotons is consistent with zero for all energies ( in agreement with the hypothesis of weak interparticle correlations ) .
although the precise track - by - track identification in the na49 experiment is much less reliable than the statistical one , the similar analysis with identified particles was prepared also for the na49 data . in order to identify particles
the combined information on the ionization energy loss ( @xmath31 ) and total momentum ( bethe - bloch curves ) was used .
the energy dependence of @xmath0 measure was calculated separately for charged pions and protons .
[ fipt_energy_identif_p_cut ] shows that the obtained na49 results are in qualitative agreement with those from the urqmd model , namely @xmath0 for pions ( all charged and negatively and positively charged separately : solid curves ) and for antiprotons ( dashed curve in the middle panel ) are consistent with zero , whereas @xmath0 versus energy for protons only ( dashed curve in the right panel ) follows a structure observed for protons in the urqmd model ( strong decrease with energy ) .
the quantitative comparison of @xmath0 values - for positively charged particles and for protons only - is rather difficult because bethe - bloch identification leads to random losses of particles .
the latter was found @xcite to reduce strongly the magnitude of correlations .
also an additional cut on total momentum ( @xmath32 gev / c ) , applied to eliminate the region of intersection of bethe - bloch curves , reduces the magnitude of @xmath0 values .
it has been already stressed in many articles , that a _ limited _ acceptance can influence the observed values of fluctuation measures .
( the na49 experiment measured @xmath4 fluctuations with a specific choice of azimuthal angle - ( @xmath20 ) curves defined in @xcite . ) therefore , as the next step , it should be verified whether the energy dependence of @xmath0 exhibits qualitatively the same structure when using complete azimuthal angle acceptance .
the dashed lines in fig .
[ fipt_urqmd_4pi ] represent the energy dependence of @xmath0 for forward - rapidity region , when no ( @xmath20 ) acceptance restrictions have been used .
it is evident that the qualitative structure of the energy dependence of @xmath0 is very similar also when using complete azimuthal angle acceptance .
the @xmath0 values have been also calculated using forward - rapidity ( anti-)protons only - also without ( @xmath20 ) acceptance restrictions ( solid lines in fig .
[ fipt_urqmd_4pi ] ) .
again , ( anti-)protons follow the behavior observed in fig .
[ fipt_urqmd_pi_p_k ] , thus indicating that the limited ( @xmath20 ) acceptance indeed does not influence the observed effects significantly . in the studied forward - rapidity region ( with @xmath17 gev / c cut ) the total multiplicity of final state protons is similar for all five energies and varies between 42.4 and 47.6 .
additionally , the @xmath0 values have been evaluated for the so - called ` @xmath33 ' acceptance which means that only transverse momentum cut ( @xmath17 gev / c ) has been used in the analysis ( dotted lines in fig .
[ fipt_urqmd_4pi ] ) . however , in this `` complete '' kinematic acceptance the observed effect of interest seems to disappear . instead
, very small but negative @xmath0 values are obtained for lower sps energies .
so far , one has learned that the effect of increased transverse momentum fluctuations arises probably from protons only .
now , one can try to establish whether those are newly produced protons , spectators or participants .
it has been shown in @xcite that for lower sps energies the na49 acceptance for @xmath4 fluctuations analysis extends to the beam spectator domain .
therefore a sample of particles used in the analysis can be contaminated by beam particles .
thus one can speculate that the increase of @xmath0 values may be somehow attributed to the presence of beam protons .
then , the measured correlation should be even stronger for the detector with complete ( ` @xmath33 ' ) acceptance because some of the beam particles - spectators or elastically rescattered protons - might hit into the midrapidity region or even into the backward hemisphere .
moreover , also participating neutrons ( as neutral particles are not registered in the detector ) should increase the total @xmath0 value .
this hypothesis can be easily tested by use of the urqmd events .
the results of these calculations are presented in fig .
[ fipt_urqmd_4pi_protons_neutrons ] , where @xmath0 is calculated for ` @xmath33 ' acceptance ( @xmath34 gev / c ) for protons only ( left ) and for protons and neutrons ( middle ) .
the total multiplicity of protons is very similar for all five energies and varies between 167.1 and 169.9 , whereas the total multiplicity of a mixture of protons with neutrons varies between 358.9 and 366.1 .
the very strong fluctuations observed for protons only ( @xmath0 at the level of 130 mev / c ) are exactly two times enlarged when neutrons are included .
however , the @xmath0 values for several selectios of newly produced particles such as pions or antiprotons remain very small even for ` @xmath33 ' acceptance ( fig .
[ fipt_urqmd_4pi_protons_neutrons ] - right ) , although the total multiplicity of all pions is much higher than the multiplicity of protons and neutrons together and strongly depends on the energy .
this result seems to indicate that the dominant effect originate not from the newly produced particles but from beam / target nucleons .
within the na49 acceptance ( forward rapidity ) the possible contamination from beam particles is the highest for 20@xmath14 gev interactions and therefore a test was carried out to check how the @xmath0 measure behaves when beam particles region is rejected .
the na49 results for 20@xmath14 gev pb+pb collisions are presented in fig . [ fipt_yproton_cut ] , where the @xmath0 measure was evaluated as a function of an upper @xmath35 cut ( @xmath35 is the particle rapidity calculated in the center - of - mass reference system assuming proton mass ) . in fig .
[ fipt_yproton_cut ] the preliminary na49 data are compared to predictions of the urqmd model with the same acceptance restrictions . it is observed - both in the data and in the urqmd model - that @xmath0 decreases when the number of rejected `` beam particles '' increases .
it has been also observed that although @xmath0 values measured with exclusion of the beam spectator region are low , the @xmath0 values obtained for `` spectator only '' region are also rather low , whereas the whole magnitude of @xmath4 fluctuations can be reproduced only by using the whole rapidity interval .
it suggests that the source of fluctuations is not concentrated at very forward rapidities only ( beam spectator domain ) but there might be a strong correlation between particles from beam spectator domain and particles at lower rapidities ( even in the production region ) .
the above effect has been observed both in the na49 data and in the urqmd events .
figure [ back_forward ] presents @xmath0 obtained for ` @xmath33 ' protons for the urqmd 20@xmath14 gev interactions .
the full points represent the increasing acceptance around the midrapidity region .
the open points correspond to the increasing `` gap '' around midrapidity and thus selection of particles closer to beam / target spectator regions .
indeed , it can be observed that the maximal magnitude of @xmath0 ( about 130 mev / c ) is obtained for a complete rapidity acceptance only , whereas @xmath0 values either in the spectator domain or in the midrapidity region are low .
the natural explanation of the effect described above consists in event - by - event impact parameter fluctuations or - more precisely - in correlation between the number of protons ( nucleons ) in the forward hemisphere and the number of protons ( nucleons ) in the production region . for more central events
the number of forward - rapidity protons is smaller , and consequently , the number of protons in the production region is higher .
the situation is opposite for less central collisions .
the existence of those different event classes results in the increased @xmath0 values for positively charged particles .
a test of the hypothesis of the correlation between protons in different rapidity windows can be found in fig .
[ pn_percent ] .
two kinds of the urqmd particle samples : protons only and a mixture of protons and neutrons were used to obtain the @xmath0 dependence on the fraction of the total inelastic cross section of nucleus+nucleus collisions ( @xmath16 ) .
the urqmd calculations correspond to ` 4@xmath36 ' acceptance in both 20@xmath14 and 158@xmath14 gev pb+pb interactions .
the urqmd results show a significant reduction of @xmath0 values when the centrality is restricted .
the original @xmath0 value for protons only at the level of 130 mev / c ( for 7.2% most central interactions at 20@xmath14 gev ) is reduced to the value consistent with zero when the centrality is restricted to 1% most central pb+pb interactions .
the similar effect was also observed within the na49 acceptance region .
figure [ fipt_percent_inel ] presents the @xmath0 values - for 20@xmath14 gev interactions - plotted as a function of the fraction of the total inelastic cross section ( @xmath16 ) .
the preliminary na49 data were compared to the predictions of the urqmd model with the same kinematic restrictions . both the na49 data ( points ) and the urqmd events ( lines ) show a significant reduction of @xmath0 values when the centrality is restricted to more central events .
the effect observed in fig .
[ pn_percent ] encourages us to study minimum bias interactions where impact parameter fluctuations from event to event should be maximal .
the @xmath0 values , obtained for minimum bias 20@xmath14 gev urqmd events ( 30k events ) , are demonstrated in table [ minbias ] .
results are shown for several particle selections at ` 4@xmath36 ' acceptance .
transverse momentum fluctuations for newly produced particles ( pions ) are consistent with zero , whereas for protons and for a sample of protons and neutrons - as it could be expected - exhibit extremely high @xmath0 values .
relatively high @xmath0 value can be also obtained when protons are combined with positively charged kaons and pions , whereas in the case of central data ( fig .
[ fipt_urqmd_4pi ] ) at ` 4@xmath36 ' acceptance , the presence of pions and kaons in a sample turned out to be sufficient to wash out the correlation originating from protons .
@xmath37 & @xmath38 @xmath39 & 1506 @xmath40 5.0 & 167.1 & 6.2 @xmath39 and @xmath41 & 2230.9 @xmath40 7.1 & 402.5 & 16.3 @xmath42 and @xmath43 & -0.3 @xmath40 1.4 & 157.5 & 171.8 @xmath43 & 2.2 @xmath40 1.5 & 83.5 & 90.7 @xmath42 & -1.2 @xmath40 1.4 & 74.1 & 81.3 @xmath44 & 1106.7 @xmath40 4.7 & 335.8 & 187.1 @xmath45 & 2.7 @xmath40 1.5 & 86.1 & 93.8 @xmath46 & 1290.6 @xmath40 4.1 & 249.6 & 93.3 figures [ back_forward ] and [ pn_percent ] together with the corresponding figures within the limited na49 acceptance ( fig .
[ fipt_yproton_cut ] and fig .
[ fipt_percent_inel ] ) seem to confirm the hypothesis that the increased @xmath0 values can be indeed connected with event - by - event fluctuations of the number of protons in the forward hemisphere and the number of protons that are closer to the production region .
there are at least two possible ways to eliminate this trivial source of correlations .
the first - and the most natural one - consists in centrality restrictions , however it can significantly reduce the event statistics .
the second one relies on applying an additional rapidity cut that would reject the beam spectator domain ( see the vertical lines in fig .
[ fipt_yproton_cut ] ) .
the na49 experiment chose the second ( rejection ) method and an additional @xmath47 cut was applied for each energy , namely for each produced particle its @xmath47 was required to be lower than @xmath48 , where @xmath49 is the beam rapidity in the center - of - mass reference system .
the final na49 results on the @xmath0 fluctuation measure , as a function of energy , are shown in fig .
[ fipt_energy1_rap_cut_3panels ] .
three panels represent all charged , negatively charged , and positively charged particles , respectively .
the points correspond to the data ( with statistical and systematic errors ) and the lines to the predictions of the urqmd model .
when additional cut on @xmath47 is applied , no significant energy dependence of @xmath0 measure can be observed for all three charge selections and @xmath0 values are consistent with the hypothesis of independent particle production ( that is @xmath50 ) .
the above analysis confirms that the preliminary na49 results , i.e. , the increased @xmath0 values at lower sps energies were indeed due to protons only .
the basic origin of this effect is the correlation between the number of protons ( generally nucleons ) in the forward hemisphere and the number of protons closer to the production region , caused by event - by - event impact parameter fluctuations . at the end of this analysis
it is worthwhile to find out what is the origin of the final - state protons in the urqmd model . in principle
, it should be clarified what kind of protons are rejected by an additional cut on @xmath47 , applied by the na49 experiment .
figure [ colln ] shows the distributions of the number of collisions of final state protons , generated by the urqmd model for the 7.2% most central pb+pb interactions at 20@xmath14 gev .
four panels correspond to different acceptance regions .
the peaks at zero ( upper panels ) represent real spectators , i.e. , protons that do not experience any collision .
their fraction is obviously higher when one restricts rapidity acceptance to the regions closer to the rapidity of the beam ( fig .
[ colln ] upper , right ) .
the mean number of collisions is generally smaller for the acceptance characteristic of the beam spectators ( right panels ) .
[ colln ] ( lower , left ) shows protons in the acceptance selected by the na49 experiment , whereas fig .
[ colln ] ( lower , right ) - protons that were finally excluded from the na49 acceptance by an additional cut on @xmath47 .
it is quite surprising that in the two lower panels there are no spectators ( @xmath51 ) at all .
it would suggest that the effect of increased @xmath0 values was connected with participants only and that spectators did not play any important role .
this is , however , true only inside the na49 acceptance and it will be shown below ( table [ tabhistory ] ) that when the ` 4@xmath36 ' acceptance is considered , both real spectators and participants , i.e. , elastically scattered protons ( precisely , the correlation between multiplicities of those two ) are responsible for the observed @xmath0 values . during the collision process
the projectile and target nucleons participate both in elastic and in inelastic interactions . in principle , @xmath52 resonances can be produced as an effect of inelastic collisions .
afterward , @xmath52s decay into pions and nucleons .
technically - within the urqmd model - for each final state particle one can attribute an integer number called the `` i d of parent process '' , which describes the last process the particle was involved in .
therefore , it is easy to disentangle spectator protons from those originating from elastic or inelastic processes .
figure [ history ] presents the distributions of i d of parent process ( last process the particle was involved in ) obtained for the final - state protons generated in the urqmd model for the 7.2% most central pb+pb interactions at 20@xmath14 gev .
different panels correspond to different acceptance regions - similarly to fig .
[ colln ] .
below , one can find a list of ids that significantly contribute to the histogram .
the i d of parent process equal zero denotes real spectators ( @xmath51 ) but also elastically scattered protons ( scattering with all barions and mesons is possible ) . for protons with i
d of parent process equal zero the number of collisions ( @xmath53 ) has a high peak at zero ( real spectators ) and a tail up to about 10 - 15 , whereas the mean value of @xmath53 is about 0.5 .
all protons originating from decays ( for example , from resonances ) has i d of parent process equals 20 . as we are considering central collisions ( 7.2% )
the majority of protons come from decays with exception of very forward - rapidity regions where spectators begin to dominate ( fig .
[ history ] upper , right ) .
the i d of parent process equals 1 means protons generated in the reaction @xmath54 , ( nucleon ) , @xmath55 , ...
( nucleon resonances ) , @xmath56 i d equals 14 means protons produced in inelastic interactions ( without string excitation ) such as : @xmath57 , or @xmath58 .
the parent history i d labeled as 30 denotes process @xmath59 . and , finally , ids equal 15 and 27 correspond to processes through 2 strings and process through 1 string , respectively
. table [ tabhistory ] shows the @xmath0 values for ` 4@xmath36 ' protons obtained for urqmd 7.2% most central pb+pb interactions at 20@xmath14 gev , where several choices of number of collisions ( @xmath53 ) or i d of parent process have been applied .
the @xmath0 value obtained for spectators ( @xmath51 ) is low but it is low also when the spectators are rejected .
the original magnitude of @xmath0 ( about 130 mev / c ) can be reproduced only when both spectators and `` not spectators '' are used in the analysis .
it simply indicates that there should be some kind of correlation between spectators and `` not spectators '' .
the same observation can be achieved for i d of parent process equal and not equal zero .
the opposite situation takes place when protons originating from decays are considered .
the @xmath0 value evaluated for protons coming from decays is rather low : however , if such protons are rejected from the analysis , - the @xmath0 does not decrease but even increases !
it means that protons from decays ( i.e. newly produced particles ) do not generate the dominant correlation in the analysis .
it is in agreement with intuitive predictions , because also newly produced particles such as pions , kaons , or antiprotons lead to @xmath0 value consistent with zero .
finally , it has been observed that not only protons from decays but also protons created in a list of inelastic processes ( labeled as 1 , 14 , 15 , 27 , 30 ) result in a @xmath0 values consistent with zero .
therefore it can be concluded that the whole effect of the increased transverse momentum fluctuations comes from beam / target particles only ( both real spectators and elastically scattered beam / target protons ) .
those particles can hit different rapidity windows - depending on the centrality of a given event .
consequently , the sample composed by a mixture of events with different impact parameters generates increased event - by - event @xmath4 fluctuations .
the impact parameter range in the 7.2% most central pb+pb interactions is wide enough to produce significantly increased @xmath0 values .
@xmath37 & @xmath38 . [ cols="^ " , ] & 2.7 @xmath40 1.1 & 14.0 & 4.3
it has been presented how event - by - event impact parameter fluctuations influence transverse momentum fluctuations . in principle , it has been explained that the preliminary na49 results - showing significantly increased @xmath0 values at lower sps energies - are connected with the trivial effects of impact parameter fluctuations .
it suggests that a precise selection of very narrow centrality bins should be crucial not only for the analysis of multiplicity fluctuations @xcite but also for transverse momentum fluctuations and probably for other kinematic characteristics .
it was predicted that either approaching the phase boundary of the qcd phase diagram or hadronization close to the critical point should lead to increased values of transverse momentum fluctuations .
therefore it is very important to distinguish between interesting physical effects and trivial ones as those connected with impact parameter fluctuations .
below , one can find several methods to eliminate the effects of event - by - event impact parameter fluctuations . 1
. one can perform the analysis for all three charge selections ( all charged , negatively charged , positively charged ) and check whether the same structure ( not necessarily the magnitude ! ) of @xmath0 versus energy or centrality ( system size ) appears for all possible charge selections ( see the na49 results in @xcite ) .
if the previous method is not possible it is safer to restrict the analysis to negatively charged particles only , instead of all charged or positively charged ones .
3 . the experiments with reliable particle identification can try to reject protons from the sample and measure @xmath0 for newly produced particles only ( pions or a mixture of pions , kaons , etc . ) .
4 . one can also probe the midrapidity region only ( production domain ) or at least one can check whether the selected acceptance does not extend to the beam / target spectator domain .
the best and the least controversial method relies on a drastic centrality restriction
. however , due to insufficient statistics , this method can not be applied by every experiment .
it has been already mentioned that due to the specific choice of the rapidity region ( midrapidity only ) the rhic experiments are essentially free of the effects of impact parameter fluctuations , even though in most cases the results are obtained by use of all charged particles registered in the detectors .
the same remark concerns also the ceres experiment at cern sps that measured transverse momentum fluctuations in 40@xmath14 , 80@xmath14 , and 158@xmath14 gev pb+au interactions by use of all charged particles - but produced in midrapidity region only @xcite . in contrary , the acceptance of the na49 experiment - selected for the analysis of transverse momentum fluctuations - extends to the beam spectator domain . however , the na49 experiment performs the analysis by use of all three charge selections and therefore it was possible to separate and finally eliminate the trivial effects connected with the impact parameter fluctuations .
i am indebted to the authors of the urqmd model for the permission to use their code in my analysis .
i am very grateful to marek gadzicki for his interesting comments and suggestions concerning this analysis .
i also appreciate the editorial remarks from ewa skrzypczak and staszek mrwczyski .
finally i thank jarek grebieszkow for the software support . | the preliminary na49 results on the energy dependence of transverse momentum fluctuations over the whole super proton synchrotron ( sps ) energy range exhibit an unexpected effect .
the @xmath0 fluctuation measure - used by the na49 experiment - manifests a different behavior for different charge combinations .
whereas the @xmath0 is consistent with zero and independent of energy for negatively charged particles , it significantly increases for both all charged and positively charged particles at lower sps energies .
the string - hadronic ultrarelativistic quantum molecular dynamics model ( urqmd ) is applied here to explain this effect .
the urqmd simulations show that the number of protons is strongly correlated with impact parameter and that the event - by - event impact parameter fluctuations are responsible for the event - by - event transverse momentum fluctuations of positively charged and all charged particles where protons are included .
the observations presented in this article are important for all experiments measuring event - by - event fluctuations , especially for those using all charged particles in the analysis .
the results can be also crucial for detectors with acceptances extending to the beam / target spectator domains .
* influence of impact parameter fluctuations * * on transverse momentum fluctuations * * katarzyna grebieszkow * + warsaw university of technology + e - mail : |
in this paper we discuss the measurement of the w mass from direct reconstruction of the invariant mass of the decay products in the channels ww@xmath3qqqq ( 4q ) and ww@xmath4qq .
preliminary results are presented for data collected in aleph@xcite during 1998 with an integrated luminosity of 174.2@xmath5 at 188.63gev .
events are preselected to remove radiative returns to the z and clustered into four jets using the ` durham - pe ` algorithm , defining @xmath6 .
events are vetoed if a charged track in a jet carries more than 90% of the jet energy or if there is more than 95% of electromagnetic energy in a @xmath7 cone around any particle .
a neural network with fourteen input variables ( nn14 ) is used to perform the final selection .
training is performed with an independent sample of the standard monte - carlo ( ` koralw`@xcite hadronised with ` jetset`@xcite ) and comparable samples of qq@xmath8 and zz events , both generated using ` pythia`@xcite , to simulate the background .
events with nn@xmath9 are used to extract the w mass@xcite .
the total charged energy and multiplicity are used to preselect events with a further cut on the total longitudinal momentum and visible energy to remove radiative returns to the z. the lepton candidates are chosen to be more energetic and isolated than the other charged tracks and are identified as an electron or muon in the detector .
the energy of electron candidates are corrected for possible bremsstrahlung photons detected in the electromagnetic calorimeter .
the ` durham - pe ` algorithm is used to force two jets from objects not used to reconstruct the lepton defining @xmath12 .
a probability for an event to come from a signal process is determined from monte carlo reference samples using the lepton energy and isolation and the event total transverse momentum .
selected events are required to have a probability greater than 0.4 .
a similar preselection to that used for electron and muon events is applied with additional constraints to remove events with i ) energy around the beam line , ii ) isolated , energetic photons and iii ) those already selected as electron or muon candidate events .
a tau jet is constructed from one or three charged tracks and two other jets are forced using the ` jade ` algorithm .
the tau jet must be that jet most anti - parallel to the missing momentum vector and isolated from the other jets . as with electron and muon candidate events , a probability is calculated and the same cut applied .
to improve the invariant mass resolution a four constraint kinematic fit is applied to the four jets to conserve energy ( as provided by lep ) and momention .
corrections are applied to the jets to take into account particle losses in the detector .
the masses from the three combinations of di - jets formed from the fitted jets are rescaled according to @xmath14 , where @xmath15 and @xmath16 are the measured jet energies .
a pairing algorithm is applied to the di - jet combinations to select that which most closely corresponds to a ww pair .
the combination with the smallest mass difference between rescaled masses is chosen provided one mass lies in the window 60 to 86 gev/@xmath2 and the other 74 to 86 gev/@xmath2 .
a binned monte - carlo reweighting procedure is employed to find the value of @xmath0 which best fits the mass distributions .
events from a ` koralw`@xcite sample with equivalent background are reweighted with a cc03 matrix element to provide a two - dimensional probability density function for the minimisation using a single parameter , @xmath0 .
variable binning controlled by the density of monte - carlo is employed optimised to produce a stable result .
the w - width is allowed to vary with @xmath0 according to the standard model . in order to check that the procedure does not introduce any biases ,
monte carlo samples are generated with @xmath0 in the range 79.35 and 81.35 gev/@xmath2 . treating these samples as data , no significant offsets in the masses measured are found .
a two constraint fit@xcite is applied to each event by minimising a @xmath18 constructed from the deviations of selected parameters of the jets and leptons from their true values and demanding that the hadron and lepton invariant masses are equal .
the single fitted mass obtained for each event must lie in the window 74 to 94.5 gev/@xmath2 . a reweighting procedure similar to that employed in the four quark channel is used to fit the mass distribution .
fixed binning is used for the electron and muon channels and variable binning is retained for the tau channel .
@p5.0cmlllll source & 4q & e & @xmath19 & @xmath20 + + calorimeter calibrations & 30 & 27 & 14 & 19 + charged particle tracking & & 7 & 3 & 3 + jet corrections & 8 & 14 & 4 & 7 + parton fragmentation & 35 & 25 & 25 & 30 + initial state radiation & 10 & 5 & 5 & 5 + lep energy & 17 & 17 & 17 & 17 + + reference mc statistics & 10 & 16 & 15 & 23 + background contamination & 10 & 8 & 1 & 25 + colour reconnection & 25 & & & + bose - einstein effects & 50 & & & + total & 77 & 47 & 37 & 53 + +
the systematic errors summarised in table 1 .
uncertainties due to the detector are determined using data taken at the z at intervals during the high energy running .
particles not seen by the detector cause discrepancies in the jet finding , the effect is estimated by matching jets built from monte carlo tracks before and after passing them through the detector simulation .
the error due to parton fragmentation effects is measured by determining the mass shift when ` herwig`@xcite rather than ` jetset ` is used for hadronisation .
the effect of initial state radiation is estimated by comparing the use of 1st and second order matrix elements .
the lep energy error is that given by lep .
colour reconnection between parton pairs in the qqqq channel is studied using variants on ` jetset ` or ` herwig ` .
the bose - einstein effect has been studied using z - peak data@xcite , the systematic error is determined from the shift in the mass measured when the parameters obtained at the z peak are applied to monte carlo events using a modified ` jetset ` .
this effect is applied between to w decay products in the qqqq channel . .w mass measurement results at 189 gev [ cols="<,<,<,<,<,<",options="header " , ] +
the mass distributions obtained are shown in figure 1 .
the preliminary results for 189 gev are given in table 2 for each channel where the be - cr error is that for the bose - einstein and colour reconnection systematics added in quadrature .
results for data taken at 172 and 183 gev have been published @xcite , combining these and the results above gives @xmath21 = 80.561 @xmath1 0.095(stat . )
@xmath1 0.050(syst . )
@xmath1 0.056(be - cr ) gev/@xmath2 @xmath22 = 80.343 @xmath1 0.089(stat . )
@xmath1 0.041(syst . )
gev/@xmath2 . 9 aleph collaboration , _ aleph : a detector for electron - positron annihilations at lep _ , nucl inst .
a294 ( 1990 ) 121 .
m.skrzypek , s.jadach , w.placzek and z.was , comp.phys.commun .
94 ( 1996 ) 216 .
t.sjstrand , comp.phys.commun .
82 ( 1994 ) 74 .
aleph collaboration , _ measurement of w - pair production in @xmath23 collisions at 189 gev _ ,
contribution to hep99 , abstract 6_405 ; http//alephwww.cern.ch/. aleph collaboration , _ measurement of the w - mass by direct reconstruction in @xmath23 collisions at 172 gev _
, phys.lett .
b 422 ( 1998 ) 384 . aleph collaboration , _ measurement of the w - mass in @xmath23 collisions at 183 gev _ , phys.lett .
b 453 ( 1999 ) 121 .
g.marchesini et al .
, , comp.phys.commun . 67 ( 1992 ) 465 .
aleph collaboration , _ measurement of the w mass in @xmath23 collisions at production threshold _ ,
b 401 ( 1997 ) 347 .
aleph collaboration , _ bose einstein correlations in w pair decays _
, contribution to hep99 , abstract 1_388 ; http//alephwww.cern.ch/. | the w - boson mass has been measured using the aleph detector at lep .
preliminary results from data taken in 1998 are added to previous measurements to give @xmath0
= 80.411 @xmath1 0.064(stat . )
@xmath1 0.037(syst . )
@xmath1 0.022(be - cr ) @xmath1 0.018(lep ) gev/@xmath2 .
glas - ppe/199919november 1999 |
seeger s mass formula @xcite was given in 1961 , with its constants fitted to ground - state ( g.s ) binding energies of some 488 nuclei available at that time .
the temperature t - dependence of these constants was later introduced by davidson _
@xcite on the basis of thermodynamical considerations of the nucleus .
these constants , however , need be fitted again since a large amount of data on experimental g.s .
binding energies @xcite , and their theoretically calculated values @xcite for , not - yet observed , neutron- and proton - rich nuclei have now become available .
furthermore , the t - dependence of the constants , in particular the pairing constant @xmath9 , need be looked in to because of their recent un - successful use in calculating the decay properties of some excited compound nuclear systems @xcite-@xcite .
note that our aim here is not to obtain a new set of constants for seeger s mass formula , but simply to include the t - dependence on experimental binding energies @xmath6 .
for this purpose , a readjustment of only two of the four constants , the bulk constant @xmath0 and the neutron - proton asymmetry constant @xmath1 , are enough to obtain the @xmath6 within @xmath21.5 mev .
a similar job was first done in @xcite for nuclei up to z=56 , and then in @xcite up to z=97 , but is redone here with an improved accuracy and up to z=118 .
thus , the domain of the work is extended to neutron - deficient and neutron - excess nuclides where @xmath6 are not available , but theoretical binding energies @xmath7 are available @xcite .
these re - fitted constants have been successfully used in the number of recent calculations @xcite-@xcite for studying the decay of hot and rotating compound nucleus ( cn ) formed in heavy ion reactions over a wide range of incident centre - of - mass ( c.m . )
energies .
a brief outline of the seeger s mass formula , and the methodology used to workout the temperature - dependent binding energies , are presented in section ii .
possible applications of the liquid drop energy in heavy ion reaction studies are also included in this section .
the calculations and results are given in section iii , together with the table of fitted constants , which could be of huge importance for people working in the relevant area of nuclear physics .
finally , the results are summarized in section iv .
according to the strutinsky renormalization procedure , the binding energy @xmath10 of a nucleus at temperature t is the sum of liquid drop energy @xmath11 and shell corrections @xmath12 @xmath13 where @xmath3 is the semi - empirical mass formula of seeger @xcite , with t - dependence introduced by davidson _
@xcite , and @xmath5 taken as the `` empirical '' formula of myers and swiatecki @xcite , also made t - dependent to vanish exponentially , @xmath14 with @xmath15=1.5 mev @xcite .
seeger s liquid drop energy @xmath3 , with its t - dependence due to davidson _
_ , is @xmath16{}\nonumber\\ & & + \bigl(\frac{z^{2}}{r_0(t)a^{\frac{1}{3}}}\bigr ) \bigl[1-\frac{0.7636}{z^{\frac{2}{3}}}-\frac{2.29}{[r_0(t)a^{\frac{1}{3}}]^{2}}\bigr]{}+\delta(t){\frac{f(z , a)}{a^{\frac{3}{4 } } } } , \label{eq:3}\end{aligned}\ ] ] with @xmath17 and , respectively , for even - even , even - odd and odd - odd nuclei , @xmath18 seeger s constants of 1961 are @xcite : @xmath19 and the pairing energy @xmath20=33.0 mev from @xcite . in the following , the bulk constant @xmath21(0 ) , and the neutron - proton asymmetry constant @xmath1 , are found enough to be readjusted/ refitted to obtain the @xmath6 .
[ = @xmath22 in mev as a function of neutron number n for z=97 , calculated by using the experimental data ( solid circles ) @xcite , theoretical data ( open circles ) @xcite , with newly fitted constants ( crosses and down open triangles ) and with the 1961 seeger s constants @xcite ( hollow squares ) . ] ) ] calculated for the decay of @xmath23ni@xmath24 formed in @xmath25s+@xmath26 mg reaction at t=3.60 mev for @xmath27=0 and 36 @xmath28 , and also at t=0 for @xmath27=0 @xmath28 . ] the t - dependence of the constants in eq . ( [ eq:3 ] )
were obtained numerically by davidson _
@xcite from the available experimental information on excited states of 313 nuclei in the mass region 22@xmath29250 by determining the partition function @xmath30 of each nucleus in the canonical ensemble and making a least squares fit of the excitation energy @xmath31 to the ensemble average @xmath32 the constants @xmath33 , @xmath34 , @xmath35 , @xmath36 and @xmath9 are given in fig .
@xcite for t@xmath374 mev , extrapolated linearly for higher temperatures .
however , @xmath9 is constrained to be positive definite at all temperatures , and with @xmath9=0 for t@xmath382 mev .
also , for the bulk constant @xmath33 , instead , an empirically fitted expression using fermi gas model is obtained , as @xmath39 for shell effects @xmath5 , the empirical formula of myers and swiatecki @xcite is @xmath40 \label{eq:7}\ ] ] where @xmath41 with @xmath42 or @xmath43 , and @xmath44 .
@xmath45 are the magic numbers 2 , 8 , 14 ( or 20 ) , 28 , 50 , 82 , 126 and 184 for both neutrons and protons .
the constants @xmath46=5.8 mev and @xmath47=0.26 mev . note that the above formula is for spherical shapes , but the missing deformation effects in @xmath5 are included here to some extent via the readjusted constants of @xmath3 since we essentially use the experimental binding energies split in to two contributions , @xmath3 and @xmath5 , for reasons of adding the t - dependence on it . finally , as an application of the two components [ @xmath11 and @xmath12 ] of the ( t - dependent ) experimental binding energy in the field of heavy - ion reactons , we define the collective fragmentation potential @xmath49+\sum_{i = 1}^{2}[\delta u_i]\exp(-t^2/t^2_0){}+v_c(r , z_i,\beta_{\lambda i},\theta_i , t)\nonumber\\ & & + v_p(r , a_i,\beta_{\lambda i},\theta_i , t){}+v_{\ell}(r , a_i,\beta_{\lambda i},\theta_i , t ) , \label{eq:8}\end{aligned}\ ] ] where the nuclear proximity @xmath50 , coulomb @xmath51 and the angular - momentum @xmath27-dependent @xmath52 potentials are for deformed and oriented nuclei and are also t - dependent . for details , see , e.g. , ref .
@xcite . based on @xmath53 at fixed r and t , and the scattering potential @xmath54 at fixed @xmath55 and @xmath56
, we calculate the cn decay cross - section by using the dynamical cluster - decay model ( dcm ) of gupta and collaborators @xcite-@xcite , worked out in terms of the decoupled collective coordinates of mass ( and charge ) asymmetry @xmath57 [ @xmath58 and relative separation r. in terms of these coordinates , using @xmath27 partial waves , the cn decay cross section is defined as @xmath59 where the preformation probability @xmath60 , refering to @xmath55 motion , is the solution of stationary schrdinger equation in @xmath55 at a fixed r , and @xmath61 , the wkb penetrability refers to r motion , both quantities carrying the effects of angular momentum @xmath27 , temperature t , deformations @xmath62 and orientations @xmath63 degrees of freedom of colliding nuclei with c.m .
energy @xmath64 .
@xmath65m$ ] , is the reduced mass , with m as the nucleon mass .
( [ eq:9 ] ) is applicable to the decay of cn to light particles ( lps , a@xmath374 , z@xmath372 ) , intermediate mass fragments ( imfs , 2@xmath37z@xmath3710 ) , the fusion - fission fragments and the quasi - fission ( q.f . )
process where the incoming channel does not loose its identity , i.e. , @xmath66=1 for qf .
the @xmath67 could be fixed for the vanishing of the fusion barrier of the incoming channel , or the light particle cross - section @xmath680 , or else defined as the critical @xmath69}/\hbar$ ] . , using eq .
( [ eq:9 ] ) , for the decay of compound system @xmath23ni@xmath24 formed in @xmath25s+@xmath26 mg reaction at t=3.60 mev , taking pairing constant @xmath70=0 and 9.5 mev , compared with experimental data @xcite . ] ( mev ) as function of temperature t ( mev ) , readjusted empirically for temperatures [email protected] mev ( solid line and solid dots ) , compared with the original curve ( dashed line ) due to davidson _
_ @xcite . ]
table 1 gives the newly fitted constants of seeger s
@xmath3 for the experimental binding energy @xmath6 @xcite , and the theoretical @xmath7 values @xcite where the experimental data were not available .
interestingly , only the bulk constant @xmath0 , working as an overall scaling factor , and the asymmetry constant @xmath1 , controlling the curvature of the experimental parabola , are required to be re - adjusted .
the role of these re - fitted constants is illustrated in fig .
[ fig:1 ] for z=97 nuclides .
we notice in fig .
[ fig:1 ] an excellent agreement between the present fits ( crosses and down open triangles ) corresponding to experimental ( solid circles ) @xcite and theoretical data ( open circles ) @xcite , respectively .
the fits are with in 0 - 1.5 mev of the available @xmath6 or @xmath7 data . also plotted in fig .
[ fig:1 ] are the results of calculations using the old 1961 seeger s constants ( hollow squares ) , showing the requirement and extent to which the fitting can clearly improve upon the older results . next , we consider an application of the re - adjusted @xmath3 with an idea to impress upon the need and to propose here atleast a partially modified variation of the pairing constant @xmath70 with temperature t , as compared to that of davidson _ et al .
[ fig:2 ] shows the fragmentation potential v(a ) for the decay of @xmath23ni@xmath24 ( a complete mass spectrum ) into light particles ( lps ) and intermediate mass fragments ( imfs ) at t=3.60 mev for two different @xmath27 values ( @xmath27=0 and 36 @xmath28 ) , compared with one at t = 0 for @xmath27=0 @xmath28 .
we notice that at t = 0 for @xmath27=0 @xmath28 , the pairing effects are very strong since all the even - even fragments lie at potential energy minima . on the other hand ,
if we include temperature effects as per prescription of davidson _ et al . _
( dashed line in fig . [ fig:4 ] ) , we find that @xmath70=0 mev in @xmath3 for t@xmath382 mev , and hence in fig .
[ fig:2 ] for t=3.60 mev , @xmath70=0 mev , the odd - odd fragments like @xmath71b , @xmath72n , @xmath73f , etc .
, become equally probable as the even - even fragments , since minima are now equally stronger .
the same result was obtained earlier in @xcite for the decay of @xmath23ni@xmath24 at t=3.39 mev , since there too @xmath70=0 mev was used from davidson _
et al._. however , if we empirically choose @xmath70=9.5 mev for t=3.60 mev ( for the best fit to imfs data in fig .
[ fig:3 ] ) , the situation becomes again favourable .
in other words , fig .
[ fig:2 ] for t=3.60 mev , @xmath70=9.5 mev shows once again that the even - even fragments , like @xmath74c , @xmath75o , etc .
, are equally favoured as odd - odd @xmath72n , @xmath73f , etc .
it is important to note that in this experiment @xcite on @xmath25s+@xmath26mg@xmath76ni@xmath77 , only the imfs are measured , and theoretically lps are more prominent at lower @xmath27-values whereas imfs seem to supersede them at higher @xmath27-values , as is also evident from fig .
[ fig:2 ] .
the calculated decay cross - sections @xmath78 for imfs at t=3.60 mev , for both @xmath70=0 and 9.5 mev cases are shown in fig .
[ fig:3 ] , compared with experimental data @xcite .
we notice in this figure that better comparisons are obtained for the case of @xmath790 calculations , contrary to earlier results in fig .
13 of @xcite for @xmath70=0 mev , but supporting the one in fig . 7 of @xcite for @xmath70=9.5 mev .
similar calulations , supporting non - zero @xmath70 values at t@xmath382 mev , are also reported for @xmath23ni@xmath24 at t=3.39 mev in fig .
7 of @xcite , and for fusion - fission cross - section in @xmath80ba ( fig .
2(b ) in @xcite ) , and the possible @xmath72c clustering in @xmath81o and @xmath82ne nuclei @xcite . these calculations
lead us to modify the variation of @xmath70 as function of t , as shown in fig .
[ fig:4 ] ( solid line throgh solid dots ) .
apparently , many more calculations are needed for fig .
[ fig:4 ] to represent a true @xmath9 . 0.5 mm [ cols="^,^,^,^,^,^,^,^,^,^,^,^,^,^,^,^",options="header " , ]
in view of the large data for ground state ( g.s . ) binding energies having become available and to be able to include the t - dependence on binding energies , we have re - fitted two of the constants , the bulk @xmath0 and neutron - proton asymmetry @xmath1 , of seeger s mass formula . the experimental g.s .
binding energies or theoretical binding energies for neutron- and proton - rich nuclei , where data are not yet available , are fitted within @xmath21.5 mev , and up to z=118 nuclei .
the method used is the strutinsky renormalization procedure to define the g.s . binding energy as a sum of the liquid drop energy and the shell correction .
taking shell correction from the empirical formula of myers and swiatecki , the two constants of seeger s liquid drop energy are fitted to obtain the experimental or theoretical binding energy .
the fitted constants of liquid drop energy have been used for understanding the dynamics of excited compound nuclear systems , which point out to the inadequacy of the variation of pairing energy constant @xmath70 with temperature t. as per the given @xmath8 variation of davidson _ et al .
_ , @xmath70=0 mev for t@xmath382 mev .
however , the recent compound nucleus decay calculations suggest that @xmath790 for t@xmath38 2 mev and hence clearly indicate the need for re - evaluation of the t - dependence of seeger s constants .
a new dependence of @xmath8 is suggested on the basis of already published calculations for compound nucleus decay studies .
need for further studies are clealy indicated .
the authors are thankful to prof .
s. k. patra for his interest in this work
. m.k.s . is thankful to csir , new delhi , and r.k.g . to department of science and technology , govt . of india , for financial support for this work in the form of research projects .
p. a. seeger , nucl . phys . * 25 * , ( 1961 ) 1 .
n. j. davidson , s. s. hsiao , j. markram , h. g. miller and y. tzeng , nucl .
a * 570 * , ( 1994 ) 61c .
g. audi , a.h .
wapstra and c. thiboult , nucl .
a * 729 * , ( 2003 ) 337 .
p. moller , j. r. nix , w. d. myers and w. j. swiatecki , at .
. data tables * 59 * , ( 1995 ) 185 .
r. k. gupta , s. k. arun , r. kumar and niyti , int .
( irephy ) * 2 * , ( 2008 ) 369 .
r. kumar and r.k .
gupta , phys .
c * 79 * , ( 2009 ) 034602
. m. bansal , r. kumar and r. k. gupta , contri .
efes - in2p3 conf . on `` many body correlations
: from dilute to dense nuclear systems '' , feb . 15 - 18 , 2011 ,
m. balasubramaniam , r. kumar , r. k. gupta , c. beck and w. scheid , j. phys .
* 29 * , ( 2003 ) 2703 .
b. b. singh , m. k. sharma and r. k. gupta , phys .
c * 77 * , ( 2008 ) 054613 .
r. k. gupta , r. kumar , n. k. dhiman , m. balasubramaniam , w. scheid and c. beck , phys .
c * 68 * , ( 2003 ) 014610 .
r. k. gupta , m balasubramaniam , r. kumar , d. singh and c. beck , nucl .
a*738 * , ( 2004 ) 479c .
r. k. gupta , m. balasubramaniam , r. kumar , d. singh , c. beck and w. greiner , phys .
c * 71 * , ( 2005 ) 014601 .
r. k. gupta , m. balasubramaniam , r. kumar , d. singh , s. k. arun and w. greiner , j. phys .
g : nucl . part . phys . * 32 * , ( 2006 ) 345 .
b. b. singh , m. k. sharma , r. k. gupta and w. greiner , int .
e * 15 * , ( 2006 ) 699 .
r. k. gupta , s. k. arun , d. singh , r. kumar , niyti , s. k. patra , p. arumugam and b.k .
sharma , int .
e * 17 * , ( 2008 ) 2244 .
r. k. gupta , niyti , m. manhas , s. hofmann and w. greiner , int .
e * 18 * , ( 2009 ) 601 .
s. k. arun , r. kumar and r.k .
gupta , j. phys .
phys . * 36 * , ( 2009 ) 085105 .
s. kanwar , m. k. sharma , b. b. singh , r. k. gupta and w. greiner , int .
e * 18 * , ( 2009 ) 1453 .
r. k. gupta , niyti , m. manhas and w. greiner , j. phys .
phys . * 36 * , ( 2009 ) 115105 .
r. k. gupta , s. k. arun , r. kumar and m. bansal , nucl .
a * 834 * , ( 2010 ) 176c .
m. k. sharma , g. sawhney , s. kanwar and r. k. gupta , mod .
a * 25 * , ( 2010 ) 2022 .
r. k. gupta , _ clusters in nuclei _ , lecture notes in physics , * 818 * , ( 2010 ) 223 , ed .
c. beck , springer - verlag berlin heidelberg .. m. k. sharma , s. kanwar , g. sawhney and r. k. gupta , aip conf . proc . * 1265 * , ( 2010 ) 37 .
niyti , r. k. gupta and w. greiner , j. phys .
phys . * 37 * , ( 2010 ) 115103 .
w. myers and w. j. swiatecki , nucl .
a * 81 * , ( 1966 ) 1 .
a. s. jensen and j. damgaard , nucl .
a * 203 * , ( 1973 ) 578 .
s. debenedetti , _ nuclear interactions _ , new york , wiley , ( 1964 ) .
s. j. sanders , d. g. kovar , b. b. back , c. beck , d. j. henderson , r. v. f. janssens , t. f. wang , and b. d. wilkins , phy .
c * 40 * , ( 1989 ) 2091 ; phys .
* 59 * , ( 1987 ) 2856 . | seeger s semi - empirical mass formula is revisited for two of its constants ( bulk constant @xmath0 and neutron - proton asymmetry constant @xmath1 ) readjusted to obtain the ground - state ( g.s . )
binding energies of nuclei within a precision of @xmath21.5 mev and for nuclei up to z=118 .
the aim is to include the temperature t - dependence on experimental binding energies , and not to obtain the new parameter set of seeger s liquid drop energy @xmath3 .
our proceedure is to define the g.s .
binding energy @xmath4 , as per strutinsky renormalization procedure , and using the empirical shell corrections @xmath5 of myers and swiatecki , fit the constants of @xmath3 to obtain the experimental binding energy @xmath6 or theoretically calculated @xmath7 if data were not available .
the t - dependence of the constants of @xmath3 , is introduced as per the work of davidson _ et al .
_ , where the pairing energy @xmath8 is modified as per new calculations on compound nucleus decays .
the newly fitted constants of @xmath3 at t=0 are made available here for use of other workers interested in nuclear dynamics of hot and rotating nuclei . |
the classical big - bang model based on einstein s general relativity does not explain the origin of the initial singularity , which is expected to be understood at a higher dimensional string theory model .
it is known that the horava - witten theory @xcite relates 11-dimensional supergravity on the orbifold @xmath1 with strongly coupled heterotic @xmath2 string theory .
then , this link suggests that as one probes higher energies , our 4-dimensional universe would first go through an intermediate regime where the orbifold dimension becomes visible , thus making our world appear 5-dimensional @xcite , and only at energies of order the string scale the universe would look 11-dimensional .
the intermediate 5-dimensional energy regime has led to the conjecture that our universe is a brane in a higher dimensional world .
these horava - witten inspired `` braneworlds '' have a topology defined by a line times a non - compact space with two branes at the boundaries of the line element , which is usually characterized as the orbifold @xmath1 . due to the compactness of the orbifold direction ,
one brane has positive tension while the other has the opposite negative one . by construction these spacetimes
are supersymmetric and this unbroken supersymmetry of the static background solution guarantees the local stability of the model @xcite .
however , the situation changes when the background solution becomes time - dependent .
the idea of having moving branes has naturally led to the suggestion that the hot big bang is the result of a brane collision @xcite , or even that our universe undergoes an endless sequence of cosmic epochs which begins with a universe expanding from a big bang and ending with a contraction to a big crunch @xcite .
colliding braneworld scenarios offer an alternative point of view for many long - standing problems and puzzles such as those connected with the existence of gravitational shortcuts ( apparently faster than light signals ) @xcite .
these issues refer to the different perspective of higher dimensional theories and conventional 4-dimensional ones .
in fact , the proposed observational tests of braneworld scenarios must be able to distinguish between higher and lower dimensional effects .
moreover , the former effects must appear as corrections of the latter in these models .
one of the approaches used to study these collisions is through low energy supergravity limits .
since we lack a complete m - theory formulation , the solutions come from the 5-dimensional supergravity theory obtained after a dimensional reduction of the higher dimensional scenario . in the case of
type ii - b theory exact solutions to this problem have been found in @xcite , where the motion of the d3 branes towards one another leads to the complete disappearance of the universe in a spacetime singularity . in the context of heterotic brane models
an exact solution of the equations of motion derived from the same horava - witten lagrangian has been found @xcite .
as in the previous case , the setup obtained from a dimensional reduction to 5 dimensions evolves to a singularity where the entire spacetime is annihilated .
this can be taken as an indication of an inherent classical instability in braneworld models consisting of a positive and a negative tension brane .
however , it has recently been proposed a way to avoid the annihilation by introducing an arbitrarily small amount of matter on the negative tension brane @xcite .
this alternative scenario is specially appealing for the ekpyrotic and cyclic universe models @xcite due to the momentary disappearance of the orbifold dimension caused by the collision and the subsequent bounce after which both branes continue to expand into the future .
in this paper we study the stability of the solution in @xcite by introducing a scalar perturbation .
general perturbations of solutions to the einstein equations are very well known in the literature , see e.g. @xcite and further references therein . in the present case we deal with perturbations of time - dependent solutions , a more delicate issue @xcite .
although we know from previous discussions that there exists an instability @xcite , our motivation is to check the possibility of obtaining some information about the signatures we would expect before the spacetime collapses .
the paper is organized as follows . in section 2
we describe the time - dependent horava - witten solution under study . in the next section
we introduce the scalar field perturbation , we solve the corresponding klein - gordon equation in the bulk , and we discuss the field evolution in the setup . finally , we present our conclusions in section 4 .
one of the most popular models connecting m - theory with phenomenology is the horava - witten model @xcite . in this setup
the 11-dimensional spacetime is the product of a calabi - yau compact manifold and a 5-dimensional spacetime consisting of two parallel 3-branes or domain walls , one with negative tension and the other with positive one .
when we perform a dimensional reduction from eleven to five dimensions , we obtain a 5-dimensional supergravity theory which admits an exact static supersymmetric solution of the form @xcite @xmath3 where @xmath4 and @xmath5 is a constant , while @xmath6 is a scalar field characterizing the size of the internal calabi - yau space .
the equations of motion for the metric and @xmath6 can be obtained from the lagrangian @xmath7 in this scenario a second domain wall is introduced at @xmath8 , taking @xmath9 as a periodical coordinate with period @xmath10 and making the @xmath11 identification , @xmath12 . the model based upon this solution is not physically realistic because it is static , while our universe is time - dependent . led by this motivation many attempts to introduce dynamics in this model have been made .
remarkable examples are the ekpyrotic model @xcite , where the big bang is regarded as the result of the collision of an external brane with our universe , and the cyclic universe @xcite , where the distance between the branes oscillates in time . however , recently chen _ et al .
_ @xcite found that the lagrangian ( [ lag ] ) also admits a time - dependent 3-brane solution given by @xmath13 with @xmath14 where @xmath15 and @xmath16 is an arbitrary constant . if we turn off the time - dependence in this solution , _
i.e. _ , set @xmath17 , we can go back to the time - independent solution ( [ static])-([st - param ] ) by performing the coordinate transformation @xmath18 and rescaling the constant @xmath19 of the original work of lukas _ et al . _
@xcite , where @xmath20 .
in fact , this transition between the solutions can be better visualized if we set @xmath21 , which is always possible since it is an integration constant .
the solution ( [ metric])-([param ] ) represents a bulk spacetime with two 3-branes with negative and positive tensions as boundaries at @xmath22 and @xmath8 , respectively .
the interval @xmath23 is of @xmath1 type , where @xmath24 occupies the strip @xmath25 and the @xmath11 symmetry corresponds to the identification @xmath26 . by choosing @xmath27 as in @xcite
we notice that @xmath28 is a positive but decreasing quantity when @xmath29 .
the spacetime is bounded in the range @xmath30 , but the proper length of this interval is time - dependent .
thus , the separation between the branes decreases as @xmath31 becomes positive , _
i.e. _ , the universe is contracting . when @xmath31 turns out to be positive , @xmath28 vanishes along @xmath32 .
this line represents a singularity which first develops on the negative tension brane ( @xmath22 ) at @xmath17 and travels along the orbifold dimension reaching the positive tension brane ( @xmath8 ) after a finite time @xmath33 .
the spacetime can not be extended beyond this point .
thus , before the branes effectively collide the power - law singularity wraps the whole spacetime .
although we know that an instability exists in the present model , we still want to check the possibility of describing its existence before the spacetime collapses .
we consider a perturbation in the form of a scalar field obeying the massive klein - gordon equation @xmath34 or equivalently @xmath35 using the metric ( [ metric ] ) in spherical coordinates and the derivatives of @xmath28 , @xmath36 and @xmath37 , eq.([kg ] ) can be rewritten as [ kg-1 ] \ { - h^-1/2 _ t ^2 - h h^-3/2 _
t + h^-1/2 _ r ^2 + h^-1/2 _ r + .
+ h^-1 _ y ^2 + h^-2 ( y ) _ y - m^2 } = 0 .
now we decompose the scalar field @xmath38 as ( t , r,,,y ) = z ( t , r , y ) y_m ( , ) , where the spherical harmonics obey the equation _
( _ y_m ) + _ ^2 y_m = -(+1 ) y_m .
thus , the function @xmath39 satisfies the equation [ kg-2 ] -_t ^2 z - _ t z + _
r ^2 z + _ r z - z + h^-1/2 _ y ^2 z + + + h^-3/2 ( y ) _ y z - m^2 h^1/2 z = 0 .
@xmath40 in order to separate the radial part we write @xmath41 as @xmath42 .
substituting back into ( [ kg-2 ] ) we obtain [ radial ] _ r ^2
r + _ r r + ( ^2 - ) r = 0 , and [ psi ] _
t ^2 + _ t - _ y ^2 - _ y + ( ^2 + m^2 h^1/2 ) = 0 , where @xmath43 is a constant .
the solution to equation ( [ radial ] ) is given by r(r ) = j ( + , r ) + y ( + , r ) , where @xmath44 and @xmath45 are bessel functions of first and second kind , respectively , and @xmath46 and @xmath47 are integration constants .
the solution of equation ( [ psi ] ) must be obtained by numerical methods .
we must thus introduce some values for the parameters @xmath16 , @xmath48 , @xmath43 , and @xmath49 .
following @xcite we choose @xmath16 to be negative so that our description of the system begins at a negative time and evolves to a positive time . in order to simplify the calculation we choose the simplest positive values for @xmath48 and @xmath43 , @xmath50 and @xmath51 , as well as a positive scalar field mass @xmath49 . the potential term @xmath52 is shown in fig.([pot ] ) .
we see that it is positive - definite although it possesses some singular derivatives along the line @xmath53 .
our results for the evolution of the field @xmath54 are shown in figs.[y0]-[y100 ] .
@xmath55 .quasinormal frequencies at @xmath22 .
[ cols="^,^,^,^,^,^,^",options="header " , ] [ t100 ] fig.[y0 ] provides information on the stability of the system before @xmath17 , when the singularity develops on the negative tension brane .
we can see that for earlier times the system is stable .
the oscillations show a quasinormal phase where we can calculate quasinormal frequencies ( see table [ t0 ] ) , which show a strong dependence on the value of the parameters @xmath56 and @xmath48 when @xmath57 .
we notice that the real part of the frequency increases with @xmath56 and @xmath48 , as well as with @xmath49 . since we calculated the frequencies in the stability range where the oscillations are almost uniform , the imaginary part appears to be very small , and its sign just suggests a slight change in the amplitude of the oscillation ( positive sign indicates an increase , while negative sign indicates a decrease ) .
however , as soon as the scalar field detects the formation of the singularity , the oscillations increase in amplitude and frequency , revealing the instability of the system .
once the singularity has formed on the negative tension brane , it travels along the extra dimension in the direction of the positive tension brane . analizing our results at a point between the two branes as shown in fig.[y50 ]
we observe that the frequency of the scalar field oscillation increases gradually until encountering the singularity .
table [ t50 ] shows some quasinormal frequencies calculated by taking the most stable oscillations .
after a time @xmath33 the singularity reaches the positive tension brane . in fig.[y100 ] we see how unstable the system becomes at this point .
the quasinormal frequencies shown in table [ t100 ] provide evidence of this behaviour , since their positive imaginary part becomes larger than in the previous cases .
the singularity eventually wraps both branes and the space between them , which constitutes the bulk , smearing the actual brane collision .
in this paper we have considered the time - dependent horava - witten solution found by chen _
we studied the stability of the model by introducing a massive scalar perturbation .
we found that before any singularity appears , _
i.e. _ , for @xmath29 , the model evolves without any instability .
for such a time - dependent solution it was possible to calculate quasinormal frequencies at different positions along the extra coordinate .
we chose 3 representative points to show our results , namely , at the position of each brane and between the two branes .
the quasinormal frequencies show a strong dependence on the parameters of the model for a massive scalar field .
the larger the parameters @xmath56 and @xmath48 are , or alternatively the larger the mass of the scalar field @xmath49 is , the larger the real part of the fundamental frequency is .
this dependence disappears when the mass of the scalar field vanishes . at @xmath17 a curvature singularity forms on the negative tension brane .
this singularity travels along the extra dimension and hits the positive tension brane at @xmath33 , before the two branes actually collide . despite knowing the instability of this model that leads to a big crunch
, we have shown that it is possible to detect it .
the evolution of the scalar field used for this purpose can provide some information about the parameters of the model and the extension of the extra dimension . a complete picture of the instability should also include gravitational perturbations .
however , since both the metric and the potential are time - dependent , such a procedure is beyond the aims of the present paper .
we wish to thank elcio abdalla , bin wang , and chi yong lin for enlightening discussions .
this work has been supported by fundao de amparo pesquisa do estado de so paulo * ( fapesp ) * , brazil , and fondo nacional de desarrollo cientfico y tecnolgico * ( fondecyt ) * , chile , under grant 3070009 . | we consider scalar perturbations in the time - dependent horava - witten model in order to probe its stability .
we show that during the non - singular epoque the model evolves without instabilities until it encounters the curvature singularity where a big crunch is supposed to occur .
we compute the frequencies of the scalar field oscillation during the stable period and show how the oscillations can be used to prove the presence of such a singularity .
= 1000 pacs numbers : 04.50.-h @xmath0 04.60.cf @xmath0 11.25.wx @xmath0 11.25.yb |
in this section we set our notational conventions .
we will use the conventions of refs .
@xcite for our spinors .
these conventions are based on the kleiss - stirling @xcite weyl spinors augmented as described in refs .
@xcite with the rules for controlling their complex phases , or equivalently , the three axes of the fermion rest frame in which the spin of that fermion is quantized .
we sometimes refer to this fermion rest frame as the global positioning of spin ( gps ) frame and to the rule for determining it as the gps rule .
the resulting conventions for the fermion spinors are then called the gps spinor conventions .
see refs .
@xcite for more details .
let us now turn to the kinematics .
the process under discussion here is shown in fig .
[ fig : graphs ] , the one - loop virtual correction to the hard bremsstrahlung process @xmath14 , for @xmath2 .
we will treat both the initial state radiation ( isr ) case and the final state radiation ( fsr ) case .
we denote the four momenta and helicity of the @xmath15 , @xmath16 , @xmath17 , and @xmath18 as @xmath19 and @xmath20 , @xmath21 , respectively .
we denote the charge of @xmath17 by @xmath22 in units of the positron charge @xmath23 .
the rest mass of fermion @xmath17 is denoted by @xmath24 .
the photon momentum and helicity will be denoted by @xmath25 and @xmath26 . with our gps conventions for spinors , we induce the following polarization vectors for photons : @xmath27 where the auxiliary 4-vector @xmath28 is exploited here to simplify our expressions as needed .
it satisfies @xmath29 . the second choice with @xmath30 , as defined in ref .
@xcite , is already an example of this exploitation
it often leads to simplifications in the resulting photon emission amplitudes . the calculations which we present
have been done using the program form of ref .
@xcite . for the @xmath31-channel case
, we presented similar results in ref .
@xcite in connection with the respective @xmath32 corrections needed for the @xmath33 total precision tag achieved in ref .
@xcite for the lep1/slc luminosity process in the monte carlo event generator bhlumi4.04 in ref .
just as in the latter case , here a considerable effort is needed to simplify our initial raw form output in order to make a practical application of the respective results in the context of a monte carlo environment such as the mc in ref .
@xcite . as in ref .
@xcite , we only present the final simplified expressions in this paper for the sake of clarity .
our metric is that of bjorken and drell in ref .
@xcite and we effect our gauge invariant calculation in the t hooft - feynman gauge . with these preliminary remarks ,
we turn now in the next section to the calculation of our process of interest .
in this section we calculate the exact virtual correction to @xmath34 , @xmath2 .
we proceed in analogy with our results on the virtual correction for the @xmath31-channel dominated low angle bhabha scattering process with a single hard bremsstrahlung in ref .
@xcite .
specifically , we express the exact amplitude for one real and one virtual photon emitted from the electron lines in the process using the gps conventions @xcite . in ref .
@xcite , the corresponding @xmath31 channel result was obtained for electron line emission . here , from the latter result ,
we first obtain the respective initial - state @xmath8 channel result by crossing the outgoing electron line with the incoming positron line , and replacing the respective final state by @xmath35 , while adding also @xmath36 boson exchange .
the results are translated into gps conventions .
then , in subsection [ form - factors ] , we provide the detailed form factors appearing in the initial state amplitudes .
the corresponding final state amplitudes are presented in subsections [ final - state ] and [ finalstate - formfactors ] . in this subsection , we define notation and set up the exact contribution for one real photon and one virtual photon emitted from the electron lines in the process . the amplitude for real plus virtual photon emission from the initial state may be written @xmath37 where the real photon emission amplitude is @xmath38 , and the factors @xmath39 contain spinor dependence .
they will be specified in the next section . in gps conventions , the amplitude @xmath38 for the initial state radiation of a single photon is given by @xmath40 & = & { eq_e\over 2kp_1 } { \hbox{$\overline{v}$}}(p_2,\lambda_2 ) { \hbox{${\rm\bf m}_1$}}({\hbox{$\not\hbox{\kern-2.3pt $ p$}$}}_1 + m - { \hbox{$\not{k}$ } } ) { \hbox{$\not{\epsilon}$}}_\sigma^\star u(p_1,\lambda_1 ) \nonumber\\ & + & { eq_e\over 2kp_2 } { \hbox{$\overline{v}$}}(p_2,\lambda_2 ) { \hbox{$\not{\epsilon}$}}_\sigma^\star ( -{\hbox{$\not\hbox{\kern-2.3pt $ p$}$}}_2 + m + { \hbox{$\not{k}$ } } ) { \hbox{${\rm\bf m}_1$}}u(p_1,\lambda_1)\ , \end{aligned}\ ] ] where @xmath41 is the annihilation scattering spinor matrix .
the form factors may be obtained from the corresponding @xmath31-channel result in ref .
@xcite for electron line emission .
specifically , the @xmath8 channel result can be obtained by crossing the outgoing electron line with the incoming positron line , and replacing the final state by @xmath35 .
we also include the effects of @xmath36 exchange in the @xmath8 channel .
our previous calculations of @xmath31-channel bremsstrahlung@xcite used the chinese magic conventions@xcite for the photon polarizations . the gps version of the magic photon polarization vector is related to the chinese magic conventions@xcite by @xmath42 the purpose of this change is to recover the more standard convention of defining photon polarization in terms of incoming states .
the choice of magic polarization vector affects the amplitude ( [ iniamp ] ) only through the definition of @xmath38 .
the remaining factors may thus be obtained directly by crossing from our previous @xmath31-channel results .
the magic choice of auxiliary vector for initial state radiation is @xmath43}$ ] , with the definition @xmath44 } & = & \left\ { \begin{array}{ll } q_0\\ q_1\\ q_2 \end{array } \right\}\quad\hbox{if}\quad\left\ { \begin{array}{ll } \mu_1 = \mu_2\\ \mu_0 = \mu_1 = -\mu_2\\ \mu_0 = \mu_2 = -\mu_1 \end{array } \right . .\end{aligned}\ ] ] using the magic polarization vector in ( [ iniamp ] ) and neglecting fermion masses gives @xmath45 where the photon-@xmath36 propagator is @xmath46 and @xmath47 } , { h\left[{0\atop \sigma } { p_3\atop \lambda_3 } { p_4\atop \lambda_4}\right]}\right ) .\ ] ] we now turn to calculating the form factors and spinor factors . it remains to describe the form factors and spinor factors needed to compute @xmath48 .
the spinor factors @xmath49 are given by @xmath50 \label{i2 } i_2 & = & { \sqrt{2}\sigma s_{-\lambda_1}(p_1,k ) s_{\lambda_1}(p_2,k ) s_{-\lambda_3}(p_4,k ) s_{\lambda_3}(p_3,k ) \over s_{-\sigma}(p_1,p_2 ) s_{-\sigma}(p_4,p_3 ) i_0 } \end{aligned}\ ] ] where the spinor product is @xmath51 .
the factors @xmath49 are crossed versions of in ref .
@xcite .
we will begin by writing the dominant term @xmath52 .
expressions can be found in ref .
@xcite for all of the scalar integrals needed for the form factors , which were previously calculated using the ff package @xcite , which implements the methods of ref .
the integrals in ref .
@xcite are not quite adequate , because of the possibility that @xmath53 . however , it was possible to analytically continue when necessary , and to reproduce the numerical results of the ff package .
thus , an expression for the form factors in terms of logarithms and dilogarithms is now available .
details on the @xmath8 channel version of the scalar integrals used in ref .
@xcite may be found in the appendix . for @xmath54 , using @xmath55 , @xmath56 with @xmath57 , the infrared yfs factor @xmath58 and @xmath59 the second expression is preferred for calculating @xmath60 when @xmath61 may be small . for @xmath62 , @xmath63 and @xmath64
are interchanged in ( [ f0 ] ) .
the coefficients of the spinor terms in ( [ iniamp ] ) are , for @xmath65 , @xmath66 and @xmath67 the coefficients @xmath68 are @xmath8-channel versions of @xmath69 in ref .
@xcite . for @xmath62 , @xmath63 and @xmath64
are interchanged in ( [ f1 ] ) and ( [ f2 ] ) .
the leading log limit is obtained by finding which terms give rise to the leading powers of the ` big logarithm ' @xmath6 when the above expressions are integrated over @xmath63 and @xmath64 .
these come from collinear terms where @xmath63 or @xmath64 go to zero . in the collinear limits , when averaged over the azimuthal angle , only the @xmath52 terms remain to order @xmath70 and @xmath6 , _ i.e. _ , to order nll . using the identities
@xmath71 the nll limit of the form factor @xmath52 is found to be @xmath72 without mass corrections .
mass corrections we have calculated primarily without any approximations , however , in the following we shall present them in the approximation @xmath73 . in particular , in this approximation , we checked by explicit calculation that the result which we obtain for the mass corrections in fact agrees with that implied by the prescription in ref .
this prescription is valid for the spin - averaged differential distribution in the limit @xmath73 , but since mass terms are located in the separate ( helicity conservation violating ) spin amplitudes , it is not difficult to `` undo '' the spin summation .
the technique of ref .
@xcite was originally applied to tree level photon emissions .
following the appendix b of ref .
@xcite we can apply it also to our case of emission of one virtual and one real photon .
taking advantage of the freedom , which we have for presenting mass terms in the @xmath73 approximation , the introduction of the mass correction leads to a replacement of @xmath52 by @xmath74 , where @xmath75 \right\}\end{aligned}\ ] ] with yfs infrared factor @xmath76 mass corrections first appear at order nll , and to this order , @xmath77 only the ll part of @xmath52 contributes to the mass correction , to order nll . the result ( [ mass - ini ] ) gives the complete effect of the mass corrections for the isr neglecting the terms that are suppressed by higher powers of @xmath78 as usual .
the amplitudes for final state radiation can be obtained by crossing the incoming electron with the outgoing @xmath18 , and the incoming positron with the outgoing @xmath17 .
thus , @xmath79 , @xmath80 , @xmath81 , and @xmath82 in the results of the previous sections . the final state radiation ( fsr ) amplitude can be written in analogy with the isr result ( [ iniamp ] ) , @xmath83 the form factors @xmath84 and the spinor factors @xmath85 are final - state analogs of those in the previous section , and will be defined in the next subsection .
the amplitude @xmath86 for the final state radiation of a single photon can be obtained from the initial state amplitude @xmath38 by crossing .
crossing leads to spinors with negative energy .
a consistent choice of branches gives @xmath87 then we obtain @xmath88 & = & { eq_f \over 2kp_4 } { \hbox{$\overline{u}$}}(p_3,\lambda_3 ) { \hbox{$\overline{\rm\bf m}_1$}}({\hbox{$\not\hbox{\kern-2.3pt $ p$}$}}_4 - m + { \hbox{$\not{k}$ } } ) { \hbox{$\not{\epsilon}$}}_\sigma^\star v(p_4,\lambda_4 ) \nonumber\\ & - & { eq_f \over 2kp_3 } { \hbox{$\overline{u}$}}(p_3,\lambda_3 ) { \hbox{$\not{\epsilon}$}}_\sigma^\star ( { \hbox{$\not\hbox{\kern-2.3pt $ p$}$}}_3 + m + { \hbox{$\not{k}$ } } ) { \hbox{$\overline{\rm\bf m}_1$}}v(p_4,\lambda_4),\end{aligned}\ ] ] where @xmath89 the magic polarization vector for final state radiation is @xmath90}$ ] . using this in ( [ fsramp ] )
gives , in the massless limit , @xmath91 with propagator ( [ prop ] ) and @xmath92 given again by ( [ i0 ] ) .
the spinor factors @xmath93 appearing in @xmath94 are given by @xmath95 as before , let .
we can obtain the form factor @xmath96 for by substituting @xmath97 in ( [ f0 ] ) , and for by interchanging @xmath98 and @xmath99 in ( [ rsub ] ) .
( since , we have . )
then , for @xmath100,@xmath101 with @xmath102 the expression ( [ rbar ] ) is obtained from ( [ ra2 ] ) using the dilogarithm identity @xmath103 the imaginary parts in ( [ fbar0 ] ) were obtained by assuming the @xmath104 terms in ( [ f0 ] ) came from a small positive imaginary part on @xmath8 or @xmath105 . for @xmath106 , @xmath107 instead : @xmath98 and @xmath99 are interchanged .
the coefficients of the spinor terms in ( [ finamp ] ) are , for @xmath100 , @xmath108 @xmath109 for @xmath106 , @xmath98 and @xmath99 are interchanged in ( [ fbar1 ] ) and ( [ fbar2 ] ) .
the nll limit is obtained as in the initial state radiation case , except that now the collinear limits are when @xmath98 or @xmath99 become small . only the form factor
@xmath96 survives to order nll , and using the identities @xmath110 we find @xmath111 without mass corrections .
spin - averaged mass corrections can be obtained from the initial state case ( [ mass - ini ] ) by crossing .
the result is that @xmath112 , where @xmath113 where @xmath114 again , mass corrections first appear at order @xmath115 , and to this order , @xmath116 the result ( [ mass - fin ] ) gives the complete effect of fsr mass corrections neglecting terms suppressed by higher powers of @xmath117 as usual .
this section translates our amplitudes into differential cross sections , and sets up comparisons with other related results . the initial state differential cross section for emitting one real photon
may be written @xmath118 where the summed , squared real photon amplitude leads to @xmath119\ , \ ] ] where the invariants @xmath120 , @xmath121 may be written in yfs3-style @xcite effective angle notation : @xmath122\nonumber\\ u_i & = ( p_i - p_{j+2})^2 & = - { 1\over4}\beta_f\beta_i s(1-r_i ) [ 2 + \cos(\theta_{1i } ) + \cos(\theta_{2i})]\end{aligned}\ ] ] with @xmath123 or @xmath124 , and @xmath125 we will be setting these mass factors to unity in the following , and adding mass corrections at the end via ( [ mass - ini ] ) or ( [ mass - fin ] ) .
the coefficients @xmath126 are defined in terms of the standard vector and axial vector fermion couplings @xmath127 and @xmath128 , and , by @xmath129 where @xmath130^{-1 } , \nonumber\\ \chi_2 & = & { s'}^2 [ ( s'-m_z^2)^2 + ( s'\gamma_z / m_z)^2]^{-1}. \end{aligned}\ ] ] the initial state differential cross section for real plus virtual photon emission may be expressed as @xmath131 .\ ] ] if is convenient to rewrite ( [ iniamp ] ) as @xmath132 in terms of a virtual correction factor @xmath133 the differential cross section ( [ dsisr11 ] ) can then be written in terms of a spin - averaged virtual correction factor @xmath134 times the cross section for pure real initial state radiation : @xmath135 where @xmath136 + v_{-}^{\lambda_1}[f_0-f_1+\lambda_1(f_2-f_3)]\right\}\over ( f_0+f_1)(u_1 ^ 2 + u_2 ^ 2 ) + ( f_0-f_1)(t_1 ^ 2 + t_2 ^ 2 ) } \end{aligned}\ ] ] with @xmath137 and @xmath138 in the nll approximation , where the @xmath139 and @xmath140 terms in ( [ iniamp ] ) may be neglected , we may use the relation @xmath141 .
then ( [ vfac ] ) simplifies to @xmath142 & = { f_0[{\hbox{$\hat{f}$}}_0(r_1,r_2)(1+\cos^2\theta_{11 } ) + { \hbox{$\hat{f}$}}_0(r_2,r_1)(1+\cos^2\theta_{22 } ) ] + 2f_1[{\hbox{$\hat{f}$}}_0(r_1,r_2)\cos\theta_{11 } + { \hbox{$\hat{f}$}}_0(r_2,r_1)\cos\theta_{22 } ] \over f_0[(1-r_1)^2(1+\cos^2\theta_{11 } ) + ( 1-r_2)^2(1+\cos^2\theta_{22 } ) ] + 2f_1[(1-r_1)^2\cos\theta_{11 } + ( 1-r_2)^2\cos\theta_{22}]}\end{aligned}\ ] ] with @xmath143 if we further drop the dependence on @xmath144 in ( [ t - u - def ] ) , letting @xmath145 for fixed @xmath146 , then we get a simpler approximation .
the angle dependence can be factored out of the cross - section ( [ dsisr11 ] ) , leading to @xmath147 with the definition @xmath148 , \ ] ] the total born cross section @xmath149 and approximate spin - averaged form factor @xmath150 it can be shown that this approximation is valid to order nll .
the nll expression ( [ nll - ini ] ) then leads to the spin - averaged form factor @xmath151 with mass corrections given by ( [ mass - ini ] ) .
this form is useful for comparison with other results on the differential cross section , as we will see in the next section .
an analogous expression can be found for the final state emission cross section .
the spin - averaged version of the final state form factor ( [ nll - fin ] ) is @xmath152 with mass corrections given by ( [ mass - fin ] ) .
we may compare our spin - averaged initial state radiation form factor with one published in ref .
@xcite . in our notation , this result may be written as @xmath153 with @xmath154 \nonumber\\ & + & [ 1+(1-r_2)^2]\left\ { \ln^2(r_1 ) + \ln^2{1-r_1-r_2\over 1-r_2 } - \ln^2{r_1\over 1-r_1-r_2}\right .
\nonumber \\ & & \qquad \qquad \qquad + \left .
2\,{\hbox{sp}}\left({1-r_1-r_2\over 1-r_2}\right ) + 2\,{\hbox{sp}}(r_1+r_2 ) - { \pi^2\over 3}\right\}\nonumber\\ & + & { r_1 r_2 ( 1-r_1-r_2)\over 1-r_2 } \left({1\over 1-r_2 } + 1 - { 3r_2\over r_1}\right ) \ln{r_1\over 1-r_1-r_2 } \nonumber\\ & - & { 2r_1 r_2 ( 1-r_1-r_2)\over r_1+r_2 } \left({(1-r_1-r_2)\over r_1+r_2 } - { 3\over r_1}\right ) \ln(1-r_1-r_2 ) \nonumber\\ & + & r_2(3r_2 + 2r_1)\ln(r_1 ) - { 2r_1 r_2 \over r_1+r_2 } + { r_1 r_2 \over
1-r_2 } + r_2(1-r_2 ) .\end{aligned}\ ] ] the nll limit of this expression may be obtained by summing the two collinear limits where @xmath155 separately . carrying this out leads precisely to our spin - averaged nll expression ( [ nll - ini - avg ] ) .
thus , we agree with ref .
@xcite to nll order .
note that the expression @xmath156 does not include mass corrections , which were not calculated in ref .
@xcite . by integrating out the separate dependence on @xmath157 in favor of the variable ,
we may obtain a result which can be compared to ref .
@xcite . to begin
, we consider the pure real photon isr cross section , and work in the approximation where the effective angles @xmath144 in ( [ t - u - def ] ) are replaced by a common angle @xmath146 .
then @xmath158 in terms of @xmath159 defined by ( [ hdef ] ) .
integrating @xmath63 and @xmath64 with the constraint @xmath160 gives @xmath161 where is the kinematic minimum value of @xmath63 or @xmath64 .
the result of the integral is , exactly , @xmath162 .\ ] ] mass corrections , obtained by the prescription of ref .
@xcite , have the effect of replacing @xmath159 by in ( [ realdcs ] ) , where @xmath163 integrating the mass term gives @xmath164 the total mass - corrected real photon emission cross section is then @xmath165 in the notation of ref .
@xcite , where this result matches the real part of ( [ f0 ] ) .
the real plus virtual isr cross section is @xmath166 where we use the nll expression ( [ nll - ini - avg ] ) for the virtual form factor , and will add mass corrections later .
doing the integral and keeping all infrared terms and terms of order @xmath70 and @xmath6 gives @xmath167 \nonumber\\ & + & { 1\over2 } \left({q_e^2 e^2\over 4\pi^2}\right)^2\sigma_0 l \left(1+z^2\over 1-z\right ) \big\ { - l\ln z + 2\ln z \ln(1-z ) \nonumber\\ & + & 2\,{\hbox{sp}}(1-z ) + 3\ln z - \ln^2 z + { z(1-z)\over 1 + z^2}\big\},\end{aligned}\ ] ] where we use the notation of ref .
@xcite for @xmath168 the mass correction is obtained by multiplying the pure real mass correction by the single virtual photon form factor evaluated in terms of @xmath105 rather than @xmath8 .
thus , we add to the differential cross section a mass term @xmath169 where @xmath170 integrating over @xmath63 and @xmath64 with @xmath171 and keeping only infrared terms and terms of order @xmath6 , we obtain @xmath172 adding the mass corrections and using the notation of ( [ realdcs - total ] ) gives the complete real plus virtual cross section at order nll , @xmath173\big\ } & .
\end{aligned}\ ] ] this result agrees precisely with the terms in ( 2.26 ) of ref.@xcite through order @xmath174 .
we illustrate the agreement we have found above in figs . 2 - 5 , for the case . in fig .
[ fig : figs - b1 ] , we show the complete @xmath175 distribution for our exact result , our nll and ll approximate results , the result of igarashi and nakazawa _ et al . _
@xcite , and the result of berends _ et al .
_ @xcite .
what we see is that there is a very good general agreement between all of these results . to better assess the difference between them , we plot in fig .
[ fig : figs - bdif ] the difference between the respective @xmath176 and @xmath177 results .
again we see very good agreement except for the hardest possible photons , where then the @xmath178 result differs significantly from the others .
to isolate the respective predictions for the nll effect , we plot in fig .
[ fig : figs - bnll ] the respective differences between our ll @xmath176 result and the other four results .
we see that there is again very good agreement but , at the level of @xmath179 , the result of ref .
@xcite is somewhat smaller in magnitude than the other three nll results in the @xmath36 radiative return regime above a cut of 0.75 . finally , in the fig .
[ fig : figs - bnnll ] we isolate the size of the three nnll results by plotting the difference between our nll result and our exact result , the result from ref .
@xcite and the result from ref .
we again see that the most pronounced difference in the results occurs for the regime above @xmath180 where the result of ref .
@xcite differs by @xmath179 from the other two for @xmath181 and differs from the exact result by @xmath182 for @xmath183 .
the result from ref .
@xcite differs from the exact result by @xmath184 for @xmath183 but is essentially indistinguishable from it for smaller values of @xmath185 .
we conclude that our exact result for the @xmath176 correction @xmath186 has a total precision tag of @xmath182 .
its nll effect has already been implemented in the mc in ref .
@xcite .
we have also made the analogous study to figs .
2 - 5 for 500 gev .
we find very similar results , with the total precision tag of @xmath187 .
in this paper , we have presented exact results for the virtual correction to the process @xmath188 for the @xmath189 .
the results are already in use in the mc in ref .
@xcite in connection with the final lep2 data analysis .
we have compared our results with those in refs .
@xcite and in general we find very good agreement , both at 200 gev and at 500 gev .
for example , the size of the nnll correction is shown to be at or below the level of @xmath190 for all values of the energy cut parameter .
our results are fully differential and are therefore ideally suited for mc event generator implementation .
this has been done in the mc in ref .
@xcite . to compare our results with the results in refs .
@xcite , we have partially integrated them accordingly . while the results in ref .
@xcite are also fully differential , they lack the complete mass corrections that our results do have . in this way
, one sees that our results are in fact unique .
they are an important part of the complete @xmath176 corrections to the @xmath191 production process needed for precision studies of such processes in the final lep2 data analysis and in the future tesla / lc physics .
two of the authors ( s.j . and b.f.l.w .
) would like to thank prof .
g. altarelli of the cern th div . and prof .
d. schlatter and the aleph , delphi , l3 and opal collaborations , respectively , for their support and hospitality while this work was completed
. b.f.l.w . would like to thank prof . c. prescott of group
a at slac for his kind hospitality while this work was in its developmental stages .
after we had submitted this paper , we became aware of related work by g. rodrigo , a. gehrmann - de ridder , m. guilleaume and j. h. kuhn , hep - ph/0106132 .
these authors also agree with the analogous results of ref .
@xcite , when the photon azimuthal angle is integrated and the photon polarization is summed for the isr process .
previously , in ref .
@xcite , the analogous exact virtually corrected photon cross sections were expressed in the @xmath31 channel using scalar integral functions which were calculated by a numerical package ff described in ref .
@xcite . in the present paper
, we have expressed these functions directly in terms of logarithms and dilogarithms .
this appendix will give the expressions for the individual scalar integrals in the @xmath8 channel . the notation for the scalar integrals will match ref .
@xcite , with kinematic notation defined in sections [ preliminaries ] and [ exact - virtual ] of this paper .
the scalar integrals with two denominators are denoted @xmath192 , and the ones appearing in the form factors are @xmath193 where the first argument is the square of the momentum through the diagram , and the remaining arguments are the masses of the two lines .
these functions are uv divergent , but only the following finite combinations are needed : @xmath194 the mass term in ( [ b13dif ] ) has been dropped when applying this expression , since the mass corrections are added explicitly to the massless limit of the calculation using the prescription of ref .
@xcite .
the integrals for three and four denominators are obtained from the appendix of ref .
@xcite . for three denominators
, we need the expressions @xmath195 where the first three arguments are the squares of the external momenta , and the next three are the masses of the three lines , in cyclic order .
the results are @xmath196 the expression ( [ c123 ] ) is an analytic continuation of the one in ref .
@xcite , as required since @xmath197 is possible , and the expression ( [ c134 ] ) drops mass terms .
in particular , the photon mass regulator @xmath198 is dropped whenever possible .
the expression with four denominators is @xmath199 all of these expressions have been checked for agreement with the ff package .
the function @xmath200 defined in equations ( [ ra2 ] ) and ( [ rb2 ] ) and appearing in the virtual photon factors is the ir - finite combination @xmath201 this completes the appendix
. 99 m. kobel _ et al .
_ , hep - ph/0007180 , in _ reports of the working groups on precision calculations for lep2 physics _ , cern-2000 - 09 , eds .
s. jadach , g. passarino and r. pittau , ( cern , geneva , 2000 ) p. 269 .
s. jadach , b.f.l .
ward and z. was , _ phys .
_ * d63 * ( 2001 ) 113009 ; _ comput .
* 130 * ( 2000 ) 260 ; and references therein .
d. bardin _ et al .
_ , _ comput .
_ * 133 * ( 2001 ) 229 .
berends , w.l .
van neerven and g.j.h .
burgers , _ nucl .
_ * b297 * ( 1988 ) 429 and references therein .
m. igarashi and n. nakazawa , _ nucl .
phys . _ * b288 * ( 1987 ) 301 .
s. jadach , b.f.l .
ward , and z. was , `` global positioning of spin gps scheme for half - spin massive spinors '' , 1998 , preprint hep - ph/9905452 , cern - th/98 - 235 , submitted to _ eur .
j. phys .
c_. s. jadach , b.f.l .
ward and z. was , _ phys .
lett _ * b449 * ( 1999 ) 97 .
r. kleiss and w.j .
stirling , nucl.phys .
* b262 * ( 1985 ) 235 ; phys .
lett . b*179 * ( 1986 ) 159 .
vermaseren , the symbolic manipulation program form , versions 1 - 3 , available via anonymous ftp from nikhefh.nikhef.nl .
s. jadach , m. melles , b.f.l .
ward and s.a .
yost , _ phys .
_ * b377 * ( 1996 ) 168 .
a. arbuzov _ et al .
_ , _ phys .
lett . _ * b383 * ( 1996 ) 238 ; s. jadach , m. melles , b.f.l . ward and s.a .
yost , _ phys .
lett . _ * b450 * ( 1999 ) 262 .
s. jadach , w. placzek , e. richter - was , b.f.l .
ward and z. was , _ comput .
* 102 * ( 1997 ) 229 .
bjorken and s.d .
drell , _ relativistic quantum fields _ , ( mcgraw - hill , menlo park , 1965 ) .
z. xu , d .- h .
zhang and l. chang , _ nucl .
phys . _ * b291 * ( 1987 ) , 392 . g.j
. van oldenborgh and j.a.m .
vermaseren , _ zeit .
phys . _ * c46 * ( 1990 ) , 425 . f.a .
berends , p. de causmaecker , r. gastmans , r. kleiss , w. troost and t.t .
wu , _ nucl .
phys . _ * b264 * ( 1986 ) 243 , 265 .
s. jadach and b.f.l .
ward , phys .
* b274 * ( 1992 ) 470 .
the yfs3 matrix element is the eex3 calculation in ref .
[ fig : graphs ] . initial state radiation graphs for @xmath202 with one virtual and one real photon , for @xmath2 .
[ fig : figs - b1 ] .
this is the @xmath175 distribution for the yfs3ff mc ( yfs3ff is the eex3 matrix element option of the mc in ref .
@xcite ) , as a function of energy cut @xmath203 .
it is divided by the born cross - section . | = 0.55 cm we present the exact @xmath0 correction to the process @xmath1 , @xmath2 , for isr@xmath3fsr at and beyond lep2 energies .
we give explicit formulas for the completely differential cross section . as an important application ,
we compute the size of the respective sub - leading corrections of @xmath4 to the @xmath5 cross section , where @xmath6 is the respective big logarithm in the renormalization group sense so that it is identifiable as @xmath7 when @xmath8 is the squared @xmath9 cms energy .
comparisons are made with the available literature .
we show explicitly that our results have the correct infrared limit , as a cross - check .
some comments are made about the implementation of our results in the framework of the monte carlo event generator mc .
= 0.6 cm currently , the final lep2 data analysis is in its beginning stages , and the desired total precision tags on the important lep2 physics processes @xmath10 , @xmath2 , are already called out in the lep2 mc workshop in ref . @xcite .
it has been demonstrated in ref .
@xcite that the monte carlo ( mc ) event generator program @xmath11
@xcite , hereafter referred to as mc , and the semi - analytical program zfitter @xcite realize these precisions ( @xmath12 ) in most channels for inclusive cross sections and that for the fully differential distributions , the mc again meets most of the requirements for the lep2 final data analysis . in this paper , we present exact results on the @xmath0 correction to the single hard bremsstrahlung processes @xmath13 , @xmath2 .
this correction is an important contribution to the differential distributions as they are realized in the mc which allows the very demanding precisions just cited to be achieved .
specifically , the exact results for the @xmath0 corrections to @xmath8-channel annihilation hard bremsstrahlung processes under study here were also considered in refs .
@xcite .
we differ from these results as follows .
concerning ref .
@xcite , the entire result was given only for the case in which the photon angle variables are all integrated out ; here , we give the fully differential results . with regard to ref .
@xcite , the completely differential results were given as well but the mass corrections were omitted . in our work ,
the masses of the electrons and positrons and the masses of the final state fermion and anti - fermion are exactly taken into account in contrast to the literature .
thus , by comparing with the two calculations in refs .
@xcite as we do here , we get a measure of the size of the mass corrections as well as cross checks on both our differential and our integrated results .
our work is organized as follows . in section [ preliminaries ] , we set our notational conventions . in section
[ exact - virtual ] , we present our exact amplitudes for the @xmath0 virtual corrections to initial - state and final - state real radiation . in section [ cross - section ] , we derive the differential cross - sections corresponding to these amplitudes in a form useful for comparisons . in section
[ comparisons ] , we compare these results with those in refs .
@xcite while illustrating our results as they are used in the mc in ref .
@xcite .
section [ conclusions ] contains our summary remarks .
the appendix contains technical details about the scalar integrals . |
the heisenberg ferromagnet is defined by the partition function @xmath0 = \int \prod [ d \omega(x ) \exp(-s)\label{eq:1}\ ] ] where @xmath1 ^ 2 \qquad \vec n^2(x ) = 1\label{eq:2}\ ] ] and @xmath2 is the element of solid angle for the orientation of @xmath3 in colour space .
the model presents a 2nd order phase transition at @xmath4@xcite . for @xmath5
there is an ordered phase , with order parameter the magnetization @xmath6 ; for @xmath7 , @xmath8 ( disordered phase ) .
we shall describe the system from a dual point of view , and show that in the disordered phase vortices condense .
the model will be viewed as a @xmath9 dimensional euclidean field theory .
vortices will be labelled by the integer valued topological charge of the 2 dimensional spacial configurations , which is a conserved quantity , and defines a @xmath10 symmetry of the system .
a disorder parameter @xmath11 will be constructed , which detects spontaneous breaking of this @xmath10 symmetry , or condensation of vortices .
the situation is analogous to the condensation of monopoles in the confining phase of gauge theories@xcite or to the condensation of abelian vortices in @xmath12@xcite . to check our construction
we shall extract from the numerical determination of @xmath11 at the phase transition the known critical indices .
usually the colour frame to which the direction of @xmath3 is referred is a fixed frame , independent of @xmath13 @xmath14 a body fixed frame can be defined@xcite by three unit vectors @xmath15 @xmath16 and @xmath17 .
the frame is defined up to an arbitrary rotation around @xmath18 . since @xmath19 @xmath20 or @xmath21 @xmath22 are the generators of the @xmath23 symmetry group .
eq.([eq:3 ] ) is nothing but the definition of parallel transport . from eq.([eq:3 ] ) it follows @xmath24\vec\xi_i = 0 $ ] or , by completeness of @xmath25 @xmath26 = \vec t\cdot\vec f_{\mu\nu}(\omega ) = 0 \label{eq:4}\ ] ] @xmath27 @xmath28 is a pure gauge , apart from singularities .
the general solution of eq.([eq:3 ] ) is then , for @xmath18 @xmath29 where @xmath30 is the value of @xmath31 at infinity , and the dependence on the path @xmath32 is trivial , because @xmath33 is a pure gauge , eq.([eq:4 ] ) .
this is true apart from singularities .
we will show that such singularities exist .
the current @xmath34 is identically conserved @xmath35 if we look at the theory as the euclidean version of a field theory , with euclidean time on the 3 axis , the corresponding conserved quantity is @xmath36 which is nothing but the topological charge of the 2 dimensional configurations of the theory .
@xmath37 can assume positive and negative integer values .
by use of eq.([eq:3 ] ) and eq.([eq:4 ] ) it is easy to show that @xmath38 where the path @xmath32 is the contour of the region in the 2 dimensional space ( @xmath39 = const . )
where eq.([eq:4 ] ) holds . since @xmath40 eq.([eq:5 ] ) shows that @xmath41 is not always a pure gauge .
this can be explicitely checked on a configuration corresponding to a static 2 dimensional instanton propagating in time ( vortex ) .
the conserved current @xmath42 identifies a @xmath10 symmetry .
we will show that this symmetry is wigner in the ordered phase @xmath5 , and is spontaneously broken in the disordered phase .
let @xmath43 be a @xmath44 dependent singular rotation creating a vortex of charge @xmath45 at the site @xmath46 in a 2 dimensional configuration .
@xmath47 the creation operator of a vortex @xmath48 at site @xmath46 , time @xmath49 , will be defined as @xmath50\bigr\ } \label{eq:7}\end{aligned}\ ] ] we measure the correlator @xmath51 by cluster property @xmath52 @xmath53 signals spontaneous breaking of the @xmath10 symmetry ( [ eq:4a ] ) , or condensation of vortices . by use of the definition ( [ eq:7 ] )
it is easy to see that @xmath54}{z[s]}\ ] ] where @xmath55 is obtained from @xmath56 , eq.([eq:2 ] ) by the replacement @xmath57 ^ 2 \hskip-5pt&\to&\hskip-5pt \left[r_q^{-1}(\vec x,0)\vec n(\vec
x,1 ) - \vec n(\vec x,0)\right]^2 \\
\left[\delta_0\vec n(\vec x , x_0)\right]^2 \hskip-5pt&\to&\hskip-5pt \left[r_q^{-1}(\vec x,0)\vec n(\vec x , x_0 + 1 ) - \vec n(\vec x , x_0)\right]^2\end{aligned}\ ] ] and that this really amounts to have a vortex propagating from @xmath58 to @xmath59@xcite . instead of @xmath60 itself it proves convenient@xcite to study the quantity @xmath61 .
as @xmath62 from eq.([eq:8 ] ) we have for @xmath63 , @xmath64 . since @xmath65 , @xmath66 $ ] .
@xmath67 is easier to measure and contains all the information on the transition .
the behaviour of @xmath67 is shown in fig.1 . for @xmath68 , @xmath69 finite limit consistent with zero , or @xmath70 , which means condensation of vortices . for @xmath71 @xmath67
can be evaluated in perturbation theory and behaves as @xmath72 @xmath73 is the lattice size .
eq.([eq:9 ] ) implies that @xmath74 for @xmath71 in the thermodynamical limit @xmath75 . around @xmath76 a finite size scaling analysis
can be performed @xmath77 0.1 in * fig.1 * 0.1 in and since @xmath78 , we get the scaling law@xcite @xmath79 the scaling law is verified , fig.2 , and allows to extract @xmath76 and @xmath80 @xmath81\\ \nu & = & 0.70 \pm 0.02\ ; [ 0.698]\end{aligned}\ ] ] they agree with the values determined from @xmath82 , which are indicated in parentheses@xcite .
-10pt 0.1 in * fig.2 *
the phase transition to disorder in 3d heisenberg model is produced by condensation of topological solitons .
a disorder parameter can be defined and out of it the critical indices can be determined .
this work has partially supported by murst .
a. di giacomo is grateful to ec , tmr project , erbfmx - ct97 - 0122 for financing the partecipation to the conference .
9 e. brezin , j. zinn - justin , _ nucl .
phys . _ * b257 * , 867 , ( 1985 ) .
a. di giacomo , g. paffuti , _ phys .
rev . _ * d56 * , 6816 , ( 1997 ) .
l.del debbio , a.di giacomo , g.paffuti and p.pieri , _ phys .
lett._*b 355 * ( 1995 ) 255 . g. di cecio , a. di giacomo , g. paffuti , m. trigiante , _ nucl .
phys . _ * b489 * , 739 , ( 1997 ) . a. di giacomo , m. mathur , _ phys .
_ * b400 * , 129 , ( 1997 ) .
p. peczak , a.l .
ferrenberg , d.p .
landau , _ phys .
rev . _ * 43 * , 6087 , ( 1991 ) . | the 3d heisenberg model is studied from a dual point of view in terms of 2d solitons ( vortices ) .
it is shown that the disordered phase corresponds to condensation of vortices in the vacuum , and the critical indices are computed from the corresponding disorder parameter . |
supermassive black holes ( bhs ) are now considered to be a common feature of galaxies which have a bulge .
furthermore , a number of observational properties of the host galaxy correlate with the bh mass . among the strongest of these correlations
is the relationship between the bh mass and velocity dispersion within the galactic bulge ( ferrarese & merritt 2000 ; gebhardt et al . 2000 ) : @xmath1 , where @xmath2 ( tremaine et al . 2002 ) .
since the velocity dispersion is measured well outside of the bh `` influence , '' this correlation indicates that an intimate relationship exists between the bh and the dynamical structure of the host galaxy .
we assume that a galaxy forms by the extended collapse of a `` halo '' composed of collisionless matter and that simultaneously the central black hole is growing proportionally to the halo as matter continues to fall in .
this is equivalent to the assumption of multidimensional self - similarity ( carter & henriksen 1991 ; henriksen 1997 ) , and macmillan & henriksen ( 2002 ) show that , in this case , the mass @xmath3 inside any surface in the halo is related to the velocity of the particles by @xmath4 where @xmath5 is the averaged velocity . since the system is self - similar , this relation will apply at both the `` bulge '' mass scale as well as the bh mass scale .
the quantities @xmath6 and @xmath7 are scales in space and time , respectively , and are related to the power - law index of the initial density perturbation @xmath8 in spherical infall models of halo growth ( henriksen & widrow 1999 ) : @xmath9 this power - law index @xmath8 depends on the spectral index @xmath10 of the primordial power spectrum , @xmath11 , by @xmath12 . there is therefore a direct link between the initial power spectrum and the predicted relationship between the bh mass and the velocity dispersion ; for agreement with observations , we must have @xmath13 , so that @xmath14 and @xmath15 .
we determine below that the addition of sufficient angular momentum breaks the self - similarity in the central regions .
this creates a `` mass reservoir '' around the bh that does grow in proportion to the galaxy mass , and from which the bh grows more slowly by collisional interactions between clumps of matter .
provided most of this mass may be absorbed by the bh on a cosmic timescale , the proposed relation should still hold .
the initial conditions for the simulations follow those of henriksen & widrow ( 1999 ) : @xmath16,\ ] ] where @xmath17 & r < r_c \\ a(1 - b)(r / r_c)^{-\epsilon } & r \ge r_c . \end{array } \right.\ ] ] this initial density profile is that of a power - law function of radius , modified at small radii to account for suppression of small scale fluctuations . the constant @xmath18 and
we choose @xmath19 .
the particles are started in the unperturbed hubble flow .
our initial numerical work began with a simple shell code , which is used to follow spherically symmetric shells of matter in an einstein - de sitter universe ; each shell evolves according to @xmath20 where @xmath21 is the total mass enclosed by the shell at radius @xmath22 .
the angular momentum @xmath23 is given by @xmath24 , evaluated at the initial shell positions .
thus the particles are allowed to have , in general , elliptical orbits , and the shells follow the evolution of their radial coordinates . to extend the spherically symmetric numerical results , we consider black hole growth in an n - body system of @xmath25 particles .
we use a modified version of piet hut s code ( http://www.ids.ias.edu/ piet / act / comp / algorithms / starter/ ) this code was modified to include a softening length to handle close encounters between particles . in both simulations ,
a central bh has been included that can `` eat '' particles which fall below the schwarzschild radius , @xmath26 , and grow in mass accordingly . also , we choose a power - law index for the initial perturbation of @xmath14 , so that @xmath27 and @xmath28 , as observed .
the shell code simulations are run , with an initial bh mass of @xmath29 , until @xmath30 , at which time all matter is essentially `` eaten '' by the bh in the @xmath31 case .
other values of the specific angular momentum are also explored .
the bh mass , as a function of time , is shown in figure [ growth]a for j = 0 , 0.01 , 0.1 , 0.2 , and 0.5 .
macmillan & henriksen ( 2002 ) predict that the bh mass should grow as @xmath32 .
this is indeed the asymptotic behavior for the case where @xmath31 , as shown in the figure .
this logarithmic slope , however , decreases as the specific angular momentum increases ; for @xmath33 , the mass is increasing as @xmath34 . for larger values of @xmath35 ,
the growth essentially `` shuts off '' at some time , indicating that shells with high initial angular momentum are never reaching the centre of the system .
three different cases are explored with the n - body simulations .
the first is an initial spherically symmetric system . in the second and third cases , spherical symmetry
is removed by giving each of the particles a random displacement , with amplitude @xmath36 .
systems are considered in which the amplitudes are about 0.8% and 1.6% of the initial extent of the system .
results are shown in figure [ growth]b for the black hole mass as a function of time .
for the case with no initial random displacements ( @xmath37 ) , the growth is approximately that expected , although at later times the black hole growth slows down due to a lack of infalling particles . for the more general cases where @xmath38 , black hole growth remains self - similar , but with a much lower logarithmic slope : @xmath39 or a little greater . in these systems ,
the particles have enough angular momentum that they do nt fall straight into the central black hole , and rather form a `` core '' at the centre which slowly feeds the black hole growth . to test the prediction that @xmath28 , the velocities of some fraction of the particles are averaged ( equating @xmath40 ) .
the predictions of macmillan & henriksen ( 2002 ) apply to a fixed scaled position vector @xmath41 , implying that we should average all particles within some comoving radius .
results are shown in figure [ msigma]a for a comoving radius of @xmath42 .
although the case where @xmath37 may have a regime where it follows the predicted @xmath43 relation , the cases with @xmath38 clearly follow no relation at all ; their average velocity is about @xmath44 for most black hole masses . of interest , however , are the results if we take a fixed radius , rather than a comoving one , and average the velocities of the particles within that radius . the plot of @xmath45 versus @xmath5 is shown in figure [ msigma]b .
all three cases clearly show a trend that goes like @xmath28 . for systems
in which there is no angular momentum or deviation from spherical symmetry initially , the numerical work , with both the spherical shell code and the more general n - body code , confirms the predicted black hole growth : @xmath46 .
however , for the simulations in which there was some deviation from a simple spherical infall of matter , results were contrary to what was expected . in particular , the n - body results show a black hole growth approximately proportional to the time . however , it is apparent from the simulations that the black hole growth in these cases is not dominated by the infalling matter , as it is in the initially spherical case .
rather , the particles form a `` core '' about the centre , and so the growth is fed in a different manner .
dimensional arguments suggest that the schwarzschild radius should scale with time as @xmath47 , where @xmath48 is the speed of light , a constant of the system .
if this is the case , then we get that the black hole mass will grow as @xmath49 , which is approximately what is observed .
note , however , that this breaks the symmetry of the system , since the mass on a larger scale is predicted to grow as @xmath50 .
the numerically derived @xmath0 relationship , which here compared the black hole growth with the average velocity of particles within some radius @xmath51 , only gave the expected relation , @xmath28 , when a fixed , rather than comoving radius was considered .
furthermore , the cases for which the growth rate was much shallower than expected still showed the same trend as the other case ; that is , @xmath28 regardless of how the black hole grew with time .
as stated above , however , this result is for a fixed radius . from the definition of @xmath41 and @xmath52 in macmillan & henriksen ( 2002 ) , we see that @xmath53 .
fixing @xmath51 requires @xmath41 to then depend on time . recalling that we re considering a density profile that goes as @xmath54 , and @xmath8
is related to the quantity @xmath55 by equation ( [ eq : relate ] ) , we can write the mass as a function of @xmath56 : @xmath57 thus , as a function of the ( fixed ) radius , we have @xmath58 .
similarly , we can write the self - similar velocity , assuming it is in its virialized state , as a function of @xmath56 : @xmath59 so that the radial dependence of the velocity takes the form @xmath60 .
combining these two equations for mass and velocity and eliminating @xmath51 gives us the familiar relation given by equation ( [ eq : predict ] ) .
thus , taking the mass and velocity at a fixed radius @xmath51 gives the same relation regardless of the time dependence of either .
we consider black hole growth in a system which forms as the black hole is growing .
if this is the case , then the system has multidimensional self - similarity , and the black hole should grow as a power - law function of time , @xmath61 .
furthermore , the velocities of the particles of the system will also evolve in a similar manner , and we predict that @xmath62 .
numerically , we confirm that the black hole does indeed grow as a power - law function of time ; however , it only follows the predicted relationship if the initial system has little or no angular momentum . for systems with
non - spherical initial conditions , the black hole grows as @xmath39 or a little greater .
furthermore , we test the@xmath0 relation with n - body simulations . although results do not show a strong relation if we take an averaged velocity within a comoving radius , the predicted relation , that @xmath28 , is clearly present if we take a fixed radius and average the particles within that .
this can be explained by considering the self - similar variable @xmath41 to be a function of time such that the time dependence of the radial coordinate is fixed . | this poster discusses a possible explanation for the relationship between the mass of the central supermassive black hole and the velocity dispersion in the bulge of the host galaxy .
we suppose that the black hole and the dark matter halo are forming simultaneously as matter falls in , and a self - similar system then exists in which the mass and the velocities of the system evolve as power - law functions of time .
this leads naturally to a relationship between the black hole mass and the velocities in the halo which , with a reasonable choice of cosmological parameters , is in good agreement with the observed relationship .
we also confirm this relationship with more robust numerical results .
[ 1996/06/01 ] a&a |
the description of radiating bhs by hawking @xcite offered the first physically relevant `` peep '' on the mysteries of quantum gravity .
after more than forty years of intensive research in this field ( see @xcite and @xcite for a recent review with an extensive reference list ) various aspects of the problem still remain to be properly explained .
for example , a satisfactory description of the terminal stage of bh evaporation remains still to be understood .
so far , the string theory seems to be the best candidate for self - consistent , ultraviolet completion of gravity at the planck scale . on the other hand , following bekenstein s idea , the bh entropy is formally identified with the area of the event horizon in planck units .
thus , combining hawking s definition of bh temperature with the area law gives a consistent thermodynamical description of semi - classical bhs dynamics .
however , to give a physical ground to this interpretation one needs a proper identification of bh micro - states , i.e. a complete statistical interpretation of entropy .
beforehand one can say that the number of micro - states has to grow exponentially with the area , but the physical meaning of these micro - states is elusive in this semi - classical quantum framework .
contrary to the classical gas thermodynamics , where the entropy is a function of the volume enclosing the system , in the bh case entropy is a function of the area of the object indicating that the eventual bh micro - states should lie somewhere on the horizon surface instead of in the internal volume + interestingly enough , in string theory the degeneracy of excitation states also shows the same exponential growth .
this suggests an identification among highly excited string states and bh micro - states .
however , a more precise stringy formulation of bhs requires a non - perturbative formulation of string theory itself , including the interaction with higher dimensional objects , i.e. d - branes .
so far , this approach provides the correct entropy counting bh micro - states at least in a limited number of cases , but not yet in general .
+ in this chapter we would like to review a `` phenomenological '' approach taking into account the most fundamental feature of string theory or , more in general , of quantum gravity , whatever its origin , which is the existence of a _ minimal length _ in the space - time fabric @xcite .
this length is , so far , identified with the planck length , or the string length , but nothing prevents it from being much longer , even down to the tev region . a simple and effective way to keep track of the modifications of bh geometries , due to
the presence of the minimal length , is to solve the einstein equations with an energy momentum tensor describing non point - like matter .
+ at first glance , one could think of modifying the four dimensional einstein action to incorporate minimal length effects . to solve the modified field equations
can be quite complicated .
this is the case , where minimal length is introduced in the einstein theory through the , so - called , `` star - product '' , i.e. the standard product between fields is replaced by a non - commutative product operation making even the simplest field theory non - local and impossible to solve .
the only way to handle non - local field theories is in a truncated perturbative expansion suppressing non - locality and introducing spurious derivative interactions . in the resulting approximate theory ,
the effect of the minimal length as a natural short - distance cut - off is lost and paradoxically infrared divergences appear even in the massive case and mix with the persisting ultraviolet divergences .
a typical example of a `` ... cure killing the patient ! ''
. + the star - product approach , while formally correct is of little practical use .
for example , bh perturbative solutions of `` starred - product '' einstein equations , manifest the same pathological short distance behavior of the corresponding standard solutions , in spite of introducing a short - distance cut - off @xcite .
+ we argued that an alternative approach is to implement the minimal length _ only _ in the matter source .
the line of reasoning is the following .
the metric field is a geometrical structure defined over an underlying manifold and the curvature is a measure of the strength of the metric field .
thus , it is the response to the presence of a mass - energy distribution .
the minimal length is an intrinsic property of the manifold itself , rather than a super - imposed geometrical structure .
it affects gravity in a subtle , indirect way : it smears matter eliminating point - like objects as physical sources of the gravitational field . as a matter of fact
, this idea goes hand - in - hand with the fundamental idea of string theory : `` particles '' are not matter - points , rather they are extended , one dimensional objects , whose length is of the order of the planck length , or any other fundamental length scale . following the above reasoning , we conclude that in general relativity the effects of the minimal length can be taken into account by _ keeping the standard form _ of the einstein tensor in the l.h.s . of the field equations and introducing a _ modified energy - momentum tensor _ as a source in the r.h.s .
+ at this point one may object that the textbook approach defines bhs as `` matter - free '' solutions of the field equations . even in the case of the reissner - nordstrm bh , the energy momentum tensor _ only _ describes the electric field .
+ this way of introducing bhs is , at least , misleading and contradicts the fundamental ideas of general relativity that the geometry of space - time is determined by the energy - momentum distribution.the _ only _ globally defined matter - free solution of the einstein equations is minkowski space - time .
furthermore , solving `` in vacuum '' faces the problem of _ a posteriori _ determination of various integration constants .
a crystal clear example of this difficulty is the kerr solution , where integration constants are fixed by comparison with the geometry of a slowly rotating sphere , while everybody is aware that the rotating sphere is _ not _ the source of the the kerr gravitation field !
+ furthermore , the symmetry of the `` vacuum '' solution , has to be imposed _ a priori _ ,
since there is noting in the field equations that can do the job for you .
+ all the above mentioned ambiguities disappear once a proper matter source is introduced in the einstein equations .
a well known example , in quantum field theory , of difficulties related to modeling elementary particles as structureless point like objects , is the appearance of ultraviolet infinities .
when computing one - loop feynman graphs , one finds infinite , unphysical , amplitudes for measurable processes .
the simplest `` cure '' of the infinity disease is to introduce a suitable short - distance cut - off , which in coordinate - space is simply a `` minimal length '' .
once divergent parts of the amplitudes are subtracted away , the ( arbitrary ) cut - off is replaced by sum physically meaningful quantity through a `` re - normalization '' , i.e. re - parametrization , of the theory .
+ following this philosophy , we propose to re - normalize the curvature singularity in a bh geometry in a similar fashion .
+ first , let us cut - off the singularity by introducing a minimal length into the line element .
+ second , relate the minimal length to the radius of the smallest physically meaningful,_extremal _ bh . +
this two - step procedure will be first applied to the simples case of schwarzschild bh . in order to remove the curvature singularity in @xmath3
, we introduce the simplest kind of short - distance cutoff through the replacement @xmath4 where , @xmath5 is , for the moment , an arbitrary length scale . according with the framework one has in mind
, @xmath5 can be thought of as the characteristic length scale of the underlying quantum theory of gravity ( whatever it is ) , i.e. the planck length , the string length , etc . in the technically much more involved approach of `` asymptotically safe '' quantum gravity , a similar `` cutoff identification ''
was introduced in @xcite for the same purpose.,[multiblock footnote omitted ] by skipping all the approach - dependent technicalities we write a regularized schwarzschild metric as @xmath6 where @xmath7 from the equation ( [ massar ] ) one sees the @xmath8 at large distance , i.e. @xmath9 , while in the opposite limit @xmath10 . as a consequence the metric approaches the standard form of the schwarzschild line element at large distance , and the desitter metric @xmath11 at short distance , with an effective cosmological constant @xmath12 .
the desitter geometry is known to be curvature singularity - free . as a matter of fact
, we have replaced a physically meaningless infinite curvature point with a central desitter condensate , where the curvature can be very high but finite .
the physical mechanism leading to the disappearance of the singularity is clear , as the desitter vacuum exerts a _ negative _ pressure balancing the gravitation pull towards the center .
thus , close to the origin , collapsed matter is supported by the negative pressure of the desitter vacuum and can attain an equilibrium configuration .
+ step two , requires to identify the physical meaning of @xmath5 through an in depth study of the bh geometry .
+ the eventual horizons are obtained from the equation[multiblock footnote omitted ] : @xmath13 plot of the function in equation ( [ horeq0 ] ) , width=302 ] the function ( [ horeq ] ) has a single minimum in @xmath14 describing an `` _ _ extremal _ _ '' bh in the sense that for any @xmath15 , the radius of the outer horizon , @xmath16 , is larger than @xmath5 .
the extremal bh can be understood as the thermodynamical equivalent of the `` minimal volume '' of a real gas @xmath17 . as it known , the entropy of the gas is of the form @xmath18 leading to @xmath19 for @xmath20 . in the bh case ,
the role of the gas with volume @xmath17 is played by the extremal bh which is expected to have zero entropy in agreement with the third law of thermodynamics . + a second , inner horizon @xmath21 , exists for any non - extremal configuration ( see fig.([mass ] ) ) . @xmath22 and @xmath21 merge into @xmath23 for @xmath24 . for @xmath25 no bh exists
thus , @xmath5 has been promoted from an arbitrary parameter with length dimension , to radius of the horizon of the minimal size extremal bh .
@xcite , there is no physical way to probe sub - planckian distances @xcite . ]
+ contrary to the case of the `` bare '' schwarzschild bh , the renormalized metric admits and extremal configuration even in the absence of electric charge and angular momentum .
+ the existence of a de sitter region at short distance , indicates the presence of a non - trivial energy - momentum tensor sourcing the renormalized metric ( [ rbh ] ) .
this energy - momentum tensor must approach a cosmological form at short - distance .
the energy density for a spherically symmetric mass distribution is given by @xmath26 let us remark that the invariant energy dentity is finite everywhere . in particular @xmath27 and this is a necessary condition to avoid the appearance of a curvature singularity in the origin .
+ the complete energy momentum tensor encodes energy and pressure distributions of an anisotropic fluid @xcite given by @xmath28 where , @xmath29 , @xmath30 , and @xmath31 follows from the energy - momentum tensor covariant divergence - free condition .
@xmath32 is finite everywhere as well .
thus , we have a regular source leading to the everywhere finite curvature geometry described by the line element ( [ rbh ] ) .
the thermodynamical description of bh starts with the computation of the hawking temperature defined as the surface gravity of the killing horizon over @xmath33 : @xmath34 in our case , @xmath35 and ( [ th ] ) leads to @xmath36 first , we notice that @xmath37 vanishes at the extremal configuration , @xmath38 , which eliminates the pathological behavior of the schwarzschild solution at the final phase of the evaporation . @xmath37 increases with increasing radius up to a maximum value at @xmath39 , then approaches zero as @xmath40
. + plot of the hawking temperature as a function of the horizon radius.,width=302 ] another important thermodynamical quantity assigned to the bh is its entropy .
after the bekenstein / hawking results , it is customary to identify the entropy with @xmath41 of the horizon area in planck units
. suppose we have a planckian bh with radius and mass given by @xmath42 where , @xmath43 and @xmath44 are the planck length and mass . continuing the previous discussion , the extremal bhs correspond to the molecular volume . in a statistical description
the extremal bh should correspond to a macro - state realized by a _
single _ micro - state and , thus , has zero entropy .
+ apart from the general assumption of the validity of the area law , one should be able to calculate it in the same way as it is calculated for a real gas , i.e. from the first law of thermodynamics . in the case of a bh
, the first law is given by : @xmath45 from ( [ 1law ] ) one can derive that @xmath46^{-1}_{r = r_+ } dr_+ \label{ds}\ ] ] for the metric ( [ rbh ] ) one finds @xmath47 where , @xmath48 for large bhs , far away from extremality , @xmath49 , we recover the standard area law @xmath50 , but as the `` quantum '' regime is approached volume corrections become important and can not be neglected . +
it is crucial to remark that the integration of ( [ ds ] ) , leading to ( [ entropy ] ) , is bounded from below by the radius @xmath23 of the extremal configuration as there are no smaller bhs .
+ thus , we obtain the generalized `` area law '' which also gives the form of the entropy satisfying the third law .
it contains further corrections induced by the presence of the minimal length . for the `` bare '' schwarzschild metric
there is no extremal , minimal size , configuration and the corresponding entropy is simply on fourth of the area . assuming that this is a `` universal '' behavior for all bhs , leads to an inconsistency with the third law , i.e. zero - temperature extremal bhs with non - zero entropy .
thus , even in the absence of a minimal length , whenever a bh admits an extremal configuration , the correct way to calculate the entropy is by integrating the first law from the minimal radius of the extremal configuration and not from zero . the correct definition of entropy is given by the difference of the area of the non - extremal configuration and the extremal one . in this subsection we are going to study eventual phase transitions of regular neutral bhs .
we follow the analogy with finite temperature quantum field theory where different vacuum phases are studied in terms of the stationary points of effective potential .
the order parameter is the vacuum expectation value of some scalar field operator expressed as a function of the temperature .
different phases correspond to different vacuum expectation values .
+ in our case off - shell free energy , @xmath51 plays the role of the effective potential while the order parameter is represented by the radius of the bh .
different phases correspond to different size bhs .
the advantage of using the _ off - shell free energy _ is that @xmath52 is a free parameter describing the evolution from non - equilibrium , @xmath53 , to @xmath54 states when the bh is in equilibrium with the surrounding thermal bath .
@xmath55 we look for the extremal of @xmath51 which corresponds to equilibrium configurations .
@xmath56 the extrema of ( [ fofff ] ) are solutions of the condition @xmath57 equation ( [ fprimo ] ) determines the free parameter @xmath52 as a function of @xmath22 @xmath58 proving the general property that the extrema of @xmath59 corresponds to bhs in thermal equilibrium with the surrounding heat bath
. + if we were able to invert ( [ teq ] ) we could find the way in which the order parameter @xmath22 evolves by varying @xmath52 , as in the finite temperature quantum field theory .
unfortunately , the equation is fourth order and can not be easily and transparently solved .
however , we can obtain simple analytic solutions by considering the near - extremal and large radius limits for low temperature bhs .
+ for near extremal configurations we find @xmath60 while , for large bhs , away from extremality , we find @xmath61 at low temperature @xmath62 and @xmath63 are the local minimum and local maximum of @xmath59 .
thus , near - extremal bhs are classically stable , while large bhs are unstable and decay either towards extremality or grow indefinitely without ever reaching the equilibrium .
+ the two extrema merge at a critical temperature @xmath64 where both ( [ fprimo ] ) and @xmath65 vanish .
one finds that this happens for @xmath66 and @xmath67 plot of the off - shell free energy for different values of the temperature @xmath52 .
, width=302 ] figure ( [ free ] ) summarizes the main conclusions of the previous analysis of phase transitions . at zero temperature
there is a single extremal bh in equilibrium with the surrounding vacuum . as the temperature increases , even slightly ,
the free energy develops a local minimum and a local maximum , the first corresponds to a small , near - extremal , classically stable bh , while the second is a large classically unstable bh .
in fact , the free energy is unbounded from below for large bhs which are colder than the surrounding heat bath . as a consequence they continue to grow never reaching an equilibrium configuration .
this behavior follows from the `` competition '' between the internal energy end entropy entering @xmath59 with opposite sign .
@xmath68 increases linearly with @xmath22 while the entropy increases with @xmath69 .
thus , the negative entropy term dominate over the positive internal energy contribution beyond some critical size .
this pathological behavior will be cured in the next section by introducing an anti de sitter background whose negative ( inward ) pressure will stop this unlimited expansion . +
a further increase in @xmath52 leads to a critical value , @xmath64 , where the maximum and the minimum merge into an inflexion point .
remarkably , @xmath70 , where only one unstable bh exists .
+ for @xmath71 bhs do not exists .
+ the critical behavior of the system can also be inferred from the form of the heat capacity : @xmath72\left [ r_+^3 -l_0 ^ 3 \left(\ , 5 - 3\sqrt{3}\,\right ) \right]}\ ] ] @xmath73 diverges for @xmath74 , it is positive for @xmath75 , and is negative for @xmath76 .
the near - extremal region has @xmath77 and bhs show the normal , stable , thermodynamical behavior , while in the region @xmath78 the anomalous behavior described above takes place , i.e. increasing the total mass @xmath68 lowers the temperature triggering a limitless growth .
the results obtained in the previous discussion indicate that there is a minimal size ( extremal ) bh even of planckian dimension , as well as the corresponding minimal area .
thus , we are led to an interesting conjecture , i.e. an holographically improved `` bohr quantization '' of the bh . instead of quantizing the mass of the bh , we rather quantize its area in terms of the minimal area of the extremal configuration @xmath79 it follows from ( [ an ] ) that the radius of the horizon increases according with @xmath80 by inserting ( [ rn ] ) into ( [ horeq ] ) we obtain the quantized mass spectrum as : @xmath81 for large @xmath82 .
highly excited bh states have mass given by @xmath83 and the difference between successive mass levels vanishes as @xmath84 .
this is the region where the thermal picture for the hawking radiation makes sense since the mass levels become practically continuous . + on the other hand , in the truly quantum regime transitions occur discontinuously through emission of single quanta @xmath85 given by the mass difference between nearby levels : @xmath86\equiv\hbar \omega_n\ ] ] in this picture , the final stage of bh decay resembles more the discontinuous spectra of the atomic transitions , than a thermal radiation from an hot body .
the unboundedness from below of the free energy , we found in the previous section , is specific to the asymptotically flat boundary conditions satisfied by the the metric ( [ rbh ] ) . to cure
the bh instability towards a limitless growth one introduces a negative cosmological constant in the einstein equations and solve them with the energy - momentum tensor ( [ source ] ) . + from a physical point of view , a negative cosmological constant represents a positive ( inward pushing ) vacuum pressure @xmath87 making an unbounded inflation energetically disfavored . +
a further motivation for studying ads bhs is that in higher dimensions they have a pivotal role in the implementation of the @xmath88 duality @xcite . this kind of duality offers a powerful tool to tackle non - perturbative features of a variety of physical systems ranging from the quark - gluon plasma @xcite to fluids @xcite and super - conductors @xcite .
the strong - coupling regime of a conformal field theory living on the flat boundary of @xmath89 is mapped by duality into the weak - coupling quantum string theory ( quantum gravity ) in the @xmath90 bulk .
this amazing spin - off of string theory connects @xmath91 physics in flat space - time to quantum gravity in @xmath89 and provides a beautiful realization of the holographic principle @xcite .
+ sticking to our toy - model , we find the line element @xmath92 where @xmath93 and @xmath94 plot of the function in ( [ horeq ] ) .
, width=302 ] we have introduced a new free parameter in the model , i.e. the vacuum pressure , which will play , together with the temperature , an important role in determining the phases of the system .
+ the two length scale @xmath5 and @xmath95 rule the short and large distance behavior of the metric , respectively .
the short distance form of the metric is again of the type @xmath96 where @xmath97 as the three parameters @xmath68 , @xmath5 , @xmath95 are free , @xmath98 can be positive , negative or even zero .
in any case the metric is singularity free being either desitter , anti desitter , or minkowski . +
the existence of horizons can be seen plotting the bh mass given by @xmath99 as a function of @xmath22 .
the plot is given in figure ( [ mad ] ) @xmath100\ ] ] the temperature is given as @xmath101 plot of the hawking temperature for different values of the vacuum pressure .
the inflexion point marks the transition between low / high pressure regimes .
, width=302 ] the entropy turns out to be @xmath102
@xmath103 \left(\ , 1 + \frac{r_+^2}{a^2}\,\right ) \nonumber\\ & & - \frac{\pi t } { g_n}\left[\ , r_+^2 - l_0 ^ 2 + l_0 ^ 3 \left(\ , 1 + \frac{3 l_0 ^ 2}{a^2 } \,\right ) \left(\ , \frac{1}{l_0 } -\frac{1}{r_+ } \,\right)\,\right ] \label{foff}\end{aligned}\ ] ] plot of the off - shell free energy for different values of the temperature @xmath52 in the low - pressure regime showing the existence of different phases . the critical curve , for @xmath104 is shown in magenta . ,
width=302 ] let us note that ( [ foff ] ) for large @xmath22 can be approximated with @xmath105 equation ( [ large ] ) shows how large radius configurations are energetically disfavored as they imply a net increase of the free energy .
the negative increase from the entropy term is no more able to compensate for positive increase of internal energy .
@xmath106 \label{fpprimo}\end{aligned}\ ] ] as previously discussed , @xmath107 .
the extrema of free energy indicate existence of both _
multiple _ and _ single _ regular bhs for different values of the temperature .
the alternation of single / multiple states is the signature of a first order phase transition , as in finite temperature quantum field theory , to which we refer .
+ it turns out that single / multiple bh transitions occur at the inflection points of free energy ( extremal points of @xmath37 ) .
thus , the following scenario is in place in the low - pressure regime @xmath108 : + 1 .
bh is in the _ frozen single state_. the only ground state is the extremal configuration with @xmath110 .
2 . @xmath111 .
bhs are in the _ cold single _ state of radius @xmath112 .
this is due to the effect of the minimal length @xmath5 .
3 . at @xmath113
an inflexion point appears in @xmath114 at @xmath115 .
4 . @xmath116 .
new local minimum develops and the system splits into two co - existing states .
the small near - extremal bh is energetically favored .
5 . @xmath117
the two minima become degenerate and the system is in a _
mixed state_. both bhs have the same free energy .
@xmath118 large bhs become stable , while near - extremal bhs are only locally stable .
@xmath119 the near - extremal minimum merges with the local maximum .
there is a new transition from multiple to a single bh state .
@xmath120 there is _ high temperature , single , stable _ bh . the above scenario describes _
first order _ phase transitions from single to multiple bhs at @xmath113 and @xmath119 .
plot of the off - shell free energy for different values of the temperature @xmath52 in the high - pressure regime , @xmath121 .
, width=302 ] one can see that the first order phase transitions take place in the low vacuum pressure regime , only .
in fact , the two parameters of the theory have `` opposite '' effects i.e. @xmath122 dominates short - range distance behavior and lowers @xmath123 , while @xmath124 dominates long - range region of @xmath22 and thus raises @xmath125 .
+ it is reasonable to expect that , at the certain point , these two opposing effects will meet creating an _ inflexion _ point of the temperature .
the confirmation of our conjecture is shown in fig.([tempads ] ) .
+ in terms of the free energy the same effect can be seen in the figure([fasi2 ] ) , where a single minimum exists beyond the inflexion point @xmath126 .
varying @xmath52 only lowers the _ position _ of the _ single _ minimum . +
a complex phase structure is a general feature of several types of bh where some characteristic length scale is present in the metric @xcite .
in the first part of this chapter we introduced an `` ad hoc '' method of eliminating curvature singularities using ideas usually adopted in quantum field theory , i. e. introducing a suitable short - distance cut - off . however , there is another approach in qft achieving the same goal but having a profound physical significance .
for example , uv infinities in feynman diagrams ( beyond the tree - level ) can be eliminated by replacing the bare ( euclidean ) propagator with an exponentially damped one @xcite @xmath127 the meaning of the above substitution is to replace a divergent two - point green function ( in coordinate space ) with a regular one solving the inhomogeneous equation @xmath128 equation ( [ reggreen ] ) shows that physical particles with propagator ( [ regprop ] ) are not matter `` points '' , but are smeared gaussian energy distributions . in quantum mechanics gaussian wave packets represents minimal uncertainty states , i.e. the closest one can get to a classical particle . + the approach of substituting point - like particles with gaussian matter distributions has been carried out in a number of papers in order to describe quantum mechanics and in qft in _ coordinate _ non - commutative space(time ) , characterized by @xmath129= i \theta^{\mu\nu } \label{comm}\ ] ] where @xmath5 can be related to lorentz invariant quantity @xmath130 it is already widely accepted that space - time at short distances is no more modeled by a smooth manifold but something completely different .
our ignorance about the space - time planckian phase leaves room for different hypothesis , e.g. string , loop , fractal , non - commutative , foamy phases , etc .
, all sharing the existence of a characteristic length scale .
the introduction of @xmath5 as the minimal width of a gaussian distribution is motivated by non - commutativity of coordinates ( [ comm ] ) , much like the non - commutativity of phase space coordinates in qm is characterized by @xmath131 . + by accepting this idea , one wonders how to implement it in the case of gravity . at first glance , one could think of modifying the very definition of the metric tensor to incorporate @xmath5 in the space - time fabric , e.g. replacing the ordinary product of functions by the star - product .
this approach , apart heavily complicating einstein equations , has the basic flaw that any perturbative expansion in theta , truncated at a finite order , leads to the loss of non - locality of the original theory .
the resulting solutions contain all the pathologies of the commutative theory ( curvature singularities ) in spite of having introduced @xmath5 from the very beginning .
+ alternatively , we argued that instead of changing the space - time geometry the effects of @xmath5 can be implemented through the matter source .
the line of reasoning is the following .
metric field is a geometrical structure defined over an underlying manifold .
curvature measures the strength of the metric field , i.e. is the response to the presence of a mass - energy distribution . on the other hand ,
energy - momentum density determines space - time curvature .
thus , we conclude that in general relativity the effects of @xmath5 can be taken into account by keeping the standard form of the einstein tensor in the l.h.s . of the field equations and introducing a modified energy - momentum tensor as a source in the r.h.s .
+ thus , we choose the mass density of a static , spherically symmetric , smeared , particle - like gravitational source as @xcite @xmath132 by solving einstein equations with ( [ ro ] ) as a matter source , we find the line element . ]
@xmath133 where , @xmath134 strictly speaking , the density ( [ ro ] ) is non - vanishing everywhere , even if it quickly drops below any measurable value at few orders of @xmath5 .
however , this may rise the question if a bh can be formed by such a smeared distribution . in order to answer this question we evoke the _ hoop conjecture _ @xcite and adapt it to the present situation .
it means that we define a _
mean radius _ of the mass distribution as @xmath135 the hoop conjecture assumes that whenever , for a given total mass @xmath68 , @xmath136 where , @xmath137 is the radius of the eventual killing horizon determined from ( [ nostra ] ) , the metric ( [ nostra ] ) will describe a bh .
plot of the mass @xmath68 as a function of the radius @xmath137 of the horizon .
, width=302 ] the analysis of the function @xmath138 shows ( see fig . ( ) ) that it admits an inner , @xmath21 , cauchy and an outer , @xmath139 killing horizon , which merge into a single degenerate horizon , @xmath140 , in the extremal case .
the radius of the extremal bh is again @xmath141 , as in the toy - model case , while the mass is @xmath142 thus , we find @xmath143 which show that even in the extremal case the mean radius of the mass distribution is smaller that the radius of the horizon .
enough mass is confined within the horizon in order to sustain the existence of the bh .
+ furthermore , the metric ( [ nostra ] ) exhibits the short - distance behavior of the desitter form with no singularity at the origin .
this can be inferred from the behavior of the incomplete gamma function @xmath144 for small argument @xmath145 a closer look at equation ( [ ro ] ) suggests an intriguing analogy with the probability density of the the ground - state wave function of an isotropic , 3d , harmonic oscillator @xmath146
where , @xmath147 is the ground state wave function . in order to identify @xmath148 and @xmath149 with the corresponding quantities in ( [ ro ] ) , we notice that @xmath150 and the mass of the extremal bh represents the minimum energy of the equivalent harmonic oscillator , i.e. @xmath151 is the zero - point energy @xmath152 by solving the two equations ( [ omega ] ) , ( [ emme ] ) we find @xmath153 the non - extremal bh quantized configurations are assumed to be , the @xmath154 excited states of the corresponding harmonic oscillator : @xmath155 due to the spherical symmetry only even oscillator states are allowed .
thus , in this new formulation of quantum bhs , the extremal configuration is pure _ collapsed zero - point energy _ , and the excited states are non extremal bhs with discrete masses given by ( [ hspectrum ] ) . + the analogy with the harmonic oscillator @xcite has to be reconciled with the requirement that even excited bh states keep the simplest geometric structure consisting of either a single extremal configuration or a non - extremal bh having one internal cauchy horizon and one external killing horizon only . to achieve this , instead of using the complete excited harmonic oscillator wave function
, we propose to keep only the highest power of the laguerre polynomials in the energy density ) has been recently used to model the final stage of the gravitational collapse of a `` thick shell '' of matter @xcite . here
we shall use the same type of distribution in a different way . ]
@xmath156 therefore , the effective geometry describing `` quantized '' bhs is still described by the equation ( [ nostra ] ) with the exception that the continuous parameter @xmath68 is replaced by its discrete version given by ( [ hspectrum ] ) @xmath157 for each @xmath82 the existence of the horizons is given by @xmath158 however , bh being described as a quantum system one has to verify the average radius of the mass distribution is inside the corresponding horizon radius .
this is known as the quantum `` hoop conjecture '' .
plot of the ground state @xmath159 ( black ) and the first excited state @xmath160 ( red ) .
horizontal lines correspond to the quantized masses @xmath151 , @xmath161 .
solid vertical lines are the killing horizons . dashed vertical lines represents @xmath162 of the mass distribution for @xmath163 .
they show as the hoop conjecture is satisfied.,width=302 ] @xmath164 @xmath165 the figure ( [ spettro ] ) below summarizes the main characteristics of the quantized bhs described above .
the difference between the classical bh given in figure ( [ mass ] ) .
while in figure ( [ mass ] ) horizons are obtained making an arbitrary cut parallel to the @xmath137-axis , since @xmath68 is a continuous parameter , in the quantized case we have distinct graphs for any value of @xmath82 and for each curve a _
single _ horizontal line @xmath166 . only for @xmath159
there is a an extremal bh , while for @xmath167 we have non - degenerate bhs corresponding to excited states . for any value of @xmath82
the hoop conjecture , @xmath168 , is satisfied .
in this chapter we have described a generalization of the standard schwarzschild geometry which takes into account the presence of a `` minimal '' length @xmath5 which should be present in a quantum , to be , formulation of gravity .
it is widely believe , from different points of view , that such a parameter should be necessarily incorporated in a physically meaningful formulation of quantized gravity . in the first part
, we have restricted ourself to simple , toy - model , of regular bh in order to display novel features of such a theory in a relatively transparent way .
it has been claimed @xcite that the way in which @xmath5 is introduced is not essential and the same features are common to different models .
in fact , we showed in the second part that a more complicated and realistic model shares basic features with the simpler toy - model .
the effects of @xmath5 , in both models , are : * to replace the curvature singularity in @xmath3 by a regular desitter core ; * to introduce a lower limit to the bh mass in the form of an extremal configuration , even in the absence of charge and angular momentum ; * to identify the radius of the extremal configuration with @xmath5 , in agreement with the uv self - complete quantum gravity hypothesis , where sub - planckian distances are screened by ( extremal ) bh formation ; * to provide a consistent final state for bh decay through hawking radiation in the form of zero temperature extremal bh remnant ; * to remove unphysical negative heat capacity during the final stage of bh evaporation ; * to introduce corrections to the area law which is strictly valid for point - like matter sources only . from the technical point of view
, @xmath5 was introduced first in the source term in the einstein equations through the properly chosen smeared matter distribution . in this picture
@xmath5 measures the de - localization , of the otherwise point - like source , induced by quantum gravitational fluctuations of the underlying space - time manifold .
the important advantage of this approach is that one can obtain _ exact _ solutions of the einstein equations .
alternative approaches , e.g. based on star - product , can not achieve the same goal since one is forced to perform truncated perturbative expansion in @xmath5 suffering from the same pathologies @xmath5 is expected to have cured . +
the use of a gaussian energy distribution suggests an intriguing analogy with the ground state of the quantum harmonic oscillator .
this has led us to envisage a possible bohr - like quantization scheme of our classical regular solution .
this has been described in the last part of this chapter , where the following quantum picture emerges : * the ground state is an extremal bh of size @xmath5 , with mass @xmath151 , which is the analogue of the harmonic oscillator zero - point energy .
* excited states are non - extremal bhs with mass @xmath169 .
* the semi - classical hawking picture of thermally radiating bhs remains valid for large @xmath82 , while it becomes modified for small @xmath82 . in the truly quantum regime
the continuous thermal spectrum turns into a discontinuous spectrum corresponding to the transitions between low lying mass levels .
* is it self - consistent to interpret the above excited states as bhs ?
+ our answer is yes , because they satisfy the quantum version of the `` hoop - conjecture '' , i.e. the mean radius of the mass distribution is always smaller that the corresponding radius of the horizon . in view of the quantum picture
described above , one can envisage a quantum @xmath170 of mass @xmath171 as a `` bound - state '' of @xmath172 quanta of energy @xmath173 .
the ground - state , or `` _ _ zero - point _ _ '' bh is given by the extremal configuration with @xmath159 and @xmath174 .
+ in the last part of this chapter , we were interested in the quantum aspects of regular bhs and have not described the thermodynamical features of the metric ( [ nostra ] ) , which however can be found in @xcite .
in section ( [ toy ] ) we regularized the newtonian potential through the substitution ( [ uno ] ) . as the coulomb potential `` suffers '' from the same `` illness '' in @xmath3 ,
one is tempted to regularize it in the same way .
, i.e. we propose the following substitution in the presence of a minimal length , the attractive coulomb potential develops a minimum at @xmath179 and a central hard core in an analogous manner as the newtonian potential in the gravitational case .
+ a more physically motivated regularization starts from replacing a point - like charge density with a smeared gaussian like distribution given by | semi - classical description of bh , as it was originally introduced by hawking and bekenstein in the early seventies , where classical solutions of the einstein equations are coupled to quantum matter fields , opened for the first time a window with a glance on the quantum aspects of gravity .
the surprising properties showed by semi - classical description of black holes ( bhs ) called for a full quantum description for such objects . in particular , the understanding of bh entropy in terms of micro - states , as well as the behavior during the final phase of evaporation , can not be properly described in semi - classical terms .
quantum gravity has a long story of doomed attempts to cure the theory from non - renormalizability and quantum anomalies . in the eighties
green and schwarz brilliantly solved these problems in the framework of @xmath0 or @xmath1 super - symmetric quantum string theories . up to now
, string theory is the best candidate for a grand unified theory including gravity at the quantum level in a consistent way . accepting that the building blocks of matter are one - dimensional extended objects , characterized by a length scale of order @xmath2 , still the relation between fundamental strings and bhs remains to be determined .
the exponential increase of string states degeneracy with the string excitation energy resembles the exponential increase of bh states with the increase of bh mass .
this and other results suggest a correspondence between bhs and highly excited strings , or string balls .
however , a more precise stringy formulation of bhs requires a non - perturbative formulation of the theory itself , including the interaction with higher dimensional objects , i.e. d - branes .
so far , this approach provides the correct entropy counting bh micro - states only in a limited number of cases , but not in general . in this chapter
we would like to review a `` phenomenological '' approach taking into account the most fundamental feature of string theory or , more in general , of quantum gravity @xcite , whatever its origin , which is the existence of a minimal length in the space - time fabric @xcite .
this length is generally identified with the planck length , or the string length , @xcite but it could be also much longer down to the tev region . a simple and effective way to keep track of the effects the minimal length in bh geometries is to solve the einstein equations with an energy momentum tensor describing non point - like matter . the immediate consequence is the absence of any curvature singularity . where textbook solutions of the einstein equations loose any physical meaning because of infinite tidal forces , we find a de sitter vacuum core of high , but finite , energy density and pressure .
an additional improvement regards the final stage of the bh evaporation leading to a vanishing hawking temperature even in the neutral , non - rotating , case . in spite of th simplicity of this model we are able to describe the final stage of the bh evaporation , resulting in a cold remnant with a degenerate , extremal , horizon of radius of the order of the minimal length . in this chapter
we shall describe only neutral , spherically symmetric , regular bhs although charged , rotating and higher dimensional bhs can be found in the literature @xcite semi - classical approach to quantum black holes euro spallucci and anais smailagic department of theoretical physics , university of trieste , strada costiera 11 , 34014 trieste , italy , and sezione infn di trieste , strada costiera 11 , 34014 trieste , italy |
all cosmic processes like the big bang , the formation and evolution of galaxies , the creation and the growth of massive black holes and the heating of the universe due to large - scale clustering will imprint characteristic radiation patterns on the electromagnetic spectrum of the universe .
the structure of the universe at the epoch of decoupling is frozen into the 2.7 k cosmic microwave background ( cmb ) .
star and galaxy light , produced after the dark ages by thermonuclear fusion , is mainly confined to the near - infrared , optical and ultraviolet spectral bands .
active black holes in galactic nuclei ( agn ) , which accrete matter and efficiently convert gravitational energy into radiation , typically shine in a very broad energy band , from radio to gamma wavelengths . the role of dust is also very important : it absorbs optical , uv and x - ray light , is warmed up and re - radiates the energy in the far - infrared band .
the extragalactic background radiation , i.e. the total cosmic energy density spectrum , therefore gives a fossil record of all radiation processes in the universe , integrated over cosmic time .
in the recent years , the extragalactic background spectrum has been determined with relatively high precision over a very broad range of the electromagnetic spectrum , apart from some inaccessible regions .
figure [ echos ] shows a compilation of most recent determinations of the cosmic energy density spectrum from radio waves to high - energy gamma rays the `` echos from the past '' .
apart from the cmb peak , which clearly dominates the energy budget of the universe , three distinct components can be identified in the spectral energy distribution : the cosmic far - infrared background ( cib ) , the cosmic optical / uv background ( cob ) and the cosmic x - ray / gamma - ray background ( cxb ) .
the cosmic x - ray background ( cxb ) radiation was the first extragalactic background emission to be announced@xcite .
in the seventies , x ray satellites like _ uhuru _ , _
ariel v _ and _ heao1 _ , which were able to scan the whole sky , produced high - quality xrb maps revealing a large degree of isotropy which immediately led to the conclusion that the origin of the xrb has to be mainly extragalactic and might therefore be of cosmological interest .
gamma - ray satellites like _ sas-3 _ , _ cos - b _ and finally the compton gamma ray observatory extended the background measurements to the highest - energy gamma rays ( see @xcite for a review ) .
the energy spectrum of the cxb has a characteristic hump at about 30 kev and a long tail towards higher energies . at energies below 1 kev diffuse galactic emission
starts to dominate the sky . at 1/4 kev a large fraction of the background determined by shadowing
experiments has been resolved by deep _ rosat _ surveys@xcite . in the sub - mm to uv range the extragalactic background light
is completely swamped by other background components , i.e. the cmb , the interplanetary dust emission , the interstellar dust emission , scattering from the interplanetary dust ( zodiacal light ) , galactic starlight and scattering by interstellar dust . in the far - infrared range
different groups have recently detected the long - sought residual cosmic far - infrared background ( cib ) signal@xcite by carefully modelling the other components in the high - frequency tail of the cosmic microwave background ( see fig .
[ echos ] ) . in the near - infrared to uv band , on the other hand ,
the hubble space telescope deep field images are so sensitive , that one can safely assume that most of the cosmic optical background ( cob ) has been resolved into discrete sources@xcite . in the mid - infrared range the bright zodiacal light does not allow a direct detection of the extragalactic background light and one has to resort to galaxy population synthesis models and relatively uncertain upper limits@xcite . at higher frequencies , in the extreme ultraviolet range
, the interstellar absorption basically inhibits the detection of the extragalactic light and the background level can only be estimated by interpolating between the uv and the soft x - ray measurements .
the extragalactic background light seems to be dominated by two distinct humps , one in the optical / near - infrared range ( cob ) , which could very well be produced by the red - shifted star light in all of the distant galaxies , and another one in the fir / sub - mm range ( cib ) , which is very likely produced by dust emission in distant galaxies .
the fact that the dust peak contains at least as much or even more energy than the stellar peak indicates that a large amount of light in the early universe had to be absorbed by dust .
however , what fraction of the dust is heated by starlight or by agn accretion power remains to be determined .
the x - ray range is one of the few regions of the electromagnetic spectrum , where on one hand the sky emission is dominated by the extragalactic background and on the other hand modern imaging telescopes are sensitive enough to resolve a substantial fraction of the background into discrete sources .
the aim of deep x - ray surveys is to resolve as much of the background as possible , then to obtain reliable optical identifications and redshifts of the x - ray sources and finally for all classes of objects to determine luminosity functions and their cosmological evolution ( see e.g. @xcite ) .
since the emitted light , integrated over cosmic time can be compared directly to the total x - ray background , important constraints on the cosmological evolution and possible residual diffuse intergalactic emission can be obtained . a large number of deep
, medium - deep and shallower surveys have been performed with _ rosat _ ( see @xcite for a review ) .
the _ rosat _
deep surveys in the lockman hole ( see [ hri]a ) have so far been the deepest x - ray observations ever performed ( until the deep _ chandra _ and _ xmm _ surveys will be completed ) .
they reach a source density of @xmath0 at the faintest flux limit , where 7080% of the soft x
ray background has been resolved into discrete sources@xcite .
deep r - band images have been obtained at the keck telescope . in the lockman hole
deep survey most optical counterparts have magnitudes in the range r=1823 . with the excellent hri positions typically only one or two counterparts are within the x ray error circle .
spectra of the optical counterparts of the x - ray sources were mostly obtained with the keck telescope .
currently 85 out of 91 x ray sources are spectroscopically identified . about 80% of those turned out to be active galactic nuclei@xcite , among them the highest redshift x
ray selected qso at z=4.45@xcite .
most agn are luminous qsos and sy1 galaxies with broad emission lines .
however about 16% show only narrow emission lines and are classified as seyfert 2 galaxies@xcite .
there is a group of x - ray sources in the lockman hole , which have optical counterparts fainter than r=24 , for which optical spectroscopy was not possible , even with the keck telescope . for these one
has to resort to nir photometry and photometric redshifts .
it turns out that the spectroscopically unidentified objects typically have very red r - k colours@xcite .
figure [ hri]b shows a correlation diagram between the r - k colours and redshifts for different types of objects identified in the lockman hole in comparison with normal galaxy colours predicted from evolutionary models .
the colours of type-1 agn are relatively blue and scatter widely , while type-2 agn and cluster galaxies show on average much redder colours , correlating well with redshift in the region expected for normal galaxies .
this indicates that the optical colours of type-2 agn are dominated by the light from the host galaxy , probably because the optical nucleus is obscured .
a population of heavily obscured agn is thus already showing up in the soft x - ray band . based on r - k colours and photometric redshifts
it is likely that the as yet unidentified sources are dust shrouded agn in the redshift range 1 - 2 .
there is , however , an interesting dichotomy in the redshift distribution of the sources .
unobscured objects ( solid circles ) are found to @xmath1 , while the obscured objects ( open circles , pluses ) are restricted to @xmath2 . there is a negative k - correction effect for obscured objects at soft x - rays : at higher redshifts it is easier to detect obscured x - ray sources , because their unobscured rest - frame hard x - ray emission is shifted into the observed soft x - ray band .
the fact that we do not find obscured objects at high redshifts could indicate a dependence of obscuration on intrisic source luminosity .
high luminosity x - ray sources to some degree could be able to clean out their environment from obscuring material .
the x ray background has a significantly harder spectrum than that of local unobscured agn .
this led to the assumption that a large fraction of the background flux is due to obscured agn , as originally proposed by setti and woltjer @xcite .
models following the unified agn schemes , assuming an appropriate mixture of absorbed and unabsorbed agn spectra folded with cosmological agn evolution models , can in principle explain the shape of the background spectrum over the whole x ray band ( e.g. @xcite ) , but gave problems with a number of new observational constraints@xcite .
apart from the agn cosmological evolution@xcite , the distribution of absorption column densities among different types of agn and as a function of redshift is one of the major uncertainties .
the standard x ray background population synthesis models ( e.g. @xcite ) assume that the absorption distribution , which has been determined observationally only for local seyfert galaxies , is independent of x ray luminosity and redshift .
in particular , the fraction of type-2 qsos , obscured high - luminosity x - ray sources , should be as high as that of seyfert-2 galaxies .
this assumption , however , has been proven wrong by recent observations ( see e.g. fig . [ hri]b , @xcite ) .
the current x - ray background models therefore still have rather limited predictive power . a significant improvement in our understanding of the hard x - ray background is to be expected from optical identifications of complete samples of sources selected at faint fluxes in the hard x ray band .
deep surveys performed with _ asca_@xcite and _ bepposax_@xcite have recently resolved about 30% of the harder x - ray background , but due to very limited angular resolution their optical identification is tedious and in the case of faint optical counterparts almost impossible . upcoming deep surveys with the _ chandra _ observatory ( _ axafi _ ) and _ xmm _ with very high sensitivity and good positional accuracy in the hard band together with optical identifications from the vlt and the keck telescopes are expected to yield a solid statistical basis to disentangle these various effects and lead to a new , unambiguous population synthesis for the x ray background .
an immediate consequence of the obscured background models is that the radiation produced by accretion processes in agn emerges completely unabsorbed only at energies well above 10 kev , thus producing the observed maximum of the xrb energy density at @xmath3 kev .
comparison between the background energy density at 30 kev and at 1 kev leads to the suggestion that most ( 8090% ) of the accretion power in the universe might be absorbed , implying a very large solid angle of the obscuring material as seen from the central source .
fabian et al .
@xcite suggest that circumnuclear starburst regions are responsible for the large covering factor
. they may be both triggering and obscuring most of the nuclear activity .
recently fabian and iwasawa @xcite have shown that these agn background synthesis models can explain the mass distribution of dark remnant black holes in the centers of nearby galaxies by conventional accretion which is largely hidden by obscuration .
the obscured background synthesis models has important consequences for the current attempts to understand black hole and galaxy formation and evolution
. the absorbed agn will suffer severe extinction and therefore , unlike classical qsos , would not be prominent at optical wavelengths .
the light of the optical counterparts of distant absorbed x - ray sources should therefore be dominated by the host galaxy , which indeed seems to be the case for a significant number of faint x - ray sources ( see fig [ hri]b ) .
if most of the accretion power is being absorbed by gas and dust , it will have to be reradiated in the fir range and be redshifted into the sub
agn could therefore contribute a substantial fraction to the cosmic fir / sub mm background which has already partly been resolved by deep scuba surveys .
the background population synthesis models have recently been used by almaini et al .
@xcite to predict the agn contribution to the sub
mm background and source counts . depending on the assumptions about cosmology and agn space density at high redshifts
they predict that a substantial fraction of the sub mm source counts at the current scuba flux limit could be associated with active galactic nuclei .
interestingly , the first optical identifications of scuba sources indicate a significant agn contribution@xcite .
another , largely independent line of arguments leads to the conclusion that accretion processes may produce an important contribution to the extragalactic background light .
dynamical studies@xcite come to the conclusion that massive dark objects , most likely dormant black holes , are ubiquitous in nearby galaxies .
there is a correlation between the black hole mass and the bulge mass of a galaxy : @xmath4 .
since gravitational energy release through standard accretion of matter onto a black hole is producing radiation about 100 times more efficiently than the thermonuclear fusion processes in stars , the total amount of light produced by accretion in the universe should be of the same order of magnitude as that produced by stars . a more detailed treatment following this argument comes to the conclusion that the bolometric agn contribution should be about 1/5 of the stellar light in the universe@xcite .
the ultraluminous iras galaxy ngc 6420 is a starburst galaxy with a double nucleus@xcite with some evidence for a weak active nucleus .
its position in the mid - infrared diagnostic diagram based on iso spectroscopy @xcite indicates that its bolometric luminosity , which is dominated by the far - infrared dust emission , should be mainly heated by star formation processes .
recently , however , _ bepposax _
hard x - ray spectroscopy revealed an underlying very luminous , though heavily obscured active galactic nucleus , which could well be responsible for the major fraction of the bolometric luminosity .
figure [ sed ] shows the spectral energy distribution ( sed ) of ngc 6240 in a @xmath5 representation from the radio to the hard x - ray band .
the sed peaks in the far - infrared dust band and shows the mid - infrared policyclic aromatic hydrocarbonic emission bands as well as the synchrotron radio emission typical for starburst galaxies .
the optical starlight peak is lower than the dust emission peak - similar to the overall extragalactic energy distribution . in the hard x - ray range
the nonthermal agn continuum emission is severely absorbed below 15 kev and strong reprocessed iron line is seen .
the uv to soft x - ray nuclear continuum is completely obscured .
the seds of three agn from the _ rosat _ deep survey in the lockman hole , which have been observed over a broad wavelength range from radio to x - rays have been superposed on this diagram .
these are ( see @xcite ) : 32a , a qso / sy1 at @xmath6=1.33 ( crosses ) , 12a a sy2 at @xmath6=0.99 ( filled circles ) and 14z , a very red object , probably a sy2 galaxy , with @xmath7@xcite .
the radio detections and upper limits at 20 cm and 6 cm are from @xcite and @xcite .
the upper limit in the millimeter - range for 14z has been obtained with iram@xcite . in the x - ray range
the _ rosat pspc _ spectra of these three sources have been fitted with a power law with galactic or intrinsic absorption . for a comparison on the same absolute scale the lockman seds
have been shifted to the redshift of ngc 6240 and scaled by the square of their luminosity distances .
the two absorbed _ rosat _ sources ( 12a and 14z ) have seds surprisingly similar to ngc 6240 .
the galaxy 12a may have a relatively stronger starburst component , indicated by its higher radio luminosity .
the main difference lies in the x - ray range , where the luminosity at hard x - ray energies is very similar for all 3 sources , while the intrinsic absorption is clearly different : @xmath8 is about @xmath9 for ngc6420 , @xmath10 for 14z and @xmath11 for 12a .
in contrast , the sed for the radio - quiet qso 32a , while fitting relatively well to the other sources in the radio band , is clearly dominated by the unobscured light of the active nucleus in the optical and x - ray range .
this figure is a nice illustration of the obscured agn model for the x - ray background and its potential relevance to the fir / sub - mm background , since it shows that x - ray - selected high - redshift agn can indeed have seds similar to local ulirgs .
unfortunately there are no observations yet , which are sensitive enough both in the x - ray band and in the sub - mm band to obtain a better quantitative estimate of the fraction of the far - infrared background produced by agn .
future joint sub
mm / x ray deep surveys will therefore be a very powerful tool to disentangle the different processes dominating the universe at high redshift .
i thank my collaborators in various projects , in particular f. bertoldi , p. ciliegi , r. giacconi , i. lehmann , m. schmidt , t. soifer , d. thompson , j. trmper , g. wilson and g. zamorani for the permission to show some material in advance of publication .
this research has made use of the nasa / ipac extragalactic database ( ned ) which is operated by the jet propulsion laboratory , california institute of technology , under contract with the national aeronautics and space administration .
this work has been supported in part by dlr grant 50 or 9908 0 . | deep x - ray surveys have shown that the cosmic x - ray background ( xrb ) is largely due to accretion onto supermassive black holes , integrated over cosmic time .
however , the characteristic hard spectrum of the xrb can only be explained if most agn spectra are heavily absorbed .
the absorbed agn will suffer severe extinction and therefore , unlike classical qsos , will not be prominent at optical wavelengths .
most of the accretion power is being absorbed by gas and dust and will have to be reradiated in the fir / sub - mm band .
agn could therefore contribute a substantial fraction to the recently discovered cosmic fir / sub - mm background .
here it is shown that a number of high - redshift absorbed x - ray sources selected in the _ rosat _ deep survey of the lockman hole have broad - band spectral energy distributions very similar to the local ulirg ngc6240 , lending additional support to the background models for the obscured universe . |
the effect of photon creation from vacuum in a cavity with rapidly varying geometrical or material properties , called nowadays as dynamical casimir effect ( dce ) , attracted attention of many researchers for a long time : see recent reviews @xcite .
the first experimental results for the open strip - line waveguide with time - dependent boundary conditions , simulating a single oscillating ideal mirror in the one - dimensional space , were reported recently in @xcite .
it is quite probable that new experiments in other geometries will be done soon , too .
therefore , the problem of the back action of different detectors on the rate of photon generation becomes actual .
the very first studies were performed in @xcite under the condition that the field - detector coupling is much stronger than the effective coupling between the selected resonance field mode in the cavity and moving boundary ( due to strong limitations on the attainable amplitude of surface oscillations in real materials ) . however , since that time there were proposed several schemes on simulating the dce by means of changing not the positions of boundaries but their material properties @xcite . in such cases the effective velocity of boundaries can be increased by several orders of magnitude , therefore the situations where the modulation depth of the instantaneous cavity eigenfrequency is bigger than the normalized field - detector coupling coefficient can be considered now quite realistic , as well .
there are two simple models of detectors .
one of them describes the detector as a two - level `` atom '' @xcite .
it can be applied for the experimental setup proposed in @xcite .
this case was extensively studied in different regimes in @xcite , and generalizations to the three- and multi - level `` atoms '' were considered in @xcite . here
we consider another simple model , where the detector is represented by a harmonic oscillator tuned in resonance with the selected field mode @xcite .
this model seems to be adequate to the case of the so called mir experiment @xcite , where the microwave quanta created via the dce are supposed to be detected by means of a small antenna put inside the cavity . since the inductive antenna ( a wire loop ) used in that experiment is a part of a lc - contour , it can be reasonably approximated as a harmonic oscillator .
therefore we study the quantum system described by the hamiltonian ( in dimensionless units ; in particular , we assume @xmath0 ) @xmath1 .
\label{htot}\ ] ] here @xmath2 and @xmath3 are the quadrature component operators of the selected resonance field mode , whereas @xmath4 and @xmath5 are the quadrature component operators of the detector ( antenna ) .
the interaction part of the hamiltonian is assumed to be proportional to the product @xmath6 @xcite in view of the standard minimal coupling term @xmath7 , assuming that operator @xmath2 is proportional to the vector potential of the field mode .
the instantaneous cavity eigenfrequency @xmath8 is chosen in the form @xmath9 , \label{parfreq}\ ] ] where @xmath10 is the unperturbed mode frequency and @xmath11 is the modulation frequency .
the real modulation depth @xmath12 is assumed to be small , as well as the real coupling coefficient @xmath13 .
the influence of all other cavity modes can be neglected if the spectrum of cavity eigenfrequencies is not equidistant @xcite . if @xmath14 , then one can diagonalize hamiltonian ( [ htot ] ) , introducing two normal modes , whose frequencies are given by the exact formula @xmath15 , so that for @xmath16 we have @xmath17 we assume that the modulation frequency @xmath11 is twice bigger than one of the two splitted normal frequencies .
we choose @xmath18 and normalize the frequencies and dynamical variables in such a way that @xmath19 .
our goal is to study the influence of the dimensionless ratio @xmath20 on the number of excitations ( `` photons '' ) in the field mode @xmath21 and in the detector @xmath22 for arbitrary ( small or big ) values of this parameter .
the paper is organized as follows . in sec .
[ sec - inv ] we derive the main dynamical equations and obtain their approximate analytical solutions in the weak coupling and weak modulation regimes , comparing these solutions with results of numerical calculations . in sec .
[ sec - pgv ] we use these solutions to calculate the time - dependent photon mean numbers for the initial vacuum state in both the cavity and detector modes . in sec .
[ sec - fluc ] we calculate the variances of the photon number operator , the photon distribution functions and the degree of squeezing in each mode .
the results are discussed in sec .
[ sec - con ] .
hamiltonian ( [ htot ] ) is a special case of generic quadratic hamiltonians ( see ref .
@xcite for a detailed treatment ) , which can be written as @xmath23 here @xmath24 is a symmetrical @xmath25 matrix , which can be splitted in four @xmath26 blocks as @xmath27 where the tilde ( ) means matrix transposition . in the case
involved we have @xmath28 for any quadratic homogeneous hamiltonian , the heisenberg operators @xmath29 are related to the initial operators @xmath30 by means of a linear transformation : @xmath31 ( the summation over repeated greek indices is implied ) , where coefficients @xmath32 form the time - dependent symplectic matrix @xmath33 .
it is more convenient , however , to use the inverse matrix @xmath34 .
( [ solheis ] ) can be rewritten as @xmath35 showing that @xmath36 is the operator integral of motion in the schrdinger picture .
matrix @xmath37 satisfies the following differential matrix equation and initial condition : @xmath38 where @xmath39 means the @xmath40 identity matrix and the elements of matrix @xmath41satisfy the commutation relation @xmath42=-\delta_{ij}$ ] .
moreover , the following symplectic identities are immediate consequencies of eq .
( [ qlam ] ) : @xmath43 writing matrices @xmath37 and @xmath44 in the block form , @xmath45 we obtain from eqs .
( [ qlam ] ) the equations ( for @xmath46 ) d_1/dt = _
1 b_3 - _ 2 , d_2/dt = _ 1 b_4 - _ 2 b_2 , [ eq - lam12 ] and similar equations for the blocks @xmath47 and @xmath48 . excluding matrices @xmath49 and @xmath48 we arrive at identical second - order equations for matrices @xmath50 and @xmath47 : -_1+_1,3_3 = 0 , [ e1 ] where _ 1 = b_3-b_2 = 2 cc 0 & -1 + 1 & 0 , [ n1 ] _ 3 = b_4 -b_3 b_2 = cc ^2(t ) -(2)^2 & 0 + 0 & 1 .
[ n3 ] the difference between @xmath51 and @xmath52 is in the initial conditions : @xmath53 we suppose that @xmath54 . in such a case ,
( [ e1 ] ) can be solved with a sufficient accuracy analytically with the aid of the method of multiple scales @xcite . for this purpose
we introduce a formal small parameter @xmath55 , writing @xmath56 , @xmath57 [ with @xmath58 and assuming that the solutions depend on the set of _ scaled times _ , @xmath59 , @xmath60 , @xmath61 , which can be considered as _
independent variables_. this means that the time derivatives @xmath62 and @xmath63 can be expressed as follows : @xmath64 also , we assume that matrices @xmath65 can be written in the form @xmath66 for small enough values of parameter @xmath55 it is sufficient to calculate the first term of the above expansion , taking into account the dependence on times @xmath67 and @xmath68 only
. therefore we neglect the terms of the second order with respect to @xmath55 in eq .
( [ e1 ] ) , replacing exact matrix ( [ n3 ] ) by the approximate form _ 3 cc 1 + 4_0(2t_0 - 2_0 t_1 ) & 0 + 0 & 1 . [ n3a ] omitting the details of cumbersome calculations , we bring here the general structure of solutions only .
it appears that the elements @xmath69 of each matrix @xmath70 ( where @xmath71 and @xmath72 ) can be written as ( remember that we put @xmath19 ) @xmath73 the meaning of symbols is as follows .
the `` fast time '' @xmath74 appears in two oscillating functions @xmath75 and @xmath76 only .
the amplitudes of these fast oscillations are modulated by combinations of hyperbolic and trigonometric functions , which depend on the `` dimensionless slow time '' @xmath77 ( which is proportional to @xmath68 ) .
namely , hyperbolic functions depend on variable @xmath78 , whereas trigonometric functions depend on @xmath79 . in turn ,
coefficients @xmath80 , @xmath81 and @xmath82 depend on the ratio @xmath20 : @xmath83 the factors @xmath84 and @xmath85 are both positive for any values of @xmath86 and satisfy the identity @xmath87 .
all other constant coefficients in eq .
( [ lambdas ] ) , such as @xmath88 , @xmath89 , and so on , also depend on the ratio @xmath86 .
explicit forms of coefficients with @xmath90 and @xmath91 are given in appendix . since
matrix @xmath92 is of the order of @xmath55 , we can write , in view of eq .
( [ eq - lam12 ] ) , approximate formulas @xmath93 .
consequently , within the chosen accuracy ( i.e. , neglecting terms proportional to @xmath55 in all the amplitude coefficients ) the elements of matrices @xmath94 can be obtained from the respective formulas ( [ lambdas ] ) for @xmath95 by means of replacements @xmath96 and @xmath97 .
we have checked the accuracy of approximate analytical solutions ( [ lambdas ] ) , comparing them with exact numerical solutions of eq .
( [ e1 ] ) .
it appears that the coincidence is quite good for @xmath98 .
an example of comparing analytical and numerical solutions is given in fig .
[ fig - anal - numer ] . in the case of @xmath99 , @xmath100 with @xmath101 and @xmath102 . ]
knowing matrices @xmath37 or @xmath44 ( [ laml ] ) one can calculate immediately the time evolution of the symmetric covariance matrix @xmath103 , whose elements @xmath104 are the central second - order statistical moments .
namely , @xmath105 for the initial vacuum states of both modes matrix @xmath106 is proportional to the unity matrix : @xmath107 .
consequently , in view of eqs ( [ laml ] ) and ( [ covte ] ) , the mean number of quanta in the @xmath108th mode ( @xmath109 ) can be expressed in terms of diagonal elements of the sum of products of matrices @xmath110 and @xmath111 ( since the first - order mean values @xmath112 are equal to zero in the case under study ) : _ m = 12(_k=1 ^ 4 _ k _ k)^(mm ) -12 .
[ nmlam ] these mean numbers do not depend on the `` fast time '' @xmath74 .
their explicit expressions are as follows : @xmath113\mathbb{s}_{\mu } ^{2 } \notag \\ & + \left [ \eta-1 + \left(1+\mathfrak{u}_{m}\right)\beta^{2}\left(\eta -\beta^{2}\right)\right]\mathtt{s}_{\nu } ^{2 } \notag \\ & -2\beta^2\left(\mathbb{c}_{\mu } \mathtt{c}_{\nu } -1\right ) + 2\beta \mathfrak{u}_{m } \mathbb{s}_{\mu } \mathtt{s}_{\nu } \big\ } , \label{gpmn}\end{aligned}\ ] ] where @xmath114 and @xmath115 .
we have introduced the short notation @xmath116 typical regimes of excitations of the two modes are illustrated in fig .
[ lnmpn ] , where we plot the photon mean numbers in the cavity mode and the detector as functions of the `` dimensionless slow time '' @xmath117 for different values of parameter @xmath86 .
higher values of @xmath86 imply on a more intense photon generation for a given time .
$ ] for the cavity ( 1 ) and detector ( 2 ) with @xmath118 ( _ s _ ) , @xmath119 ( _ f _ ) and @xmath120 ( _ h _ ) . note the time delay in increasing the number of quanta in the detector for @xmath119 and @xmath121 . ]
we see that for @xmath122 the numbers of excitations in each mode grow exponentially after a short transient time , but the number of quanta in the detector mode is significantly smaller than in the cavity mode for any fixed value of @xmath117 . for smaller values of @xmath86
we observe some `` beats '' between the two modes .
however , these `` beats '' exist for limited time intervals , since only the first line in formula ( [ gpmn ] ) is important for @xmath123 . thus we see that asymptotically the number of quanta in the field mode is always bigger than in the detector ( since @xmath124 while @xmath125 )
. the ratio can be very big if @xmath126 : in this case @xmath127 . on the other hand ,
@xmath128 if @xmath129 , and @xmath130 for @xmath131 ( in the asymptotical regime @xmath123 )
. for @xmath132 we have @xmath133 , @xmath134 , and @xmath135 , so that @xmath136 and @xmath137 . therefore the numbers of excitations in the both modes coincide ( provided the slow time variable @xmath117 is not very small ) , with corrections of the order of @xmath86 : @xmath138 moreover , the main term in ( [ n1tr ] ) does not depend on the field - detector coupling coefficient in this case .
( [ n1tr ] ) is in agreement with the results obtained in @xcite by using the method of slowly varying amplitudes ( although the argument of the @xmath139 function in @xcite was twice bigger than here due to a misprint in the definition of parameter @xmath81 in that paper ) .
for @xmath122 we have @xmath140 , @xmath141 , and @xmath142 . in this case
@xmath143 , \ ] ] so that @xmath144 if @xmath145 .
if @xmath146 , then @xmath147 , @xmath148 , and @xmath149 , so that @xmath150 and @xmath151 . therefore @xmath152 + \emph{o}(\beta^{-3 } ) .
\label{n - bigbeta}\end{aligned}\ ] ] the number of photons in the field mode coincides with the result @xcite obtained in the absence of any detector , up to corrections of the order of @xmath153 .
the number of excitations in the detector is about @xmath154 times smaller , and this leads to a significant time delay in appearance of excitations in the detector : while the photons in the field mode appear after time @xmath155 , the detector begins to `` feel '' their presence after the time @xmath156 .
this time delay exists also for very small times , since in the limit @xmath157 eq .
( [ gpmn ] ) can be simplified as follows : @xmath158
it is interesting to know , besides the mean numbers of quanta , their variances @xmath159 , the degree of squeezing of quadrature components and the photon distribution functions ( pdf ) @xmath160 , where @xmath161 means the @xmath108th fock state and @xmath162 is the statistical operator describing the mixed ( due to the interaction ) state of the mode . for the initial vacuum states of the field mode and detector
the time - dependent statistical operator is gaussian .
therefore we can use the results obtained in ref .
@xcite for the most general gaussian states .
in our case the first - order mean values of the quadrature components are equal to zero , so that @xmath163 where @xmath164 is the _ invariant uncertainty product _ ( iup ) of the @xmath108th mode , which satisfies the schrdinger robertson uncertainty relation @xcite @xmath165 for any quantum state . besides , it is easy to verify that @xmath166 for the states with zero first - order mean values of the quadrature components . therefore the variances of the photon number in the gaussian states with zero first - order mean values of the quadrature components must obey the inequality @xmath167 . the equality sign is attained for _ thermal _
( i.e. , mixed ) quantum states . on the other hand , for pure squeezed vacuum states with @xmath168
we have _ ^(sqzvac ) = 2 ( + 1 ) .
[ sigsqz ] the calculations lead us to the identical expressions for the functions @xmath169 for @xmath170 and @xmath171 @xmath172 this result may seem surprising at first glance , since other statistical properties ( such as the mean number of quanta or degree of squeezing studied in the next subsection ) in each mode are different . but
this is the consequence of the property of iup for the gaussian states : for these states iup determines the quantum purity according to the formula @xmath173 . on the other hand
, it is known that for any _ pure _ bipartite quantum state the purities of each part are identical . since the initial state of the total system
is chosen to be pure ( vacuum ) , it remains pure for any instant of time in the absence of dissipation ( assumed in this paper ) .
therefore @xmath174 .
simple formula for variances can be obtained if @xmath132 : @xmath175 comparing it with ( [ n1tr ] ) , we conclude that quantum states of the both modes are non - thermal , although they become highly mixed with the course of time , since the purity in this case diminishes as @xmath176^{-1}$ ] .
the limit ( [ sigsqz ] ) is achieved approximately for the field mode if @xmath126 : @xmath177 the invariant uncertainty product in this case equals _
m^(1 ) = 14 + 14 ^ 2 .
[ d - bigbeta ] comparing eqs .
( [ n - bigbeta ] ) and ( [ d - bigbeta ] ) , we conclude that quantum states of the both modes become significantly mixed when the mean number of excitations in the detector exceeds the unit value .
for @xmath178 we have @xmath179 , whereas @xmath180 . consequently , if @xmath181 , then @xmath182 for @xmath109 , like in the squeezed vacuum state , even if the states are highly mixed . on the other hand ,
the photon statistics in the two modes are quite different for @xmath183 , especially in the short - time limit @xmath184 .
frequently the photon statistics is characterized by the mandel parameter @xmath185 . however , in the case concerned this parameter is not very useful , since it is always greater than unity and grows unlimitedly with time in the dce regime .
a more convenient parameter is the ratio @xmath186 $ ] , since it varies between @xmath187 and @xmath188 for the gaussian states with zero mean values of quadrature components . the evolution of functions @xmath189 and @xmath190 for different values of parameter @xmath86 is illustrated in fig .
[ fig - rat ] .
$ ] for the cavity ( 1 ) and detector ( 2 ) with @xmath118 ( _ s _ ) , @xmath119 ( _ f _ ) and @xmath120 ( _ h _ ) . ] the difference is clearly seen in the limit @xmath184 , when eq .
( [ dmt ] ) can be simplified as follows : _ m ( ) = 14 + ^2 ^4 + _ o_(^8 ) .
[ dmtsmall ] therefore the contribution of @xmath191 can be neglected in eq .
( [ sig - n ] ) with @xmath170 , in view of eq .
( [ n1small ] ) , so that @xmath192 . but
@xmath193 has the same order of magnitude as @xmath194 for @xmath157 , according to eqs .
( [ n2small ] ) and ( [ dmtsmall ] ) .
therefore @xmath195 .
more precisely , @xmath196 + \frac{7}{6}\beta^2\tau^4 + \emph{o}(\tau^{6 } ) , \nonumber \\
z_2(\tau ) & = 1+\frac{\tau^2}{9 } \big [ 4 - \frac{19}{45}\tau^2 \big ] +
\frac{128}{135}\beta^2\tau^4 + \emph{o}(\tau^{6 } ) .
\nonumber\end{aligned}\ ] ] the degree of squeezing is characterized by the _ invariant squeezing coefficient _
( isc ) @xcite ( which is equivalent to the _ principal squeezing _ introduced in @xcite ) = . [ isc
] this is twice the minimal value of variance of any quadrature component taken over the period of fast oscillations .
for @xmath178 we have the asymptotical formula @xmath197 \notag \\ & \times \left\{2\eta^2\left[\beta^2\left(\eta+\beta^2\right)\left(1+\mathfrak{u}_{m}\right ) + 1+\eta \right]\right\}^{-1}.\end{aligned}\ ] ] in the special case of @xmath132 we obtain @xmath198 , in agreement with @xcite .
a very high degree of squeezing can be obtained in the field mode if @xmath146 : then @xmath199 .
however , there is practically no squeezing in this limit case in the detector mode : @xmath200 .
for intermediate values of parameter @xmath86 the isc does not go asymptotically to some limit value , but it exhibits slow oscillations in time with the frequency @xmath201 . for example , for @xmath122
we obtain _ m^(as)_=1= .
this function oscillates between the values @xmath202 and @xmath203 for @xmath170 ( the field mode ) .
but it is three times bigger for the detector mode .
it is not difficult to calculate the minimal value of the coefficient @xmath204 as function of slow time @xmath205 for arbitrary values of parameter @xmath86 : _ 2^(as)_min= 12 + .
moreover , the minimum of this function with respect to @xmath86 also can be found analytically .
it is achieved for @xmath206 , being equal to @xmath207 .
consequently , the detector mode can not be strongly squeezed .
the pdf of the gaussian states was derived in @xcite .
for zero mean values of quadrature components @xmath208 and @xmath209 it can be expressed in terms of the legendre polynomials @xmath210 as ( k)= p_k ( ) , [ dist0 ] where y _ = 1 + 4 2(1 + 2 ) . [ dpm ] the behavior of pdf as function of @xmath211 depends on the value of the argument of the legendre polynomial .
if this argument is close to zero , then strong oscillations of function @xmath212 are observed , since legendre polynomials of zero argument turn into zero for odd values of @xmath211 .
otherwise @xmath212 changes slowly and monotonously .
this happens for @xmath131 , as was shown in @xcite .
if @xmath146 , then the argument of the legendre polynomials for the first mode tends asymptotically to the small quantity @xmath213 , so that the pdf shows typical oscillations of the squeezed vacuum state .
but for the second mode the argument of the legendre polynomials goes asymptotically to the value @xmath214 , which is not small .
therefore the pdf of the detector mode is monotonous function without oscillations ( that agrees with the absence of squeezing in this mode ) , which can be well approximated ( for @xmath215 ) by the universal dependence derived in @xcite ( k ) . in particular , considering formally @xmath211 as a continuous variable ( i.e. , using the euler maclaurin summation formula ) , we obtain the following value for the probability of detecting any value @xmath211 smaller than the mean value @xmath216 : ( k < |n ) = ( /2)0.68 , [ k - erf ] where @xmath217 is the error function . the distributions @xmath218 and @xmath219 for @xmath122 and @xmath220 are shown in fig . [ fig - py ] .
they are rather different , especially for small values of @xmath211 .
the photon mean numbers are @xmath221 and @xmath222
. nonetheless , the total probabilities to measure the photon number smaller than the @xmath223 are practically the same for each mode : @xmath224 and @xmath225 , in full agreement with eq .
( [ k - erf ] ) . for the cavity @xmath226 and detector @xmath227 in the case of @xmath120 and @xmath220 .
we have shown that statistical properties of the detector quantum state ( variances of the photon numbers , photon distribution function , and the degree of quadrature squeezing ) can be quite different from that of the field mode .
this can be important for planning the experiments on dce and analysing their results .
in particular , the discovered time delay in the increase of the mean number of quanta in the detector indicates that influence of losses in the detector ( which are not taken into account in the present study ) can be essential , even if losses in the cavity can be neglected .
but we leave this problem for the further studies , as well as the influence of different initial states ( e.g. , thermal or squeezed ) of the field mode and detector on the rate of generation of the casimir photons .
all other coefficients of matrices @xmath50 and @xmath47 are equal to zero .
the coeficients of matrices @xmath49 and @xmath48 are given by formulas @xmath232 , @xmath233 , and similar ones for other lower case and capital symbols .
d. a. r. dalvit , p. a. maia neto , and f. d. mazzitelli , in _
casimir physics _ , edited by d. dalvit , p. milonni , d. roberts , and f. da rosa , lecture notes in physics vol .
834 ( springer , berlin , 2011 ) , p. 419 .
a. v. dodonov and v. v. dodonov , phys .
a * 85 * , 015805 ( 2012 ) .
a. v. dodonov and v. v. dodonov , phys .
a * 85 * , 055805 , 063804 ( 2012 ) ; _ ibid . _ * 86 * , 015801 ( 2012 ) . c. braggio , g. bressi , g. carugno , f. della valle , g. galeazzi , and g. ruoso , nucl .
instr . and meth . a * 603 * , 451 ( 2009 ) .
v. v. dodonov and v. i. manko , _ invariants and the evolution of nonstationary quantum systems _ , edited by m. a. markov , proceedings of lebedev physics institute vol .
183 ( nova science , commack , new york , 1989 ) . | we consider the problem of photon creation from vacuum inside an ideal cavity with vibrating walls in the resonance case , taking into account the interaction between the resonant field mode and a detector modeled by a quantum harmonic oscillator .
the frequency of wall vibrations is taken to be twice the cavity normal frequency , modified due to the coupling with the detector .
the dynamical equations are solved with the aid of the multiple scales method .
analytical expressions are obtained for the photon mean numbers and their variances for the field and detector modes , which are supposed to be initially in the vacuum quantum states .
we analyze different regimes of excitation , depending on the ratio of the modulation depth of the time - dependent cavity eigenfrequency to the coupling strength between the cavity mode and detector .
we show that statistical properties of the detector quantum state ( variances of the photon numbers , photon distribution function , and the degree of quadrature squeezing ) can be quite different from that of the field mode . besides , the mean number of quanta in the detector mode increases with some time delay , compared with the field mode . |
fly is the tree n - body code we design , develop and use to run very big simulations of the large scale structure of the universe using parallel systems mpp and smp .
fly uses the leapfrog numerical integration scheme for performance reasons , and incorporates fully periodic boundary conditions using the ewald method .
the i / o data format is integrated with the astromd package .
+ astromd is an analysis and visualization tool specifically designed to deal with the visualization and analysis of astrophysical data .
astromd can find structures having a not well defined shape or symmetries , and performs quantitative calculations on a selected region or structure .
astromd makes use of virtual reality techniques which are particularly effective for understanding the three dimensional distribution of the fields , their geometry , topology and specific patterns .
the display of data gives the illusion of a surrounding medium into which the user is immersed .
the result is that the user has the impression of travelling through a computer - based multi - dimensional model which could be directly hand - manipulated .
the fly code , written in fortran 90 and c languages , uses the one - side communication paradigm : it has been developed on the cray t3e using the shmem library .
fly is based on the following main characteristics .
it adopts a simple domain decomposition , a grouping strategy and a data buffering that allows us to minimize data communication .
fly does not split the domain with orthogonal planes , but the domain decomposition is done by assigning an equal number of particles to each processor .
the input data particles is a sorted file containing the fields of position and velocity , so that particles with a near tag number are also near in the physical space , and the arrays containing the tree properties are distributed using a fine grain data distribution . during the tree walk procedure ,
fly builds a single interaction list ( il ) to be applied to all particles inside a grouping cell ( @xmath1 ) .
this reduces the number of the tree accesses to build the il .
we consider a hypothetical particle we call _ virtual body _ ( vb ) placed in the center of mass of the @xmath1 : the vb interaction list @xmath2 is formed by two parts : @xmath3 where @xmath4 includes the elements more distant than a threshold parameter from vb and @xmath5 includes the elements near vb . using the two lists it is possible to compute the force @xmath6 of each _ p _ particle in @xmath1 as the sum of two components : @xmath7
the component @xmath8 is computed only once for vb , and it is applied to all the _ p _ particles , while the @xmath9 component is computed separately for each particle .
the size of the @xmath1 , and the tree - level where it can be considered , is constrained by the maximum allowed value of the overall error of this method . in this sense
the performance of fly is a level - based code .
the data buffer is managed with a policy of a simulated cache in the local ram .
every time the pe has to access a remote element , at first it looks for the local simulated cache and , if the element is not found , the pe executes the get calls to down - load the remote element and stores it in the buffer . in a simulation with 16-million - particles
clustered , with 32 pes and 256 mbytes of local memory , without the use of the simulated cache , the pes execute about @xmath10 remote gets .
this value , using the data buffering , decreases at @xmath11 remote gets , with an enormous advantage in terms of scalability and performance .
1 shows the code performance on cray t3e system , obtained by running simulations with 32 and 64 pes , and fly scalability , considering the case of 16,777,216 particles , where a speed - up factor of 118 is reached using 128 pes .
the highest performance obtained in a clustered configuration is a positive effect of the grouping characteristic .
the obtained results show that fly has a very good scalability and a very high performance and it can be used to run very big simulations .
astromd is developed using the visualization toolkit ( vtk ) by kitware , a freely available software portable on several platforms which range from the pc to the most powerful visualization systems , with a good scalability .
data are visualized with respect to a box which can describe the whole computational mesh or a sub - mesh .
astromd can find structures having a not well defined shape or symmetries , and performs quantitative calculations on a selected region or structure .
the user can choose the sample of the loaded particle type ( i.e. stars , dark matter and gas particles ) , the size of the visualization box and the starting time from which he wants to show the evolution of the simulation .
density _ entries control the visualization of the iso - surfaces .
these are calculated on a grid whose resolution can be user - selectable . to allow the user to investigate with more accuracy a subset of the visualized system it is possible to use a cubic sampler .
if the sampler is selected ( _ show / hide _ menu ) , visualized in the scene and enabled ( _ sampler _ menu ) , all the computations are performed only inside the region of the sampler ( fig .
astromd can also show the evolution in time of the simulated system over all the interval of time for which the data are available , performing interpolations at intermediate frames . during the evolution the updated time of evolution
is displayed in the _ time _ entry of the _ cloud _ section .
finally snapshots of the images displayed can be created using the button _ take it _ in the _ screenshot _ menu .
astromd is developed by the visit ( visual information technology ) laboratory at cineca ( casalecchio di reno - bologna ) in collaboration with the astrophysical observatory of catania .
antonuccio - delogu , v. , becciani , u. , pagliaro , a. , van kampen , e. , colafrancesco , s. , german , a. and gambera , m. 2000 , submitted to barnes , j.e . and hut , p. 1986 , nature , 324 , 446schroeder , w. , ken , m. , lorensen , b. 1999 , `` the visualization toolkit '' , prentice hall becciani , u. , antonuccio - delogu , v. 2000 , j. comput .
phys . , 163 , 118 earnshaw , r.a . , watson , d. 1993 , `` animation and scientific visualization '' , academic press ltd .
qinn , t. , stadel , c. tipsy http://www-hpcc.astro.washington.edu/tools/tipsy salmon , j. and warren , m.s .
1997 , proc . of the eight conf . on parallel processing for scientific computing , siam 1997 | we present a new code allowing to evolve three - dimensional self - gravitating collisionless systems with a large number of particles @xmath0 . *
fly * ( fast level - based n - body code ) is a fully parallel code based on a tree algorithm .
it adopts periodic boundary conditions implemented by means of the ewald summation technique .
fly is based on the one - side communication paradigm to share data among the processors that access remote private data avoiding any kind of synchronism .
the code was originally developed on a cray t3e system using the _ shmem _ library and it was ported on sgi origin 2000 and on ibm sp ( on the latter making use of the _ lapi _ library ) .
version 1.1 is an open source _ freely available code_. + fly data output can be analysed with astromd , an analysis and visualization tool specifically designed to deal with the visualization and analysis of astrophysical data .
astromd can manage different physical quantities
. it can find out structures without well defined shape or symmetries , and perform quantitative calculations on selected regions .
is a _ freely available _ code . |
measurements of diboson production cross sections at the tevatron test the electroweak sector of the standard model and have been used to place limits on models of physics beyond the standard model @xcite .
diboson measurements are also useful in the context of searches for the standard model higgs boson at the tevatron . in this presentation
we focus on recent diboson results that are relevant to the higgs searches . the search for the higgs boson at the tevatron involves searching for a very small signal in overwhelming backgrounds .
sophisticated analysis techniques are often used to exploit small differences between signal and background events .
the searches also gain power from increasing signal acceptance and dividing events into several regions depending on their signal - to - background ratios .
some of the diboson searches and measurements presented below take advantage of similar techniques while others use somewhat simpler strategies .
comparison of the results derived with different techniques is a useful test of the analysis strategies employed in the higgs searches .
the measurements presented here are performed in @xmath4 collision data with @xmath3 tev collected by either the cdf ii or detector .
the detectors are described in detail elsewhere @xcite @xcite .
@xmath0 , @xmath1 , and @xmath2 production have all been observed at the tevatron at the 5@xmath5 level in events where both bosons decay leptonically
. table [ tab : lep_sum ] shows recent measurements of the cross sections for each of the two experiments .
the measured cross sections agree well between the two experiments and with the standard model predictions .
the observation of @xmath2 production and the measurement of the @xmath0 production cross section are discussed in more detail below .
.[tab : lep_sum]summary of diboson cross sections measured in leptonic channels at the cdf and detectors . [ cols="<,^,^,^ " , ] cdf reported the first observation of diboson production where one boson decays to leptons and the other to hadrons @xcite in 3.5 @xmath7 .
events with very large missing transverse energy ( @xmath860 gev ) and exactly two jets were used for this observation .
no veto on events with a lepton in the final state was imposed .
the analysis was therefore sensitive to a sum of @xmath9 , and @xmath2 processes .
one challenge in this channel was understanding the qcd multi - jet background ( mjb ) .
an event with many jets can have large fake @xmath10 because of mismeasurement of the jet energies , often stemming from instrumental effects , making this background difficult to model .
the size of the mjb background was significantly reduced by imposing cuts on the @xmath10 significance and the angle between the @xmath10 and the jets .
the remaining mjb was modeled with a data sample enriched in multi - jet events , selected by finding events with a large difference in direction between their track - based missing transverse momentum and their calorimeter - based missing transvere energy .
the second large background stemmed from electroweak processes ( @xmath11+jets , @xmath12+jets , and @xmath13-quark production ) .
these backgrounds were modeled using monte carlo .
the uncertainty on the model was evaluated with data , using @xmath14jet events .
the signal cross section was extracted by a fit to the dijet mass spectrum with signal and background templates .
the fit is shown in figure [ fig : metjets_cdf ] .
the fitted cross section was found to be @xmath15(stat)@xmath16(syst)@xmath17(lumi ) , with the dominant systematic uncertainty due to the jet energy scale .
the signal was observed with a significance of 5.3@xmath5 .
fit to mass dijet mass distribution to extract @xmath18 cross section in events with large missing transverse energy and two jets.,scaledwidth=48.0% ] the first evidence of @xmath20 was reported by the collaboration @xcite in 1.07 @xmath7 .
events with a well - identified electron or muon , at least two jets , and @xmath820 gev were selected .
the background due to qcd multi - jet events was reduced by requiring the transverse mass of the lepton-@xmath10 system to be larger than 35 gev .
the remaining backgrounds were dominated by @xmath11+jets production , with some qcd multi - jet , @xmath12+jets , and top quark production contributing as well .
the qcd multijet background was modeled using data with somewhat loosened lepton requirements .
the @xmath11+jets background was modeled with simulated events from algpen interfaced with the pythia parton shower .
the modeling of the @xmath11+jets background was critical to the analysis , so careful comparison between data and monte carlo was carried out .
discrepancies in the jet @xmath21 distributions and the @xmath22 distributions were observed ; the models were reweighted to agree with data .
once confident in the modeling , a random forest classifier ( rf ) was used to discriminate between signal and background events .
several kinematic variables , such as the dijet mass , were used as inputs and the rf was trained on part of the background to build a classification for each event .
the distribution of the rf output observed in data was fitted to a sum of predicted rf templates to extract the signal significance and cross section .
the fit is shown in figure [ fig : wwwz_d0 ] .
the measured cross section is @xmath23 pb , with dominant systematic uncertainties from the modeling of the @xmath11+jets background distribution and the jet energy scale .
the significance of the observed signal is 4.4@xmath5 .
fit to random forest classifier output used to extract the @xmath20 cross section.,scaledwidth=48.0% ] the same analysis was carried out using only the dijet mass distribution rather than the random forest classifier output .
since less information about the event is used , the measurement is expected to be less precise .
the dijet mass is shown in figure [ fig : wwwz_d0_mjj ] .
the result of the fit was @xmath24 , compatible with the result from the rf classifier .
dijet mass distribution for data , backgrounds , and expected @xmath25 signal in @xmath26 final state.,scaledwidth=48.0% ] the first observation of @xmath20 was presented by cdf in 2.7 @xmath7 .
events with an isolated electron or muon , exactly two jets , and @xmath820 gev were chosen .
strong cuts were imposed in events with an electron to reduce the qcd multi - jet background due to jets faking electrons . as a result
the measurement was dominated by events with muons .
validation of the background modeling was also critical for this analysis .
three regions were chosen according to the dijet mass : the signal - rich region with @xmath27 gev and two control regions with @xmath28 gev and @xmath29 gev where very little signal was expected .
good modeling was observed in each region .
some mismodeling in the dijet mass was observed when the control regions were combined , and a corresponding systematic uncertainty on the shape of the @xmath11+jets background was applied .
matrix element calulations were used to discriminate between signal and background .
the differential cross sections of signal and background processes were evaluated for each event .
a discriminant called the event probability discrminant was formed to separate signal from background .
the predicted shapes of signal and background discriminants were fit to the data to extract the diboson cross section . the data superimposed on the background templates is shown in figure [ fig : wwwz_cdf ] .
the measured cross section is @xmath30 pb where the dominant systematic uncertainty was the jet energy scale .
the significance of the signal observation was 5.4@xmath5 .
distribution of the discriminant derived from matrix elements used to extract the @xmath25 cross section in the @xmath31 final state , scaledwidth=48.0% ] a second complimentary search was carried out at cdf using a larger data sample of 3.9 @xmath7by fitting the @xmath32 spectrum .
the event selection criteria were adjusted to achieve a smoothly falling shape in the @xmath32 distribution of the backgrounds . in particular , the @xmath33 threshold on each individual jet was lowered , but the @xmath33 of the dijet system was required to be larger than 40 gev .
the diboson signal resulted in a bump on top of the background , as shown in figure [ fig : wwwz_mjj ] .
a fit with signal and background templates was carried out , and the extracted cross section was @xmath34(stat ) @xmath35(sys ) pb , corresponding to an observed significance of 4.6@xmath5 .
distribution of the dijet mass for data and signal and background models for @xmath20.,scaledwidth=48.0% ]
several measurements of diboson production cross sections have been carried out recently at the tevatron .
these measurements can be a testing ground for techniques used in searches for the standard model higgs boson .
measurements are performed in many different final states , ranging from those with several identified leptons to those with no identified leptons and two jets .
different analysis techniques are also used , from counting signal events to performing fits to kinematic quantities to techniques involving classifiers or matrix element calculations .
there is good agreement between results found at cdf and , as well as good agreement with nlo predictions .
we thank the fermilab staff and the technical staffs of the participating institutions for their vital contributions .
this work was supported by the u.s .
department of energy and national science foundation ; the italian istituto nazionale di fisica nucleare ; the ministry of education , culture , sports , science and technology of japan ; the natural sciences and engineering research council of canada ; the national science council of the republic of china ; the swiss national science foundation ; the a.p .
sloan foundation ; the bundesministerium fr bildung und forschung , germany ; the world class university program , the national research foundation of korea ; the science and technology facilities council and the royal society , uk ; the institut national de physique nucleaire et physique des particules / cnrs ; the russian foundation for basic research ; the ministerio de ciencia e innovacin , and programa consolider - ingenio 2010 , spain ; the slovak r&d agency ; and the academy of finland .
9 d. acosta _ et al .
_ ( cdf collaboration ) , phys . rev .
d * 71 * , 032001 ( 2005 ) . v. m. abazov _
_ ( collaboration ) , nucl .
methods phys .
a * 565 * , 463 ( 2006 ) .
v. m. abazov _
et al . _ ( collaboration ) , phys .
* 101 * , 171803 ( 2008 ) .
t. aaltonen _
( cdf collaboration ) , phys .
* 100 * , 201801 ( 2008 ) . the cdf collaboration , public note 9753 ( 2009 ) .
v. m. abazov _
_ ( collaboration ) , arxiv:0904.0673v1 . t. aaltonen _
( cdf collaboration ) , phys . rev .
lett . * 103 * , 091803 ( 2009 ) . v.
m. abazov _
et al . _ ( collaboration ) , phys . rev .
* 102 * , 161801 ( 2009 ) . | we present a summary of recent @xmath0 , @xmath1 , and @xmath2 measurements carried out in @xmath3 tev at the tevatron .
the observation of rare @xmath2 events and the precise measurement of the @xmath0 cross section in fully leptonic decay modes are described .
results in semi - leptonic decay modes where one boson decays to two quarks are also presented .
the measurements described are all in good agreement with the standard model and are relevant to searches for the higgs boson at the tevatron . |
our paper contains a direct elementary proof of the following series transformation : ( * landen s transformation * ) if @xmath0 , then@xmath1{x}}{1+\sqrt{x}}\right)^{2}-\frac{1^{2}\cdot 3}{2^{2}\cdot 4^{2}}\left(\frac{2\sqrt[4]{x}}{1+\sqrt{x}}\right)^{4}-\frac{1^{2}\cdot 3^{2}\cdot 5}{2^{2}\cdot 4^{2}\cdot 6^{2}}\left(\frac{2\sqrt[4]{x}}{1+\sqrt{x}}\right)^{6}-\cdots\right\}= & \\ & \\={\displaystyle}1+\left(\frac{1}{2}\right)^{2}x+\left(\frac{1\cdot 1}{2\cdot 4}\right)^{2}x^{2}+\left(\frac{1\cdot 1\cdot 3}{2\cdot 4\cdot 6}\right)^{2}x^{3}+\left(\frac{1\cdot 1\cdot 3\cdot 5}{2\cdot 4\cdot
6\cdot 8}\right)^{2}x^{4}+\cdots & \\\end{array } $ } \ ] ] @xmath2 the interest in this theorem is that it gives a direct proof the equality of the following two formulas for the perimeter of an ellipse with semi - axes @xmath3 and @xmath4 .
( * the perimeter of an ellipse * ) the perimeter , @xmath5 , of an ellipse with semi - major axes @xmath3 and @xmath4 and eccentricity @xmath6 is:@xmath7@xmath2 the first formula for @xmath5 is due to maclaurin @xcite in 1742 while the second is due to ivory @xcite in 1796 .
berndt ( @xcite , p. 147 ) points out that ivory s formula can be obtained from a form of landen s transformation to be found in erdlyi s compendium ( @xcite , p.111 , formula ( 5 ) ) , but we have not been able to encounter a direct elementary proof in the literature which transforms the first of the two formulas for @xmath5 into the second one , and which does not appeal to general transformation formulas .
ivory s own proof cleverly finesses this difficulty by showing that both series are equal to an integral .
the following identity is valid for all integers @xmath8 and real numbers @xmath9.@xmath10 the following identities between binomial coefficients are well known:@xmath11 for example , see knuth ( @xcite pp .
71 and 46 , resp . )
now solve the first identity for @xmath12 , write @xmath13 in place of @xmath9 everywhere , use the second identity to get rid of the negative sign inside the binomial coefficients , and finally multiply and divide by @xmath14 this next lemma is the novelty in our proof .
the following relation is valid for all integers @xmath15 @xmath16 we construct , in _ two different ways _ , a certain polynomial , @xmath17 , whose coefficient of @xmath18 is the sum @xmath19 defined in ( 3 ) .
observe that the _ coefficient of _ @xmath20 in @xmath21is equal to ( put @xmath22)@xmath23therefore , _ the coefficient of _
@xmath24in@xmath25is equal to@xmath26thus , _ the sum _ we are seeking to evaluae , viz.@xmath27is the _ coefficient _ of @xmath18 in the _ _ sum__@xmath28this is the * _ first form _ * of the polynomial @xmath29 but @xmath17 can also be written@xmath30which is the * _ second form _ * of @xmath17 . expanding it by the binomial theorem we obtain@xmath31we see that all the powers of @xmath32 are _
even_. thus , if @xmath9 is _ odd _ , the coefficient of @xmath18 is _ zero _ , while if it is _ even _ , the coefficient of @xmath33is equal to@xmath34 the following result is well known ( @xcite ) : ( * vandermonde s theorem * ) the following relation holds for all real numbers @xmath3 and @xmath4:@xmath35 @xmath2
our proof of * theorem 1 * is based on an idea of carlson @xcite .
however we use the fundamental * lemma 2 * , of which we have given a direct elementary proof based on the elementary properties of the binomial coefficients ( see 2 ) , while carlson proves the transformation on the basis of general expansion theorems for r - functions . "
our proof consists of two steps : the left - hand side of * theorem 1 * is rearranged into the following power series in @xmath36:@xmath37we achieve this via the binomial theorem , * lemma 1 * and * lemma 2*. the power series @xmath38 is rearranged into a power series in @xmath32 alone via the binomial theorem . applying * lemma 1 * and vandermonde s theorem , the coefficients of this series collapse down to the coefficients of the right hand side of * theorem 1*.
all subsequent series converge absolutely and uniformly for @xmath0 , so that the following series manipulations are valid in that range.@xmath39{x}}{1+\sqrt{x}}\right)^{2}-\frac{1^{2}\cdot 3}{2^{2}\cdot 4^{2}}\left(\frac{2\sqrt[4]{x}}{1+\sqrt{x}}\right)^{4}-\frac{1^{2}\cdot 3^{2}\cdot 5}{2^{2}\cdot 4^{2}\cdot 6^{2}}\left(\frac{2\sqrt[4]{x}}{1+\sqrt{x}}\right)^{6}-\cdots\right\}\\ & = { \displaystyle}(1+\sqrt{x})\sum_{m=0}^{\infty}\frac{-1}{2m-1}\left\{\frac{1}{4^{m}}\binom{2m}{m}\right\}^{2}\left\{{\displaystyle}\left(\frac{2\sqrt[4]{x}}{1+\sqrt{x}}\right)^{2}\right\}^{m}\\ & = { \displaystyle}(1+\sqrt{x})\sum_{m=0}^{\infty}\frac{-1}{2m-1}\left\{\frac{1}{4^{m}}\binom{2m}{m}\right\}^{2}{\displaystyle}\left(\frac{2\sqrt[4]{x}}{1+\sqrt{x}}\right)^{2 m } \\ & = { \displaystyle}\sum_{m=0}^{\infty}\frac{-1}{2m-1}\left\{\frac{1}{4^{m}}\binom{2m}{m}\right\}^{2}4^{m}(\sqrt{x})^{m}(1+\sqrt{x})^{-2m+1 } \\ & = { \displaystyle}\sum_{m=0}^{\infty}\frac{-1}{2m-1}\left\{\frac{1}{4^{m}}\binom{2m}{m}\right\}^{2}4^{m}(\sqrt{x})^{m}(1 + 2\sqrt{x}+x)^{-m+\frac{1}{2}}\\ & = { \displaystyle}\sum_{m=0}^{\infty}\frac{-1}{2m-1}\left\{\frac{1}{4^{m}}\binom{2m}{m}\right\}^{2}4^{m}(\sqrt{x})^{m}(1+x)^{\frac{1}{2}}(1+x)^{-m}\left(1+\frac{2\sqrt{x}}{1+x}\right)^{-m+\frac{1}{2}}\\ & = -\sum_{m=0}^{\infty}\frac{1}{2m-1}\left\{\frac{1}{4^{m}}\binom{2m}{m}\right\}^{2}4^{m}\left(\frac{\sqrt{x}}{1+x}\right)^{m}(1+x)^{\frac{1}{2}}\sum_{k=0}^{\infty}\binom{-m+\frac{1}{2}}{k}\left(\frac{2\sqrt{x}}{1+x}\right)^{k}\\ & = -(1+x)^{\frac{1}{2}}\sum_{m=0}^{\infty}\frac{1}{2m-1}\left\{\frac{1}{4^{m}}\binom{2m}{m}\right\}^{2}2^{m}\sum_{k=0}^{\infty}\binom{-m+\frac{1}{2}}{k}\left(\frac{2\sqrt{x}}{1+x}\right)^{k+m}\\ & = -(1+x)^{\frac{1}{2}}\sum_{m=0}^{\infty}\sum_{k=0}^{\infty}\frac{2^{m}}{2m-1}\left\{\frac{1}{4^{m}}\binom{2m}{m}\right\}^{2}\binom{-m+\frac{1}{2}}{k}\left(\frac{2\sqrt{x}}{1+x}\right)^{k+m}\\ & = -(1+x)^{\frac{1}{2}}\sum_{m=0}^{\infty}\sum_{k=0}^{\infty}\frac{2^{m}}{2m-1}\left\{\frac{1}{4^{m}}\binom{2m}{m}\right\}^{2}\frac{2m-1}{2m-1 + 2k}\binom{-m-\frac{1}{2}}{k}\left(\frac{2\sqrt{x}}{1+x}\right)^{k+m}\\ & \left(\text{since $ \binom{r}{k}=\dfrac{r}{r - k}\binom{r-1}{k}$}\right)\\ & = -(1+x)^{\frac{1}{2}}\sum_{m=0}^{\infty}\sum_{k=0}^{\infty}\frac{2^{m}}{4^{2m}}\binom{2m}{m}^{2}\frac{1}{2m+2k-1}\frac{{\displaystyle}\binom{2m+2k}{m+k}\binom{m+k}{k}}{{\displaystyle}\binom{2m}{m}}\frac{(-1)^{k}}{4^{k}}\left(\frac{2\sqrt{x}}{1+x}\right)^{k+m}\\ & ( \text{by \textbf{lemma 1}})\\ & = ( 1+x)^{\frac{1}{2}}\sum_{m=0}^{\infty}\sum_{k=0}^{\infty}\frac{2^{m}}{4^{2m}}\binom{2m}{m}\frac{1}{2m+2k-1}{\displaystyle}\binom{2m+2k}{m+k}\binom{m+k}{k}{\displaystyle}\binom{2m}{m}\frac{(-1)^{k}}{4^{k}}\left(\frac{2\sqrt{x}}{1+x}\right)^{k+m}\\ & ( \text{now put $ k:=n - m$ and interchange the order of summation})\end{aligned}\ ] ] @xmath40 @xmath41 \left(\frac{2\sqrt{x}}{1+x}\right)^{n } ( \text{by \textbf{lemma 2}})\\ & = ( 1+x)^{\frac{1}{2}}\sum_{m=0}^{\infty}\frac{-1}{4m-1}\frac{1}{4^{2m}}\binom{4m}{2m}{\displaystyle}\frac{1}{4^{m}}\binom{2m}{m}\left(\frac{2\sqrt{x}}{1+x}\right)^{2 m } ( \text{putting $ n=2m$}.)\end{aligned}\ ] ] putting together the two extreme members of this chain of equalities , * _ we have shown : _
* @xmath42{x}}{1+\sqrt{x}}\right)^{2}-\frac{1^{2}\cdot 3}{2^{2}\cdot 4^{2}}\left(\frac{2\sqrt[4]{x}}{1+\sqrt{x}}\right)^{4}-\frac{1^{2}\cdot 3^{2}\cdot 5}{2^{2}\cdot 4^{2}\cdot 6^{2}}\left(\frac{2\sqrt[4]{x}}{1+\sqrt{x}}\right)^{6}-\cdots\right\}\\ & = ( 1+x)^{\frac{1}{2}}\sum_{m=0}^{\infty}\frac{-1}{4m-1}\frac{1}{4^{2m}}\binom{4m}{2m}{\displaystyle}\frac{1}{4^{2m}}\binom{2m}{m}\left(\frac{2\sqrt{x}}{1+x}\right)^{2m}. \end{aligned}\ ] ] @xmath2 this completes the proof of * step 1*.
we expand the term @xmath43 by the binomial theorem .
therefore , the rhs of the equation in * theorem 3 * becomes@xmath44again , we put @xmath45 , interchange the order of summation , and observe that * _ the coefficient of _ * @xmath46 * * _ is_**@xmath47where we have used both the conclusion and the proof of * lemma 1*. so , if we now substitute this last expression:@xmath48into the rhs of ( 5 ) we obtain@xmath49{x}}{1+\sqrt{x}}\right)^{2}-\frac{1^{2}\cdot 3}{2^{2}\cdot 4^{2}}\left(\frac{2\sqrt[4]{x}}{1+\sqrt{x}}\right)^{4}-\frac{1^{2}\cdot 3^{2}\cdot 5}{2^{2}\cdot 4^{2}\cdot 6^{2}}\left(\frac{2\sqrt[4]{x}}{1+\sqrt{x}}\right)^{6}-\cdots\right\}\\ & = \sum_{n=0}^{\infty}\sum_{m=0}^{n}\frac{(-1)^{n-1}}{2n-1}\frac{1}{4^{n}}\binom{2n}{n}\binom{-n+\frac{1}{2}}{m}\binom{n}{n - m}x^{n}\\ & = \sum_{n=0}^{\infty}\frac{(-1)^{n-1}}{2n-1}\frac{1}{4^{n}}\binom{2n}{n}\left\{\sum_{m=0}^{n}\binom{-n+\frac{1}{2}}{m}\binom{n}{n - m}\right\}x^{n}\\ & ( \text{and by \textsc{vandermonde } 's theorem this is equal to } ) \end{aligned}\ ] ] | we prove landen s transformation by direct elementary transformation of infinite series . |
comets are cold icy bodies in the solar system that were formed in the solar nebula and are considered to be the signature bodies to understand the formation of the solar system .
as the comet nucleus makes its journey towards its parent star , the ices start sublimating giving rise to a mixture of gas and dust which form the coma . for comets at heliocentric distances less than 3 au ,
the visible band spectrum shows strong molecular emission bands riding on the continuum radiation scattered by the cometary dust .
studying these molecular emission bands has been an important part of the cometary study .
+ comet c/2014 q2 ( lovejoy ) , an oort cloud comet , was discovered by terry lovejoy in august 2014 using an 8 inch telescope when the comet had a visual magnitude of 15 .
the comet brightened to a magnitude of 4 in the visual band in january 2015 being closest to earth at 0.469 au on january 7 .
the comet s perihelion was on the 30th of january 2015 at a heliocentric distance of 1.29 au .
according to the data from jpl horizons database , the comet orbit has an eccentricity of 0.9976 , orbital inclination of 80.3@xmath3 , semi - major axis of 576.34 au and orbital period of 13700 years .
+ @xcite have carried out a detailed survey of 85 comets during the years 1976 - 1992 .
they have studied the variation of gas production rates and dust to gas ratio with heliocentric distances .
they have defined certain limits in the gas ratios in order to classify the comets as to whether they are depleted in carbon - chain molecules or not . according to their study ,
most of the carbon depleted comets are from the jupiter family .
they also say that cn is produced from the nucleus as well as from the dust in the comet s coma .
several spectroscopic surveys have been carried out , e.g. @xcite , @xcite , @xcite . @xcite
have spectroscopically studied five comets .
they found a linear correlation between production rates of c@xmath4 , c@xmath5 and c@xmath1 with respect to cn .
no correlation was found between the production rate ratios and heliocentric distance . @xcite
have given new statistical methods in cometary spectroscopy for the extraction and subtraction of the sky using the comet frames itself .
they have given a dynamical classification of comets by studying 26 different comets .
more recently , @xcite have compiled and reported the spectroscopic results for 130 comets observed from mcdonald observatory .
they have found remarkable similarity in the composition of most of the comets .
they quote that the carbon chain depleted comets can be from any dynamical class .
however they have not found any of them from the halley type comets ( tisserand parameter @xmath6 2 and period @xmath6 200 years ) due to significantly low number of these types in their sample . + in our study of comet c/2014 q2 , we have determined the production rates for various gas species , @xmath2 and dust to gas ratios at different heliocentric distances .
the observations were carried out with lisa spectrograph mounted on the 0.5 m(f/6.8 ) telescope ( planewave instruments cdk20 ) at the mount abu infra - red observatory ( miro ) , mount abu , india . @xcite
have given a brief overview of the telescope at the observatory . in the following we describe , briefly , the spectrograph and comet observations .
lisais a low resolution high luminosity spectrograph which is designed for the spectroscopic study of faint and extended objects .
the light from the object that passes through the slit falls on the grating .
the rest of the light is reflected towards the guiding ccd camera .
two detector cameras are placed on either side of the spectrograph : atik 314l is used for acquiring the object spectrum .
this has a chip size of 1392 x 1040 pixels where each is a square pixel of size 6.45 @xmath7 .
the light that passes through the slit is reflected by the mirror and is transformed into a parallel beam by a collimator .
this is then diffracted by the grating and then passes through the objective lens and gets converged onto the detector . as the focal length of the collimator is 130 mm and that of the camera lens is 88 mm , the image on the slit plane is multiplied by a factor of 88/130=0.68 onto the plane of the ccd .
the second detector is atik titan guiding camera .
this has a chip size of 659 x 494 pixels where each is a square pixel of size 7.4 @xmath7 .
all the reflected light that does not pass through the slit is directed to the guiding mirror .
this is then sent to the guiding optics which refocuses the light on to the guiding camera .
there are 8 different slits that are provided with the instrument .
we have used a slit width of 1.37 arcsec for january and february , and a slit width of 2.08 arcsec for march and may observations .
grating with 300 grooves / mm was used covering a total wavelength range from @xmath8 3800 @xmath9 to @xmath8 7400 @xmath9 with a wavelength scale of 2.6 @xmath9 per pixel .
the plate scale on the ccd plane is 0.57 arcsec per pixel .
the length of the slit imaged onto the ccd plane is 130 arcsec .
we carried out spectroscopic observations from mount abu infra - red observatory ( miro ) using lisa , a low resolution spectrograph mounted on the 0.5 m telescope during january to may 2015 .
the sky conditions were photometric during the observing period .
bias images were taken at regular intervals .
the slit was assumed to be uniformly illuminated . with this assumption ,
lamp flats were taken after every round of observation .
this was used for flat field corrections .
an argon - neon lamp was used for the wavelength calibration .
details of the comet observations are given in table [ comet_obs ] .
the exposure times mentioned in the table are for each individual frame .
the slit was placed at the photo - center of the comet and was manually tracked through the guiding ccd throughout the exposure time .
the observations were made using the scheme , sky - object - sky , for the proper background sky subtraction .
the standard stars ( along with their spectral types ) from iraf s catalogue of spectroscopic standardsobserved for flux calibration are as follows : hd15318(b9 ) , hd19445 ( g2v ) , hr1544(a1 ) , hd126991(g2v ) , hd129357(g2v ) , hd140514(g2v ) , hr7596(a0 ) , hd192281(o4.5 ) , hd217086(o7 ) .
these were preferably observed at a similar airmass as that of the comet .
.observational log .
columns : date , mid - ut , heliocentric distance in au , geocentric distance in au , phase angle in degree , exp.(exposure time ) in second , airmass [ cols="^,^,^,^,^,^,^",options="header " , ] [ afrho_table ] @xmath10 the gas production rate as well as @xmath2 increase after perihelion and show a decreasing trend only after february 2015 .
it is interesting to note the simultaneous increase in gas and dust which indicates an increase in the overall activity of the comet after its perihelion passage .
this kind of asymmetry has been seen in many comets .
in fact @xcite have tabulated the difference in the pre and post perihelion values for different class of comets with positive values indicating post perihelion activity .
although we do not have data points at exactly the same distance for pre and post perihelion passages , we can perhaps say that this comet may have a large positive asymmetry .
this indicates that there might be volatile material present beneath the surface of the comet or the surface of the nucleus consists of layers of ice that have different vaporization rates .
we have carried out spectral study of the comet c/2014 q2 ( lovejoy ) which shows strong emission lines of c@xmath0 , c@xmath1 and cn .
the molecular production rates and the quantity af@xmath11 are estimated .
we arrive at the following conclusions : * the comet shows a large positive asymmetry , which indicates increase in the post - perihelion activity . * considering the limits set by @xcite for the carbon abundance , c/2014 q2 is found to be in the typical class of comets . *
the dust to gas ratio remains fairly constant within the observed heliocentric range which is in agreement with @xcite * the @xmath2 values indicate that the dust distribution favours smaller dust particles near perihelion which depletes as the comet moves away .
* new haser model scale lengths were obtained which best fit the observed data .
these differ significantly from the previously quoted values .
this might be due to the comet s intrinsic nature itself or affected by the solar activity at the time of observation .
a more exhaustive study is required to confirm some of the conclusions drawn here .
this work is supported by the dept of space , govt of india .
we acknowledge the local staff at the mount abu infra - red observatory for their help .
we would also like to thank prof .
david schleicher for his invaluable inputs and suggestions .
we acknowledge with thanks prof jeremy tatum for useful discussions .
we would like to specially thank ms .
navpreet kaur and mr .
sameer for their help in the observations .
we also thank our colleagues in the astronomy & astrophysics division at prl for their comments and suggestions .
99 ahearn m.f .
, schleicher d.g .
, feldman p. , mills r. , thompson d. , 1984 , aj , 89 , 579 .
ahearn m.f.,millis r.,schleicher d.g.,osip d.j , birch p. , 1995 ,
icarus , 118 , 223 cochran a.l . ,
barker e.s . ,
gray c.l . ,
2012 icarus , 218 , 144 - 168 cochran a.l . , 1985 ,
aj , 90 , 12 farnham t l. , schleicher d g. , and ahearn m f. , 2000 , icarus 147,180 - 204 .
michael r. , and disanti michael a. , 1991 , aj , 383 , 356 - 371 fink u. , hicks m.d . , 1996 , apj , 459 , 729 - 743 fink u.,2009 , icarus , 201 , 311 - 314 gilbert a.m. , wiegert p.a .
, unda - sanzana e. , vaduvescu o. , 2010 , mnras , 85 , 220 ganesh , s. , baliyan , k. s. , chandra , s. , joshi , u. c. , kalyaan , a. , mathur , s. n. , 2013 , astronomical society of india conference series , 99 , 9 krishna swamy k.s . , `` the physics of comets '' , 3@xmath12 edition , world scientific series langland - shula l e. , smith g h. , 2011 , icarus , 213 , 280 - 322 newburn r.,spinard h.,1989 , aj , 97 , 552 schleicher d g. , 2010 , aj , 140 , 973 tatum j. , 1984 , aa , 135 , 183 - 187 | spectra of comet c/2014 q2 ( lovejoy ) were taken with a low resolution spectrograph mounted on the 0.5 m telescope at the mount abu infrared observatory ( miro ) , india during january to may 2015 covering the perihelion and post - perihelion periods .
the spectra showed strong molecular emission bands ( c@xmath0 , c@xmath1 and cn ) in january , close to perihelion .
we have obtained the scale lengths for these molecules by fitting the haser model to the observed column densities .
the variation of gas production rates and production rate ratios with heliocentric distance were studied .
the extent of the dust continuum using the @xmath2 parameter and its variation with the heliocentric distance were also investigated .
the comet is seen to become more active in the post - perihelion phase , thereby showing an asymmetric behaviour about the perihelion .
[ firstpage ] comet : general - comet : individual : c/2014 q2 - methods : observational - techniques : spectroscopic - telescopes - oort cloud |
type ia supernovae ( sne ia ) are believed to be thermonuclear runaway explosions of carbon / oxygen white dwarfs ( see , e.g. , hillebrandt & niemeyer 2000 for a review ) .
observationally , sne ia share similar light - curve shapes ; thus , traditionally the fitting of sn ia light curves is conducted with templates constructed from well - observed sne ia ( e.g. , jha et al .
2007 ; guy et al . 2007 ) .
some other attempts have also been proposed to characterize sn ia light curves with different techniques .
for example , kessler et al . (
2010 ) use the supernova photometric classification challenge , a publicly released tool with simulated sne for light - curve classification of sne and photometric redshift estimation ; bazin et al .
( 2011 ) employ the salt2 package ( described by guy et al . 2010 ) as a sn ia light - curve fitter to select sn - like events through the photometric sample of the cfht supernova legacy survey ; and kim et al .
( 2013 ) model sn light curves by training a parameterized model for the multiband light curves as arising from a gaussian process , followed by applying the results to spectrophotometric time series of sne to simultaneously standardize sne and fit cosmological parameters .
however , to date , no single functional form has been proposed with reasonable physical meaning to fit the light curves of sne ia from explosion to a few weeks after peak brightness . in this paper , we present an empirical fitting method to characterize the optical light curves of sne ia by using a mathematically analytic function .
assuming the sn ia bolometric luminosity scales as the surface area of the expanding fireball ( which is approximately a blackbody at early times , modified to some degree by line blanketing in the blue and ultraviolet ) , and given that optical wavelengths are on the rayleigh - jeans tail of its nearly thermal spectral energy distribution at typical temperatures exceeding @xmath3k ( see [ s : discussion ] ) , the optical luminosity of the sn increases quadratically with the photospheric radius ( see riess et al .
1999 ) : @xmath4 ^ 2 t , \label{eq_lvtt2}\ ] ] where @xmath5 is the photospheric radius , @xmath6 is the fireball temperature , @xmath7 is the photospheric expansion velocity , @xmath8 is the time of explosion , and @xmath9 is the time after explosion . assuming that the blackbody temperature is roughly constant at early times ( see discussion below ) , the optical luminosity is @xmath10 thus , considering only very early times ( a few days after explosion )
, the optical luminosity should be roughly proportional to the square of the time since explosion ( @xmath11 , commonly known as the @xmath12 model ; e.g. , arnett 1982 ; riess et al .
1999 ) , _ assuming the photospheric velocity does not change much _ ( but see discussion below ) .
observationally , the @xmath12 model fits well for several sne ia with early - time observations ( e.g. , sn 2011fe , nugent et al .
2011 ; sn 2012ht , yamanaka et al .
2014 ) .
however , here we aim to extend the fitting to much longer times for sne ia ( several weeks after explosion but before entering the phase dominated by cobalt decay ) , so it is inappropriate to assume that the photospheric velocity is nearly constant ; in fact , observed spectral series show that @xmath7 drops rapidly at early times and thereafter slowly but steadily decreases ( e.g. , silverman et al .
therefore , a model for the photospheric velocity evolution is required before fitting the light curves of sne ia .
zheng et al . ( 2016a ) propose a broken - power - law model to fit the photospheric velocity evolution from the explosion time to more than a month later .
the function , which is also widely used for fitting gamma - ray burst afterglows as well as early - time sn ia light curves ( e.g. , zheng et al . 2012 ; 2013 ; 2014 ) , is given by @xmath13}^{-1/s } , \label{eq_velocity}\ ] ] where @xmath7 is the photospheric velocity , @xmath14 is a scaling constant , @xmath15 is the break time , @xmath16 and @xmath17 are the two power - law indices before and after the break ( respectively ) , and @xmath18 is a smoothing parameter . applying this velocity function to equation [ eq_lvt2 ] , we find the sn ia optical luminosity to be @xmath19}^{-2/s}. \label{eq_lvtbkn1}\ ] ] to simplify this , we introduce @xmath20 and @xmath21 as @xmath22 @xmath23 and so equation [ eq_lvtbkn1 ] becomes @xmath24}^{-2/s}. \label{eq_lvtbkn2}\ ] ] equation [ eq_lvtbkn2 ] is the final function we propose to fit to sn ia light curves for a wide time range ( from explosion to more than a month later ) .
it reaches a peak value when @xmath25}. \label{eq_tpeak}\ ] ] the functional form of equation [ eq_lvtbkn2 ] is very similar to the broken power - law function equation [ eq_velocity ] , but with some variance .
an advantage of the empirical equation [ eq_lvtbkn2 ] is that there is reasonable physics behind it ; @xmath20 is considered the rising power - law index and @xmath21 the decaying index , and both are related to the photospheric velocity evolution .
the function itself is also mathematically analytic , derived directly from the photospheric velocity evolution function ( along with the simple assumption that the emission is approximately that of a blackbody ) .
we choose sn 2011fe as a case study to apply our fitting method ; it is the most ideal object for our purpose .
first , it was discovered when extremely young and the explosion time is well constrained ( nugent et al .
2011 ; li et al .
second , sn 2011fe is well observed both spectroscopically and photometrically in multiple bands after discovery ( e.g. , nugent et al . 2011 ; richmond & smith 2002 ; vinko et al .
2012 ) .
since the fitting function of equation [ eq_lvtbkn2 ] is derived directly from the photospheric velocity evolution function , we first apply the velocity fitting of sn 2011fe using equation [ eq_velocity ] , similar to the fitting of other sne given by zheng et al .
( 2016a ) .
our results are shown in figure [ 11fe_velocity_fitting ] , where the photospheric velocity is derived from the strong si ii @xmath266355 absorption line .
6355 absorption ) evolution of sn 2011fe .
top panel : the result of a power - law fit to the early - time data .
middle - top panel : the result of a linear fit to the later - time data .
middle - bottom panel : the result of a broken power - law fit to all the data .
bottom panel : the residuals for each fit .
solid points are included in the fitting while open points are not . ]
figure [ 11fe_velocity_fitting ] confirms the finding of zheng et al .
( 2016a ) that the photospheric velocity can be well described by a broken - power - law model ( equation [ eq_velocity ] ) over a relatively long time range .
a single power law ( top panel ) or a linear function ( middle - top panel ) can fit only either early - time data or later - time data independently , while a broken power law can fit all the data ( middle - bottom panel ) .
the single power - law fit at early times gives an index of -0.20 , and the single linear function at later times yields @xmath27 km s@xmath28 day@xmath28 , while the composite fit gives power - law indices @xmath29 and @xmath30 .
these results are consistent with the fits to the other sne ia ( sne 2009ig , 2012cg , 2013dy and 2016coj ) given by zheng et al .
( 2016a ) . next
, we apply multiband light - curve fitting to sn 2011fe using equation [ eq_lvtbkn2 ] .
optical data are gathered from the published literature ( nugent et al .
2011 ; richmond & smith 2002 ; vinko et al .
2012 ) , excluding a few clear @xmath31 outliers among different instruments . since sne ia
usually exhibit a shoulder in the @xmath5 band and a second peak in the @xmath32 band , we restrict the fits to earlier times in the redder bands ( @xmath5 , @xmath32 ) than in the bluer bands ( @xmath2 , @xmath33 , @xmath34 ) . given that the explosion time of sn 2011fe is well estimated ( nugent et al .
2011 ) , we fix @xmath8 in the fits .
note that sne ia are expected to have a dark phase " which lasts for a few hours to days between the moment of explosion and the first observed light ( e.g. , rabinak , livne , & waxman 2012 ; piro & nakar 2013 , 2014 ) .
however , thus far there has been no direct observational evidence for a dark phase , although some progress is starting to be made ( e.g. , shappee et al .
the quantity we determine from the data is actually the first - light time . in the following ,
we simply assume that the explosion time @xmath8 is equal to the first - light time . for sn 2011fe , we denote @xmath35 = mjd 55796.687 .
the fitting results are given in figure [ 11fe_iabkns_fitting_case1 ] and table 1 .
, with @xmath8 fixed to be @xmath35 ( mjd 55796.687 ) .
diamond - shaped data points are included in the fitting while cross - shaped ones are excluded .
the fitting results cover both a long time range ( from explosion to a few weeks after peak time ) and a large flux scale ( nearly a factor of 100 , or @xmath0mag , and even @xmath1mag in @xmath2 ) . ] as shown in figure [ 11fe_iabkns_fitting_case1 ] , by using just a single function ( equation [ eq_lvtbkn2 ] ) , the fitting results are surprisingly good over many weeks and nearly a factor of 100 in flux ( @xmath0mag but @xmath1mag in @xmath2 where we have extra data from the night of discovery ) .
most flux residuals are within the 12@xmath36 measurement uncertainty .
the fit was applied to a long time range for all filters : for the @xmath2 , @xmath33 , and @xmath34 bands , data are used up to 50 days after explosion , while @xmath5 ( 35 days ) and @xmath32 ( 31 days ) are affected by the shoulder and second peak ( respectively ) in the light curves .
interestingly , although we include data only up to 50 days in @xmath34 , the model still provides a good match the data up to about 3 months post - explosion ( though this might be just a coincidence ) .
the @xmath37 values derived from the fits ( table 1 ) are all around 2.1 ( except for @xmath33 , with a somewhat larger value but also more uncertain ) , close to the @xmath2-band rising index of 2.01 derived by nugent et al .
( 2011 ) from the first few days of data though zhang et al .
( 2015 ) find the sn 2011fe rising index to vary from 2.25 to 2.63 in different filters .
in general , a rising index around 2.1 is consistent with the commonly known @xmath12 model for most sne ia , or the @xmath38 model ( @xmath393.0 ) used by various groups ( e.g. , conley et al .
2006 ; hayden et al .
2010 ; ganeshalingam et al .
2011 ; firth et al .
2015 ) ; however , these previous studies derived the index from the very early - time data ( a few days after first - light time ) , whereas here we derive it from light - curve fitting over a long time using equation [ eq_lvtbkn2 ] .
cc|c|c|c|c 0.4 mm @xmath2 & [email protected] ( [email protected] ) & [email protected] & [email protected] & [email protected] + @xmath33 & [email protected] ( [email protected] ) & [email protected] & [email protected] & [email protected] + @xmath34 & [email protected] ( [email protected] ) & [email protected] & [email protected] & [email protected] + @xmath5 & [email protected] ( [email protected] ) & [email protected] & [email protected] & [email protected] + @xmath32 & [email protected] ( [email protected] ) & [email protected] & [email protected] & [email protected] overall , the fits to the light curves ( up to @xmath41 month after explosion ) for sn 2011fe with a single empirical function are very successful .
since sne ia share similar light - curve shapes , equation [ eq_lvtbkn2 ] also has the potential to fit most other sne ia .
thus , it is important to understand the possible physics behind this expression .
the form of equation [ eq_lvtbkn2 ] is a variant of a broken - power - law function .
the parameters @xmath42 , @xmath8 , and @xmath15 are easy to understand , while the parameters @xmath37 , @xmath43 , and @xmath18 are inherited from equation [ eq_velocity ] , the photospheric velocity evolution function . here
, @xmath20 represents the rising power - law index of the light curves and @xmath21 represents the decaying index .
the most interesting and important parameter is probably @xmath37 ( and thus the closely related @xmath44 ) .
the relation between ( @xmath37 , @xmath43 ) and ( @xmath44 , @xmath45 ) through equations [ eq_alphar_alpha1 ] and [ eq_alphad_alpha2 ] comes from the derivation of equation [ eq_lvtbkn2 ] , where @xmath44 and @xmath45 are the velocity decay indices from the broken power law that is measured directly from optical spectra
. however , there seems to be discrepancy between the @xmath44 value inferred from light - curve fitting and the @xmath44 value measured from the photospheric velocity ( through si ii @xmath266355 ) , which we denote as @xmath46 . from the light - curve fits , @xmath47 ( see table 1 , except @xmath33 ) , while from spectroscopic si ii @xmath266355 observations , @xmath46 is around @xmath48 ( see section [ s : sec_siii_velocity_fitting ] ) .
if @xmath46 is adopted , @xmath37 would be 1.6 according to equation [ eq_alphar_alpha1 ] , much smaller than our fit - derived value of @xmath49 .
one possible reason for this discrepancy could be changes in the temperature .
when deriving equation [ eq_lvt2 ] from equation [ eq_lvtt2 ] , we assumed the temperature @xmath6 to be constant , which is unlikely to be true for sne ia . for sn 2011fe , mazzali et al .
( 2014 ) found a significant increase in temperature during the first few days , a roughly constant temperature around peak brightness , and a relatively slow decay after peak .
if we fit a power law to the first few days before peak ( see figure [ 11fe_temperature_fitting ] , dashed line ) , we find a rising index of @xmath50 .
this temperature increase would contribute together with @xmath46 to the final @xmath37 value , which would be around @xmath51 ( according to equation [ eq_alphar_alpha1 ] ) , much closer to the observed value @xmath52 .
however , even though we did not consider the changes in temperature , the formulation of equation [ eq_lvtbkn2 ] itself can in principle accommodate all contributions by adjusting the @xmath37 value .
this means that technically , it is fine to fit the light curves using equation [ eq_lvtbkn2 ] , but the relation in equation [ eq_alphar_alpha1 ] would not be valid owing to other contributions such as temperature changes .
the parameter @xmath37 could still represent the early - time rising index , though @xmath44 would include not only the @xmath46 contribution but also the changes in temperature and perhaps other contributions .
for example , we use the si ii @xmath266355 line to measure the photospheric velocity , but there is considerable debate regarding whether si ii @xmath266355 really does provide an accurate photospheric velocity ( e.g. , blondin et al .
et al . 2012 )
. however , for our fitting purposes , it is reasonable to assume that there is a relation between the si ii @xmath266355 velocity and the photospheric velocity , even if the relation varies with time ; equation [ eq_lvtbkn2 ] can accommodate this through adjustments in the @xmath37 value during the fitting process .
one may also notice that similar to @xmath44 , the @xmath45 value from the light - curve fitting ( around @xmath53 ; see table 1 ) differs from the photospheric velocity fitting ( @xmath54 ; see [ s : sec_siii_velocity_fitting ] ) .
the reason could be similar to that for @xmath44 ; after peak brightness , the temperature decreases as the fireball expands , and the ejecta also become less dense .
this causes the luminosity to drop fast , with a larger absolute value for the power - law index @xmath45 ( and @xmath43 ) .
technically , one can also fit the temperature with a broken power law ( see figure [ 11fe_temperature_fitting ] , dotted line ) ; thus , both the increase and decrease of the power - law index would contribute to the light - curve fitting .
but in order to simplify the fitting procedure , and considering that equation [ eq_lvtbkn2 ] itself can accommodate the temperature changes by adjusting the @xmath37 and @xmath43 values , we did not explicitly include the temperature component in our model ; however , its contribution has been included in @xmath37 and @xmath43 during the light - curve fitting .
essentially , equation [ eq_lvtbkn2 ] represents blackbody emission from the expanding fireball of a sn ia .
the radius of the fireball is proportional to the photospheric velocity , and therefore the equation is derived directly from the photospheric velocity evolution .
although when deriving equation [ eq_lvtbkn2 ] we assumed the blackbody temperature is constant , the equation itself can accommodate the temperature changes though the interpretation becomes more complicated if trying to distinguish the contributions from different components .
nevertheless , we see that the equation has a reasonable physical explanation , and it can be used to accurately fit sn ia light curves during at least the first month after explosion . in reality , very few sne ia other than sn 2011fe are discovered sufficiently early for their explosion time to be well estimated .
however , a useful potential application from the above fitting method is that one can use equation [ eq_lvtbkn2 ] to estimate the first - light or explosion time , @xmath35 . for most sne ia ,
not discovered extremely early but instead around 1 - 2 weeks after explosion and well monitored thereafter , by fitting the light curves using equation [ eq_lvtbkn2 ] one can estimate @xmath35 .
this application will be presented by zheng , kelly , & filippenko ( 2016b ) .
we have found a function with a reasonable physical explanation that can well fit sn ia optical light curves .
it is mathematically analytic and derived directly from the photospheric velocity evolution , adopting the simple assumption that the optical emission is approximately the blackbody emission of an expanding fireball .
applying this function to the case study of sn 2011fe gives surprisingly good results . since sne ia share similar light - curve shapes , this fitting method has the potential to fit most other sne ia , providing valuable benefits for large datasets of sn ia light curves obtained with current ( e.g. , pan - starrs ; intermediate palomar transient factory ) and near - future ( e.g. , zwicky transient facility ; the large synoptic survey telescope ) facilities .
we thank isaac shivvers and melissa l. graham for useful discussions and suggestions , as well as the staff of the observatories where data were obtained .
a.v.f.s supernova group at uc berkeley is grateful for financial assistance from nsf grant ast-1211916 , the tabasgo foundation , and the christopher r. redlich fund .
research at lick observatory is partially supported by a generous gift from google . | we present a new empirical fitting method for the optical light curves of type ia supernovae ( sne ia ) .
we find that a variant broken - power - law function provides a good fit , with the simple assumption that the optical emission is approximately the blackbody emission of the expanding fireball .
this function is mathematically analytic and is derived directly from the photospheric velocity evolution . applying it to the case study of sn 2011fe gives a surprisingly good fit that can describe the light curves from explosion to a few weeks after peak brightness , as well as over a large range of fluxes ( @xmath0 mag , and even @xmath1mag in the @xmath2 band ) .
since sne ia share similar light - curve shapes , this fitting method has the potential to fit most other sne ia and characterize their properties in large statistical samples such as those already gathered and in the near future as new facilities become available . |
mechanical harmonic oscillator plays an important role in the historical development of quantum mechanics .
the harmonic oscillator problem was one of the few completely solvable problems when one began to learn quantum mechanics . due to the macroscopic nature , the experiments of mechanical harmonic oscillators did nt achieve much progress for a quite long time , however .
recently , with the development in quantum information processing , people are now searching for possible applications of mechanical harmonic oscillator in this field , and evidence for quantized displacement in a nanomechanical harmonic oscillator has been observed @xcite .
many good physical ideas about the harmonic oscillator came up , though this system is rather simple .
squeezed state was one of them .
squeezed state provides a good example of the interplay between experiment and theory in the development of quantum mechanics .
statistical properties of squeezed states have been widely investigated and the possibility of applying squeezed states to study the fundamental quantum physics phenomena , as well as to detect the gravitational radiation , has been recognized @xcite . though the idea of squeezed states originated from mechanical harmonic oscillator , the first experiment realization was the squeezed states of electromagnetic field in nonlinear quantum optics@xcite .
in the nonlinear optical experiments , three - wave and four - wave mixing were two main methods to generate squeezed states .
if one injects low energy photons into a nonlinear optical medium , the second harmonic generation may be induced and this forms a squeezed state , this procedure is called three - wave mixing . though the theory of four - wave mixing was more complicated , the first squeezed state of electromagnetic field was implemented in this system . and
recently , the theory of generating squeezed states in a high - q cavity was considered @xcite .
recently , there has been significant progress in realizing quantum optics in solid state electrical circuits .
this new subject has a nickname circuit quantum electrodynamics ( circuit qed ) " @xcite .
the extreme strong coupling limit in cavity qed has been implemented experimentally with circuit qed systems , such as superconducting charge qubit and superconducting transmission line resonator ( stlr ) system @xcite , flux qubit and quantum @xmath1 oscillator system @xcite , and phase qubit and nanomechanical resonator ( nmr ) system ( which is a mechanical harmonic oscillator ) @xcite . because of the extreme strong coupling in circuit qed , generating squeezed states in such systems became very interesting and attracted much attention @xcite . in ref .
@xcite , using two modes of the stlr coupled to a superconducting charge qubit simultaneously , the authors studied microwave parametric down - conversion and discussed the squeezed states of one mode of the stlr . in the other refs.@xcite , the authors discussed the generation of squeezed states of nmr .
the generation of squeezed states of nmr in these schemes use either the operations on the qubit or dissipation and measurement to produce the needed nonlinearity . in this paper
, we propose a scheme for generating squeezed states of nmr in circuit qed , and it is similar to the three - wave mixing in optical cavity experiments .
our proposal is based on the system consisting of a superconducting cooper - pair box ( cpb ) , also a superconducting charge qubit , a nmr and a stlr .
we find that with certain biased conditions of the cpb , the nonlinear process can be implemented and the squeezed states of nmr can easily be generated .
the nonlinear interaction can be switched on and off at will by changing the external biased flux of the cpb . by controlling the gate charge @xmath2 and/or @xmath3
, one can get different squeezed variables @xmath4 and/or @xmath5 .
compared with other schemes , our scheme is more simple and effective in generating the squeezed states . in this scheme , the state of cpb is adiabatically kept in its ground states , and does not need qubit operations so that it avoids the restriction set by the decoherence time of the qubit .
the system under consideration is shown in fig.[model ] . a stlr ( the blue line at the bottom of fig.[model ] ) is placed close to a cpb . the state of cpb is separately controlled by the gate voltage @xmath6 through a gate capacitance @xmath7 .
the cpb is also coupled to a large superconductor reservoir through two identical josephson junctions with capacitance @xmath8 and josephson energy @xmath9 , and this forms a superconducting quantum interference device ( squid ) and is also the basic configuration of superconducting charge qubit .
the squid configuration allows one to apply external flux to control the josephson energy . for superconducting charge qubits ,
the capacitance @xmath7 is much less than @xmath8 . in this regime
a convenient basis is formed by the charge states , characterized by the number of cooper pairs @xmath10 on the cpb .
in the neighborhood of @xmath11 , only two charge states @xmath12 and @xmath13 play a role , while all other charge states , having a much higher energy , can be ignored . in this case
, the hamiltonian of the cpb reads @xcite @xmath14 where @xmath15 is the charging energy , @xmath16 is the total capacitance saw by the cpb , @xmath17 is the gate charge induced by the gate voltage , and @xmath18 is the externally applied flux . for a cpb fabricated inside a stlr ,
there is not only the dc voltage @xmath6 but also a quantum part @xmath19 applied on the gate capacitance .
the quantized voltage at the antinode @xmath20 of the stlr takes it s maximum amplitude @xcite @xmath21 here , @xmath22 , with @xmath23 , @xmath24 and @xmath25 being the length , the inductance and capacitance per unit length of the stlr , respectively . at low temperatures ,
there is only one mode of the stlr , say @xmath26 , that couples to the cpb , then the quantum voltage applied on the gate capacitance beomes @xmath27 .
the hamiltonian of the joint system ( stlr and cpb ) has a spin - boson form @xmath28 now we consider the coupling between the nmr and the cpb .
a nmr is fabricated as a part of the squid , which has a length @xmath29 and width @xmath30 , as shown in fig.[model ] .
considering the small displacement of the nmr , the effective area of the squid is @xmath31 , where @xmath4 is the displacement operator and @xmath32 , with @xmath33 and @xmath34 being the frequency and mass of the nmr , respectively .
thus the flux bias of the squid is @xmath35 , where @xmath36 is the flux bias corresponding to the equilibrium position of the nmr @xcite . with the total flux bias ,
the effective josephson coupling energy of the squid becomes @xmath37 the hamiltonian of the total system , including stlr , nmr and cpb reads @xmath38 it s obviously that the josephson coupling energy eq.([potential ] ) is a nonlinear function of @xmath4 and can be expanded to @xmath39 .
\label{potential2}\end{aligned}\ ] ] in general , @xmath4 is very small , then the first and second term of eq.([potential2 ] ) can be discarded and expanded to the first order of @xmath4 , respectively . in this way
, one can get the linearized function of josephson coupling energy of @xmath4 , and there is only linear interaction terms remained in hamiltonian ( [ htotal ] ) . as pointed by zhou _
@xcite , the nonlinearity of the josephson coupling energy in @xmath4 should nt be neglected all the time , and it would be important for the generation of squeezed states .
in the past years , two cavities interacting with a two - level atom has been widely studied @xcite , they focused on the linear interaction in such systems . here , we study the nonlinear interaction in the stlr , nmr and cpb system . an important nonlinear effect is the generation of squeezed states , which is an outstanding task in quantum mechanics . in the following
we show how we realize the nonlinear interaction and how we can use the nonlinear interaction to generate squeezed states .
our method is similar to the three - wave mixing in quantum optics . in order to realize nonlinear interaction
, one can bias the squid at @xmath40 , here @xmath41 is an integer , and expand the josephson energy to the second order in @xmath4 , then the hamiltonian ( [ htotal ] ) becomes @xmath42 where @xmath43 in the following we choose the eigenenergy basis ( spanned by @xmath44 and @xmath45 ) to simplify the above hamiltonian . here , @xmath46 $ ] is the mixing angle . in the representation of eigenenergy basis of the cpb and under the rotating - wave approximation ,
the hamiltonian ( [ hamiltonian ] ) is simplified to @xmath47 where @xmath48 , @xmath49 , @xmath50 , @xmath51 and @xmath52 . assuming the detunings between the cpb and the stlr and the nmr
satisfy the large detuning limits , that is @xmath53hlich transformation @xcite on the total hamiltonian ( [ hra ] ) . dividing the hamiltonian ( [ hra ] ) into two parts @xmath54 , with @xmath55 and @xmath56
apply a unitary transformation @xmath57 with the generator @xmath58 , and expand @xmath59 to second order in @xmath60 ( @xmath61 ) , we obtain the effective hamiltonian @xmath62 if the cpb is adiabatically kept on the ground state , the effective hamiltonian becomes @xmath63 in general , @xmath64 and @xmath65 , then the effective hamiltonian can be approximated as @xmath66 in the interaction picture , the hamiltonian ( [ pdc ] ) reads @xmath67 where @xmath68 . if @xmath69 , that is @xmath70 and @xmath71 , in this situation the hamiltonian becomes @xmath72 where @xmath73 is the coupling constant between the stlr and nmr . in the parametric approximation ,
the hamiltonian ( [ hinter ] ) becomes @xmath74 where @xmath75 and @xmath76 is the amplitude and phase of the stlr which is in a coherent state .
the time evolution operator of the nmr in the interaction picture reads @xmath77 in fact , the time evolution operator ( [ top ] ) is the squeezed operator of the nmr . for a time duration @xmath78
, the squeezed operator reads @xmath79 where @xmath80 is the effective squeezed parameter . for the nmr initially in the vacuum state @xmath81 and @xmath82 , the variance in the two quadratures @xmath83 and @xmath84 , where @xmath85 and @xmath86 , can be calculated directly using the transformation @xmath87 , @xmath88 and the nmr is in the squeezed state @xmath89 if the nmr in the coherent state @xmath90 initially , one can generate the ideal squeezed state @xmath91 . to estimate the squeezed efficiency , we choose the typical parameters in current solid - state circuits experiments as follows : @xmath92 ghz , @xmath93 ghz , @xmath94 ghz , @xmath95 , @xmath96 t , @xmath97 m , @xmath98v , and @xmath99 m. we can get two different values of coupling constant @xmath100 hz with @xmath2 and @xmath101 hz with @xmath102 . in the above discussion
, we have assumed an ideal situation in which the noise and fluctuations were not included .
the noise in the cpb may not be considered in our scheme because the cpb can naturally stay in it s ground state at low temperatures for a very long time . here
we investigate the effect of a finite linewidth @xmath103 induced by phase fluctuations in driving coherent state on the squeezing properties of the nmr , as has been done in three - wave mixing in quantum optics . versus @xmath104 for different ratios ( a ) @xmath105 , ( b ) @xmath106 ,
( c ) @xmath107 , ( d ) @xmath108.,width=8 ] for the linewidth @xmath103 , and in the limit @xmath109 , the variances in the two quadratures are given by @xcite @xmath110 in fig.[noise ] , we have plotted @xmath111 versus @xmath104 for various ratios of @xmath112 .
the variances in the amplitude of @xmath4 increases due to the phase fluctuations in the driving coherent state .
there is a minimum in @xmath111 which decreases with increasing @xmath112 .
it s evident from eq.([sq ] ) that the phase fluctuation in driving coherent state would affect the squeezed efficient severely with increasing time @xmath78 due to the existence of @xmath113 .
our scheme is much different from the existing ones such as refs.@xcite . in our scheme
, it does nt need any operations on the cpb and does nt use dissipation and measurement to generate the needed nonlinearity .
it just needs to adiabatically keep the cpb in it s ground state .
our scheme can greatly decrease the effect of the decoherence of the cpb on the squeezed efficency .
the cpb plays the role of nonlinear media as in quantum optics , and our result is similar to the three - wave mixing . by controlling @xmath2 and/or @xmath114 , we can get different squeezed variables @xmath4 and/or @xmath5 . in conclusion , we proposed a scheme for generating squeezed states in solid state circuits system .
in such a system , a nmr is fabricated as a part of a squid , which consists of a cpb , and a stlr is capacitively coupled to the cpb .
the nonlinear interaction between the cpb and the nmr can be implemented by setting the external biased flux of the cpb at some certain values . by performing frhlich transformation
, we can get the nonlinear hamiltonian of parametric down - conversion of the stlr - nmr system . in our scheme
, the cpb plays the role of nonlinear media " and the squeezed states of the nmr can be generated in a manner similar to the three - wave mixing in quantum optics .
this work is supported by the national fundamental research program grant no .
2006cb921106 , china national natural science foundation grant nos .
10325521 , 60433050 , 60635040 , the srfdp program of education ministry of china , no .
20060003048 and the key grant project of chinese ministry of education no.306020 .
a. gaidarzhy , g. zolfagharkhani , r. l. badzey and p. mohanty , phys .
lett . * 94 * , 030402 ( 2005 ) ; m. o. scully and m. s. zubairy , _ quantum optics _ , cambridge university press , cambridge , 1997 .
m. f. bocko and r. onofrio , rev .
68 * , 755 ( 1996 ) .
r. e. slusher , l. w. hollberg , b. yurke , j. c. mertz , and j. f. valley , phys .
lett . * 55 * , 2409 ( 1985 ) .
l. a. wu , h. j. kimble , j. l. hall and h. wu , phys .
bf 57 , 2520 ( 1987 ) .
n. g. de almeida , r. m. serra , c. j. villas - bas and m. h. y. moussa , phys .
a * 69 * , 035802 ( 2004 ) .
a. blais , r .- s .
huang , a. wallraff , s. m. girvin , and r. j. schoelkopf , phys .
a * 69 * , 062320 ( 2004 ) .
i. chiorescu , p. bertet , k. semba , y. nakamura , c. j. p. m. harmans and j. e. mooij , nature * 431 * , 159 ( 2004 ) .
a. n. cleland and m. r. geller , phys .
lett . * 93 * , 070501 ( 2004 ) .
w. y. huo and g. l. long , arxiv : quant - ph/0702104 .
a. wallraff , d. i. schuster , a. blais , l. frunzio , r.- s. huang , j. majer , s. kumar , s. m. girvin , and r. j. schoelkopf , nature * 431 * , 162 ( 2004 ) . m. r. geller and a. n. cleland , phys . rev .
a * 71 * , 032311 ( 2005 ) .
k. moon and s. m. girvin , phys .
rev . lett .
* 95 * , 140504 ( 2005 ) .
x. x. zhou and a. mizel , phys .
rev . lett.*97 * , 267201 ( 2006 ) .
p. rabl , a. shnirman and p. zoller , phys . rev .
b * 70 * , 205304 ( 2004 ) .
l. tian and r. w. simmonds , arxiv : cond - mat/0606787 .
r. ruskov , k. schwab and a. n. korotkov , phys .
b * 71 * , 235407 ( 2005 ) .
f. xue , y. x. liu , c. p. sun and f. nori , arxiv : quant - ph/0701209 .
g. benivegna and a. messina , j. mod
. opt . * 41 * , 907 ( 1994 ) .
y. b. xie , j. mod . opt . * 42 * , 2239 ( 1994 ) . c. p. sun
, l. f. wei , y. x. liu , and f. nori , phys .
a * 73 * , 022318 ( 2006 ) .
y. makhlin , g. schn and a. shnirman , rev .
phys . * 73 * , 357 ( 2001 ) .
a. d. armour , m. p. blencowe and k. c. schwab , phys .
lett . * 88 * , 148301 ( 2002 ) .
e. k. irish and k. schwab , phys .
b * 68 * , 155311 ( 2003 ) .
i. martin , a. shnirman , l. tian and p. zoller , phys .
b * 69*,125339 ( 2004 ) . h. fr@xmath115hlich , phys . rev . * 79 * , 845 ( 1950 ) . | we propose a scheme for generating squeezed states in solid state circuits consisting of a nanomechanical resonator ( nmr ) , a superconducting cooper - pair box ( cpb ) and a superconducting transmission line resonator ( stlr ) .
the nonlinear interaction between the nmr and the stlr can be implemented by setting the external biased flux of the cpb at certain values .
the interaction hamiltonian between the nmr and the stlr is derived by performing fr@xmath0hlich transformation on the total hamiltonian of the combined system . just by adiabatically keeping the cpb at the ground state
, we get the standard parametric down - conversion hamiltonian .
the cpb plays the role of nonlinear media " , and the squeezed states of the nmr can be easily generated in a manner similar to the three - wave mixing in quantum optics .
this is the three - wave mixing in a solid - state circuit . |
the measurement of [ @xmath3/fe ] ratios of globular clusters provides important information regarding relative ages .
the @xmath3-elements ( c , o , ne , mg , si , s , ar , ca , and ti ) are synthesized predominantly by massive stars , and ejected into the interstellar medium by type ii supernovae ( sne ii ) explosions . on the other hand ,
iron - peak elements are produced by both sne ii and type ia supernovae ( sne ia ) .
enhanced values of [ @xmath3/fe ] ratios in globular clusters indicate the domination of sne ii nucleosynthesis , while lower ratios indicate increasing contribution of sne ia , which are thought to appear 10@xmath4 or more years later .
therefore , the measurement of the behavior of [ @xmath3/fe ] versus [ fe / h ] tells us how rapidly chemical enrichment proceeded in our galaxy ( wheeler , sneden , & truran 1989 ) .
the [ @xmath3/fe ] ratio is also important in deriving accurate age from comparisons of color - magnitude diagrams ( cmds ) to model isochrones .
[ @xmath3/fe ] ratios also provide a means of diagnosing the initial mass function ( imf ) and star formation rate ( sfr ) parameters ( mcwilliam 1997 ) .
an imf skewed to high - mass stars would have a higher @xmath3-element yield while a high sfr would give an enhanced [ @xmath3/fe ] at a higher [ fe / h ] since the interstellar medium will reach higher [ fe / h ] before the onset of sne ia contributions .
in the recent study of three metal - poor inner old halo globular clusters ( zinn 1993 , da costa & armandroff 1995 ) , ngc 6287 , ngc 6293 , and ngc 6541 , lee & carney ( 2002b ) claimed that [ si / ti ] ratios are anticorrelated with galactocentric distance of the old " halo globular clusters , in the sense that the silicon abundance increases and the titanium abundance decreases with galactocentric distances , and they proposed that the masses of sne ii progenitors may have been responsible for this elemental abundance gradient .
however , the impact of the skewed imf of sne ii progenitors in metal - rich globular clusters @xmath3-element abundances appears to be unclear as yet mainly due to the lack of study of metal - rich clusters .
thus more chemical abundance studies of metal - rich globular clusters near the galactic center are necessary , since the most metal - rich globular clusters are concentrated in the galaxy s central regions . palomar 6 is known to be a metal - rich globular cluster located 0.9 kpc from the galactic center . as lee & carney ( 2002a ) discussed , the metallicity of palomar 6 has been in controversy for two decades . malkan ( 1981 )
was the first to estimate the metallicity of palomar 6 , [ fe / h ] = @xmath21.30 , using a reddening - free metallicity index @xmath5 , ranking it a metal poor inner halo cluster .
zinn ( 1985 ) derived [ fe / h ] = @xmath20.74 for palomar 6 by reanalyzing malkan s photometry .
ortolani et al . (
1995 ) estimated [ fe / h ] @xmath6 @xmath20.4 based on the ( @xmath7 , @xmath8 ) - cmd . the results of minniti ( 1995b ) and bica et al .
( 1998 ) are worth noting .
minniti ( 1995b ) studied the metallicities of seven highly reddened clusters near the galactic center including palomar 6 .
he obtained low s / n ratio spectra of six red - giant branch ( rgb ) stars with 2 resolution covering @xmath9 4700 - 5400 .
he suggested that the sum of mg @xmath10 5175 , fe @xmath10 5270 , and fe @xmath10 5535 lines ( mg + 2fe ) is the optimal indicator of metallicity following faber et al .
( 1985 ) , and minniti derived [ fe / h ] = + 0.2 @xmath0 0.3 for palomar 6 based on the location of the giants in the ( mg + 2fe ) versus @xmath11 diagram ( see his figure 1 ) .
more recently , bica et al . ( 1998 ) undertook a metallicity study of the globular clusters near the galactic center with an expanded sample .
they measured the ca ii triplet @xmath9 8498 , 8542 , and 8662 using intermediate resolution ( 5.4 pixel@xmath1 ) integrated spectra and obtained [ @xmath12 = @xmath20.09 for palomar 6 .
it should be emphasized that the result of bica et al .
( 1998 ) may have a potential problem with field star contamination , since palomar 6 lies in the galactic center and the field star contamination is expected to be very high ( see , for example , figure 8 of lee & carney 2002a ) .
the abundance study of galactic bulge giants showed that the mean [ fe / h ] = @xmath20.25 ( mcwilliam & rich 1994 ) , which is comparable to that of bica et al .
( 1998 ) found for stars in or near palomar 6 .
also the metallicity estimates relying on low resolution spectra can give erroneous results in the metal - rich rgb stars due to blending with some of the notorious molecular bands , such as tio ( see , for example , plez 1998 ) and cn ( see , for example , mcwilliam & rich 1994 ) .
mcwilliam & rich ( 1994 ) have demonstrated the effect of the cn blending in the metal - rich bulge k giants in their figure 7 .
they claimed that the cn blending may cause a discrepancy in metallicity by @xmath6 0.3 dex between the results of mcwilliam & rich ( 1994 ) , who employed high resolution spectroscopy , and those of rich ( 1988 ) , who employed a similar method as minniti ( 1995 ) .
recently , stephens & frogel ( 2004 ) employed medium - resolution ( @xmath13 = 1380 1650 ) infrared ( ir ) @xmath14-band spectroscopy for 8 globular clusters near the galactic center .
they measured na i doublet ( @xmath9 21910
21966 ) , ca i triplet ( @xmath9 22450 22560 ) , co(2,0 ) band ( @xmath9 22915 23025 ) and obtained [ fe / h ] = @xmath20.52 @xmath0 0.11 for palomar 6 .
comparisons of high resolution spectra of cool stars of wallace & hinkle ( 1996 ) with the spectral regions used by stephens & frogel ( 2004 ) show that their spectral regions are vulnerable to contamination from the cn blending and their results may have problems similar to those of rich ( 1988 ) .
therefore high resolution spectroscopy free from blending with molecular bands is required . in this paper
, we present high - resolution ir spectroscopy of three rgb stars in palomar 6 . due to its high interstellar reddening value ,
@xmath15 @xmath6 1.25 1.4 mag , ir spectroscopy is essential for studying palomar 6 .
our observations were carried out at the 3.5 m nasa infrared telescope facility using cshell , the facility cryogenic ir echelle spectrograph ( tokunaga et al .
1990 ; greene et al .
1993 ) , on july 13 14 1997 for m71 and may 22 25 1998 for palomar 6 .
cshell is a long - slit spectrograph which uses a 31.6 lines / mm , 63.5@xmath16 echelle grating with narrow band circular variable filters that isolate a single order . the detector for our observations was a 256 @xmath17 256 hughes sbrc insb array , which provides a spectral coverage of @xmath6 750 km sec@xmath1 ( @xmath6 50 ) .
all of our observations were obtained with a 0.5 arcsec slit , providing a nominal resolution of 42,000 , or 7 km sec@xmath1 , using a nodding " mode , which moves the telescope between frames and stores each frame separately . because of the limited spectral coverage in a single grating setting , it was necessary to choose particular grating settings to cover spectral features of special interest .
we chose two spectral regions based on the ir atlas of arcturus ( hinkle et al .
one setting is for the two fe i lines ( centered at @xmath6 @xmath10 22395 ) and the other is for the two si i lines and a ti i line ( centered at @xmath6 @xmath10 21815 ) .
it should be emphasized that our spectral regions are free of molecular lines , such as cn or co. both spectral regions contain the neutral scandium lines , however , we do not derive the scandium abundance due to the lack of information of the hyperfine splitting of the sc i lines .
the basic photometric data for our targets using star ids of minniti ( 1995b ) are given in table 1 and the finding chart is given in figure 1 .
note that we adopt @xmath18 photometry by lee & carney ( 2002a ) in the table .
we also obtained spectra for the @xmath19co 2 - 0 bandhead for each target and for arcturus to measure the radial velocity .
this was necessary , as noted above , since field contamination toward palomar 6 is expected to be high due to its low galactic latitude .
once we confirmed probable cluster membership using the velocities for each target ( see section 3.1 ) , we obtained useful spectra for stars a , d , and g. we also obtained spectra of rapidly rotating early type stars in order to remove telluric absorption features .
the journal of observations is summarized in table 2 .
the raw data reductions were performed following the similar procedures described by carr , sellgren , & balachandran ( 2000 ) . in nodding mode ,
the data are stored with a " or b " name extensions to indicate beam polarity . subtraction of these a - beam and b - beam images removed dark current , residual images on the array , and any thermal or night sky line emission .
the beam images were combined together separately .
the residuals of the cosmic ray hit events on these combined images were removed by eye using the iraf task imedit , and then the spectrum was extracted using the iraf task apall .
separate dispersion solutions were determined for the two beams before combining them into one spectrum .
the same procedures were done for hot star spectra and then the hot star spectra were divided into the target spectra to correct for telluric absorption lines and for fringing in the spectra .
the target spectra were then continuum fitted and normalized .
in figure 2 , we show spectra of arcturus and our targets in palomar 6 for the @xmath19co 2 - 0 bandhead centered at @xmath6 @xmath10 22945 . in order to derive the heliocentric radial velocities ,
we cross - correlated our object spectra with that of arcturus ( @xmath20 = @xmath25.2 km sec@xmath1 , evans 1967 ) .
table 3 shows our heliocentric radial velocity measurements for our targets .
we also list those of minniti ( 1995b ) . in general
, our radial velocity measurements do not agree with those of minniti ( 1995b ) .
we believe that minniti ( 1995b ) misidentified star f as a cluster member since the radial velocity of this star is not consistent with cluster membership .
an independent check of the radial velocity for palomar 6 remains desirable .
however , in this paper we consider our velocity measurements are correct and stars a , c , d , and g are cluster members .
our mean heliocentric velocity of palomar 6 is + 180.6 @xmath0 3.2 km sec@xmath1 and is 20 km sec@xmath1 smaller than that of minniti ( 1995b ) .
( the error is that of the mean and is affected by both our measurement uncertainties and by the cluster s internal velocity dispersion . )
laboratory measurements of the oscillator strengths for the lines in our spectral regions do not exist .
therefore , we derived the oscillator strengths following the procedure described by carr et al .
( 2000 ) , which exploits the known abundances and stellar atmospheric parameters for both the sun and for arcturus .
we used kurucz s solar atmosphere model with @xmath21 = 1.0 km s@xmath1 and adopted @xmath22 = 4300 k , @xmath23 = 1.5 , and @xmath21 = 1.7 km s@xmath1 for arcturus ( peterson et al .
1993 ) .
we first measured the equivalent widths of the absorption lines from the solar intensity spectrum ( wallace et al .
1996 ) and then derived the corresponding oscillator strengths for each line using the solar model atmosphere ( kurucz 1994 ) and moog ( sneden 1973 ) . for collisional
damping , we considered only the van der waals force , and we adopted the unsld approximation .
it is suspected that the collisional damping constant calculated using the unsld approximation is probably too small ( e.g. steffen 1985 ) , so we artificially increased the collisional damping constants by matching the synthetic spectrum to the solar intensity spectrum of wallace et al .
( 1996 ) . to calculate the synthetic solar spectrum
, we use the macroturbulent velocity of 3.0 km s@xmath1 for the sun ( gray 1977 ) and the solar elemental abundances of anders & greevesse ( 1989 ) . the synthetic spectra with no enhancement in the collisional damping constant fail to match the line core depth and the line wing profile in particular for iron lines .
thus we enhanced the collisional damping constant by a factor of 3.5 for the iron lines and by a factor of 2 for the silicon and titanium lines .
we show the synthetic solar spectra in figure 3 and our solar oscillator strengths in table 4 .
note that our fe i oscillator strengths are slightly different from those of carr et al .
( 2000 ) , who adopted different collisional damping enhancement factors .
their damping constants were enhanced by a factor of 3 for fe i @xmath10 22386.90 and by a factor of 2 for fe i @xmath10 22399.16 .
our solar oscillator strength for ti i @xmath10 21788.63 is rather uncertain , since it is on the wing profile of the nearby si i @xmath10 21785.60 , which has a strong wing profile in the solar spectrum . in the figure , it is evident that the solar intensity spectrum shows subtle line asymmetries , in particular for iron lines , probably caused by the doppler shift due to the convective and oscillatory motions in the solar atmosphere , which can not be reproduced by the plane - parallel one dimensional model atmosphere ( asplund 2003 ) .
a comparison with an arcturus spectrum provides a crucial opportunity to examine the reliability of our solar oscillator strengths , since the chemical composition and the stellar parameters of arcturus are closer to those of our target rgb stars than those of the sun and a high resolution , high s / n ratio spectrum is also available for arcturus ( hinkle et al .
1995 ) . using the solar oscillator strengths and our equivalent width measurements for arcturus , we obtained [ fe / h ] = @xmath20.48 @xmath0 0.07 , [ si / fe ] = @xmath240.24 @xmath0 0.09 , and [ ti / fe ] = + 0.61 for arcturus .
( the errors are those of the mean . )
our iron and silicon abundances are in good agreement with those of sneden et al .
( 1994 ) , who obtained [ fe / h ] = @xmath20.47 @xmath0 0.04 and [ si / fe ] = + 0.14 @xmath0 0.07 , = @xmath20.5 @xmath0 0.1 , [ si / fe ] = + 0.4 , [ ti / fe ] = + 0.3 , @xmath22 = 4300 @xmath0 30 k , @xmath23 = 1.5 @xmath0 0.15 , and @xmath21 = 1.7 @xmath0 0.3 km sec@xmath1 for arcturus . see table 5 . ] while our titanium abundance is very high compared to that of sneden et al .
( 1994 ) , [ ti / fe ] = + 0.27 @xmath0 0.04 .
figure 4 shows a comparison between the observed arcturus spectrum ( hinkle et al .
1995 ) and the synthetic spectrum using our solar oscillator strengths and the elemental abundance ratios of sneden et al .
the figure clearly shows that the solar oscillator strength for ti i 21788.63 does not work for arcturus .
the equivalent width for the ti i 21788.63 line in the synthetic spectrum is 296 m and it is about 66 m smaller than that we measured from the arcturus spectrum of hinkle et al .
we examined the atomic spectral lines of kurucz & bell ( 1995 ) and found no strong lines near ti i 21788.63 .
we then generated the synthetic spectrum using the wavelengths and the oscillator strengths given by kurucz & bell ( 1995 ) and measured the equivalent width produced by atomic spectral lines around ti i 21788.63 .
we found that ti i 21788.25 and mn i
21788.49 lines contribute no more than 5 m to our equivalent measurement of ti i 21788.63 .
we also generated a synthetic solar spectrum using the same line list except for ti i 21788.25 and mn i 21788.49 lines near ti i 21788.63 and measured the equivalent width of 0.8 mwhich is too small to cause such a large discrepancy in the titanium abundance of arcturus .
in the synthetic solar spectrum where the spectral resolution is much higher than that of arcturus spectrum , the line cores of ti i 21788.63 and the combination of ti i 21788.25 and mn
i 21788.49 are well separated .
therefore it is not likely that our equivalent width measurement of ti i 21788.63 from the observed solar spectrum is contaminated by the previously known nearby atomic spectral lines .
we also examined cn molecule absorption lines ( wallace & hinkle 1996 ) and found no known cn spectral lines at @xmath10 21788.63 .
this leads us to suspect that there may exist unidentified atomic spectral lines very close to ti i 21788.63 . also importantly ,
the excitation potential energy of ti i 21788.63 is 1.749 ev and this line forms higher in the atmosphere where non - local thermodynamic equilibrium ( nlte ) effects may be considered . on the other hand , fe i and si i lines of our interest form deep in the atmosphere due to their high excitation potential energies and nlte effects to these lines
are less severe .
we examined the solar @xmath25 value for the ti i 21788.63 line using the giant stars i-45 and i-46 in m71 .
sneden et al .
( 1994 ) studied the same stars in the optical employing high resolution echelle spectroscopy and they obtained [ fe / h ] = @xmath20.78 @xmath0 0.04 , [ ti / fe ] = + 0.56 @xmath0 0.06 for i-45 and [ fe / h ] = @xmath20.79 @xmath0 0.06,[ti / fe ] = + 0.47 @xmath0 0.07 for i-46 . using the solar oscillator strength values and the stellar parameters of sneden et al .
( 1994 ) , we obtained [ fe / h ] = @xmath20.9 @xmath0 0.1 , [ ti / fe ] = + 1.0 for i-45 and [ fe / h ] = @xmath20.8 @xmath0 0.1 , [ ti / fe ] = + 0.8 for i-46 .
the discrepancy in the titanium abundance is large again . on the other hand ,
our silicon abundance measurement for the star i-45 is in good agreement with that of sneden et al .
our silicon abundance of the star i-46 is [ si / fe ] = @xmath20.1 and is about 0.4 dex lower than that of the star i-45 .
sneden ( private communication ) kindly noted that the silicon spectral lines for the star i-46 are too weak to measure reliably in his spectra .
if the discrepancy in the titanium abundance is caused by nlte effects , then using the oscillator strengths derived from the observed arcturus spectrum of hinkle et al .
( 1995 ) would mitigate the problem since arcturus is a metal - rich rgb star and the atmospheric structure of arcturus is more similar to our program stars than that of the sun is .
as mentioned above , ti i 21788.63 line has a low excitational potential energy and it suffers from more severe nlte effects than fe i and si i lines , which have high excitational potential energies ( ruland et al . 1980 ;
brown et al . 1983 ; hauschildt et al .
1997 ) .
we examine the nlte effects on our titanium abundance using one dwarf ( 61 uma ) and three giants ( @xmath3 sge , @xmath26 cep , and @xmath3 ori ) with solar metllicity . to avoid the possible contamination by unidentified atomic spectral lines very close to ti i 21788.63 , we chose three more ti i lines free from cn blending at @xmath10 21903.31 , 22216.74 , 22239.03 and we derived oscillator strengths using the solar spectrum and the arcturus spectrum .
for the arcturus ti i oscillator strengths , we adopt @xmath27(ti ) = 4.79 ( sneden et al .
1994 ) and our ti i oscillator strengths are shown in table 6 .
wallace & hinkle ( 1996 ) provided high - resolution ir spectra ( @xmath13 @xmath28 45,000 ) of 61 uma , @xmath3 sge , @xmath26 cep , and @xmath3 ori and the chemical compositions and stellar parameters of these four stars are previously known in optical .
table 7 shows the basic stellar parameters and iron , titanium abundances of 61 uma , @xmath3 sge , @xmath26 cep , and @xmath3 ori ( luck 1977 , cornide & rego 1984 , lambert et al .
1984 , mcwilliam 1990 , carr et al .
we then measured the equivalent widths for fe i 22386.90 , 22399.16 and ti i 21788.63 , 21903.31 , 22216.74 , 22239.03 and obtained @xmath27(fe ) and @xmath27(ti ) for the stars .
remind that fe i 22386.90 , 22399.16 and ti i 21788.63 lines are of interest in our study .
our iron abundance measurements using the solar oscillator strengths are in excellent agreement with previous measurements in optical within @xmath29(fe ) @xmath30 0.05 dex . for the titanium abundance measurements ,
the situation is somewhat different .
the titanium abundance for 61 uma is not known , but we assume that this thin disk , solar - type star has [ ti / fe ] = 0.00 , from which we deduce log @xmath31(ti ) = 4.99 . as table 7 shows , using @xmath25 values for the different ti i lines derived from solar spectrum leads to consistent titanium abundances for 61 uma , whose surface gravity is similar to the sun .
but the solar ti i @xmath25 values do not appear to yield correct abundances for the three giant stars .
table 7 shows that in those cases , the @xmath25 values derived from the arcturus spectrum provide a better match .
this may reflect unknown problems in either underlying and unidentified absorption lines or , since the behavior appears to affect all the ti i lines , problems in the adopted @xmath32 relations for giants compared to dwarfs . for our titanium abundance determination in palomar 6 , in which we rely on giant stars ,
we therefore adopt @xmath25 values derived from the arcturus spectrum .
table 8 summarizes our adopted @xmath25 values .
having good stellar parameters , such as the effective temperature and the surface gravity , is very important in any stellar abundance study , since the absolute or the relative elemental abundance scales will depend on the input stellar parameters . in optical spectroscopy ,
the effective temperature is usually determined by using weak neutral iron lines with log(w@xmath33/@xmath10 ) @xmath34 @xmath35 , lying on the linear part of curve of growth ( cog ) . since only the two very strong neutral iron lines with similar excitation potentials are available in our spectral regions , we have to rely upon photometric temperature estimates .
however , applying the photometric method using relations between color versus @xmath22 for palomar 6 still poses a problem , since the interstellar reddening value for the cluster is not well known . in the case of using the @xmath36 color index to estimate the stellar temperature , an uncertainty of 0.1 mag in @xmath15 will result in an uncertainty of @xmath6 100 k in the photometric surface temperature in the stellar temperature range of our interest .
the discrepancy between previous interstellar reddening estimates for palomar 6 is as large as @xmath37 @xmath6 0.15 mag ( lee & carney 2002a ) .
we , therefore , devise an alternative photometric temperature determination method , which is independent of interstellar reddening and the photometric zero points , using the @xmath14 magnitude differences between the horizontal branch ( hb ) stars and the rgb stars . in figure 5
, we show a plot of @xmath22 versus @xmath38 for the rgb sequences with z = 0.010 , 0.004 ( equivalently [ fe / h ] = @xmath20.5 , @xmath20.9 ) and the zero - age horizontal branch ( zahb ; kim et al . 2002 ; yi , demarque , & kim 1997 ) . the differences in @xmath38 ( i.e. , the differences in bolometric magnitudes ) between hb and rgb stars
are linearly related to the surface temperature of the rgb stars , with a slight dependence on metallicity . since the bolometric correction for rgb stars in the @xmath14 passband , @xmath39 , is linearly correlated with the effective temperature ( see , for example , bessell , castelli , & plez 1998 ) , the difference in the @xmath14 magnitude is also linearly correlated with the difference in luminosity , @xmath38 .
therefore , we can estimate the temperature of our target rgb stars by simply measuring the difference in @xmath14 magnitude between the hb and the target rgb stars , @xmath40 .
since hb stars are not located horizontally in ir cmds , it is necessary to adopt the @xmath14 magnitude at the intersection between rgb and red hb ( rhb ) in ir photometry for our hb reference level , @xmath41 ( for example , see kuchinski et al .
1995 ) . in practice , instead of using the theoretical values , we relied upon the @xmath22 versus @xmath42 diagram for rgb stars in well - studied globular clusters with similar metallicities as palomar 6 .
we show @xmath43 cmds of the metal - rich globular clusters 47 tuc ( [ fe / h ] = @xmath20.7 , frogel , persson , & cohen 1981 ) , m69 ( [ fe / h ] = @xmath20.7 , frogel , persson , & cohen 1983 ; davidge & simons 1991 ) , m71 ( [ fe / h ] = @xmath20.7 , frogel , persson , & cohen 1979),and m107 ( [ fe / h ] = @xmath21.0 , frogel , persson , & cohen 1983 ) in figure 6 .
we adopt @xmath41 magnitudes from kuchinski et al .
( 1995 ) for these template clusters and we use @xmath41 = 13.53 mag for palomar 6 ( see lee & carney 2002a ) . since the interstellar reddening values are well known for template clusters , we can derive the effective temperature for the rgb stars in template clusters using the relation given by alonso , arribas , & martinez - roger ( 1999 ) . in figure 7
, we show a plot of @xmath42 versus @xmath22 for the rgb stars in the template clusters and we obtain a linear relation between @xmath22 and @xmath42 , @xmath44 using this relation and @xmath42 values for our target rgb stars in palomar 6 listed in table 1 , we obtain @xmath22 = 3718 , 3741 , and 3667 k for stars a , d , and g , respectively .
the effective temperature estimates for palomar 6 a , d , and g using the @xmath45 versus @xmath22 relation given by alonso et al .
( 1999 ) are 3703 , 3747 , and 3597 k for @xmath15 = 1.3 mag ( lee & carney 2002a ) .
these values agree with those from the @xmath42 versus @xmath22 relation within @xmath46 @xmath6 100 k. as a cross - check , we also obtained the surface temperature for m71 rgb stars using equation ( 1 ) and we have @xmath22 = 4043 k for i-45 and @xmath22 = 4019 k for i-46 .
our surface temperatures for m71 rgb stars are in good agreement with those of sneden et al .
( 1994 ) , who obtained temperatures of 4050 k and 4000 k , respectively . to estimate the surface gravity , we rely on the model isochrones of yi , demarque , & kim ( 1997 ) and kim et al .
first we calculated the bolometric correction in the @xmath14 passband , @xmath39 , for the temperatures and surface gravities of each grid of model isochrones by interpolating the ( @xmath22 , @xmath23 , @xmath39 ) grids of bessell , castelli , & plez ( 1998 ) .
then we calculated the bolometric magnitudes ( @xmath47 ) using @xmath48 and obtained the magnitudes in @xmath14 passband ( @xmath49 ) by subtracting @xmath39 from @xmath47 for each grid of the model isochrones , assuming @xmath50 this procedure allows us to calculate @xmath42 for the model isochrones . in figure 8 , we show a plot of @xmath42 versus @xmath23 for the model isochrones with [ fe / h ] = @xmath20.9 , @xmath20.5 and [ @xmath3/fe ] = + 0.3 for 10 , 14 gyr . in the figure
, we also show the locations of rgb stars in m71 and our program stars in palomar 6 . for m71 rgb stars
, we obtained @xmath23 = 1.0 for i-45 and i-46 .
our surface gravities for m71 rgb stars are @xmath51 = 0.2 dex larger than those of sneden et al .
for palomar 6 , we obtained @xmath23 = 0.3 for palomar 6 a and d , and @xmath23 = 0.2 for palomar 6 g assuming [ fe / h ] @xmath6 @xmath21.0 for palomar 6 ( see lee & carney 2002a ) .
note that our surface gravity estimates using the model isochrones should be considered as the upper limit of surface gravities of our program stars , since the mass loss has not been included in model isochrone calculations ( kim , y. -c .
, private communication ) . since our program stars in palomar 6 lie near the tip of rgb , the mass loss rate will be large , leading to lower surface gravities . to estimate the microturbulent velocity @xmath21 requires many spectral lines , preferably fe i lines in optical spectroscopy , both on the linear and the flat parts of the cog .
since we have only two strong fe i ( or two si i ) lines and we are not able to properly estimate the microturbulent velocities for our program rgb stars . therefore we adopt the microturbulent velocity of 2 km sec@xmath1 for our program rgb stars without any verification . in figure 9 , we show comparisons of @xmath23 versus @xmath22 versus @xmath21 of rgb stars in globular clusters extensively studied in optical ( brown , wallerstein , & oke 1990 ; ivans et al . 1999 , 2001 ; norris & da costa 1995 ; sneden et al . 1994 , 2004 ) .
nine stars in these six clusters have @xmath23 @xmath30 0.3 , comparable to our program stars in palomar 6 .
those nine stars have a mean @xmath21 of 1.9 @xmath0 0.2 km sec@xmath1 .
if we include stars with @xmath23 @xmath30 0.5 , there are 23 stars and they have a mean @xmath21 of 2.0 @xmath0 0.3 km sec@xmath1 ( the error is the standard deviation ) . since these stars are very similar to our program stars in palomar 6 , our choice of @xmath21 = 2.0 km sec@xmath1 is thought to be reasonable . it should be noted that , however , the microturbulent velocity appears to increase with decreasing surface gravity in m supergiants ( see , for example , figure 4 of ramrez et al . 2000 ) . with temperature and surface gravity estimates , 72-depth plane - parallel local thermodynamic equilibrium ( lte ) model atmospheres
were computed using the program atlas9 , written and supplied by dr .
r. l. kurucz .
the model atmospheres were computed using opacity distribution functions and abundances with enhanced abundances of all the @xmath3 " elements ( o , ne , mg , si , s , ar , ca , and ti ) by 0.4 dex , assuming our target rgb stars would prove to be @xmath3-element enhanced ( see discussions below ) . the abundance analysis was performed using the program moog ( sneden 1973 ) .
since our abundance analysis is based upon the neutral lines , the elemental abundances are not sensitive to the surface gravity . on the other hand , since we rely on very strong lines , our results will depend on the choice of microturbulent velocity of the model atmosphere , which we are not able to estimate using our target spectra .
we show our equivalent width measurements in table 9 and the elemental abundances for palomar 6 and m71 in table 10 .
we also show the elemental abundances of sneden et al .
( 1994 ) in the table . in table 11 , we show estimated errors resulting from the uncertainties in the input model atmosphere @xmath52 = + 200 k , @xmath53 = @xmath20.2 , and @xmath54 = + 0.5 km sec@xmath1 , which are appropriate for our analysis . in general ,
[ fe / h ] and [ si / fe ] ratios are not sensitive to the changes in the input parameters of model atmospheres .
however , [ ti / fe ] is very sensitive to the surface temperature and the microturbulent velocity .
an uncertainty of @xmath52 = + 200 k leads to an uncertainty of @xmath6 @xmath240.3 dex and an uncertainty of @xmath54 = + 0.5 km sec@xmath1 leads to @xmath6 @xmath20.3 dex in the [ ti / fe ] ratio .
table 10 shows that our m71 abundances are consistent with those obtained by sneden et al .
, given the uncertainties and the limited number of lines available for our analyses .
the mean metallicity for palomar 6 is [ fe / h ] = @xmath21.0 @xmath0 0.1 , consistent with that derived from the slope of the rgb in the @xmath43 cmd , [ fe / h ] @xmath6 @xmath21.1 ( lee & carney 2002a ) , and is slightly more metal - poor than m71 , [ fe / h ] = @xmath20.8 @xmath0 0.1 , unless our surface temperatures or the microturbulent velocities for the palomar 6 rgb stars are inaccurate by more than @xmath55 400 k or @xmath56 @xmath20.7 km sec@xmath1 .
we show the synthetic spectra for [ fe / h ] = @xmath21.0 , @xmath20.5 , 0.0 , and the observed spectra of the palomar 6 rgb stars in figure 10 . in the figure we used an ad hoc @xmath25 value for sc
i @xmath10 22400.70 since the hyperfine splitting components for the line are not known .
the figure clearly indicates that the observed fe i absorption lines are too weak to lead to the super- or near - solar metallicity as claimed by minniti ( 1995b ) or bica et al .
( 1998 ) . figure 11 shows synthetic spectra for the star palomar 6 g using the model atmospheres with [ fe / h ] = + 0.20 , which is the metallicity of palomar 6 claimed by minniti ( 1995 ) . as can be seen in the figure , synthetic spectra
do not reproduce the observed spectrum of the star palomar 6 g until @xmath22 @xmath57 5000 k , which is an unreasonably high surface temperature for the super - metal - rich rgb star ( see also figure 6 ) . therefore , it is most likely that palomar 6 is much more metal - poor than claimed by minniti ( 1995b ) or bica et al .
( 1998 ) .
our [ si / fe ] ratio for the star i-45 shows good agreement with the results of sneden et al .
( 1994 ) , suggesting that our measurement for the silicon abundance is reliable .
it appears that the silicon abundance of the star palomar 6 g is enhanced by @xmath6 0.4 dex .
the titanium abundance is enhanced by @xmath6 0.5 dex relative to the iron abundance .
figure 12 shows synthetic spectra of the star palomar 6 g with [ si / fe ] = + 0.4 and [ ti / fe ] = + 0.5 . in their figure 9 ,
lee & carney ( 2002a ) showed the palomar 6 @xmath58 cmd drooping at the rgb - tip , which led ortolani et al .
( 1995 ) to conclude that palomar 6 has [ fe / h ] @xmath6 @xmath20.4 .
the droop seen at the tip of the rgb branch in the cmd of palomar 6 could be due to its high titanium abundance which increases the line blanketing effect by tio bands and may not due to the overall high metallicity .
in particular , since the star palomar 6 g has a low effective temperature , this effect would be large .
it should be emphasized again that our [ ti / fe ] ratio measurement is very sensitive to @xmath22 and @xmath21 of the input model atmosphere and our derived titanium abundance for palomar 6 relies on the single line measurement in a single star . as shown in table 11
, an uncertainty of @xmath46 = 200 k leads to the uncertainty of almost 0.3 dex in the [ ti / fe ] ratio .
our titanium abundance of palomar 6 relied upon a very strong single line measurement and the microturbulent velocity was not measured in an appropriate way . a change of 0.5 km sec@xmath1 in the microturbulent velocity results in @xmath6 @xmath20.3 dex change in the [ ti / fe ] ratio .
globular clusters do not define a single homogeneous population with a single history and , in particular , the galactic central regions may contain globular clusters belonging to various populations .
the globular clusters near the galactic center may provide the means by which we can improve our understanding of the processes and chronology involved in the formation of the galactic halo and the bulge .
carney , latham , & laird ( 1990 ) and wyse & gilmore ( 1992 ) argued that the halo is the ancestor to the bulge .
the idea is that the halo is metal - poor because star formation terminated early by the loss of gas from the halo , and the low angular momentum of the halo means the gas should have largely collected dissipatively into the central regions ( i.e. , into the bulge ) .
however , the definition of the bulge clusters currently being used is rather ambiguous mainly due to lack of complete kinematical information .
minniti ( 1995a ) drew attention to a set of clusters that may be associated with the galaxy s bulge , distinct from the central concentration of the spheroidal halo clusters . in practice , the metal - rich clusters located within a few kpc from the galactic center are usually categorized as the bulge population , but halo clusters formed in the metal - rich tail may also exist in the galactic central regions . in a recent study ,
dinescu et al . ( 2003 ) discussed the population status of seven globular clusters with @xmath59 @xmath34 3.2 kpc based on their actual space velocities .
one of their intriguing findings is that the metal - rich globular cluster ngc 6316 is most likely a halo cluster , in spite of its high metallicity , [ fe / h ] = @xmath20.55 .
unfortunately , palomar 6 does not as yet have a measurement of its proper motion . to assign it to the disk , bulge , or halo populations , we will have to rely only upon its metallicity , position , and radial velocity . in figure 13 ,
we show the metallicity distribution and the kinematical properties of globular clusters within 3 kpc from the galactic center using data available from harris ( 1996 ) . in addition
, we adopt the recent metallicity measurements by origlia , rich , & castro ( 2002 ) for liller 1 , carretta et al .
( 2001 ) for ngc 6528 , and cohen et al .
( 1999 ) for ngc 6553 .
the metallicity distribution shows two peaks at [ fe / h ] @xmath6 @xmath21.3 and @xmath20.4 , suggesting that at least two distinct populations exist in the galactic central regions .
the bottom panel shows a plot of @xmath60 versus @xmath61 for the metal - rich and metal - poor clusters using the relations given by frenk & white ( 1980 ) , corrected by zinn ( 1985 ) , where @xmath60 is the radial velocity observed at the sun s position by an observer at rest with respect to the galactic center and @xmath62 is the angle between the line of sight and the direction of galactic rotation at the cluster ( zinn 1993 ) .
the @xmath61 and @xmath60 values are listed in table 13 . in the figure , the mean rotation velocity , @xmath63 ,
is given by the slope of the straight line , and the line - of - sight velocity dispersion , @xmath64 , is given by the standard deviation of the points about the line .
the fit for the metal - poor clusters ( [ fe / h ] @xmath65 @xmath21.0 ) is represented by the solid line and that for the metal - rich clusters ( [ fe / h ] @xmath66 @xmath21.0 ) by the dashed line .
the metal - poor clusters have @xmath63 = + 111 @xmath0 57 km sec@xmath1 and @xmath64 = 137 @xmath0 31 km sec@xmath1 , while the metal - rich clusters have @xmath63 = + 89 @xmath0 36 km sec@xmath1 and @xmath64 = 99 @xmath0 23 km sec@xmath1 .
the rotational velocity solution of the metal - rich clusters with the lower mass bar - like kinematics is also shown in the figure by the dotted line ( burkert & smith 1997 ) .
there is no distinctive difference in mean kinematics between the two metallicity regimes , but the metallicity distribution shows that if there are two different groups based on different mean metallicities , the overlap in [ fe / h ] is considerable .
we can , nonetheless , still employ kinematics as a discriminant since more disk - like or bulge - like clusters will , on average be closer to the mean rotational solutions than would halo clusters .
the observed radial velocity of palomar 6 deviates by over 150 km sec@xmath1 from the rotational velocity solutions , suggesting that palomar 6 may have halo kinematics .
we believe that palomar 6 is probably a globular cluster that is more probably classified as belonging to the halo population rather than a bulge or disk population , based on both its kinematics and its relatively low metallicity . like other halo clusters
, its high [ @xmath3/fe ] ratio suggests that the star formation rate was very high when the cluster formed .
additional spectroscopic studies of high and low metallicity and high and low velocity clusters in the inner galaxy will be needed to unravel the complex histories of star formation .
as we discussed above , the measurement of @xmath3-element abundances of globular clusters provides important information regarding relative ages , in the sense that the enhanced value of [ @xmath3/fe ] ratios indicates the high star formation rate and old age of the system . in figure 14
, we show the silicon and the titanium abundances of globular clusters as a function of metallicity ( see lee & carney 2002b and references therein ) . in the figure
, our [ ti / fe ] ratio of palomar 6 appears to be 0.2 dex higher than the mean value of [ ti / fe ] of other globular clusters , [ ti / fe ] @xmath6 0.3 dex .
however , given the uncertainty of palomar 6 [ ti / fe ] value , the significance of this difference is hard to access without further data .
the enhanced @xmath3-element abundances of palomar 6 suggest that palomar 6 must have formed very quickly , in spite of its high metallicity .
if true , the galactic inner halo must have experienced a very rapid chemical enrichment history .
figure 15 shows [ si / ti ] ratio versus galactocentric distance , @xmath59 , of globular clusters .
the bisector linear fit to the old halo clusters ( 18 clusters ) and that to the old halo clusters with @xmath59 @xmath30 8 kpc ( 12 clusters ) are also shown in the figure . as we mentioned above , lee & carney ( 2002b ) claimed that [ si / ti ] ratios of the old halo globular clusters are anticorrelated with galactocentric distances , in the sense that the silicon abundance increases and the titanium abundance decreases with galactocentric distances .
they argued that the contributions from sne ii events with different progenitor masses could explain such a gradient . in the figure
, the [ si / ti ] ratio of palomar 6 does not appear to follow the trend claimed by lee & carney ( 2002b ) and , further , the [ si / ti ] ratio of palomar 6 appears to be lower than those of metal - rich bulge clusters ngc 6528 ( [ fe / h ] = + 0.07 , carretta et al .
2001 ) and ngc 6553 ( [ fe / h ] = @xmath20.16 , cohen et al .
the [ si / ti ] ratio of palomar 6 appears to agree with those of the metal - rich bulge k giants by mcwilliam & rich ( 1994 ) .
this may suggest that the chemical enrichment history of palomar 6 was more complex than those of other clusters in the galactic central regions .
again , given the uncertainty in the palomar 6 [ si / ti ] value , the significance of this difference is hard to assess without further data .
in this paper , we have discussed high - resolution infrared echelle spectroscopy for the globular clusters palomar 6 and m71 .
our mean heliocentric velocity of palomar 6 is + 180.6 @xmath0 3.2 km sec@xmath1 and is 20 km sec@xmath1 smaller than that of minniti ( 1995b ) .
contrary to the recent results using low resolution spectroscopy , our results have shown that palomar 6 has an intermediate metallicity with [ fe / h ] = @xmath21.0 @xmath0 0.1 and is slightly more metal - poor than m71 .
our estimated uncertainties in the surface temperatures , gravities , and microturbulent velocities for the three stars we have studied will not affect our derived [ fe / h ] value by more than 0.2 dex .
we believe that previous estimates , based on lower resolution , lower - s / n spectra were affected by unrecognized molecular line blanketing or by stars that are not members of the cluster .
the [ @xmath3/fe ] ratio of palomar 6 appears to be enhanced by @xmath6 0.3 0.5 dex in spite of its high metallicity , suggesting that the galactic inner halo must have experienced a very rapid chemical enrichment history .
jwl wishes to thank j. s. carr , c. sneden and y. -c .
kim for their kind discussions .
we thank an anonymous referee for useful comments and a careful review of the paper .
this research was supported by the national aeronautics and space administration ( nasa ) grant number go-07318.04 - 96a from the space telescope science institute , which is operated by the association of universities for research in astronomy ( aura ) , inc .
, under nasa contract nas 5 - 26555 .
we also thank the national science foundation for financial support via grants ast@xmath29619381 , ast@xmath29888156 and ast@xmath2030541 to the university of north carolina .
support for this work was also provided by the korea science and engineering foundation ( kosef ) to the astrophysical research center for the structure and evolution of the cosmos ( arcsec ) .
sbc is pleased to acknowledge nsf grants 98 - 19870 and 00 - 98619 to the university of maryland .
crcccc a & 8.733 & 1.543 & 0.262 & 16.94 & 4.797 + b & 8.541 & 1.482 & 0.269 & 16.61 & 4.989 + c & 9.770 & 1.456 & 0.262 & 17.11 & 3.760 + d & 8.855 & 1.551 & 0.284 & 16.89 & 4.675 + e & 9.632 & 1.462 & 0.249 & 17.18 & 3.898 + f & 8.267 & 1.638 & 0.322 & 17.22 & 5.263 + g & 8.467 & 1.646 & 0.315 & 17.27 & 5.063 + clccccc pal 6 & a & 22395 & 24 may 1998 & 8400 & 45 + & d & 22395 & 25 may 1998 & 4200 & 45 + & g & 22395 & 24 may 1998 & 5400 & 60 + & g & 21815 & 25 may 1998 & 4200 & 70 + & & & & & + m71 & i-45 & 22395 & 14 july 1997 & 2100 & 40 + & i-45 & 21815 & 14 july 1997 & 2400 & 50 + & i-46 & 22395 & 13 july 1997 & 2700 & 55 + & i-46 & 21815 & 13 july 1997 & 2400 & 50 + ccccc a & 185.3 & 4.0 & & yes + b & 26.7 & 1.7 & 56.4 & no + c & 173.5 & 3.6 & 243.3 & yes + d & 186.8 & 2.3 & 201.2 & yes + e & @xmath213.5 & 2.1 & & no + f & 134.5 & 3.3 & 200.0 & no + g & 176.7 & 2.2 & & yes + & & & + mean & 180.6 & & 200.6 & + ccccrrcc 22386.90 & fe i & 5.033 & & 229 & @xmath20.490 & & @xmath20.481 + 22399.16 & fe i & 5.099 & & 72 & @xmath21.275 & & @xmath21.249 + 21785.60 & si i & 6.719 & & 400 & 0.295 & & + 21825.63 & si i & 6.721 & & 293 & 0.030 & & + 21788.63 & ti i & 1.749 & & 33 & @xmath21.265 & & + ccccrcccr 21788.63 & 1.749 & & 33 & @xmath21.265 & & 361 & 5.17 & @xmath20.884 + 21903.31 & 1.739 & & 17 & @xmath21.600 & & 316 & 5.18 & @xmath21.210 + 22216.74 & 1.734 & & 10 & @xmath21.850 & & 237 & 4.91 & @xmath21.730 + 22239.03 & 1.739 & & 11 & @xmath21.810 & & 264 & 5.05 & @xmath21.555 + 22386.90 & fe i & 267 & 7.52 & & & 257 & 7.29 & & & 503 & 7.79 & & & 577 & 7.54 & + 22399.16 & fe i & 105 & 7.56 & & & 126 & 7.50 & & & 320 & 7.76 & & & 352 & 7.42 & + 21788.63 & ti i & 66 & 4.97 & 4.59 & & 122 & 5.19 & 4.81 & & 585 & 5.65 & 5.27 & & 919 & 5.42 & 5.04 + 21903.31 & ti i & 41 & 5.02 & 4.63 & & & & & & 518 & 5.71 & 5.32 & & & & + 22216.74 & ti i & & & & & & & & & 398 & 5.50 & 5.38 & & 769 & 5.31 & 5.19 + 22239.03 & ti i & 32 & 5.10 & 4.84 & & & & & & 433 & 5.60 & 5.34 & & 792 & 5.37 & 5.12 + + & @xmath67 fe i @xmath68 & & 7.54 & & & & 7.40 & & & & 7.78 & & & & 7.48 & + & @xmath67 ti i @xmath68 & & 5.03 & 4.68 & & & 5.19 & 4.81 & & & 5.62 & 5.33 & & & 5.37 & 5.12 + ccccl 22386.90 & fe i & 5.033 & @xmath20.490 & solar + 22399.16 & fe i & 5.099 & @xmath21.275 & solar + 21785.60 & si i & 6.719 & 0.295 & solar + 21825.63 & si i & 6.721 & 0.030 & solar + 21788.63 & ti i & 1.749 & @xmath20.884 & arcturus + ccrrrrrr 22386.90&fe i & 299 & 259 & 296 & & 251 & 284 + 22399.16&fe i & 130 & 142 & 120 & & 152 & 160 + 21785.60&si i & & & 226 & & 251 & 174 + 21825.63&si i & & & 194 & & 238 & 174 + 21788.63&ti i & & & 545 & & 473 & 467 + cccccccccc @xmath69 & 3718 & 3741 & 3667 & & 4043 & 4019 & & 4050 & 4000 + @xmath23 & 0.3 & 0.3 & 0.2 & & 1.0 & 1.0 & & 0.8 & 0.8 + @xmath21 & 2.00 & 2.00 & 2.00 & & 2.00 & 2.15 & & 2.00 & 2.15 + [ fe / h ] & @xmath21.0 @xmath0 0.1 & @xmath21.0 @xmath0 0.1 & @xmath21.0 @xmath0 0.1 & & @xmath20.9 @xmath0 0.1 & @xmath20.8 @xmath0
0.1&&@xmath20.78 @xmath0 0.04&@xmath20.79 @xmath0 0.06 + [ si / fe]&&&+0.4 @xmath0 0.0 & & + 0.3 @xmath0 0.1&@xmath20.1 @xmath0 0.1&&@xmath240.38 @xmath0 0.06 & + [ ti / fe ] & & & + 0.5 & & + 0.6 & + 0.4 & & + 0.56 @xmath0 0.06 & + 0.47 @xmath0 0.07 + lcccr terzan 3 & @xmath20.73 & 2.4 & @xmath20.975 & @xmath2184.92 + ngc 6235 & @xmath21.40 & 2.9 & @xmath20.085 & 94.77 + ngc 6256 & @xmath20.70 & 2.1 & @xmath20.808 & @xmath2140.45 + ngc 6266 ( m62 ) & @xmath21.29 & 1.7 & @xmath20.624 & @xmath284.90 + ngc 6273 ( m19 ) & @xmath21.68 & 1.6 & @xmath20.584 & 133.75 + ngc 6287 & @xmath22.01 & 1.7 & @xmath240.052 & @xmath2275.91 + ngc 6293 & @xmath21.99 & 1.4 & @xmath20.413 & @xmath2150.35 + ngc 6304 & @xmath20.59 & 2.1 & @xmath20.291 & @xmath2113.33 + ngc 6325 & @xmath21.17 & 2.0 & @xmath240.089 & 45.09 + ngc 6333 ( m9 ) & @xmath21.72 & 1.7 & @xmath240.975 & 262.87 + ngc 6342 & @xmath20.65 & 1.7 & @xmath240.792 & 147.42 + ngc 6355 & @xmath21.50 & 1.0 & @xmath20.070 & @xmath2167.61 + terzan 2 & @xmath20.40 & 0.9 & @xmath20.586 & 104.59 + terzan 4 & @xmath21.60 & 1.4 & @xmath20.445 & @xmath255.27 + hp 1 & @xmath21.50 & 0.8 & @xmath20.516 & 53.20 + liller 1 & @xmath20.30 & 2.6 & @xmath20.273 & 41.21 + terzan 1 & @xmath20.35 & 1.8 & @xmath20.186 & 35.57 + ton 2 & @xmath20.50 & 1.4 & @xmath20.988 & @xmath2212.01 + ngc 6401 & @xmath21.12 & 0.8 & @xmath240.843 & @xmath239.99 + palomar 6 & @xmath21.08 & 0.9 & @xmath240.343 & 200.22 + terzan 5 & @xmath20.28 & 0.7 & @xmath240.809 & @xmath267.79 + ngc 6440 & @xmath20.34 & 1.3 & @xmath240.919 & @xmath236.37 + terzan 6 & @xmath20.50 & 1.6 & @xmath20.132 & 130.21 + terzan 9 & @xmath21.00 & 0.6 & @xmath240.866 & 83.81 + ngc 6522 & @xmath21.44 & 0.6 & @xmath240.547 & @xmath27.09 + ngc 6528 & @xmath240.07 & 1.3 & @xmath240.146 & 199.37 + ngc 6541 & @xmath21.76 & 2.2 & @xmath20.816 & @xmath2202.01 + ngc 6553 & @xmath20.16 & 2.5 & @xmath240.294 & 24.90 + ngc 6558 & @xmath21.44 & 1.0 & @xmath240.043 & @xmath2186.84 + ngc 6569 & @xmath20.86 & 1.2 & @xmath240.103 & @xmath216.67 + ngc 6624 & @xmath20.42 & 1.2 & @xmath240.976 & 74.51 + ngc 6626 ( m28 ) & @xmath21.45 & 2.6 & @xmath240.432 & 58.22 + ngc 6638 & @xmath20.99 & 1.6 & @xmath240.929 & 59.40 + ngc 6637 ( m69 ) & @xmath20.71 & 1.6 & @xmath240.451 & 55.80 + ngc 6642 & @xmath21.35 & 1.6 & @xmath240.980 & @xmath28.05 + ngc 6652 & @xmath20.96 & 2.4 & @xmath240.146 & @xmath296.84 + ngc 6681 ( m70 ) & @xmath21.51 & 2.1 & @xmath240.436 & 240.28 + ngc 6717 & @xmath21.29 & 2.3 & @xmath240.941 & 82.76 + ngc 6723 & @xmath21.12 & 2.6 & @xmath240.023 & @xmath286.44 + | we present high - resolution infrared echelle spectroscopy for the globular clusters palomar 6 and m71 .
our mean heliocentric radial velocity of palomar 6 is + 180.6 @xmath0 3.2 km sec@xmath1 and is 20 km sec@xmath1 lower than that found by minniti in 1995 .
contrary to the previous metallicity estimates using low - resolution spectroscopy , our results show that palomar 6 has an intermediate metallicity with [ fe / h ] = @xmath21.0 @xmath0 0.1 and is slightly more metal - poor than m71 .
reasonable changes in the surface temperature or the microturbulent velocity of the model atmospheres do not affect [ fe / h ] at more than @xmath0 0.2 dex . in spite of its high metallicity , based on the spectrum of a single cluster member the [ si / fe ] and [ ti / fe ] ratios of palomar 6 appear to be enhanced by 0.4 dex and 0.5 dex , respectively , suggesting that the galactic inner halo may have experienced a very rapid chemical enrichment history . |
recently , a study of dependence by using copulas has been getting more attention in the areas of finance , actuarial science , biomedical studies and engineering because a copula function does not require a normal distribution and independent , identical distribution assumptions .
furthermore , the invariance property of copula has been attractive in the finance area .
but most copulas including archimedean copula family are symmetric functions so that the copula model fitting for asymmetric data is not appropriate . liebscher ( @xcite ) introduced two methods for the construction of asymmetric multivariate copulas . the first is connected with products of copulas .
the second approach generalizes the archimedean copulas .
the resulting copulas are asymmetric but are little extension of the parametric families of copulas .
this paper proposes a new generalized copula family which include asymmetric copulas in addition to all copula families available in the current literature .
we will characterize the set of all twice continuously differentiable copulas that may arise .
we will start with some basic concepts of copulas in the next section . in section 3
we characterize a class of differentiable copulas and state our main theorem . in section 4
we will discuss some well - known copulas and introduce a new class of copulas ( asymmetric in general ) in support of our main result . finally , in section 5 , we will study the dependence structure by calculating spearman s rho and kendall s tau and describe a method to maximize and minimize spearman s rho using an approximation technique for a special class of copulas that arises from our construction .
in this section we recall some concepts and results that are necessary to understand a ( bivariate ) copula . throughout this paper @xmath0
denotes the unit interval @xmath1 $ ] .
a copula is a multivariate distribution function defined on @xmath2 , with uniformly distributed marginals . in this paper , we focus on bivariate ( two - dimensional , @xmath3 ) copulas . a bivariate copula is a function @xmath4 , which satisfies the following properties 1 .
@xmath5 2 .
@xmath6 3 .
@xmath7 with @xmath8 .
the importance of copulas has been growing because of their applications in several fields of research .
their relevance primarily comes from the following theorem of sklar ( see @xcite ) : [ sklar ] let @xmath9 and @xmath10 be two random variables with joint distribution function @xmath11 and marginal distribution functions @xmath12 and @xmath13 , respectively . then there exists a copula @xmath14 such that h(x , y ) = c(f(x ) , g(y ) ) for all @xmath15 .
if @xmath12 and @xmath13 are continuous , then @xmath14 is unique .
otherwise , the copula @xmath14 is uniquely determined on @xmath16 .
conversely , if @xmath14 is a copula and @xmath12 and @xmath13 are distribution functions , then the function @xmath11 defined above is a joint distribution function with margins @xmath12 and @xmath13 .
sklar s theorem clarifies the role that copulas play in the relationship between multivariate distribution functions and their univariate margins .
a proof of this theorem can be found in @xcite .
frchet lower and upper bounds for copulas are denoted by @xmath17 and @xmath18 , respectively , and defined by @xmath19 @xmath20 for all @xmath21 then it well - known that for any copula @xmath22 , @xmath23 in this paper we will mainly concentrate on copulas which are twice continuously differentiable , i.e. , @xmath24 . with this assumption and using property 3 ( p3 ) of copulas , for all @xmath25 with @xmath8 , we have @xmath26 .
this implies @xmath27 is monotonically increasing in the second variable .
hence @xmath28 \geq 0 $ ] .
therefore @xmath29 is increasing in its first variable .
hence we deduce the following lemma : [ copula_pde ] the following two statements are equivalent : 1 .
@xmath14 is a twice continuously differentiable copula .
2 . @xmath14 satisfies the following dirichlet inequality problem : @xmath30 with boundary conditions : @xmath31 @xmath32
in this section we will start by solving the above problem stated in lemma [ copula_pde ] and then we will characterize all twice differentiable copulas .
suppose @xmath34 is a continuous real - valued function .
then inequality ( [ co_pd_in ] ) can be reformulated as follows , @xmath35 integrating twice we get @xmath36 where @xmath11 and @xmath13 are two arbitrarily real - valued functions of @xmath37 and @xmath38 , respectively . using boundary conditions
@xmath39 , we have @xmath40constant . hence @xmath14 has the following form @xmath41 now using the boundary condition @xmath42 , we have @xmath43 differentiating both sides with respect to @xmath38 , we have @xmath44 similarly using the fourth boundary condition @xmath45 , we have @xmath46 this leads to the following theorem , [ mjk ] suppose @xmath47 is a continuous real - valued function such that @xmath48 then @xmath49 is a copula .
furthermore , every twice continuously differentiable copula is of the form given in ( [ copula ] ) .
if we substitute @xmath50 for @xmath51 in equations ( [ copula1 ] ) , ( [ copula2 ] ) and ( [ copula3 ] ) , it is easy to verify that every twice continuously differentiable copula is given by ( [ copula ] ) . to prove the other direction , we have from ( [ copula ] ) , @xmath52 , which is continuous and non - negative on @xmath53 .
it is easy to check @xmath39 for all @xmath54 .
we also have , @xmath55 similarly , we can show that @xmath45 . hence by lemma [ copula_pde ] @xmath14 is a copula .
one importance of theorem [ mjk ] is the fact that in general there is no assumption of symmetry on @xmath56 and hence on @xmath22 . with the help of above theorem , in the next section we will construct a class of examples of non - symmetric copulas
it is quite easy to verify theorem [ mjk ] for well - known @xmath33 copulas by constructing the corresponding @xmath57 . in this section
we will begin with showing those renowned examples and later we will show how to construct other copulas .
archimedean copula is a very interesting class of copulas , whose investigation arose in the context of associative functions and probabilistic metric spaces ( see @xcite ) and today has also many applications in the statistical context ( see @xcite ) .
moreover , archimedean copulas are widely used in applications , especially in finance , insurance and actuarial science , due to their simple form and nice properties .
let @xmath58 $ ] be a continuous , decreasing function such that @xmath59 .
the _ pseudo - inverse _ of @xmath60 is the function denoted @xmath61}$ ] with domain @xmath62 $ ] , range @xmath0 and defined by @xmath63}(t ) = { \displaystyle}\left\{\begin{array}{lcr } \varphi^{-1}(t ) & \text{if}\quad 0\leq t \leq \varphi(0),\\ 0 & \text{if}\quad \varphi(0)\leq t \leq \infty\\ \end{array}\right.\nonumber\ ] ] let @xmath58 $ ] be a continuous , convex , strictly decreasing function such that @xmath59 .
then a copula @xmath14 is called _ archimedean _ if it can be written as @xmath64}\left(\varphi(u ) + \varphi(v)\right ) , \;\;\;\;\forall ( u , v)\in \mathbb{i}^2.\ ] ] @xmath60 is called an additive generator of @xmath14 . now assuming @xmath65 ,
let us define @xmath66}}^{''}\left(\varphi(u ) + \varphi(v)\right )
\varphi'(u ) \varphi'(v ) - 1.\ ] ] then it is evident that @xmath57 is continuous .
also one can easily show that @xmath67 and @xmath57 satisfies both ( [ copula_bd1 ] ) and ( [ copula_bd2 ] ) .
the frank copula is an archimedean copula given by @xmath68 where @xmath69 is a parameter .
its generator is given by @xmath70 then @xmath71}}^{''}(x ) = \frac{1}{\theta}\frac{e^{\theta + x } ( e^\theta-1)}{(1 - e^\theta + e^{\theta + x})^2 } , \qquad \varphi_\theta ' ( x)=\frac{\theta e^{-\theta x}}{e^{-\theta x}-1}.\ ] ] therefore we have @xmath72 one can easily verify that @xmath57 satisfies theorem [ mjk ] .
the farlie - gumbel - morgenstern ( fgm ) copula takes the form @xmath73 where @xmath74 $ ] is a parameter .
the fgm copula was first proposed by morgenstern ( 1956 ) .
the fgm copula is a perturbation of the product copula ; if the dependence parameter @xmath75 equals zero , then the fgm copula collapses to independence .
this is attractive primarily because of its simple analytical form .
fgm distributions have been widely used in modeling , for tests of association , and in studying the efficiency of nonparametric procedures .
however , it is restrictive because this copula is only useful when dependence between the two marginals is modest in magnitude .
let @xmath76 .
it can be easily verified that @xmath57 agrees with theorem [ mjk ] , and it generates the fgm copulas .
the following lemma will introduce a new class of @xmath33 copulas .
[ mjk_lem1 ] suppose @xmath77 , the space of sequences whose series is absolutely convergent , are sequences of real numbers .
also suppose @xmath57 is a real - valued function on @xmath53 defined by @xmath78 then @xmath79 @xmath80 furthermore , if @xmath81 , then @xmath82 for all @xmath83 .
notice that we can rewrite @xmath56 as , @xmath84 the first conclusion follows from the fact that the sequences @xmath77 and @xmath85 , @xmath86 . now to prove @xmath87 for all @xmath83 , first notice that @xmath88 .
also , @xmath89 for all @xmath90 .
hence we have @xmath91 therefore the additional hypothesis on @xmath57 guarantees that @xmath87 .
it is evident that @xmath92 is continuous and therefore by theorem [ mjk ] , @xmath14 , defined by @xmath93 is a copula , where @xmath57 is of the form given in lemma [ mjk_lem1 ] . noting that @xmath94 where @xmath95 and @xmath96 , or equivalently , @xmath97 it follows that @xmath98 .
similarly , if @xmath99 , @xmath100 with sequences @xmath101 , and @xmath102 , then @xmath103 will satisfy the conclusions of lemma [ mjk_lem1 ] , and will generate a copula by eq .
( [ copula ] ) .
this restatement of lemma [ mjk_lem1 ] clearly indicates that @xmath103 is obtained from products of functions in the disc algebra with vanishing zero - th moment and where the product of the @xmath104 norms of the coefficients are @xmath105 .
more precisely , [ generator ] if @xmath103 is a function on the @xmath106-torus arising from the product of functions on the unit circle , each section of @xmath103 has a vanishing zero - th moment , and the product of @xmath104 norms of fourier coefficients of components of @xmath103 is @xmath105 , then @xmath107 generates a @xmath108-copula .
this theorem provides a large class of examples from which one may construct copulas with optimal properties ( @xmath109 and @xmath110 ) .
[ four_cop_ex ] let @xmath111 , then @xmath112 define @xmath14 as follows @xmath113 then it is easy to verify that @xmath14 forms a copula and more importantly it is not symmetric .
a contour plot of @xmath14 is given in figure [ fig : fourier - cop ] . ,
the two most commonly used nonparametric measures of association for two random variables are spearman rho ( @xmath114 ) and kendall tau ( @xmath115 ) . in general
they measure different aspects of the dependence structures and hence for many joint distributions these two measures have different values .
suppose x and y are two random variables with marginal distribution functions f and g , respectively .
then _ spearman _
@xmath114 is the ordinary ( pearson ) correlation coefficient of the transformed random variables f(x ) and g(y ) , while _ kendall _ @xmath115 is the difference between the probability of concordance @xmath116 $ ] and the probability of discordance @xmath117 $ ] for two independent pairs ( x1 , y1 ) and ( x2 , y2 ) of observations drawn from the distribution . in terms of dependence properties ,
spearman @xmath114 is a measure of average quadrant dependence , while kendall @xmath115 is a measure of average likelihood ratio dependence ( see @xcite for details ) .
if @xmath9 and @xmath10 are two continuous random variables with copula @xmath14 , then kendall @xmath115 and spearman @xmath114 of @xmath9 and @xmath10 are given by , @xmath118 @xmath119 from lemma ( [ mjk_lem1 ] ) , if we define @xmath120 where @xmath121 , then the fourier copulas are given by @xmath122 \nonumber\\ & & \qquad \left[c_m\sin{(2\pi mu ) } + d_m\{1-\cos{(2\pi mv)}\}\right]\nonumber\end{aligned}\ ] ] using ( [ tau ] ) and ( [ rho ] ) , we have @xmath123 @xmath124 now since @xmath125 , we can conclude that @xmath126 @xmath127 _ fourier copulas can be generalized by writing @xmath57 as follows , @xmath128 where @xmath129 and @xmath130@xmath131 .
the latter condition here is to ensure that @xmath103 will have range in @xmath132 .
notice that for all @xmath133 , @xmath134 is equal to @xmath135 and hence no additional conditions are necessary to assure @xmath57 to be real - valued .
+ this yields @xmath136 and @xmath137 and it can be shown again that @xmath138 , and @xmath139 _ in this section we will optimize spearman s rho using an approximation method for @xmath33-copulas with @xmath57 of the form , @xmath140 , where @xmath60 and @xmath141 both are continuous real - valued functions on @xmath0 . notice that in this special case , @xmath114 can be simplified into the following form , @xmath142 du\;dv - 3\nonumber\\ & = & 12\int_0 ^ 1 \int_0^u \varphi(s ) ds\;du\int_0 ^ 1 \int_0^v \psi(t ) dt\;dv.\nonumber\end{aligned}\ ] ] this suggests that optimizing @xmath114 is equivalent to optimizing both @xmath143 and @xmath144 .
define @xmath145 and @xmath146 .
then for some positive @xmath147 , the optimization problems become , @xmath148 @xmath149 + although it apparently looks like these optimization problems can be solved independently , they are related by the fact that @xmath150 .
this implies @xmath151 . for the optimal possibility ,
we choose , @xmath152 and @xmath153 .
this is evident from the fact that both @xmath154 and @xmath155 can be positive or negative , @xmath156 will occur either if both @xmath154 and @xmath155 are maximum or if both are minimum and @xmath157 will occur if one of @xmath154 and @xmath155 is maximum and the other is minimum . geometrically , @xmath154 will be maximum if @xmath13 has the form as in figure [ fig : gmax ] and will be minimum if @xmath13 has the form as in figure [ fig : gmin ] . , minimizing @xmath155,scaledwidth=80.0% ] , minimizing @xmath155,scaledwidth=80.0% ] one can easily prove that , in order to optimize @xmath154 , @xmath158 must be equal to @xmath159 . for convenience , now onwards we will write @xmath160 for @xmath159 .
this suggests that if @xmath161 , or @xmath162 then @xmath154 will be maximum or minimum , respectively .
but in either case , @xmath13 is not differentiable at @xmath163 , and hence @xmath60 is not continuous . to avoid this
, we will approximate @xmath13 by a smooth function as follows : for arbitrarily small @xmath164 , define @xmath165 it is worth noting that @xmath166 as @xmath167 and @xmath168 .
, @xmath169 for @xmath170 , @xmath171,scaledwidth=80.0% ] , @xmath169 for @xmath170 , @xmath171,scaledwidth=80.0% ] we can similarly optimize @xmath155 by approximating maximum and minimum of @xmath11 by the following functions @xmath172 hence optimum values of @xmath114 will be obtained by approximating @xmath57 by the following functions , @xmath173 @xmath174 notice that each of @xmath175 and @xmath176 will generate a copula as it satisfies all the hypothesis of theorem [ mjk ] .
then the corresponding spearman s rho and kendall s tau are given by , @xmath177 @xmath178 @xmath179\ ] ] @xmath180\ ] ] the optimal values of @xmath114 and corresponding @xmath115 will be obtained by letting @xmath167 .
table [ table : optization ] shows how the values of @xmath114 approach the optimal values as @xmath167 and it is clear that @xmath181 and @xmath182 . .@xmath114 and @xmath115 values as @xmath183 changes . [ cols="^,^,^,^,^",options="header " , ]
we proposed a new generalized copula family which include not only asymmetric copulas but also all copula families available in the current literature .
especially , the family of fourier copulas we proposed is very useful copula family for analyzing asymmetric data such as financial return data or cancer data in bioinformatics . in our future study
, we will extend our copula method to a multivariate case and then incorporate time varying component to our proposed method . | this paper proposes a new class of copulas which characterize the set of all twice continuously differentiable copulas .
we show that our proposed new class of copulas is a new generalized copula family that include not only asymmetric copulas but also all smooth copula families available in the current literature .
spearman s rho and kendall s tau for our new fourier copulas which are asymmetric are introduced .
furthermore , an approximation method is discussed in order to optimize spearman s rho and the corresponding kendall s tau . corresponding author ] |
w.s . thanks brenno carlini vallilo and nathan berkovits for discussions , and nathan berkovits for explaining the modern covariant description of the @xmath6 picture . in our previous paper
we constructed the brst operator for the superparticle and superstring in a super yang - mills background @xcite . from the brst operator
we can get the gauge fixed hamiltonian : @xmath208 where @xmath209 and @xmath210 are the graded covariant derivatives .
notice that @xmath26 and @xmath19 are shorthand notation for @xmath211 and @xmath212 , where @xmath213 is the ghost number and @xmath214 is the ghost level .
( even level and odd level correspond to fermion and boson respectively . )
the expression @xmath215 " means ghosts only " .
now we go to the lagrangian form of the action for @xmath3 . to obtain complete results for the amplitude rules
, we need to keep terms in the hamiltonian quadratic in the background fields .
this has two unusual consequences : in the lagrangian , ( 1 ) all these terms will become linear ( as familiar from the bosonic case ) , and ( 2 ) such terms new to the supersymmetric case will appear only with @xmath216 . neglecting @xmath217 and @xmath218
we see @xmath219\\ & \rightarrow&-\f{1}{2}\dot{x}^{2}+i\pi\dot{\theta}+a\cdot \dot{x}-ia_{\alpha}\dot{\theta}^{\alpha}+w^{\alpha}[\pi_{0\alpha}+(\dot{x}-a)\cdot(\gamma\theta_{0})_{\alpha}-\f{1}{2}(\gamma\theta)_{\alpha}\cdot w\gamma\theta]\end{aligned}\ ] ] by redefining @xmath220 ( deformed only with gauge fields ) we get @xmath221\ ] ] the background terms then give the vertex operator @xmath222 where @xmath223 the fact that @xmath224 vanishes by its free field equations is related to the fact that its contraction with @xmath26 gives a @xmath225 , canceling a ( spacetime ) propagator , and thus contracting two 3-point vertices into a 4-point vertex .
thus , they originate from terms in the hamiltonian quadratic in background fields .
the string vertex operator is the same , with the @xmath54 derivative replaced with the left- or right - handed worldsheet derivative . in our previous paper
@xcite the background coupling had additional terms involving the expression @xmath226 , quadratic in ghost @xmath19 s .
these terms never contribute to amplitudes because there are no ghost @xmath26 s to cancel them .
( @xmath227 has a ghost @xmath26 , but together with a ghost @xmath19 . )
this is also true for the superstring .
the operator ( affine lie ) algebra remains simple because the currents are no more than cubic in the fundamental variables : @xmath234j_{c}(z')+g_{ab}''(z_{1}-z_{2})\mbox{\boldmath$\eta$}_{ab}+::j_{a}(z_{1})j_{b}(z_{2})::\ ] ] where @xmath11 has zero - modes @xmath14 , of which only @xmath15 and @xmath16 act nontrivially on @xmath10 , and @xmath235 is the relevant green function .
for example , @xmath236 and @xmath237 .
the various definitions are @xmath238p_{a}(z)&\equiv&\gamma^{a}_{\alpha\beta}[p_{a}(z_{\alpha})+p_{a}(z_{\beta})]\\ f _ { a\alpha\beta}[z]\omega^\beta(z)&\equiv&2\gamma_{a\alpha\beta}\omega^{\beta}(z_{a})\\ f_{\alpha a\beta}[z]\omega^\beta(z)&\equiv&-2\gamma_{a\alpha\beta}\omega^{\beta}(z_{a})\\ \mbox{\boldmath$\eta$}_{\alpha}{}^{\beta}&=&-i\delta_{\alpha}^{\beta}\\ \mbox{\boldmath$\eta$}^{\beta}{}_{\alpha}&=&i\delta_{\alpha}^{\beta}\\ \mbox{\boldmath$\eta$}_{a b}&=&-\eta_{a b}\\ & & \mbox{otherwise vanish}\end{aligned}\ ] ] and @xmath239j_{d}(z')j_{c}(z_{3})\nonumber\\ & & + g_{bc}'(z_{2}-z_{3}){f_{bc}}^{d}[z']j_{a}(z_{1})j_{d}(z')\nonumber\\ & & + ( -1)^{bc}g_{ac}'(z_{1}-z_{3}){f_{ac}}^{d}[z']::j_{d}(z')j_{c}(z_{3})::\nonumber\\ & & + g_{bc}'(z_{2}-z_{3}){f_{bc}}^{d}[z']::j_{a}(z_{1})j_{d}(z')::\nonumber\\ & & + : : j_{a}(z_{1})j_{b}(z_{2})j_{c}(z_{3})::\end{aligned}\ ] ] it is then straightforward to get ( [ simplerule ] ) , which is all that is needed in amplitude calculations .
the @xmath19 expansion of the superfields follows directly from the constraints on the ( super)field strengths @xmath240~&=&~f_{ab } \\ \{\nabla_{0 \alpha},\nabla_{0 \beta}\}~&=&~2\gamma_{a \alpha \beta } \nabla^{a}\\ ~[\nabla_{0 \alpha},\nabla_{a}]~&=&2\gamma_{a \alpha \beta}w^{\beta}\end{aligned}\ ] ] and the bianchi identities that follow from them . although in practice we perform component expansions by evaluating spinor derivatives at @xmath52 , we can also directly expand superfields in @xmath19 . in a wess - zumino gauge
we have : @xmath241 where @xmath242 indicates @xmath243 independence , and we have expanded only to constant field strengths @xmath186 and @xmath167 , which is sufficient for lower - point diagrams because of the deficiency of @xmath26 s . the vertex operator @xmath244 for the superparticle is then : @xmath245\end{aligned}\ ] ] here @xmath246 includes the physical @xmath247 .
( in terms of superfields and currents we hide this physical @xmath247 in @xmath248 and @xmath249
. then @xmath109 has only ghost number non - zero @xmath176 . )
notice that the spinor vertex is the supersymmetry generator @xmath250 , which will happen again in the superstring case . inserting plane waves for the fields , @xmath251e^{ik\cdot x}\end{aligned}\ ] ]
the superstring vertices are essentially the same
. since our theory is 1st - quantized , we should calculate amplitudes in terms of worldline green functions with periodic boundary conditions @xcite .
so our partition function with imaginary time is @xmath252 where @xmath253 and @xmath254 comes from the schwinger proper - time integral representation of the 1-loop vacuum energy @xmath255 $ ] .
since this is a 1-loop amplitude , we should impose periodic boundary conditions , on both @xmath121 and @xmath19 ( to preserve supersymmetry ) : @xmath256 this boundary condition also results in a supertrace naturally in the loop amplitude .
in this setting the color - ordered , n - point , 1-loop amplitude of the superparticle can be written as : @xmath257 the factor @xmath258 comes from @xmath259 .
the @xmath196 integration factors come from the nth order expansion of the vertex operator .
the worldline green function @xmath260 is given in ( [ worldlinegf ] ) .
examples of the kinematic factor @xmath195 are given in section 5 .
the factor @xmath261 is the trace of group generators in a given ordering .
the procedure is almost identical to the particle case .
one difference is that our green functions are now doubly periodic : @xmath262 again this periodic boundary condition in both directions is required by supersymmetry .
also , there is a topological distinction among graphs , namely planar , nonplanar , and unorientable graphs .
we will concentrate on the planar one here ; the others follow from similar considerations . we can write the color - ordered , n - point , 1-loop , superstring amplitude in a form identical to that of the particle case ( [ particleamplitude ] ) , but with the string green function given in appendix [ ir ] . after the usual change of variables @xmath263
we have @xmath264 in the bosonic case there is a factor of @xmath265^{-d+2}$ ] coming from the partition function ( and a @xmath266 from the tachyon mass ) for @xmath32 @xmath121 s and the 2 reparametrization ghosts @xmath0 and @xmath1 . in the supersymmetric case
this is canceled ( as in all superstring formulations ) by an @xmath265^{(2^{d/2})(1 - 2 + 3 - 4+\cdots)}=[f(w)]^{2^{(d-4)/2}}$ ] which comes from the infinite pyramid of spinors , in @xmath133 .
however , the regularization introduces corrections to the spinor partition function : @xmath267 the @xmath66 expansion of this partition function gives corrections to amplitudes .
for example , in the 6-point , 1-loop amplitude we expect a term @xmath268 , where the @xmath269 comes from 6 @xmath140 s and the @xmath270 comes from expansion of the spinor partition function .
the general fourier decomposition of a function in 2 dimensions with doubly periodic boundary conditions @xmath271 for real @xmath272 is @xmath273 then the @xmath274 for the green function of the differential operator @xmath275 is easily found to be @xmath276 for simplicity we can set @xmath277 by translational invariance . using schwinger proper - time parametrization
we get @xmath278 next using jacobi s transform @xmath279 we get @xmath280 then using @xmath281 we get @xmath282 also using @xmath283 we get @xmath284 now let s transform each sum into a sum over positive integers only : @xmath285 @xmath286 @xmath287 where @xmath288 we can subtract out @xmath45 for the particle from @xmath289 , which includes the part divergent as @xmath290 : @xmath291}{\sinh(\pi\tau\epsilon)}\ ] ] the remainder is @xmath292 in the @xmath290 limit ( @xmath293 ) , using @xmath294 we get for the remainder @xmath295
^ 2}{2t}-\frac12\ln|\lambda|-\ln|f(w)^{2}|+g^x_{un}\end{aligned}\ ] ] where ( assuming @xmath296 for simplicity ) @xmath297 ^ 2}{2t}\\ \theta_1(\f{iz}{2\pi}|\f{it}{2\pi})&=&-iw^{1/8}(\rho^{1/2}-\rho^{-1/2 } ) \prod_{m=1}^\infty(1-w^m\rho)(1-w^m\rho^{-1})(1-w^{m})\\ \theta'_1(0|\f{it}{2\pi})&=&2\pi w^{1/8}f^3(w),\qquad f(w)=\prod_{m=1}^\infty(1-w^m)\\\end{aligned}\ ] ] and @xmath143 is the unregularized green function with the usual @xmath77-dependent constant " added to normalize its short distance behavior to be the same as that of the tree case ( see , e.g. , @xcite ) .
term is the zero - mode behavior , and the second term is a constant that wo nt contribute to massless amplitudes ( because of derivatives and @xmath147 ; the non-@xmath299 piece is the same as for the particle ) .
the worldsheet green function for @xmath19 can be obtained by differentiating that for @xmath121 . however , to be careful about zero - modes some modification is needed . for the 1st - order differential operator @xmath300
we find the mode sum of the green function @xmath301 where the @xmath66 in the numerator is nontrivial because the second - order green function has a @xmath68 pole .
the above sum is almost identical to the second - order case except for the change @xmath302 .
therefore we can write the first - order green function as @xmath303}{\sinh(\pi\tau\epsilon)}+\frac12\sum_{n=1}^{\infty}\frac{1}{n+\epsilon}(\lambda_{n}^{n}+c.c.)\right.\nonumber\\ & & \left.\hspace{1cm}+\frac12\sum_{n , m=1}^{\infty}\frac{1}{n+\epsilon } w^{mn}_{n}(\rho_{n}^{n}+\rho_{n}^{-n}+c.c.)\right]\end{aligned}\ ] ] where @xmath304\\ \ln\rho_{n}&=&i\left[x+iy\left(1+\frac{\epsilon}{n}\right)\right]\\ w_{n}&=&e^{-2\pi\tau(1+\epsilon / n)}\end{aligned}\ ] ] hence it has a less divergent leading term followed by the expected differentiated second - order green function : @xmath305 we will concentrate on terms which give fermion contributions . in ( [ supertree ] ) only @xmath306 and @xmath307 give the @xmath308 amplitude . let s consider @xmath309 , for example . using ( [ simplerule]),([calgebra ] )
we get for the tree ( no fermion zero - mode regularization ) @xmath310\end{aligned}\ ] ] then we see in the wess - zumino gauge ( @xmath311 , etc . )
@xmath312a(1)\cdot w(2)\gamma w(3)\nonumber\\ e:(\omega dd)&:&\langle\omega^{\alpha}(1)d_{\beta}(2)\rangle ( d_{\gamma}(3)a_{\alpha}(1))w^{\beta}(2)w^{\gamma}(3)\nonumber\\ & & -\langle\omega^{\alpha}(1)d_{\gamma}(3)\rangle ( d_{\beta}(2)a_{\alpha}(1))w^{\beta}(2)w^{\gamma}(3)\nonumber\\ & = & i\left[-\f{1}{(z_{1}-z_{2})^{2}}\f{1}{z_{3}-z_{1}}+\f{1}{(z_{1}-z_{3})^{2}}\f{1}{z_{2}-z_{1}}\right]a(1)\cdot w(2)\gamma w(3)\end{aligned}\ ] ] this reduces to ( [ 3pt ] ) .
we will concentrate on the case where the fermions are at both ends .
the other case can be easily obtained by permutation .
there are two kinds of contributions : @xmath313 ( with @xmath314 ) and @xmath315 ( and corresponding @xmath316 ) .
the @xmath317 contribution gives a gp sum with the corresponding @xmath318 as usual .
the explicit formula is @xmath319}w^{\gamma}(3)\right)w^{\delta}(4)\nonumber\\ & + & \f{3}{16}tr(\gamma_{ab}\gamma_{cd})w^{\alpha}(1)f^{ab}(2)\left(\f{1}{2!}d_{[\delta}d_{\alpha]}f^{cd}(3)\right)w^{\delta}(4)\nonumber\\ b:&-&w^{\alpha}(1)\left(\f{1}{3!}d_{[\delta}d_{\alpha}d_{\gamma]}w^{\beta}(2)\right)\left(d_{\beta}w^{\gamma}(3)\right)w^{\delta}(4)\nonumber\\ & + & \f{3}{16}tr(\gamma_{ab}\gamma_{cd})w^{\alpha}(1)\left(\f{1}{2!}d_{[\delta}d_{\alpha]}f^{ab}(2)\right)f^{cd}(3)w^{\delta}(4)\nonumber\\ c:&-&w^{\alpha}(1)\left(d_{\alpha}w^{\beta}(2)\right)\left(d_{\delta}w^{\gamma}(3)\right)\left(\f{1}{2!}d_{[\beta}d_{\gamma]}w^{\delta}(4)\right)\nonumber\\ & + & \f{3}{16}tr(\gamma_{ab}\gamma_{cd})w^{\alpha}\left(d_{\alpha}w^{\beta}(2)\right)f^{ab}(3)\left(d_{\beta}f^{cd}(4)\right)\nonumber\\ d:&-&w^{\alpha}(1)\left(d_{\delta}w^{\beta}(2)\right)\left(d_{\alpha}w^{\gamma}(3)\right)\left(\f{1}{2!}d_{[\gamma}d_{\beta]}w^{\delta}(4)\right)\nonumber\\ & + & \f{3}{16}tr(\gamma_{ab}\gamma_{cd})w^{\alpha}(1)f^{ab}(2)\left(d_{\alpha}w^{\gamma}(3)\right)\left(d_{\gamma}f^{cd}(4)\right)\nonumber\\ e:&&\left(d_{\beta}d_{\gamma}w^{\alpha}(1)\right)\left(d_{\delta}w^{\beta}(2)\right)\left(d_{\alpha}w^{\gamma}(3)\right)w^{\delta}(4)\nonumber\\ & -&\f{3}{16}tr(\gamma_{ab}\gamma_{cd})\left(d_{\beta}f^{ab}(1)\right)\left(d_{\delta}w^{\beta}(2)\right)f^{cd}(3)w^{\delta}(4)\nonumber\\ f:&&\left(d_{\gamma}d_{\beta}w^{\alpha}(1)\right)\left(d_{\alpha}w^{\beta}(2)\right)\left(d_{\delta}w^{\gamma}(3)\right)w^{\delta}(4)\nonumber\\ & -&\f{3}{16}tr(\gamma_{ab}\gamma_{cd})\left(d_{\gamma}f^{ab}(1)\right)f^{cd}(2)\left(d_{\delta}w^{\gamma}(3)\right)w^{\delta}(4)\end{aligned}\ ] ] @xmath320 and @xmath321 vanish due to a gp sum . @xmath322 and @xmath323 give identical contributions , using integration by parts ( momentum conservation ) and the ( free ) @xmath186 field equation @xmath324 .
the results are given in ( [ superresult ] ) .
these results appear in the literature in forms where neither gauge invariance nor permutation symmetry ( relating ffbb and fbfb ) is manifest , which we now provide for comparison .
when written in terms of each momentum and gauge field , the results are ( before applying integration by parts ) @xmath325 each of @xmath322 and @xmath323 can then be re - expressed as @xmath326 another expression for each of @xmath327 can be obtained by absorbing the second term into the first term , and the summed result is : @xmath328 there are totally @xmath329 terms ( and also 12 corresponding @xmath330 terms ) contributing to the 4-fermion amplitude : @xmath331[\gamma\delta]&:&\f{1}{2!2!}(d_{[\delta}d_{\gamma]}w^{\alpha})w^{\beta}(d_{[\beta}d_{\alpha]}w^{\gamma})w^{\delta}\nonumber\\ & & -\f{1}{2!2!}(d_{[\gamma}d_{\delta]}w^{\alpha})w^{\beta}w^{\gamma}(d_{[\beta}d_{\alpha]}w^{\delta})\nonumber\\ & & -\f{1}{2!2!}w^{\alpha}(d_{[\delta}d_{\gamma]}w^{\beta})(d_{[\alpha}d_{\beta]}w^{\gamma})w^{\delta}\nonumber\\ & & + \f{1}{2!2!}w^{\alpha}(d_{[\gamma}d_{\delta]}w^{\beta})w^{\gamma}(d_{[\alpha}d_{\beta]}w^{\delta})\nonumber\\~ [ \alpha\gamma][\beta\delta]&:&-\f{1}{2!2!}(d_{[\delta}d_{\beta]}w^{\alpha})(d_{[\gamma}d_{\alpha]}w^{\beta})w^{\gamma}w^{\delta}\nonumber\\ & & + \f{1}{2!2!}(d_{[\beta}d_{\delta]}w^{\alpha})w^{\beta}w^{\gamma}(d_{[\gamma}d_{\alpha]}w^{\delta})\nonumber\\ & & + \f{1}{2!2!}w^{\alpha}(d_{[\alpha}d_{\gamma]}w^{\beta})(d_{[\delta}d_{\beta]}w^{\gamma})w^{\delta}\nonumber\\ & & -\f{1}{2!2!}w^{\alpha}w^{\beta}(d_{[\beta}d_{\delta]}w^{\gamma})(d_{[\alpha}d_{\gamma]}w^{\delta})\nonumber\\~ [ \alpha\delta][\beta\gamma]&:&-\f{1}{2!2!}w^{\alpha}(d_{[\alpha}d_{\delta]}w^{\beta})w^{\gamma}(d_{[\gamma}d_{\beta]}w^{\delta})\nonumber\\ & & + \f{1}{2!2!}(d_{[\gamma}d_{\beta]}w^{\alpha})(d_{[\delta}d_{\alpha]}w^{\beta})w^{\gamma}w^{\delta}\nonumber\\ & & -\f{1}{2!2!}(d_{[\beta}d_{\gamma]}w^{\alpha})w^{\beta}(d_{[\delta}d_{\alpha]}w^{\gamma})w^{\delta}\nonumber\\ & & + \f{1}{2!2!}w^{\alpha}w^{\beta}(d_{[\alpha}d_{\delta]}w^{\gamma})(d_{[\beta}d_{\gamma]}w^{\delta})\end{aligned}\ ] ] for each term there are 4 terms , which come from @xmath332^{2}$ ] . among them
only the @xmath333 term survives , and the others vanish due to a gp sum from corresponding @xmath334 terms . for each group
two terms are equal to the other 2 terms , and the resultant 6 terms are @xmath335 symmetry between any two fermion lines is somewhat obscure in this form .
but there are fierz identities which make it clear : @xmath336 @xmath337 @xmath338 using the above identities we can rewrite ( [ 4fermistep ] ) as @xmath339 now symmetry in fermion lines can be checked using fierz identity @xmath320 .
w. siegel , `` lorentz covariant gauges for green - schwarz superstrings , '' _ strings 89 _ , college station , tx , mar 13 - 18 , 1989 , eds .
r. arnowitt , r. bryan , m.j .
duff , d. nanopoulos , c.n .
pope ( world scientific , 1990 ) ; + s. j. gates , jr . , m. t. grisaru , u. lindstrm , m. roek , w. siegel , p. van nieuwenhuizen and a. e. van de ven , phys .
b * 225 * ( 1989 ) 44 ; + u. lindstrm , m. roek , w. siegel , p. van nieuwenhuizen and a. e. van de ven , nucl .
b * 330 * ( 1990 ) 19 ; + m. b. green and c. m. hull , phys .
b * 225 * ( 1989 ) 57 ; + r. kallosh , phys .
b * 224 * ( 1989 ) 273 .
f. bastianelli , g. w. delius and e. laenen , phys .
b * 229 * ( 1989 ) 223 ; + j. m. l. fisch and m. henneaux , preprint ulb - th2 - 89 - 04-rev ( jun 1989 ) ; + u. lindstrm , m. roek , w. siegel , p. van nieuwenhuizen and a. e. van de ven , j. math .
* 31 * ( 1990 ) 1761 .
w. siegel , `` covariant approach to superstrings , '' _ symp . on anomalies , geometry and topology
_ , argonne , il , mar 28 - 30 , 1985 , eds .
w.a . bardeen and a.r .
white ( world scientific , 1985 ) ; + nucl .
b * 263 * ( 1986 ) 93 . | we give a new , manifestly spacetime - supersymmetric method for calculating superstring scattering amplitudes , using the ghost pyramid , that is simpler than all other known methods .
no pictures nor non - vertex insertions are required other than the usual @xmath0 and @xmath1 ghosts of the bosonic string .
we evaluate some tree and loop amplitudes as examples .
march 28 , 2006 yitp - sb-06 - 5 .5 in * simpler superstring scattering * + kiyoung lee and warren siegel + _ c. n. yang institute for theoretical physics + state university of new york , stony brook , ny 11794 - 3840 _ + many formalisms have been introduced for calculating scattering amplitudes for superstrings .
the most practical of these have been ( covariant ) ramond - neveu - schwarz ( rns ) @xcite , ( lightcone ) green - schwarz ( gs ) @xcite , hybrid rns - gs ( h ) @xcite , and pure spinor ( ps ) @xcite .
all of these have ( at least ) two important defects : \(1 ) some kind of insertion is required .
it may be separate from the vertices , or may be combined with some vertices to put them into different pictures " .
the result is to complicate the calculations or destroy manifest symmetry .
( the only exception is tree graphs with external bosons only , where such methods make cyclic symmetry more obscure but avoid producing extra terms that cancel . )
\(2 ) supersymmetry is not completely manifest .
the most serious case is rns , where fermion vertices are much more complicated than boson ( because the spinors are not free fields , so in practice noncovariant exponentials of bosons must be used ) , and sums over spin structures ( periodic / antiperiodic boundary conditions ) must be performed in loops . in the gs and h cases
there is partial supersymmetry ( and partial 10d lorentz invariance ) , which complicates vertices for the longitudinal " directions , which are required for general higher - point calculations ; for this reason we will not consider gs and h in detail .
the most symmetric is ps , which has only an integration measure that is explicitly dependent on the spinor coordinates . in a previous paper @xcite
we introduced a new formalism for the superstring ( based on a similar one for the superparticle @xcite ) using an infinite pyramid of ghosts for the spinor coordinate ( gp ) @xcite .
a derivation was also given from a covariant action .
( the rns action is not spacetime - supersymmetry covariant .
the gs action @xcite has defied covariant quantization @xcite .
the h and ps formalisms do not follow from the quantization of an action with general worldsheet metric . )
the becchi - rouet - stora - tyutin operator found there was rather complicated , but fortunately none of the results of our previous paper will be needed explicitly here for calculation , but only for justification of the validity of our approach .
in fact , the gauge - fixed action and massless vertex operators were guessed much earlier @xcite .
( an early attempt to apply them to amplitude calculations failed because spinor ghosts were not included @xcite . )
the fact that these simple rules can be applied so naively hints that perhaps an even simpler formalism exists that implies the same rules .
there are ( at least ) two new conceptual results in this paper ( in addition to the explicit calculations ) , both of which involve the treatment of zero - modes . these allow us to evaluate trees and loops without evaluating explicit integrals or ( super)traces over these zero - modes , thereby solving the above two problems : \(1 ) in loop calculations we infrared regularize the worldsheet propagators . in principle one
should do this anyway , since ir divergences are notorious in two dimensions , especially for 2d conformal field theories , but usually such problems are avoided by examining only ir - safe quantities . in our case
such a regularization allows a simple counting of the infinite number of zero - modes arising from the ghost pyramid ( including those from the physical spinor ) , with the only result being the introduction of factors of 1/4 due to the usual summation @xmath2 .
( regularization of @xmath3 zero - modes is unnecessary ; it only replaces the momentum - conservation @xmath4-function with a sharp gaussian . )
\(2 ) in tree graphs these zero - modes do not appear separately , having been absorbed into the definition of the ( first - quantized ) vacuum .
specifically , since we do not perform explicit integration over spinor zero - modes , we also do not need to define measure factors for such integrations , make insertions of operators ( essentially dirac @xmath4-functions in those modes ) to kill those modes , nor use operators of different pictures to hide such insertions .
we do not make special manipulations to deal with such modes ; care of them is taken automatically by naively ignoring them .
although we do not analyze this vacuum ( or other ) state in detail here ( we effectively work with the old heisbenberg matrix mechanics , ignoring schrdinger wave functions ) , we explain why such behavior is implied by the standard n=1 superspace formulation of the vector multiplet .
the net result of these ideas is that the calculational rules are the most naive generalization of the rules of the bosonic string : ( 1 ) the @xmath0 and @xmath1 ghosts appear in the same way , affecting only the measure . ( 2 ) the spinor ghosts serve only to ensure correct counting of zero - modes , and give an extra factor of 1/4 to any trace of @xmath5-matrices .
( 3 ) ir regularization takes care of all ( physical and ghost ) spinor zero - modes .
( 4 ) the vertex operator for the massless states generalizes the bosonic - string one just by adding the same spin terms as in ordinary field theory or supergraphs ( to include the spinor vertices ) , taking into account the stringy generalization of the algebra of covariant derivatives @xcite .
consequently , for the case of tree graphs with external vectors only , our rules are almost identical to ( r)ns calculations in the @xmath6 picture .
we explain the advantage of this picture and why it is more relevant to the superstring . as an interesting side result ,
we show how the @xmath7 terms in the @xmath8 current algebra arise already in the superparticle .
we now present the main result of this paper , the rules themselves , with examples later .
( derivations are given in the appendices . ) here we will calculate amplitudes with only massless external states .
( we also concentrate on open strings , but the results generalize in the usual way to closed . ) to a limited extent first - quantization can be applied to particles as well as to strings : it gives only one - particle irreducible graphs ( vertices at the tree level ) , whereas for the string it gives complete s - matrix amplitudes by duality ( for given loop level and external states ) .
however , the methods are almost identical , particularly since the superparticle is the zero - modes of the superstring .
the vertex operators follow from the results of our previous paper @xcite but are basically those of @xcite with a small modification from ghosts ( as expected from the integrated vertex operators of ps @xcite ) : @xmath9 where @xmath10 are superfields and @xmath11 are 2d currents : @xmath12 @xmath13 where @xmath11 have zero - modes @xmath14 , of which only @xmath15 and @xmath16 act nontrivially on @xmath10 .
@xmath17 are the currents of @xcite , while @xmath18 is the lorentz current of the @xmath19 ghosts ( superspin " ) .
( appendix [ hl ] gives the relation of vertex operators between lagrangian and hamiltonian formalisms . ) as for the bosonic string , the integrated vertex operator is @xmath20 and the unintegrated one is @xmath21 ; the @xmath0 and @xmath1 ghosts work in exactly the same way , to keep the measure conformal .
( we could also add a term @xmath22 to the unintegrated vertex operator to avoid having to apply @xmath23 @xcite . )
the external - state superfields and the currents can be expanded in @xmath19 for evaluation in terms of 2d green functions of the fundamental variables : for example , the vertex for just the vector is then @xmath24 where @xmath25 is the lorentz current of all @xmath19 s , physical and ghost .
there are also terms higher - order in @xmath19 , but in the absence of external fermions there are no @xmath26 s to cancel the extra @xmath19 s , so such terms wo nt contribute . because of its universality , this form is useful for comparison to other formalisms .
however , when calculating general amplitudes ( including fermions ) , it is more convenient to expand neither the currents nor superfields ( thus manifesting supersymmetry ) .
this requires rules for evaluating products of arbitrary numbers of currents .
although this problem is generally intractable for arbitrary representations of arbitrary current algebras , in our case it is relatively simple : \(1 ) @xmath18 does nt act on the superfields .
it is quadratic in free fields , so the matrix element of any product of such currents is simply the sum of products of loops of them ( in 2d perturbation theory ) , from contracting the ( ghost ) @xmath19 of one with the @xmath26 of the next .
each such loop contributes the trace of the product of the @xmath5 matrices that appear sandwiched between @xmath19 and @xmath26 in @xmath27 ( where @xmath28 " means to restrict to ghosts ) .
\(2 ) the remaining currents @xmath29 , @xmath30 , and @xmath31 form a separate algebra .
although their loops " are more complicated ( since @xmath32 is cubic in free fields ) , the structure constants are so simple that no loop contains more than 4 currents : only the combinations @xmath33 , @xmath34 , @xmath35 , or @xmath36 . since @xmath32 and @xmath37 ( but not @xmath38 )
can also act on superfields , the matrix element of such currents and superfields reduces to the sum of products of these 4 types with strings of @xmath32 and @xmath37 acting on superfields .
the loops are : @xmath39\nonumber\\ \langle d_{\alpha}(1)d_{\beta}(2)d_{\gamma}(3)d_{\delta}(4)\rangle&=&2ig_{\theta}'(1,2)g_{\theta}(1,3)g_{\theta}(1,4)(\gamma^{a}_{\alpha\gamma}\gamma_{a\delta\beta}-\gamma^{a}_{\alpha\delta}\gamma_{\gamma\beta})\nonumber\\ & & + 2ig_{\theta}'(1,2)g_{\theta}(2,3)g_{\theta}(2,4)(\gamma^{a}_{\beta\delta}\gamma_{\gamma\alpha}-\gamma^{a}_{\beta\gamma}\gamma_{a\delta\alpha})\nonumber\\ & & + ig_{x}''(1,2)\left[g_{\theta}(1,3)g_{\theta}(2,4)\gamma^{a}_{\alpha\gamma}\gamma_{a\beta\delta}\right.\nonumber\\ & & \left.~~-g_{\theta}(2,3)g_{\theta}(1,4)\gamma^{a}_{\beta\gamma}\gamma_{a\alpha\delta}\right]+\mbox{perm.}\end{aligned}\ ] ] where @xmath40 refers to fully contracted operator products , and 1 " means @xmath41 " , etc .
we have distinguished the @xmath3 and @xmath42 green functions ( @xmath43 and @xmath44 ) because only @xmath44 gives zero - mode corrections , which is explained in detail in appendix [ ir ] . for n string loops , @xmath45 is a genus - n green function : for trees , @xmath46 and @xmath47 ; at 1 string loop they are jacobi theta functions and their derivatives ; etc .
the action of the currents on the fields is given by considering all possible symmetrizations of the @xmath32 s .
any symmetrization of 2 @xmath32 s ( acting on a field ) gives @xmath48\ ] ] this reduces any string of currents to sums of strings of @xmath37 s times antisymmetrized strings of @xmath32 s , which are evaluated as @xmath49}(2)p_a(3)\cdots p_b(4)a(5 ) = g_{\theta}(1,5)\cdots g_{x}'(4,5 ) ( d_{[\alpha}\cdots d_{\beta]}p_a\cdots p_b a)(5)\ ] ] where @xmath50 , and we can replace @xmath51 with the usual supersymmetry covariant derivative @xmath16 in such antisymmetrizations since final results can always be evaluated at @xmath52 by supersymmetry .
by 10d dimensional analysis , any @xmath18 loop is dimensionless , while any @xmath8 loop has dimension 2 .
this implies ( contrary to expectations , but well known from the bosonic case ) that each @xmath8 loop carries an extra factor of _ the inverse of _ @xmath53 .
( in the particle case , there is instead an inverse of @xmath54 . )
thus , the maximum number of @xmath8 loops gives the lowest power in momenta , and each loop less gives two more powers of momenta .
( one way to see the dimensional analysis is to note that each current acting on a superfield gives a @xmath55 or @xmath44 .
the same is true in a @xmath8 loop , except that 2 currents close " the loop to give a @xmath56 or @xmath57 .
thus , each @xmath8 loop introduces an extra factor of @xmath58 . on the other hand ,
closing an @xmath18 loop gives a @xmath59 instead of @xmath57 , so such loops give no extra factor . ) finally , there is the usual momentum dependence coming from green functions connecting the superfields to each other , from their @xmath3 dependence only : for the usual plane waves , @xmath60 with units @xmath61 for the string .
the final result for an amplitude is given as a
kinematic factor " times a scalar function of momentum invariants , expressed as an integral over the worldsheet positions of the vertices .
the kinematic factor is expressed , by the above procedure , as a sum of products of superfields , representing external state wave functions .
the string rules have effectively already performed covariant @xmath19 integration , so these superfields may be evaluated at @xmath52 .
( as in the usual superspace methods , where @xmath19 expansion and integration is replaced by the action on the lagrangian " of the product of all supersymmetry covariant derivatives @xmath16 , supersymmetry guarantees that all @xmath19 dependence cancels , up to total @xmath3 derivatives . )
the evaluation of spinor derivatives follows from the ( linearized ) constraints on the gauge covariant superspace derivatives , and their bianchi identities @xcite .
the result is @xmath62}w^\beta\end{aligned}\ ] ] the result is also ( linearized ) gauge invariant ( except for @xmath63 , as explained above ) , so one may use a wess - zumino gauge where @xmath64 at @xmath52 .
( a review of gauge covariant derivatives appears in appendix [ ce ] . ) in evaluation of tree graphs there is the usual @xmath4-function for conservation of total momentum from the zero - modes of @xmath3 , but @xmath19 effectively has no zero - modes : the effect of the @xmath19 ghosts is to mimic gs where , unlike momentum , the 8 surviving fermionic variables of the lightcone are self - conjugate , and thus have no vanishing eigenvalues .
thus there is no residual integration over @xmath19 zero - modes ( unlike ps ) . in loops
there is the usual summation over @xmath19 zero - modes in the sum over all states , but the ghosts again mimic gs by effectively reducing the number to 8 from the physical 32 ( @xmath19 and conjugate @xmath26 ) , using the sum @xmath65 when counting the number of @xmath19 s at successive ghost levels ( alternating in statistics ) .
application of this rule requires infrared regularization of the 2d green functions to remove " the zero - modes : the factor in the partition function from these zero - modes is the ir regulator @xmath66 to the power @xmath67 ( from the 16-valued spinor index on the @xmath19 s ) .
since the @xmath42 green function goes as @xmath68 ( + the usual finite expression + @xmath69 ) , the amplitude vanishes until 4-point .
thus the power of the regulator counts zero - modes .
the 1/4 rule also applies in @xmath5-matrix algebra .
amplitudes involve traces of products of @xmath5-matrices .
these matrices are the same at each ghost level ( except that chirality , as well as statistics , alternates with ghost level ) , so the net effect of the ghosts appears only when taking a trace : applying the usual @xmath5-matrix identities , the trace is reduced to @xmath70 , again reproducing gs .
the difference from gs is that the @xmath5-matrices are for 10 dimensions , so the result is lorentz covariant , and the usual 10d levi - civita tensor is produced ( where appropriate ) instead of spurious 8d @xmath66-tensors .
for example , anomalies can be found from 6-point graphs .
( details of the regularized green functions are given in appendix [ ir ] . )
we begin by proving that the trees with external bosons are identical to those obtained from ( the ns sector of ) rns .
this is most obvious in the @xmath6 picture .
although this picture was the original one to be used in ( r)ns amplitude calculations , it was immediately replaced with the @xmath71 picture @xcite .
we refer here to the picture for the physical coordinates ( @xmath72 ) , and not just the ghosts : for example , vector vertices have always been @xmath73 except for two @xmath74 vertices , while in the @xmath6 picture all vertices are @xmath73 . in the proof of equivalence @xcite ,
starting from the @xmath6 picture , one pulls factors of ( the @xmath75 modes of ) @xmath76 ( worldsheet supersymmetry generator ) off of two unintegrated @xmath73 vertices to turn them into @xmath74 vertices , then collides the @xmath45 s to produce ( the 0 mode of ) a worldsheet energy - momentum tensor @xmath77 , which gives a constant acting on a physical state .
( with ghosts the approach is similar , with @xmath45 replaced with the picture - changing operator , which is simply the operator product of the gauge - fixed @xmath45 with @xmath78 in terms of the bosonized ghost @xmath79 . )
the resulting rules are then the same as the rules for the bosonic string , including the factors of @xmath1 for the three unintegrated vertices , except that the @xmath80 vertex has the extra spin term .
the @xmath81 and @xmath5 ghosts are completely ignored ; the vacuum used is in what is usually called the @xmath82 picture " , so the zero - modes of @xmath5 ( or @xmath79 ) are already eliminated .
( what is usually called picture changing " in the modern covariant formalism would start with the @xmath71 picture , introduce two factors of picture changing times inverse picture changing , use the picture changing to change the two @xmath74 vertices , and use the inverses to change the initial and final vacuua .
unfortunately , the inverse has an overall factor of @xmath1 , so in the new vacuum @xmath83 @xcite , and the @xmath5 s pick out the @xmath74 terms again in two of the unintegrated vertices @xmath84 .
thus such transformations preserve the @xmath71 picture as far as the physical sector is concerned . ) historically , the @xmath6 picture was introduced first because : ( 1 ) it is more similar to the bosonic string , and ( 2 ) cyclic symmetry is manifest ( no need to bother with picture changing ) . the @xmath71 picture was then chosen because the physical - state conditions were more obvious .
although in modern language the brst conditions are clear in either picture , it s interesting to examine the differences in the pictures if the ghosts are ignored , since the ghosts differ in different formulations of the superstring , but all formulations have similar integrated vertices .
then the ground state of the @xmath71 picture is the physical " tachyon , at @xmath85 , while in the @xmath6 picture it s an unphysical " tachyon at @xmath86 .
furthermore , the @xmath6 picture has an additional ancestor " trajectory 1/2 unit higher than the leading physical trajectory .
these disadvantages " were noticed in the days before gliozzi - scherk - olive projection . on the other hand , for the superstring
this projection eliminates the physical " tachyon as well as the ancestor trajectory .
so the only remaining additional unphysical state of the @xmath6 picture is its vacuum , while gso projection has eliminated the vacuum of the @xmath71 picture altogether !
this suggests that any comparison of the rns formulation to others would be easier in the @xmath6 picture .
the proof of equivalence of the @xmath6 vector trees to the vector trees of our formalism is then simple : one only has to note that the operator algebra of the vertices is identical .
but the vertices are identical in form ; only the explicit representation of the spin current is different .
so one only has to check the equivalence of the two current algebras . since they are both ( 10d ) lorentz currents , quadratic in free fields , this means just checking that the central charge is the same .
( the same method has been used for comparing ps to the @xmath71 picture @xcite . )
the reason the result for the central charge is the same is that the gp result is the same as the gs result : the @xmath5-matrix algebra is the same except for a trace , which is 1/4 as big in the lightcone as for a covariant spinor , but gp again gets a factor of 1/4 from summing over ghosts .
( as we ll see below , similar arguments apply in loops , unless one gathers enough spin currents to produce a levi - civita tensor . ) the calculations in the @xmath6 picture ( and gp ) are somewhat harder than the @xmath71 picture because two vertices have been replaced with ones that generate more terms , which cancel . also ,
rns bosonic trees are simpler than ps or gp because integration over the vector fermion @xmath87 effectively does all @xmath5-matrix algebra
. however , tree amplitudes with fermions are much harder in rns than ps or gp ( and increase in difficulty as the number of fermions increases ) .
ps is still simpler than gp , because @xmath19 integration takes the place of the change in the two vertices , and so also avoids generating extra terms .
so , for trees rns is the easiest for pure bosons , ps is easiest with fermions , and gp is a bit harder than both .
however , gp requires fewer rules , since all vertices are the same , so it produces more terms at an intermediate stage but is easier to program " .
this feature is a peculiarity of tree graphs : at the loop level we ll see that gp maintains the simplest rules , while rns produces extra terms that cancel ( because supersymmetry is not manifest ) . at first sight these rules for gp might seem peculiar because there is no explicit integral over spinor zero - modes , as expected in known superspace approaches .
the answer can be seen from examining the simpler ( and better understood ) case of 4d n=1 super yang - mills . since the vacuum of the open bosonic string can be identified with a constant yang - mills ghost ( or gauge parameter ) ,
we examine the ghost superfield @xmath79 , and look at @xmath88 , a supersymmetric condition . since this superfield is chiral , and supergraphs prefer unconstrained superfields , we write @xmath89 in terms of a general complex superfield @xmath90 . then clearly @xmath91 in our case .
this is still supersymmetric because of the gauge invariance @xmath92 .
furthermore , this @xmath90 has a nice norm , @xmath93 . in hilbert - space notation
we thus write the norm and supersymmetry as @xmath94 so the vacuum is supersymmetry invariant up to a brst triviality , and the norm includes zero - mode integration , but the extra zero - modes are absorbed by the vacuua , and no insertions are required .
( we could also use @xmath95 to define @xmath96 $ ] , @xmath97 . )
this supersymmetry of the vacuum is enough to ensure the amplitudes transform correctly , since the vertex operators are superfields times supersymmetry invariant currents , and the vacuum and vertex operators ( integrated and unintegrated ) are brst invariant .
( the unintegrated vertex operators we have used are brst invariant only after including terms higher - order in ghost @xmath19 s , which do nt contribute to amplitudes for massless external states , and probably not for massive ones either , because of the absence of ghost @xmath26 s to cancel them . )
the fact that the vacuum is half - way " up in the @xmath19 expansion was also found for the expansion in the spinor ghost coordinates in a lightcone analysis of the brst cohomology for the gp superparticle @xcite .
note that this choice of vacuum is relevant only for trees ; at 1 loop one effectively does a ( super)trace over all states rather than a vacuum expectation value , so the vacuum is irrelevant .
( we assume a similar situation will occur at higher loops , but we have not checked yet . )
as we will see below , one important affect on this vacuum choice for trees , which does not affect loops , is that : _ for tree graphs only , background fields are always evaluated in the wess - zumino gauge . _
if this vacuum structure can be better understood , it might be possible to find an analog of the @xmath71 picture for gp , avoiding the production of extra canceling terms , making it the simplest formalism even for trees . as an attempt at formulating such a picture
, one can consider this picture for rns : for pure bosons , two vertices must be in the @xmath82 picture " , so it is convenient to consider those two as the initial and final states , using the vertex operators on the initial and final vacuua .
generalizing only those 2 states to include fermions , we can write their vertex operators as in gp , but now identifying the currents with @xmath98 the first two are the usual for the spinor ( in the @xmath99 picture ) and vector ( in the @xmath82 picture ) , while the last can be identified as that for the spinor in the @xmath100 picture ( also with conformal weight 1 ) if we use the supersymmetric gauge " @xmath101 instead of the wz gauge .
( hitting @xmath102 with picture changing produces @xmath103 . ) these currents satisfy almost the same algebra as the usual ones ( including @xmath104 , to leading order ) ; the only exceptions are @xmath105 and @xmath106 . as a guess for the gp analog
, we can then try to construct a new @xmath8 for this picture that depends only on the ghosts .
unfortunately ( the simplest guess for ) this construction seems not to work , apparently because the dependence on the wz gauge hasnt been eliminated , and is incompatible with the supersymmetric gauge . as explained in section 2
, we prefer the superfield formalism for the calculation of amplitudes with fermions .
this includes the all - vector amplitude in the same calculation .
the only nonvanishing operator products for the 3-point tree , after applying the landau gauge condition ( @xmath107 ) , are : @xmath108 where the @xmath109 contraction is the usual @xmath5 trace .
the other contributions , like @xmath110 and @xmath111 , all vanish using @xmath112 in the wess - zumino gauge .
we give some details of the calculation in appendix [ superamplitude ] . notice that f and h combine to give the gp sum @xmath113 . from these combinations
we find the manifestly supersymmetric 3-point tree amplitudes for vectors and spinors @xmath114 where @xmath115 are the vectors and @xmath116 the spinors .
( note that we use the usual anticommuting fields for the spinors ; numerical evaluation involves fermionic functional differentiation , replacing these fields with the usual commuting wave functions , and may introduce signs if not all terms have the same ordering . )
this result applies to both the superparticle and superstring . in the string case
there is also a factor of @xmath117 from the green functions , but this is canceled as usual with the inverse factor from the conformal measure obtained from @xmath118 . the kinematic factor in supersymmetric amplitudes is closely related to the spinor zero - mode problem , which is the most important problem in the lorentz covariant superparticle and superstring .
if we naively integrate over zero - modes of the infinite pyramid of spinors with no vertex attached , we find @xmath119 .
so we need to regularize the zero - mode integration . in appendix [ ir ]
we derive the 2d green function with a 2d regularization mass , but it turns out that the zero - mode behavior of the 2d green function is exactly that of the 1d one .
so we will concentrate on the 1d case here . to do this ir regularization
we introduce small mass terms in the superparticle free action ( for 1d proper time " coordinate @xmath54 ) @xmath120 now we can fix the measure of zero - modes for @xmath121 and @xmath19 without ambiguity . for @xmath121 , neglecting the laplacian term , which vanishes for zero - modes , @xmath122 where @xmath77 is the range of @xmath54 ( at 1 loop , the period ) .
here we used @xmath123 .
therefore our zero - mode measure for @xmath121 is @xmath124 however , this bosonic zero - mode does not appear explicitly , since this always gives momentum conservation thanks to the vertex operators . similarly for @xmath19
we see @xmath125 where @xmath126 " stands for fermionic and bosonic spinor respectively .
then our zero - mode measure for a spinor is @xmath127 in our case we have an infinite pyramid of spinors and hence we get @xmath128 where we used coherent - state regularization for the ambiguous sum @xmath129 : @xmath130&=&\int \frac{d^{2}z}{\pi}e^{-z^{*}z}\langle z|(a^{\dag}a+1)(-1)^{a^{\dag}a}|z\rangle\nonumber\\ & = & \int \frac{d^{2}z}{\pi}e^{-|z|^{2}}(\langle z|a^{\dag}a|-z\rangle+\langle z|-z\rangle)\nonumber\\ & = & \int \frac{d^{2}z}{\pi}(-|z|^{2}+1)e^{-2|z|^{2}}\nonumber\\ & = & -\f{1}{4}+\f{1}{2}=\f{1}{4}\end{aligned}\ ] ] more intuitively @xmath131 so at @xmath132 we get 1/2 and 1/4 respectively . therefore in @xmath133 we get effectively @xmath134 for zero - modes .
so our complete spinor measure with non - zero modes is @xmath135 the significant role of this effective power will be clear after we discuss the green function .
the regularization @xmath136 explains how we can get a physical @xmath137 spinor contribution out of 2 covariant 16-component spinors @xmath26 and @xmath19 . because we can not project covariant spinors into physical spinors in a covariant way
, we need to add infinitely many ghosts to achieve this 1/4 reduction in amplitudes .
we summarize the results of appendix [ ir ] here .
we find the regularized 1d green functions for @xmath3 and @xmath19 @xmath138}{\sinh(\xi t/2)}\nonumber\\ g^{\theta}(z)&=&i(-\partial_{z}+\epsilon)\left[\frac{1}{2\epsilon}\frac{\cosh[\epsilon ( |z|-t/2)]}{\sinh(\epsilon t/2)}\right]\end{aligned}\ ] ] the @xmath66 correction to @xmath139 in @xmath140 is nontrivial because it multiplies a green function with a @xmath68 term .
it is convenient to expand the green functions in @xmath66 when we calculate scattering amplitudes : @xmath141 where @xmath142 and @xmath143 and @xmath144 are the usual @xmath145 green functions with periodic boundary conditions , normalized to @xmath146 .
the extra constant will not contribute to massless amplitudes because of derivatives and @xmath147 . because the mass ( re)moves zero - modes , the usual fudges of the massless green functions are eliminated : there is no freedom to add constants ( dependent on @xmath77 , but not @xmath54 ) to @xmath45 , and the @xmath4 function in its equation of motion is not modified to @xmath148 to preserve
charge conservation " .
but the latter property is restored upon expansion in the regulator : @xmath149 similarly we will do this expansion for superstring green functions .
the details are given in appendix [ ir ] .
however , expansion of @xmath150 is unnecessary , because in vertex operators @xmath121 appears only as @xmath151 ( and as an argument of the superfields ) , and any contraction involving this vertex operator is always finite .
( the derivative kills the potentially divergent @xmath152 term . ) for this reason @xmath121 regularization gives only energy - momentum conservation and is irrelevant to amplitude corrections .
but @xmath66 expansion of @xmath153 is crucial , as we will see in the next section .
here we give simple examples .
the only differences from standard first - quantization of a loop of a scalar particle or bosonic string can be associated with
kinematic factors " that may also depend on the positions of the vertices on the worldline / sheet .
( for a summary of the standard analysis of the other factors , see appendix [ loop ] . )
collecting the results of the zero - mode measure and the green function zero - mode behavior , _ the amplitude is zeroth order in @xmath66 . _
since there is an @xmath134 in the measure , we should pick up an @xmath154 in the integrand of the path integral .
for example , one sub - diagram of the n - point 1-loop amplitude is proportional to @xmath155 then to evaluate this amplitude we should expand each @xmath153 and collect terms with @xmath154 .
we now notice that every @xmath140 gives @xmath156 . for @xmath157
there are not enough powers of @xmath158 and so their amplitudes just vanish .
there is no zero - mode behavior for any contraction involving @xmath159 because of the derivative .
therefore @xmath37 contractions start to contribute only at @xmath160 ( a black dot in fig.[diagram ] ) .
the first nonvanishing amplitude is at @xmath161 .
however , this is just the case where every @xmath140 from the @xmath162 s contributes @xmath156 .
so the integration is trivially done for @xmath163 and only its spin algebra matters .
there are two kinds of diagrams : the case where all 4 points are connected , and the case where each pair of points is connected separately ( fig .
[ diagram ] ) .
these two diagrams have opposite sign .
each closed contraction should be traced over all ghost pyramid spinors to give @xmath136 .
therefore we get for @xmath164 , omitting external field factors , @xmath165\nonumber\\ & & + \frac{1}{16}\left[tr(\gamma^{ab}\gamma^{cd})tr(\gamma^{ef}\gamma^{gh})+\mbox{2 permutations}~\right]\end{aligned}\ ] ] using the mathematica code tracer.m we evaluate this gamma - matrix trace to find @xmath166[cd][ef][gh])\nonumber\\ & & -\f{1}{2}\left[(\delta^{ac}\delta^{bd}-\delta^{ad}\delta^{bc})(\delta^{eg}\delta^{fh}-\delta^{eh}\delta^{fg})\right.\nonumber\\ & & \quad\quad+(\delta^{ae}\delta^{bf}-\delta^{af}\delta^{be})(\delta^{cg}\delta^{dh}-\delta^{ch}\delta^{dg})\nonumber\\ & & \quad\quad+\left.(\delta^{ag}\delta^{bh}-\delta^{ah}\delta^{bg})(\delta^{ce}\delta^{df}-\delta^{cf}\delta^{de})\right]\end{aligned}\ ] ] this is the well - known kinematic factor for both tree and 1-loop .
we can also express this results in terms of @xmath167 as @xcite @xmath168\end{aligned}\ ] ] which can be interpreted as graviton " , dilaton " , and axion " as far as lorentz ( and not gauge ) structure is concerned .
( in the nonplanar case , it actually corresponds to those poles for color singlets in the 1 + 2=3 + 4 channel . ) here we again prefer the superfield formalism as explained in section 2 .
however , the 4-point one - loop case is dramatically simplified due to the ir regularization . consider the 4 types of fully contracted operators again and then notice that they can have only limited @xmath68 factors , since each @xmath140 gives such a factor while @xmath169 does nt : @xmath170 this means that except for @xmath109 they appear at best from the 6-point at 1 loop .
so the only contractions for this amplitude are from @xmath171 and @xmath172 .
we also need to consider the case where 4 @xmath173 s act on the superfields
. then we can directly write down the kinematic factor for the manifestly supersymmetric , 4-point , 1-loop amplitude @xmath174 } w^{\alpha}(1)w^{\beta}(2)w^{\gamma}(3)w^{\delta}(4)\nonumber\\ & & + \f{3}{32}tr(\gamma^{ab}\gamma^{cd})d_{[\alpha}d_{\beta]}f_{ab}(1)f_{cd}(2)w^{\alpha}(3)w^{\beta}(4)+\mbox{perm.}\nonumber\\ & & + \hat{k}_{4}(f^{4})\end{aligned}\ ] ] @xmath175 is the same as @xmath163 above except that the ( super)traces do nt include the physical @xmath176 .
of course , this missing contribution comes from @xmath177 @xmath178 @xmath179 @xmath180 plus different permutations of the @xmath173 s .
also , the missing contribution for the @xmath181 terms comes from @xmath182}w^{\beta})w^{\gamma}(d_{[\gamma}d_{\beta]})w^{\delta}$ ] plus different permutations .
note that this result already has the same form as the 4d n=1 supergraph calculation for n=4 super yang - mills @xcite ( if we rewrite it in majorana notation for comparison ) , where there @xmath183 already , so @xmath18 terms are unnecessary to produce @xmath184 .
( there the @xmath185 comes from overall @xmath19 integration , the @xmath173 s of the @xmath186 s being killed by loop-@xmath19 integration . )
we give here the fermion part of the result of ( [ 4pt1loopsuper ] ) and leave details to appendix [ superamplitude ] .
@xmath187 where @xmath188}$ ] .
( the @xmath189 $ ] means to sum over permutations with signs to antisymmetrize . )
the second form of the ffbb amplitude can be interpreted as axion " and traceless graviton " terms .
( using the fermion field equation and symmetry , the former term is totally antisymmetric in @xmath190 and a total curl on the fermions , as the @xmath191 factor is then for the bosons , while the latter term is symmetric and traceless in @xmath192 . )
we have written these amplitudes in manifestly gauge invariant form .
note that the complete 4-point amplitude is totally symmetric in all 4 external lines .
( this was clear from the original form ( [ 4pt1loopsuper ] ) . )
this means that not only are the specific cases listed above separately symmetric between boson lines and between fermion lines ( if we had used wave functions instead of fermionic fields then they would be antisymmetric ) , but the amplitudes for other arrangements of fermions and bosons are obtained simply by permutation .
the usual representations are given in appendix [ superamplitude ] . in principle
there is no difficulty to evaluate higher - point diagrams .
some new terms occur compared to the @xmath161 case .
first of all , @xmath159 can contribute from one vertex , acting on a field , which is indicated by a black dot in fig .
[ diagram ] .
( all the other vertices contribute contractions between @xmath193 and @xmath194 from @xmath25 . )
terms of the green function higher - order in the @xmath66 expansion start to appear and thus @xmath195 has @xmath196 dependence .
we give a schematic diagram for various types of contractions in fig .
[ diagram ] .
notice that our diagram exactly coincides with earlier covariant rns results @xcite .
there can also be corrections from the fermion partition function because of regularization .
for example , this correction in the 6-point amplitude is proportional to @xmath197 ( see appendix [ partition ] ) .
first we will consider the part of the amplitude that does nt have a black dot in fig .
[ diagram ] .
let s call the graphs without and with a black dot @xmath198 and @xmath199 respectively . since the 5-point amplitude has 5 sides we should choose @xmath200 from exactly one side .
this is true for both the pentagon and triangle + ellipse graphs .
the difference between them is the gamma matrix trace factor .
so we can write down the part of @xmath201 for a given group - index ordering ( the @xmath202th vector has @xmath203 ) as : @xmath204\nonumber\\ & & + \frac{1}{16}[~g^{\theta}_{0}(z_{2}-z_{1 } ) tr(\gamma^{a_{1}b_{1}}\gamma^{a_{2}b_{2}}\gamma^{a_{3}b_{3}})tr(\gamma^{a_{4}b_{4}}\gamma^{a_{5}b_{5}})+11~\mbox{permutations}~]\nonumber\\\end{aligned}\ ] ] then we can write @xmath205 where @xmath163 was given in subsection 5.2 .
@xmath201 and @xmath206 complete the 5-point planar amplitude .
totally antisymmetric @xmath66-tensor terms vanish because the 5 external momenta are not independent .
notice that the light - cone gs calculation reduces to our results after heavy algebra @xcite , and rns needs a spin - structure sum to produce this result @xcite .
we postpone the @xmath207-point amplitudes to another paper , which will be interesting because of the anomaly cancellation issue .
one good thing in our covariant formalism is that we have a totally antisymmetric @xmath66-tensor naturally in the hexagon amplitude , where we have enough momenta to have a nonvanishing result , contrary to the 5-point case .
there are many avenues of further study , in particular : \(1 ) many types of diagrams can be calculated . at the tree level ,
diagrams with many fermions have not yet been explicitly evaluated in any formalism .
new algebraic methods for the current algebra might be useful . at the 1-loop level , little has been done with fermions or higher - point functions .
alternative ir regularization schemes could be considered .
the 2-loop 4-vector calculation would be a good test , and nothing more than that has been done at 2 loops , and nothing at all at higher loops .
\(2 ) the hilbert space needs to be studied covariantly , especially the vacuum , to completely justify the naive manipulations we have made for tree graphs
. it would be useful to find the relation of these methods to supergraphs , where explicit zero - mode integrations appear ( both in loops , corresponding to @xmath26 zero - modes , and an overall integral for @xmath19 zero - modes . ) massive vertex operators for physical states are expected to also be relatively simple , as the spinor ghosts should appear again in a minimal way ( as opposed to the more complicated structure of the brst operator ) .
the analogy to second - quantized ghost pyramids ( e.g. , for higher - rank forms ) might be useful : there ghosts beyond the first generation ( i.e. , the usual faddeev - popov ghosts ) appear only at 1 loop , to define the measure .
\(3 ) closer relations to other formulations might exist .
an analog to the @xmath71 picture of rns might further simplify tree calculations .
the many similarities with ps suggests it might be a particular gauge choice of gp that truncates the ghost spectrum . |
the experimental discovery of bose - einstein condensation @xcite in dilute systems of trapped alkali - metal atoms , such as rubidium ( @xmath0 ) , lithium ( @xmath1 ) , sodium ( @xmath2 ) and ytterbium ( @xmath3 ) , has spurred a renewed interest into the investigation of macroscopic quantum phenomena and interference effects , allowing for a deeper understanding of the conceptual foundations of quantum mechanics @xcite .
this fascinating research area has been growing up thanks to the high degree of experimental manipulation and control @xcite .
interference between condensates released in a potential with a barrier was first observed in 1997 @xcite and that paved the way for further investigations on the problem of bose condensates in a double well potential .
then josephson oscillations have been observed in one dimensional optical potential arrays @xcite .
a single bosonic josephson junction was produced for the first time in 2005 with @xmath0 atoms and its dynamics was experimentally investigated both within tunneling as well as self - trapping regime @xcite@xcite@xcite .
more recently , mixtures of @xmath4 and @xmath5 atoms have been produced and experimentally investigated @xcite as well , whose intraspecies scattering lengths could be tunable via magnetic and optical feshbach resonances .
furthermore the realization of heteronuclear mixtures of @xmath5 and @xmath6 atoms with tunable interspecies interactions @xcite paved the way to the exploration of double species mott insulators and , in general , of the quantum phase diagram of two species bose - hubbard model @xcite .
the interplay between the interspecies and intraspecies scattering produces deep consequences on the properties of the condensates , such as the density profile @xcite and the collective excitations @xcite .
however , the wide tunability of such interactions makes a bec mixture a very interesting subject of investigation , both from experimental and theoretical side as a mean of studying new macroscopic quantum tunneling phenomena as well as the interplay between quantum coherence and nonlinearity .
indeed novel and richer behaviours are expected in such a multicomponent bec . on the theoretical side ,
a bosonic josephson junction with a single species of bec has been widely investigated by means of a two - mode approximation @xcite@xcite@xcite , within the classical as well as the quantum regime . in the classical regime , characterized by large particle numbers and weak repulsive interactions
, the gross - pitaevskii equation provides a reliable description . within the two mode approximation it reduces to two generalized josephson equations which describe the time evolution of the relative phase and the population imbalance between the wells @xcite and
differ from their superconducting counterpart @xcite by the presence of a nonlinear term which couples the variables .
because of such a term , a bosonic josephson junction exhibit a variety of novel phenomena which range from @xmath7-oscillations to macroscopic quantum self - trapping ( mqst ) @xcite .
while the @xmath8-oscillations , as well the usual josephosn ones , deal with a symmetric oscillation of the condensate about the two wells , the mqst phenomenon is characterized by a broken symmetry phase with a population imbalance between the wells . in the quantum regime ,
characterized by smaller values of the particle number and strong interactions , an increasing of phase fluctuations is observed together with the suppression of number fluctuations .
furthermore the time evolution is characterized by phase collapse and revival @xcite .
the quantum behaviour of bosonic josephson junctions has been deeply investigated by means of the usual quantum phase model @xcite@xcite@xcite as well as by starting from a two - mode bose - hubbard hamiltonian @xcite@xcite@xcite . in this context
the phase coherence of the junction has been characterized by studying the momentum distribution @xcite@xcite .
the generation and detection of schroedinger cat states has been investigated as well ; indeed the presence of such a kind of states reflects in the strong reduction of the momentum - distribution contrast @xcite@xcite .
more recently such a theoretical analysis has been successfully extended to a binary mixture of becs in a double well potential @xcite@xcite @xcite@xcite .
the semiclassical regime in which the fluctuations around the mean values are small has been deeply investigated and found to be described by two coupled gross - pitaevskii equations . by means of a two - mode approximation such equations can be cast in the form of four coupled nonlinear ordinary differential equations for the population imbalance and the relative phase of each species .
the solution results in a richer tunneling dynamics . in particular , two different mqst states with broken symmetry have been found @xcite , where the two species localize in the two different wells giving rise to a phase separation or coexist in the same well respectively .
indeed , upon a variation of some parameters or initial conditions , the phase - separated mqst states evolve towards a symmetry - restoring phase where the two components swap places between the two wells , so avoiding each other .
recently , the coherent dynamics of a two species bec in a double well has been analyzed as well focussing on the case where the two species are two hyperfine states of the same alkali metal @xcite . in this paper
we study the quantum behaviour of a binary mixture of bose - einstein condensates ( bec ) in a double - well potential starting from a two - mode bose - hubbard hamiltonian .
we analyze in detail the small tunneling amplitude regime where number fluctuations are suppressed and a mott - insulator behaviour is established .
we perform a perturbative calculation up to second order in the tunneling amplitude and study the stationary states and the dynamics of the two species bosonic josephson junction .
finally , the dynamical generation of schroedinger cat states is investigated starting from an initial coherent spin state and shown to affect the time - dependent population imbalance and momentum distribution@xcite@xcite .
we focus on the contrast in the momentum distribution between the two wells and show how it vanishes for a two - component cat state
. that could be interesting in view of the experimental realization of macroscopic superpositions of quantum states @xcite@xcite .
the paper is organized as follows . in section 2
we introduce our model hamiltonian within the two - mode approximation and define the various parameters .
then we adopt the angular momentum representation and focus on the small tunneling amplitude regime . in section 3
we apply perturbation theory in the tunneling amplitude to our hamiltonian and find analytical expressions for the energy eigenvalues and eigenstates up to second order .
section 4 and 5 are devoted to the study of the quantum evolution of the number difference of bosons between the two wells in correspondence of two different initial conditions : completely localized states and coherent spin states . in the first case
both the short and the long time dynamics is studied and a rich behaviour is evidenced , ranging from small amplitude oscillations and collapses and revivals to coherent tunneling . in the second case the short - time scale evolution of number difference is determined and a more irregular dynamics is evidenced , with suppression of the dominant frequency when the number of bosons increase .
then , schroedinger cat states are shown to generate as a result of the time - evolution of an initial coherent state when the tunneling between the two wells is suppressed , and their influence on the contrast in the momentum distribution is studied .
finally , in section 6 some conclusions and outlooks of this work are presented .
a binary mixture of bose - einstein condensates @xcite@xcite loaded in a double - well potential is described by the general many - body hamiltonian : @xmath9 where @xmath10 @xmath11 are the hamiltonians for bosons of species @xmath12 and @xmath13 respectively and @xmath14 is the interaction term between bosons of different species . for dilute mixtures one can replace the interaction potentials @xmath15 , @xmath16 and @xmath17 with the effective contact interactions : @xmath18 where @xmath19 and @xmath20 are the intraspecies coupling constants of the species @xmath12 and @xmath13 respectively , @xmath21 and @xmath22 being the atomic masses and @xmath23 , @xmath24 the @xmath25-wave scattering lengths ; furthermore @xmath26 is the interspecies coupling constant , where @xmath27 is the reduced mass and @xmath28 is the associated @xmath25-wave scattering length . in this way the hamiltonian ( [ m1])-([m4 ] ) can be rewritten as : @xmath29 @xmath30 here @xmath31 is the double well trapping potential and , in the following , we assume @xmath32 ; @xmath33 @xmath34 , @xmath35 are the bosonic creation and annihilation operators for the two species , which satisfy the commutation rules : @xmath36 & = & \left [ \psi _ { i}^{+}\left ( \overrightarrow{r}\right ) , \psi _ { j}^{+}\left ( \overrightarrow{r}^{\prime } \right ) \right ] = 0 , \label{m8 } \\ \left [ \psi _ { i}\left ( \overrightarrow{r}\right ) , \psi _ { j}^{+}\left ( \overrightarrow{r}^{\prime } \right ) \right ] & = & \delta _ { ij}\delta \left ( \overrightarrow{r}-\overrightarrow{r}^{\prime } \right ) , \text { \ \ \ \ }
% i , j = a , b , \label{m9}\end{aligned}\ ] ] and the normalization conditions : @xmath37 @xmath38 , @xmath35 being the number of atoms of species @xmath12 and @xmath13 respectively .
the total number of atoms of the mixture is @xmath39 .
now a weak link between the two wells produces a small energy splitting between the mean - field ground state and the first excited state of the double well potential and that allows us to reduce the dimension of the hilbert space of the initial many - body problem .
indeed for low energy excitations and low temperatures it is possible to consider only such two states and neglect the contribution from the higher ones , the so called two - mode approximation @xcite @xcite@xcite . in this way , by taking into account for each of the two species @xmath12 and @xmath13 the mean - field ground states @xmath40 , @xmath41 and the mean - field excited states @xmath42 , @xmath43 , the wave functions @xmath44 , @xmath35 , can be rewritten as : @xmath45 where @xmath46 , @xmath35 , and @xmath47 , @xmath48 and @xmath49 , @xmath50 ( @xmath51 , @xmath52 and @xmath53 , @xmath54 ) are the creation ( annihilation ) operators for a particle of the species @xmath12 , @xmath13 in the ground and the excited state respectively .
they satisfy the usual bosonic commutation relations @xmath55
= \left [ b_{i},b_{j}^{+}\right ] = \delta _ { ij}$ ] .
furthermore @xmath56 , @xmath35 are assumed real for simplicity and such that @xmath57 , which simplifies the calculations .
let us change the basis and switch to the atom number states in such a way that the expectation value of the population of the left and right well can be defined .
the new annihilation operators are @xmath58 , @xmath59 and @xmath60 , @xmath61 for the species @xmath12 and @xmath13 respectively , so that the wave functions ( [ m11 ] ) become : @xmath62 by substituting equations ( [ m12 ] ) into the hamiltonian ( [ m6])-([m7 ] ) , after some algebra we obtain its second quantized version within the two - mode approximation : @xmath63 where @xmath64 , @xmath35 , is the number of atoms of species @xmath12 and @xmath13 respectively , expressed as a sum of numbers of atoms in the left and right well .
the parameters are defined as follows : @xmath65 @xmath66 @xmath67 @xmath68 @xmath69 @xmath70 @xmath71 @xmath72 @xmath73 ; \text { \ \ \ } i , j = a , b;\text { \ \ \ } i\neq j. \label{m22}\ ] ] in eq .
( [ m13 ] ) , the terms proportional to @xmath74 , @xmath35 , describe tunneling of particles of species @xmath12 and @xmath13 from one to the other well while the terms proportional to @xmath75 , @xmath35 , deal with the local interaction within the two wells and the terms proportional to @xmath76 correspond to additional two - particle processes .
finally the terms proportional to @xmath77 and @xmath78 couple the two species and then various constant terms follow , which we will drop for simplicity .
in this paper we focus on the small tunneling amplitude regime where number fluctuations are suppressed and a mott - insulator behaviour is established , so it is convenient to introduce the angular momentum representation for the species @xmath12 and @xmath13 as follows : @xmath79 where the operators @xmath80 , @xmath81 , @xmath82 , obey to the usual angular momentum algebra and the following relations hold : @xmath83 in particular , the components @xmath84 , @xmath35 give the difference of the number of bosons of the species @xmath85 , @xmath86 and @xmath87 , occupying the two minima of the double well potential , i. e. the population imbalances , which are experimentally observable quantities .
thus hamiltonian ( [ m13 ] ) can be cast in the following form : @xmath88 where the constant terms have been dropped .
let us now simplify the notation by introducing the following parameters : @xmath89 and rewrite the hamiltonian ( [ m24 ] ) as : @xmath90 within the experimental parameters range it is possible to show that @xmath91 , @xmath35 , and @xmath92 @xcite@xcite , then in the following we put @xmath93 and @xmath94 , which corresponds to neglecting the spatial overlap integrals between the localized modes in the two wells . in this way the binary mixture of becs within two - mode approximation maps to two ising - type spin model in a transverse magnetic field . in the following we will focus on the symmetric case @xmath95 and @xmath96 because it allows us to perform analytical calculations while capturing many relevant phenomena characterizing the physics of the system .
so the model hamiltonian ( [ m24 ] ) becomes : @xmath97 @xmath98 @xmath99 where , in the small tunneling amplitude regime , @xmath100 is considered as a perturbation .
the total hamiltonian commutes with @xmath101 and @xmath102 , which leads to the conservation of total angular momentum with quantum numbers @xmath103 and @xmath104 respectively .
so the whole hilbert space has finite dimension , equal to @xmath105 , thus it depends on the number of bosons of the species @xmath12 and @xmath13 respectively .
the whole basis @xmath106 is given by the eigenvectors of @xmath107 ( @xmath108 ) and @xmath109 ( @xmath110 ) with @xmath111 and @xmath112 .
as a first step we need to diagonalize the unperturbed hamiltonian ( [ m28 ] ) , which can be done by performing the following @xmath113 rotation on the operators @xmath107 , @xmath109 : @xmath114 while an analogous rotation needs to be carried out on @xmath115 , @xmath116 entering the perturbation ( [ m29 ] ) . as a result we get : @xmath117 which , by defining @xmath118 , @xmath119 , and @xmath120 , @xmath121 , @xmath122 , @xmath123 , can be cast in the final form : @xmath124 in the following section we will find analytical expressions for the eigenvalues and the eigenvectors up to second order by performing perturbation theory in the tunneling amplitude .
in the present section we apply second - order perturbation theory to the hamiltonian of eq .
( [ m32 ] ) in the small tunneling amplitude limit , which allows us to derive analytical expressions for the stationary states of the system . in order to pursue this task
let us rewrite eq .
( [ m32 ] ) in dimensionless form by assuming @xmath125 as unit of energy : @xmath126 where @xmath127 and @xmath128 , then take @xmath129 as unperturbed hamiltonian and @xmath130 as a small perturbation term . here
@xmath131 , @xmath82 , obey the usual angular momentum algebra and the following relation holds : @xmath132 where : @xmath133 in principle , the rotated basis @xmath134 of the unperturbed hamiltonian ( [ m34 ] ) is given by the eigenvectors of @xmath135 ( @xmath136 ) and @xmath137 ( @xmath138 ) with @xmath139 and @xmath140 , whose corresponding eigenvalues are @xmath141 .
the presence of the operator @xmath142 , which does not commute with the perturbation term @xmath143 , makes the problem of finding eigenvalues and eigenvectors of the full hamiltonian ( [ m33 ] ) within perturbation theory much more involved . in order to simplify the treatment and carry out analytical calculations while retaining the relevant phenomenology , we concentrate on the particular case of a binary mixture where the two species are equally populated , i. e. @xmath144 , and have the same population imbalance between the two wells , i. e. @xmath145 .
this situation allows us to describe the quantum dynamics of the system in correspondence of the mqst regime , for which we need a completely localized initial state
. that fixes @xmath146 while @xmath147 could be an even or odd integer depending on @xmath148 even or odd , and leads to the following zero - order eigenvalues : @xmath149 .
each eigenvalue is two - fold degenerate , with the only exception of the ground state for @xmath150 even , @xmath151 , which is nondegenerate .
the two - dimensional subspace of degeneracy is spanned by the states @xmath152 ( where @xmath153 ) and the corresponding zero - order eigenvectors are : @xmath154 by switching on the perturbation term ( [ m35 ] ) it is possible to show that the degeneration is lifted starting from the levels with smaller @xmath155 ; in general the double degeneracy of the zero - order eigenvalues @xmath156 will be lifted at the @xmath157-th order of perturbation theory @xcite . by applying perturbation theory @xcite up to order @xmath158 , we obtain the following corrected eigenvalues : @xmath159 @xmath160 @xmath161 where @xmath162 . furthermore , for @xmath163
even , the nondegenerate ground state @xmath164 belongs to the symmetry class of @xmath165 .
the corresponding eigenvectors , up to order @xmath158 , are given in the appendix .
in the following sections we use the analytical expressions of energy eigenvectors derived in the appendix , see eqs .
( [ m41])-([m48 ] ) , in order to study the quantum evolution of @xmath166 , that is the number difference of bosons of species @xmath12 and @xmath13 between the two wells of the potential .
in this section we investigate the quantum evolution of the number difference of bosons of species @xmath12 and @xmath13 between the two wells assuming a completely localized state as initial condition .
that could be interesting in order to elucidate the quantum behavior of the system in correspondence of the classical mqst regime and to put in evidence new phenomena including quantum coherence in a multicomponent system .
in such a case we will study both the short and the long time dynamics : as a result a rich behaviour emerges , ranging from small amplitude oscillations and collapses and revivals to coherent tunneling .
although such a physics is well known for the single component bose josephson junction , in our case the dynamics shows that the two species can coexist in the same potential well as if there would be an attractive interaction between them . as a first step
let us recall the general formula which gives the time evolution of the mean value of @xmath167 @xcite : @xmath168 where @xmath169 is the dimensionless time , the sums are over all the eigenvectors @xmath170 , being @xmath171 or @xmath172 , and @xmath173 are the projections of the initial state @xmath174 on the basis @xmath175 : @xmath176
so it is clear how the knowledge of eigenvalues and eigenvectors is enough in order to study the quantum evolution of @xmath137 , the bohr frequencies involved , @xmath177 , and the corresponding weights @xmath178 .
let us now study the dynamics of the system when all the bosons of species @xmath179 and @xmath13 are initially contained in one of the two wells of the potential , say the right one , and then the imbalances of the two species coincide , so that @xmath180 , @xmath181 , @xmath182 , @xmath183 ; furthermore the two species are equally populated , i.e. @xmath184 .
that implies @xmath146 and @xmath185 in our _ center of mass _ rotated basis .
the corresponding initial condition is : @xmath186 in order to investigate the short timescales evolution we need to keep terms up to second order in the tunneling amplitude @xmath187 when we compute the weights in eq .
( [ m49 ] ) .
we find that : @xmath188 , \label{m52}\ ] ] where the frequency involved is : @xmath189 at short timescales small amplitude oscillations with frequency @xmath190 around the initial condition ( @xmath191 , @xmath192 ) are observed and that coincides with a strongly self - trapped regime . in order to investigate the dynamics at longer timescales we have to take into account also the small splittings @xmath193 and @xmath194 of the two higher pairs of quasidegenerate eigenvalues which provide two further frequencies ( see ref .
@xcite for the derivation ) : @xmath195 the whole result is : @xmath196 \right .
\nonumber \\ & & \left .
+ 2\cos ( \omega _ { \mu } \tau ) \cos ( \frac{\omega _ { 1}}{2}\tau ) -\cos ( \omega _ { 1}\tau ) -\cos ( \omega _ { 0}\tau ) \right ] , \label{m55}\end{aligned}\ ] ] and , by putting @xmath197 , the short timescale dynamics , eq .
( [ m52 ] ) , is recovered .
summarizing , at longer timescales the two species bosons are still localized in the initial potential well but the quantum dynamics exhibits collapses and complete revivals . indeed the coefficient @xmath198 , which multiplies the higher frequency term @xmath199 , gives rise to the beat , which is responsible for the observed collapses and revivals at timescales fixed by @xmath200 , as shown in fig .
[ fig : dynamics ] . finally , at very large timescales determined by the frequency @xmath201 all the bosons tunnel coherently back and forth between the two traps ; only the first term @xmath202 is responsible of such a coherent tunneling , since all harmonic functions containing the frequency @xmath200 and @xmath203 are small in amplitude and proportional to @xmath158 , thus they are unable to transfer bosons from one trap to the other .
the tunneling dynamics within macroscopic quantum self - trapping regime described above is analogous to that of the @xmath204-mode fixed point obtained by the gross - pitaevski approach @xcite , where the two species localize in the same well despite the repulsive interaction between them .
let us finally note that , despite the explicit dependence on @xmath205 of the frequencies ( [ m53])-([m54 ] ) , the different physics related to the three time scales described above is simply due to the energy splitting introduced by the renormalized tunneling for small @xmath77 .
thus in the case of a mixture of becs with equal population the dynamics remains similar to that of a single component bec , apart the coexistence of the two species in the same well .
as for the experimental detection of the long timescales phenomena ( collapses / revivals and coherent tunneling ) , since the time for their appearance is abruptly increased with @xmath163 , this implies a rapid decrease of the characteristic frequencies rendering more difficult the observation of the intermediate and long time behavior in current bec experiments .
indeed pure condensates consisting of @xmath206 atoms of @xmath5 loaded in a double well have been recently realized @xcite@xcite thus rendering the detection of the intermediate time behavior possible .
mixtures with a number of atoms ranging from @xmath207 and @xmath208 ( @xmath209 and @xmath6 @xcite ) to @xmath210 and @xmath211 ( @xmath4 and @xmath5 @xcite ) have also been recently realized , but in this case very small characteristic frequencies are implied .
however , these phenomena may be relevant for molecular systems where the number of vibrational excited quanta is small .
is @xmath212 and the boson number is @xmath213,title="fig : " ] is @xmath212 and the boson number is @xmath213,title="fig : " ] is @xmath212 and the boson number is @xmath213,title="fig : " ] in the next section we will further investigate the dynamics of the system by assuming as initial state a simple coherent state and then study the formation of a particular superposition of such coherent states , the so called schroedinger cat states .
in this section we choose as initial condition a simple coherent spin state @xcite and study the short - time scale evolution of number difference ; in this way a more complex dynamics will appear . finally , we study the generation of schroedinger cat states ; in particular , we focus on the contrast in the momentum distribution and show how it vanishes for a two - component cat state .
let us start by considering as initial condition the following coherent spin state @xcite : @xmath214 where the coefficient @xmath215 is : @xmath216 and @xmath217 and @xmath218 are two angles characterizing the superposition .
the time evolution of the mean value of @xmath137 up to first order in the tunneling amplitude @xmath187 is given by : @xmath219 + c_{2}\sin \left ( \omega _ { e}\tau \right ) \right .
\nonumber \\ & & \left .
+ \sum_{n=0or1/2}^{n_{2}/2 - 1}\frac{n_{2}!}{\left ( \frac{n_{2}}{2}% + n\right ) !
\left ( \frac{n_{2}}{2}-n\right ) ! } \frac{n_{2}-2n}{2\left ( 2n+1\right ) } a_{n}\right ] , \label{m58}\end{aligned}\ ] ] where @xmath220 -\frac{1}{\tan ^{2n+1}\left ( \frac{\theta } { 2}\right ) } \left [ \cos \left ( f_{n}\tau -\phi \right ) -\cos \left ( \phi \right ) \right ] \label{m59}\ ] ] with frequencies @xmath221 .
furthermore the coefficients @xmath222 and @xmath223 are given by : @xmath224 -\left ( \frac{n_{2}}{2}% + 1\right ) \left [ \tan \left ( \frac{\theta } { 2}\right ) -\frac{1}{\tan \left ( \frac{\theta } { 2}\right ) } \right ] \right\ } , \label{m60}\ ] ] @xmath225 \left [ \left ( \frac{n_{2}}{6}-\frac{% 1}{3}\right ) \sin \left ( 3\phi \right ) -\left
( \frac{n_{2}}{2}+1\right ) \sin \left ( \phi \right ) \right ] , \label{m61}\ ] ] for @xmath163 even , and @xmath226 , \label{m62}\ ] ] @xmath227 , \label{m63}\ ] ] for @xmath163 odd , respectively . finally , for @xmath163 even , the zero - order mean value @xmath228 is given by : @xmath229 \left [ \cos ( \omega _ { e}\tau ) -1% \right ] + 2\sin \left ( 2\phi \right ) \sin \left ( \omega _ { e}\tau \right ) \right\ } , \label{m64}\end{aligned}\ ] ] where the dominant frequency @xmath230 is equal to @xmath231 .
the corresponding expression for @xmath163 odd is : @xmath232 \left [ \cos ( \omega _ { e}\tau ) -1\right ] -2\sin \left ( \phi \right ) \sin \left ( \omega _ { e}\tau \right ) \right\ } , \label{m65}\end{aligned}\ ] ] where @xmath233 .
as one can see , the dominant frequency is gradually suppressed with the number of bosons @xmath234 @xcite .
this is clearly seen in fig .
[ fig : st - dynamics ] where the boson number difference between the two traps is plotted for different @xmath235 ( even ) values as a function of the dimensionless time @xmath236 .
one also notices a decrease of the oscillation amplitude at increasing @xmath235 .
the effect of @xmath205 is instead shown in fig .
[ fig : st - dynamics - lambda ] where the short - time dynamics of the boson number difference is analyzed for two values of the interspecies interaction . when @xmath205 increases the amplitude of the oscillations decreases .
the detection of the mixture dynamics is thus more favorable for values of @xmath205 smaller than unity .
is @xmath237 and @xmath213(black - dashed line ) , @xmath238(red - dotted line ) , @xmath239(blue - straight line ) , while @xmath240 and @xmath241 . ] . .
the value of @xmath187 is @xmath242 while @xmath243(black - dashed line ) and @xmath244(green - straight line ) , while @xmath240 and @xmath245 . ] .
let us consider the coherent spin state ( [ m56 ] ) ; the expectation value of the hamiltonian ( [ m33 ] ) on such state is given by : @xmath246 where @xmath247 and has the maximum value for @xmath248 .
this result also corresponds to the mean - field result for the energy .
now , starting from the coherent spin state ( [ m56 ] ) we are interested in looking for schroedinger cat states .
such states are quantum superposition of macroscopic states and their realization has already been suggested for a single species bose - josephson junction in @xcite .
also in the case of a bose - josephson junction with binary mixtures one might realize cat states from the time - evolution of an initially coherent state following a sudden rise of the barrier between the two wells .
thus we consider at time @xmath249 a zero inter - well coupling @xmath187 , i.e. the time evolution is governed by the hamiltonian @xmath250 in eq .
( [ m34 ] ) . for each basis vector @xmath251 of the coherent state ( [ m56 ] ) ,
the time - evolution is given by @xmath252 , where @xmath253 is the so - called revival time such that @xmath254 .
considering now the times @xmath255 , @xmath256 integer , the time evolution of the coherent state is governed by the factor @xmath257 which satisfies the property @xmath258 , depending on the parity of @xmath256 . for the choice of even @xmath256 , a discrete fourier transform leads to the cat state : @xmath259 i.e. a superposition of @xmath256 coherent states , where @xmath260 . in particular , the cat state affects the momentum distribution .
this dependence could be important to probe experimentally their existence .
in particular , when considering the two - component cat state , i.e. for the choice @xmath261 , one obtains that the contrast in the momentum distribution , i.e. the expectation value of @xmath262 on the unperturbed state , vanishes@xcite .
furthermore , the amplitude of the intervals of time in which the contrast is zero increases with increasing @xmath163 as clearly shown in fig .
[ fig : contrast ] . it should be noted that despite the close similarity in the behavior of the contrast between the single component bjj and the double one , the mixture will be a better candidate for the creation and detection of cat states .
in fact their creation time is @xmath263 and since for repulsive interaction between the two species and @xmath264 we get @xmath265>1 $ ] , such time can be made short enough to render their detection more favorable .
for example by fixing the ratio of @xmath266 interaction to @xmath267 interaction to be 2.13 , a parameter accessible in the jila setup@xcite , the detection time is twice smaller than the case of a single component bec . for @xmath268 and @xmath245 and even number of bosons .
the red line is for @xmath213 , the blue one for @xmath269 and the black one for @xmath270 .
the interval in which the contrast is zero increases with increasing @xmath235 . ]
in this paper we investigated the quantum dynamics of a bose josephson junction made of a binary mixture of becs loaded in a double well potential within the two - mode approximation .
we focused on the small tunneling amplitude limit and adopted the angular momentum representation for the bose - hubbard dimer hamiltonian .
perturbation theory up to second order in the tunneling amplitude enabled us perform analytical calculations in the symmetric case where @xmath271 and @xmath96 . in this way we obtained the energy eigenvalues and eigenstates
, whose knowledge is mandatory in order to investigate the quantum evolution of the number difference of bosons between the two potential wells . in order to study the quantum dynamics more easily and analitycally
, we restricted to the case in which the two species are equally populated and imposed the condition of equal population imbalance of the species @xmath12 and @xmath13 between the two wells .
we concentrated on the two following initial conditions : completely localized states and coherent spin states , and found a rich and complex behaviour , ranging from small amplitude oscillations and collapses and revivals to coherent tunneling . finally , we considered the generation of schroedinger cat states and pointed out their influence on the momentum distribution through the vanishing of the contrast .
we showed that the creation time can be rendered short enough in the case of a mixture in order to render their detection more favorable .
that could be crucial in order to build up an experimental protocol to produce and detect cat states within such systems .
we stress that in this work we have chosen to study the symmetric case .
this allowed us to obtain analytical results , while giving rise to the relevant phenomenology which characterizes the physics of the junction .
the general case of different couplings between the two bosonic species and/or different populations needs to resort to numerical calculations and will be the subject of a future publication @xcite .
another interesting issue which deserves further investigation is a careful analysis of the quantum manifestations of the self - trapping transition and in general of the mqst phenomenon in this more general context .
the complex dynamics of the generalized bose josephson junctions investigated in the present paper could be experimentally testable within the current technology .
for instance , the jila group recently @xcite succeeded in producing a mixture of @xmath4 and @xmath5 atoms , whose interactions are widely tunable via feshbach resonances . in particular
it is possible to fix the scattering length of @xmath5 as well as the interspecies one and to tune the scattering length of @xmath4 .
that allows one to explore the parameter space in a wide range and also to realize the symmetric regime @xmath271 .
because of the high degree of experimental control , such a setup could be employed to reproduce the phenomenology described in this work .
the authors would like to thank m. salerno for driving their attention on the topic of bose josephson junctions and e. orignac and a. minguzzi for discussions and for a critical reading of the manuscript .
the eigenvectors of the full hamiltonian ( [ m33 ] ) , up to order @xmath158 , are : @xmath273 } { 2}}% \left| 0,2^{+}\right\rangle , \label{m41}\ ] ] @xmath274 \right ) \left| 0,1^{-}\right\rangle + \frac{k}{6\lambda } % \sqrt{\left [ j_{2}\left ( j_{2}+1\right ) -2\right ] } \left| 0,2^{-}\right\rangle \nonumber \\ & & + \frac{k^{2}}{96\lambda ^{2}}\sqrt{\frac{\left [ j_{2}\left ( j_{2}+1\right ) -2\right ] \left [ j_{2}\left ( j_{2}+1\right ) -6\right ] } { 2}}\left| 0,3^{-}\right\rangle , \label{m42}\end{aligned}\ ] ] @xmath275 \right ) \left| 0,1^{+}\right\rangle -\frac{k}{\lambda } % \sqrt{\frac{j_{2}\left ( j_{2}+1\right ) } { 2}}\left| 0,0\right\rangle + \frac{k}{6\lambda } \sqrt{\left [ j_{2}\left ( j_{2}+1\right ) -2\right ] } \left| 0,2^{+}\right\rangle \nonumber\\ & & + \frac{k^{2}}{96\lambda ^{2}}\sqrt{\frac{\left [ j_{2}\left ( j_{2}+1\right ) -2\right ] \left [ j_{2}\left ( j_{2}+1\right ) -6\right ] } { 2}}\left| 0,3^{+}\right\rangle , \label{m43}\end{aligned}\ ] ] @xmath276 \right ) \left| 0,\frac{1}{2}% ^{\pm } \right\rangle + \frac{k^{2}}{48\lambda ^{2}}\sqrt{\left [ j_{2}\left ( j_{2}+1\right ) -\frac{3}{4}\right ] \left [ j_{2}\left ( j_{2}+1\right ) -\frac{% 15}{4}\right ] } \left| 0,\frac{5}{2}^{\pm } \right\rangle \nonumber \\ & & + \left [ \frac{k}{4\lambda } \sqrt{j_{2}\left ( j_{2}+1\right ) -\frac{3}{4}}%
\mp \frac{k^{2}}{16\lambda ^{2}}\sqrt{\left [ j_{2}\left ( j_{2}+1\right ) + % \frac{1}{4}\right ] \left [ j_{2}\left ( j_{2}+1\right ) -\frac{3}{4}\right ] } % \right ] \left| 0,\frac{3}{2}^{\pm } \right\rangle , \label{m44}\end{aligned}\ ] ] @xmath277 where the coefficients are defined as : @xmath278 @xmath279 @xmath280 99 m. h. anderson , j. r. ensher , m. r. matthews , c. e. wieman , e. a. cornell , _ science _ * 269 * ( 1995 ) 198 ; c. c. bradley , c. a. sackett , j. j. tollett , r. g. hulet , _ phys
* 75 * ( 1995 ) 1687 ; k. b. davis , m. o. mewes , m. r. andrews , n. j. van druten , d. s. durfee , d. m. kurn , w. ketterle , _ phys .
* 75 * ( 1995 ) 3969 .
m. albiez , r. gati , j. folling , s. hunsmann , m. cristiani , m. k. oberthaler , _ phys .
lett . _ * 95 * ( 2005 ) 010402 ; r. gati , m. albiez , j. folling , b. hemmerling , m. k. oberthaler , _ appl .
b _ * 82 * ( 2006 ) 207 .
j. catani , l. de sarlo , g. barontini , f. minardi , m. inguscio , _ phys .
a _ * 77 * ( 2008 ) 011603(r ) ; p. buonsante , s. m. giampaolo , f. illuminati , v. penna , a. vezzani , _ phys . rev .
* 100 * ( 2008 ) 240402 .
a. smerzi , s. fantoni , s. giovanazzi , s. r. shenoy , _ phys .
* 79 * ( 1997 ) 4950 ; s. raghavan , a. smerzi , s. fantoni , s. r. shenoy , _ phys . rev . a _ * 59 * ( 1999 ) 620 ; s. giovanazzi , a. smerzi , s. fantoni , _ phys .
lett . _ * 84 * ( 2000 ) 4521 . | we study the quantum behaviour of a binary mixture of bose - einstein condensates ( bec ) in a double - well potential starting from a two - mode bose - hubbard hamiltonian .
we focus on the small tunneling amplitude regime and apply perturbation theory up to second order .
analytical expressions for the energy eigenvalues and eigenstates are obtained .
then the quantum evolution of the number difference of bosons between the two potential wells is fully investigated for two different initial conditions : completely localized states and coherent spin states . in the first case
both the short and the long time dynamics is studied and a rich behaviour is found , ranging from small amplitude oscillations and collapses and revivals to coherent tunneling . in the second case the short - time scale evolution of number difference
is determined and a more irregular dynamics is evidenced . finally , the formation of schroedinger cat states is considered and shown to affect the momentum distribution . |
solar activity can be seen through the evolution of sunspots number in quasi - periodic oscillatory series with periods going from 8 to 15 years and with a mean period of 11 years . due to the change of magnetic field polarity in solar hemispheres alternatively
each cycle , the period is rather 22 years .
the quasi - periodic evolution of many activity phenomena is still a unsolved key problem in solar physics ( along with , e.g. , heating of the solar chromosphere and corona , and solar flares ) .
an important issue of this understanding is due to the influence of solar activity over the terrestrial climate ( ( * ? ? ?
* archibald 2006 ) ) .
many physical models were proposed to understand the basic mechanisms ( e.g. , ( * ? ? ?
* benevolenskaya 1998 ) ) producing solar / stellar activity and its quasi - cyclic evolution . from sunspots time series , mathematical approaches have highlighted several other hidden periods other that the well - apparent 11-years period .
these works contributed to provide some observed parameters to constraint theoretical models , and to predict the future evolution of the sunspots number ( e.g. , ( * ? ? ?
* clilverd 2003 ) ; ( * ? ? ?
* sello 2003 ) ) . in this work
we follow another mathematical approach , investigating the influence of the past evolution of the sunspots number over the present one .
a natural and simple way to carry out this analysis is to assume that the susnpots number have a temporal delay behaviour .
thus , the relationship between present and past values becomes non - linear and could prove that the past amplitude influence the present one . as for the abovementioned methods , this one also allows some predictions about future solar cycles by using intermediate parameters to characterize the general past evolution .
we consider a temporal delay behavior ( eq . [ eq1 ] ) to rely the present evolution of a phenomena with some particular events in its past evolution .
the variation of @xmath0 in time is not only related with its current value @xmath1 but also with its past values @xmath2 , where @xmath3 is our temporal delay .
the simplest way is to assume only one past temporal influence and some proportionality between the variables ( e.g. , ( * ? ? ?
* murray 1993 ) ) , as follows : @xmath4 because we use yearly values of the relative sunspots number , given by sdic ( http:// www.sidc.be/sunspot-data/ ) , @xmath5 is equal to @xmath6 .
the present and past values of @xmath0 are already known and we assumed some values for the parameter @xmath3 ( @xmath7 ; fig .
[ fig : ratiofactor ] , left panel ) .
so we can determine constants @xmath8 and @xmath9 ( eq . [ eq2 ] ) from the correlation between the sunspots number ratio and the past sunspots number ( fig .
[ fig : ratiofactor ] right panel ) .
the @xmath3 values which minimize the correlation between the sunspots number ratio and the past spots number , are 7 or 8 years ( fig .
[ fig : ratiofactor ] ) . to test the accuracy of our method
, we applied it on the solar cycle 23 ( table [ tab : cycle23 ] and fig .
[ fig : evolspots ] , left panel ) .
our predictions are consistent with the observations , except for the unusually long period of low activity at the end of this cycle .
we also characterized the next solar cycle @xmath11 ( table [ tab : prediction ] and fig .
[ fig : evolspots ] ) using our modelling of the solar cycle @xmath12 as well as the spots number observed until @xmath13 and @xmath14 .
the next maximum sunspots number would take place between @xmath15 and @xmath16 and should be close to 60 ( fig .
[ fig : evolspots ] , right panel ) as for the solar cycle @xmath17 . using
the solar minimum occurred in @xmath18 ( @xmath19 from sdic ) , the predicted maximum is consistent with the value from the linear least - square fit on the both axis ( red line on fig .
[ fig : rminrmax ] ) and is a few smaller than the range given by ( * ? ? ?
* brajsa ( 2009 ) ) .
moreover the next solar minimum should not occur before 2019 or even 2020 .
lcccccc panel of & & & + fig .
[ fig : evolspots ] & & & + & epoch & number & epoch & number & epoch & number + left & @xmath13 & @xmath27 & @xmath28 & @xmath29 & @xmath30 & @xmath31 + middle & @xmath13 & @xmath32 & @xmath33 & @xmath34 & @xmath35 & @xmath36 + right & @xmath37 & @xmath38 & @xmath33 & @xmath39 & @xmath40 & @xmath41 + these preliminary results are encouraging because we find a similar delay as that observed between the geomagnetic activity and solar cycle peaks ( * ? ? ?
* ( hathaway & wilson 2006 ) ) linked probably with magnetic activity by some kind of histeresis mechanism . at present , we can successfully reproduce some previous solar cycles ( e.g. , epoch and solar maximum ) .
we plan to improve our predictive method including the influence of delays for which there is a good correlation .
this more detailed description could allow to solve the overestimation of the next solar minimum and to reproduce the asymmetries observed during the previous cycles .
our final aim is to obtain a reliable prediction of the whole solar activity by identifying the fundamental lower and high activity precursors of its present and future evolution .
all these improvements are still needed to be able to predict future low solar activity periods ( e.g. , maunder / dalton minimum ) . | the so - called is generally characterized by the quasi - periodic oscillatory evolution of the photospheric spots number .
this quasi - periodic pattern has always been an intriguing question .
several physical models were proposed to explain this evolution and many mathematical data analysis were employed to determine the principal frequencies noticeable in the measured data .
both approaches try to predict the future evolution of the solar activity and to understand the physical phenomena producing these cycles . here
we present the analysis of the sunspots number evolution using the time - delay approach .
our results show than the solar cycle can also be characterized by this behavior implying the influence of the past evolution over the present one , suggesting an histeresis mechanism , linked probably with magnetic activity . |
on @xmath0 , consider @xmath1 a bounded potential and the schrdinger operator @xmath2 defined by , for @xmath3 , @xmath4 the potentials @xmath1 we will deal with are * @xmath1 periodic ; * @xmath5 random e.g. anderson model i.e. the entries of the diagonal matrix @xmath1 are independent identically distributed non constant random variable .
the spectral theory of such models has been studied extensively ( see e.g. @xcite ) and it is well known that * when @xmath1 is periodic , the spectrum of @xmath6 is purely absolutely continuous ; * when @xmath5 is random , the spectrum of @xmath6 is almost surely pure point i.e. the operator only has eigenvalues ; moreover , the eigenfunctions decay exponentially at infinity .
pick @xmath7 .
the main object of our study is the operator @xmath8 when @xmath9 becomes large . here
, @xmath10 is the integer interval @xmath11 and @xmath12 if @xmath13 and @xmath14 if not .
+ clearly , the essential spectrum of @xmath15 is that of the discrete laplace operator , that is , @xmath16 $ ] , and it is purely absolutely continuous . outside
this absolutely continuous spectrum , @xmath15 has only discrete eigenvalues associated to exponentially decaying eigenfunctions .
+ we are interested in the resonances of the operator @xmath15 . these can be defined as the poles of the meromorphic continuation of the resolvent of @xmath15 through @xmath17 , the continuous spectrum of @xmath15 ( see theorem [ thr:1 ] and e.g. @xcite ) .
the resonance widths , that is the imaginary part of the resonances , play an important role in the large time behavior of @xmath18 , especially the resonances of smallest width that give the leading order contribution ( see @xcite ) .
.1 cm as @xmath19 , @xmath15 converges to @xmath6 in the strong resolvent sense .
thus , it is natural to expect that the differences in the spectral nature between the cases @xmath1 periodic and @xmath1 random should reflect into differences in the behavior of the resonances . as we shall see now ,
this is the case .
assume that @xmath1 is @xmath20-periodic ( @xmath21 ) and does not vanish identically .
consider @xmath2 and let @xmath22 be its spectrum and @xmath23 be its integrated density of states ( see section [ sec : periodic - case ] and e.g. @xcite ) .
[ thr:22 ] there exists @xmath24 , a discrete ( possibly empty ) set of energies in @xmath25 such that , for @xmath26 a compact interval s.t .
@xmath27 , there exist @xmath28 , @xmath29 , @xmath30 a neighborhood of @xmath31 and @xmath32 non vanishing , real analytic function on @xmath1 such that , for @xmath33 s.t .
@xmath34 , one has 1 .
for any interval @xmath35 , one has @xmath36,\ z\text { resonance of } h_l\}}{2l}=\int_j dn(e)+o(1);\ ] ] 2 .
let @xmath37 the resonances of @xmath15 in @xmath38 $ ] ordered by increasing real part ; then , @xmath39 after rescaling their width by @xmath9 , resonances are nicely interspaced points lying on an analytic curve ( see fig .
[ fig:2 - 3 ] ) .
we give a more precise description of the resonances in theorem [ thr:2 ] and propositions [ pro:2 ] and [ pro:3 ] .
in particular , we describe the set of energies @xmath24 and the resonances near these energies : there , the resonances lie further away from the real axis , the maximal distance being of order @xmath40
. theorem [ thr:22 ] only describes the resonances closest to the real axis . in section [ sec : periodic - case ] , we also give results on the resonances located deeper into the lower half of the complex plane .
.5 cm assume that @xmath5 is the anderson model i.e. its entries are i.i.d .
distributed uniformly on @xmath41 $ ] .
consider @xmath42 .
let @xmath43 be its almost sure spectrum , @xmath44 , its density of states ( i.e. the derivative of the integrated density of states , see section [ sec : periodic - case ] and e.g. @xcite ) and @xmath45 , its lyapunov exponent ( see section [ sec : random - case ] and e.g. @xcite ) .
the lyapunov exponent is known to be continuous and positive ( see e.g. @xcite ) ; the density of states satisfies @xmath46 for a.e .
@xmath47 ( see e.g. @xcite ) .
+ define @xmath48 .
we prove [ thr:29 ] pick a compact interval @xmath49 . for @xmath50 , @xmath51-a.s .
, one has @xmath52 \right\}\\=\int_i\min\left(\frac{c}{\rho(e)},1\right)n(e)de .
\end{gathered}\ ] ] thus , most of the resonances of @xmath53 below @xmath31 are exponentially close to the real axis : more precisely , they live in @xmath54 where @xmath55 and @xmath56 .
+ one can also describe the resonances locally .
therefore , fix @xmath57 such that @xmath58 .
let @xmath59 be the resonances of @xmath53 .
we first rescale them .
define @xmath60 consider now the two - dimensional point process @xmath61 we prove [ thr:23 ] the point process @xmath62 converges weakly to a poisson process of intensity @xmath63 in @xmath64 $ ] . in the random case ,
the structure of the ( properly rescaled ) resonances is quite different from that in the periodic case ( see fig .
[ fig:2 - 3 ] ) .
the real parts of the resonances are scaled in such a way that that their average spacing becomes of order one . by theorem [ thr:23 ] ,
the imaginary parts are typically exponentially small ( in @xmath9 ) ; they are rewritten on a logarithmic scale so as to become of order @xmath63 too .
theorem [ thr:23 ] is the analogue for resonances of the well known result on the distribution of eigenvalues and localization centers for the anderson model in the localized phase ( see e.g. @xcite ) . +
as in the case of the periodic potential , theorem [ thr:23 ] only describes the resonances closest to the real axis . in section [ sec : random - case ] , we also give results on resonances located deeper into the lower of the complex plane .
.2cmbesides proving theorems [ thr:22 ] and [ thr:23 ] , the goal of the paper is to describe the statistical properties of the resonances and relate them ( the distribution of the resonances , the distribution of the widths ) to the spectral characteristics of @xmath2 , possibly to the distribution of its eigenvalues ( see e.g. @xcite ) .
+ as they can be analyzed in a very similar way , we will discuss three models : the model @xmath15 defined above , its analogue on the half - line as well as the `` half - infinite '' model on @xmath0 , that is , @xmath65 where @xmath1 is chosen as above , periodic or random .
+ the scattering theory or the closely related questions of resonances for the operator or for closely related one - dimensional models has already been discussed in various works both in the mathematical and physical literature @xcite .
we will make more comments on the literature as we will develop our results .
+ though in the present paper we restrict ourselves to discrete models , it is clear that continuous one - dimensional models can be dealt with essentially using the methods developed in the present paper .
we now turn to our main results a number of which were announced in @xcite .
pick @xmath66 a bounded potential and , for @xmath67 , consider the following operators : * @xmath68 on @xmath69 with dirichlet boundary conditions at 0 ; * @xmath70 on @xmath0 ; * @xmath71 defined in .
[ rem:3 ] for the sake of simplicity , in the half line case , we only consider dirichlet boundary conditions at @xmath14 . but the proofs show that these are not crucial ; any self - adjoint boundary condition at @xmath14 would do and , mutandi mutandis , the results would be the same .
+ note also that , by a shift of the potential @xmath1 , replacing @xmath9 by @xmath72 , we see that studying @xmath73 is equivalent to studying @xmath74 on @xmath0 .
thus , to derive the results of section [ sec : introduction ] from those in the present section , it suffices to consider the models above , in particular , @xmath75 . for the models @xmath76 and @xmath73 , we start with a discussion of the existence of a meromorphic continuation of the resolvent , then , study the resonances when @xmath1 is periodic and finally turn to the case when @xmath1 is random
. + the existence of a meromorphic continuation of the resolvent of @xmath71 depends on the nature of @xmath1 ; so , it will be discussed when specializing to @xmath1 periodic or random .
one proves the well known and simple [ thr:1 ] the operator valued functions @xmath77 and @xmath78 ) admit a meromorphic continuation from @xmath79 to @xmath80 \cup[2,+\infty)\right)$ ] ( see fig . [ fig:1 ] ) with values in the operators from @xmath81 to @xmath82 .
+ moreover , the number of poles of each of these meromorphic continuations in the lower half - plane is at most equal to @xmath9 .
the resonances are defined as the poles of this meromorphic continuation ( see fig .
[ fig:1 ] ) .
we assume that , for some @xmath83 , one has @xmath84 let @xmath85 be the spectrum of @xmath86 acting on @xmath69 with dirichlet boundary condition at @xmath14 and @xmath22 be the spectrum of @xmath87 acting on @xmath0 .
one has the following description for these spectra : * @xmath22 is a union of intervals @xmath88 $ ] for some @xmath89 ( @xmath90 ) ( see e.g. @xcite ) ; the spectrum of @xmath91 is purely absolutely continuous and the spectral resolution can be obtained via a bloch - floquet decomposition ; * on @xmath69 ( see e.g. @xcite ) , one has * * @xmath92 and @xmath22 is the a.c .
spectrum of @xmath6 ; * * the @xmath93 are isolated simple eigenvalues associated to exponentially decaying eigenfunctions .
the operators @xmath94 ( for @xmath95 ) admit an integrated density of states defined by @xmath96\}}{\#(\llbracket - l , l\rrbracket\cap\bullet)}.\ ] ] here , the restriction of @xmath97 to @xmath98 is taken with dirichlet boundary conditions ; this is to fix ideas as it is known that , in the limit @xmath19 , other self - adjoint boundary conditions would yield the same result .
+ the integrated density of states is the same for @xmath99 and @xmath91 ( see e.g. @xcite ) .
it defines the distribution function of some probability measure on @xmath22 that is real analytic on @xmath100 .
let @xmath101 denote the density of states of @xmath99 and @xmath91 , that is , @xmath102 .
+ [ rem:7 ] when @xmath9 get large , as @xmath76 tend to @xmath99 in strong resolvent sense , interesting phenomena should take place near energies in @xmath85 .
+ on the other hand , let @xmath103 be the shift by @xmath104 steps to the left i.e. @xmath105 is the potential @xmath1 shifted @xmath104 times to the left i.e. @xmath106 . then , for @xmath107 when @xmath19 , @xmath108 tend to @xmath91 in strong resolvent sense .
thus , interesting phenomena should happen near energies in @xmath22 .
+ as for @xmath109 , it converges in strong resolvent sense to @xmath71 that we study in section [ sec : half - line - pert ] . in @xmath110 , clearly , @xmath76 and @xmath73 have only discrete eigenvalues .
we will now describe what happens for the resonances near @xmath16 $ ] . in our study of the resonances of @xmath111
, it will be useful to consider the translates of @xmath91 restricted to the negative axis .
more precisely , on @xmath112 ( where @xmath113 ) , consider the operator @xmath114 with dirichlet boundary condition at @xmath14 .
let @xmath115 .
+ one knows that , as for @xmath99 , @xmath116 and it is purely absolutely continuous ( see e.g. ( * ? ? ?
* chapter 7 ) ) .
@xmath117 may also have discrete eigenvalues in @xmath118 .
+ in the case of the whole line , we shall need another auxiliary operator namely @xmath119 considered on @xmath69 with dirichlet boundary conditions at @xmath14 .
its spectral properties are similar to those of @xmath120 .
[ rem:6 ] before describing various properties of the auxiliary operators and some objects related to them , let us explain why they naturally occur in this study
. therefore , define @xmath15 to be the operator @xmath76 restricted to @xmath121 with dirichlet boundary conditions at @xmath9 .
note that @xmath15 is also the operator @xmath109 restricted to @xmath121 with dirichlet boundary conditions at @xmath14 and @xmath9 .
+ the operator @xmath122 is the strong resolvent limit of @xmath15 when @xmath19 .
if @xmath123 denotes the translation by @xmath124 that unitarily maps @xmath125 into @xmath126 , then , @xmath127 converges in the strong resolvent sense to @xmath120 when @xmath19 and @xmath128 . indeed , as @xmath1 is @xmath20 periodic , @xmath129 . to describe the resonances of @xmath130
, we introduce the function @xmath131 : for @xmath132 , define @xmath133 then , one has [ pro:1 ] for @xmath134 , @xmath135 is a positive measure that is absolutely continuous on @xmath22 . moreover ,
its density , say , @xmath136 is real analytic on @xmath100 and there exists @xmath137 a positive real analytic function such that , on @xmath100 , one has @xmath138 . for @xmath139 ,
define @xmath140 the existence and analyticity of the cauchy principal value @xmath141 on @xmath100 is guaranteed by the analyticity of @xmath142 in @xmath100 ( see e.g. @xcite ) .
moreover , one has @xmath143 in the lower half - plane @xmath144im@xmath145 , define the function @xmath146 where * in the first formula , the function @xmath147 is the analytic continuation to the lower half - plane of the determination taking values in @xmath148 $ ] over the interval @xmath149 $ ] ; * in the second formula , the branch of the square root @xmath150 has positive imaginary part for @xmath151 .
the function @xmath152 is analytic in @xmath144im@xmath145 and in a neighborhood of @xmath25 .
moreover , one proves [ pro:4 ] @xmath152 vanishes identically if and only if @xmath1 vanishes identically .
moreover , if @xmath153 , there exists @xmath154 and @xmath155 such that @xmath156 .
thus , if @xmath153 ( which we assume from now on ) , in @xmath144im@xmath145 and on @xmath25 , @xmath157 has only finitely many zeros , each of finite multiplicity .
+ recall that @xmath158 on @xmath69 with dirichlet boundary condition at @xmath14 .
define the function @xmath159 as the distribution function of the measure defined by @xmath160 in the same way as we have defined @xmath161 , @xmath162 and @xmath163 from @xmath120 , one can define @xmath164 , @xmath165 and @xmath166 from @xmath122 .
they also satisfy proposition [ pro:1 ] , relation and proposition [ pro:4 ] .
+ for the description of the resonances , it will be convenient to define the following functions @xmath167 and @xmath168 one has [ pro:5 ] pick @xmath95 .
let @xmath26 be a compact interval .
+ there exists a neighborhood of @xmath31 such that , in this neighborhood , the function @xmath169 is analytic and has a positive imaginary part .
+ the function @xmath170 ( resp .
@xmath171 ) takes the value @xmath172 only at the zeros of @xmath163 ( resp .
@xmath173 ) .
we shall see that the the zeros of @xmath174 play a special role for the resonances of @xmath111 .. 1 cm we start with a description of resonance free regions near the real axis . [ thr:5 ] let @xmath31 be a compact interval in @xmath17 . then , 1 . if @xmath175 ( resp .
@xmath176 ) , then , there exists @xmath177 such that , for @xmath9 sufficiently large , @xmath76 ( resp .
@xmath73 ) has no resonances in the rectangle @xmath178 ; 2 .
if @xmath179 , then , there exists @xmath177 such that , for @xmath9 sufficiently large , sufficiently large , @xmath76 and @xmath73 have no resonances in the rectangle @xmath180 ; 3 .
fix @xmath181 and assume the compact interval @xmath31 to be such that @xmath182 and @xmath183 ( @xmath184 are defined in the beginning of section [ sec : periodic - case ] ) : 1 .
if @xmath185 then , there exists @xmath50 such that , for @xmath9 sufficiently large such that @xmath186 , @xmath76 has a unique resonance in the rectangle @xmath187 ; moreover , this resonance , say @xmath188 , is simple and satisfies @xmath189 and @xmath190 for some @xmath191 independent of @xmath9 ; 2 . if @xmath192 then , there exists @xmath50 such that , for @xmath9 sufficiently large such that @xmath186 , @xmath76 has no resonance in the rectangle @xmath193 .
so , below the spectral interval @xmath17 , there exists a resonance free region of width at least of order @xmath194 ; it is of order @xmath63 . for @xmath195 , if @xmath186 , each discrete eigenvalue of @xmath99 that is not an eigenvalue of @xmath117 generates a resonance for @xmath76 exponentially close to the real axis ( when @xmath9 is large ) . when the eigenvalue of @xmath120 is also an eigenvalue of @xmath196
, it may also generate a resonance but only much further away in the complex plane , at least at a distance of order @xmath63 to the real axis .
let @xmath197 be the eigenvalues of @xmath15 , that is , of @xmath76 ( or @xmath73 ) restricted to @xmath121 with dirichlet boundary conditions listed in increasing order ; they are described in theorem [ thr:16 ] . + pick a compact interval @xmath26 . for @xmath95 and @xmath198 , for @xmath9 large , define the complex number @xmath199\ ] ] where the determination of @xmath200 is the inverse of the determination @xmath201 mapping @xmath202 onto @xmath203 . + note that , by proposition [ pro:5 ] , for @xmath9 sufficiently large , we know that , for any @xmath204 such that @xmath198 , one has @xmath205 and @xmath206 \in(0,+\infty)\setminus\{1\}.\ ] ] thus , @xmath207 is well defined by . moreover , as the zeros of @xmath208 are of finite order , one checks that @xmath209 where the constants are uniform for @xmath204 such that @xmath198 . + we prove the [ thr:2 ] pick @xmath95 and @xmath134 .
let @xmath26 be a compact interval .
+ then , there exists @xmath210 and @xmath29 such that , for @xmath33 satisfying @xmath211 , for each @xmath198 , there exists a unique resonance of @xmath111 , say @xmath212 , in the rectangle @xmath213 + i\left[-\eta_0,0\right]$ ] ; this resonance is simple and it satisfies @xmath214 .
this result calls for a few comments .
first , the picture one gets for the resonances is the following ( see also figure [ fig:10 ] ) .
as long as @xmath215 stays away from any zero of @xmath208 , the resonances are nicely spaced points as the following proposition proves .
[ pro:2 ] pick @xmath95 and @xmath134 .
let @xmath26 be a compact interval not containing any zero of @xmath208 .
+ then , for @xmath9 sufficiently large , for each @xmath198 , the resonance @xmath212 admits a complete asymptotic expansion in powers of @xmath194 and one has @xmath216 where the remainder term is uniform in @xmath204 .
proposition [ pro:2 ] implies theorem [ thr:22 ] . in theorem [ thr:22 ]
, we chose @xmath217 .
the set @xmath24 of exceptional points in theorem [ thr:22 ] is exactly the set of zeros of the function @xmath218 in @xmath17 .
near the zeros of @xmath218 , the resonances take a `` plunge '' into the lower half of complex plane and their imaginary part becomes of order at most @xmath40 .
indeed , theorem [ thr:2 ] and imply [ pro:3 ] pick @xmath95 and @xmath134 .
let @xmath219 be a zero of @xmath208 of order @xmath220 .
+ then , for @xmath221 , for @xmath9 sufficiently large , if @xmath204 is such that @xmath222 , the resonance @xmath212 satisfies @xmath223 where the remainder term is uniform in @xmath204 such that @xmath222 . ) , height=151 ] when @xmath224 , the asymptotic shows that there can be a `` resonance '' phenomenon for resonances : when the two functions @xmath163 and @xmath166 share a zero at the same real energy , the maximal width of the resonances increases ; indeed , the factor in front of @xmath40 is proportional to the multiplicity of the zero of @xmath173 .
+ the resonances found in theorem [ thr:2 ] are not necessarily the only ones : deeper into the lower complex plane , one may find more that we describe in the next section .
they are related the zeros of @xmath163 when @xmath225 and @xmath173 when @xmath224 .
we now study what happens below the line @xmath144im@xmath226 for @xmath76 and @xmath73 .
+ the functions @xmath163 and @xmath166 are analytic in the lower half plane and , by proposition [ pro:4 ] , they do nt vanish in an neighborhood of @xmath227 .
hence , for @xmath31 chosen as in theorem [ thr:2 ] , the function @xmath208 has only finitely many zeros in @xmath228 .
+ we prove [ thr:3 ] pick @xmath95 and @xmath134 .
pick @xmath31 as in theorem [ thr:2 ] .
let @xmath229 be the zeros of @xmath208 in @xmath228 .
+ there exists @xmath230 such that , for @xmath9 sufficiently large s.t .
@xmath128 , one has , * in the rectangle @xmath231 $ ] , the only resonances of @xmath76 and @xmath109 are those given by theorem [ thr:2 ] ; * in the strip @xmath232 $ ] , the resonances of @xmath233 lie in @xmath234 * in @xmath235 , the number of resonances ( counted with multiplicity ) is equal to the order of @xmath236 as a zero of @xmath208 .
we see that the total number of resonances below @xmath25 that do not tend to the real axis is finite .
fix @xmath21 . on @xmath0
, we now consider the operator @xmath237 where @xmath238 for @xmath239 and @xmath240 for @xmath241 .
we prove [ thr:25 ] the resolvent of @xmath71 can be analytically continued from the upper half - plane through @xmath242 to the lower half plane .
the resulting operator does not have any poles in the lower half - plane or on @xmath242 .
+ the resolvent of @xmath71 can be analytically continued from the upper half - plane through @xmath243 ( resp .
$ ] ) to the lower half plane ; the poles of the continuation through @xmath243 ( resp .
$ ] ) are exactly the zeros of the function @xmath245 when continued from the upper half - plane through @xmath243 ( resp .
$ ] ) to the lower half - plane . using the same arguments as in the proof of proposition [ pro:4 ]
, one easily sees that the continuations of the function @xmath246 to the lower half plane through @xmath243 and @xmath244 $ ] have at most finitely many zeros and that these zeros are away from the real axis . +
this also implies that the spectrum on @xmath71 in @xmath16\cup\sigma_{\mathbb{z}}$ ] is purely absolutely continuous except possibly at the edges of @xmath22 and @xmath247 .
we now turn to the random case .
let @xmath5 where @xmath248 are bounded independent and identically distributed random variables .
assume that the common law of the random variables admits a bounded compactly supported density , say , @xmath249 . + set @xmath250 on @xmath69 ( with dirichlet boundary condition at @xmath14 to fix ideas ) .
let @xmath251 be the spectrum of @xmath252 .
consider also @xmath253 acting on @xmath0 .
then , one knows ( see e.g. @xcite ) that , @xmath51 almost surely , @xmath254+\text{supp}\,g\ ] ] one has the following description for the spectra @xmath251 and @xmath255 : * @xmath51-almost surely , @xmath256 ; the spectrum is purely punctual ; it consists of simple eigenvalues associated to exponentially decaying eigenfunctions ( anderson localization , see e.g. @xcite ) ; one can prove that , under the assumptions made above , the whole spectrum is dynamically localized ( see e.g. @xcite and references therein ) ; * for @xmath252 ( see e.g. @xcite ) , one has * * @xmath51-almost surely , @xmath257 ; * * @xmath43 is the essential spectrum of @xmath252 ; it consists of simple eigenvalues associated to exponentially decaying eigenfunctions ; * * the set @xmath258 is the discrete spectrum of @xmath252 ; it may be empty and depends on @xmath51 .
it is well known ( see e.g. @xcite ) that the integrated density of states of @xmath6 , say @xmath259 , is defined as the following limit @xmath260\}}{2l+1}.\ ] ] the above limit does not depend on the boundary conditions used to define the restriction @xmath261 .
it defines the distribution function of a probability measure supported on @xmath43 . under our assumptions on the random potential , @xmath262
is known to be lipschitz continuous ( @xcite ) .
let @xmath263 be its derivative ; it exists for almost energy .
if one assumes more regularity on @xmath249 the density of the random variables @xmath264 , then @xmath101 can be shown to exist everywhere and to be regular ( see e.g. @xcite ) .
+ one also defines the lyapunov exponent , say @xmath265 as follows @xmath266 where @xmath267 for any @xmath268 , @xmath51-almost surely , the lyapunov exponent is known to exist and to be independent of @xmath51 ( see e.g. @xcite ) .
it is positive at all energies .
moreover , by the thouless formula @xcite , it is positive and continuous for all @xmath268 and it is the harmonic conjugate of @xmath269 .. 1cmfor @xmath95 , we now define @xmath270 to be the operator @xmath271}$ ] .
the goal of the next sections is to describe the resonances of these operators in the limit @xmath19 .
we again start with a description of the resonance free region near @xmath17 . as in the periodic case ,
the size of the @xmath270-resonance free region below a given energy will depend on whether this energy belongs to @xmath272 or not .
we prove [ thr:4 ] fix @xmath95 .
let @xmath31 be a compact interval in @xmath17 .
then , @xmath51-a.s .
, one has 1 . for @xmath95 ,
if @xmath273 , then , there exists @xmath177 such that , for @xmath9 sufficiently large , there are no resonances of @xmath270 in the rectangle @xmath274 ; 2 .
if @xmath275 , then , for @xmath276 , there exists @xmath29 such that , for @xmath277 , there are no resonances of @xmath270 in the rectangle @xmath278 where * @xmath279 is the maximum of the lyapunov exponent @xmath265 on @xmath31 * @xmath280 3 .
pick @xmath281 and assume that @xmath282 and @xmath283 , then , there exists @xmath50 such that , for @xmath9 sufficiently large , @xmath284 has a unique resonance in @xmath285 ; moreover , this resonance , say @xmath188 , is simple and satisfies @xmath286 and @xmath287 for some @xmath288 independent of @xmath9 . when comparing point ( 2 ) of this result with point ( 2 ) of theorem [ thr:5 ] ,
it is striking that the width of the resonance free region below @xmath43 is much smaller in the random case ( it is exponentially small in @xmath9 ) than in the periodic case ( it is polynomially small in @xmath9 ) .
this a consequence of the localized nature of the spectrum i.e. of the exponential decay of the eigenfunctions of @xmath289 .
we will now see that below the resonance free strip exhibited in theorem [ thr:4 ] one does find resonances , actually , many of them .
we prove [ thr:6 ] fix @xmath95 .
let @xmath31 be a compact interval in @xmath290 . then , 1 . for any @xmath291 , @xmath51-a.s .
, one has @xmath292 2 . for @xmath293 such that @xmath46 and @xmath294 ,
define the rectangle @xmath295 where @xmath296 is defined in theorem [ thr:4 ] , and @xmath297 then , for @xmath293 such that @xmath46 and @xmath298 $ ] , @xmath51-a.s .
, one has @xmath299 3 . for @xmath50 , @xmath51-a.s .
, one has @xmath300
\right\}\\=\int_i\min\left(\frac{c}{\rho(e)},1\right)n(e)de .
\end{gathered}\ ] ] the striking fact is that the resonances are much closer to the real axis than in the periodic case ; the lifetime of these resonances is much larger .
the resonant states are quite stable with lifetimes that are exponentially large in the width of the random perturbation .
point ( 3 ) is an integrated version of point(2 ) .
let us also note here that when @xmath224 , point ( 3 ) of theorem [ thr:6 ] is the statement of theorem [ thr:29 ] .
+ note that the rectangles @xmath301 are very stretched along the real axis ; their sidelength in imaginary part is exponentially small whereas their sidelength in real part is of order @xmath63 ( in @xmath9 ) .
+ to understand point ( 2 ) , rescale the resonances of @xmath270 , say , @xmath302 as follows @xmath303 then , the rectangle @xmath304 is mapped into * the rectangle @xmath305 if @xmath294 , * the rectangle @xmath306 if @xmath307 .
thus , the denominator of the quotient in is just the volume of the rescaled @xmath301 .
so , point ( 2 ) states that in the limit @xmath308 and @xmath309 small and @xmath9 large , the rescaled resonances become uniformly distributed in the rescaled rectangles .
we see that the structure of the set of resonances is very different from the one observed in the periodic case ( see fig .
[ fig:2 - 3 ] ) .
we will now zoom in on the resonance even more so as to make this structure clearer .
therefore , we consider the two - dimensional point process @xmath310 defined by @xmath311 where @xmath312 and @xmath313 are defined by . + we prove [ thr:7 ] fix @xmath314 such that @xmath46 .
then , the point process @xmath310 converges weakly to a poisson process in @xmath315 $ ] with intensity @xmath63 .
that is , for any @xmath316 , if @xmath317 resp .
@xmath318 , are disjoint intervals of the real line @xmath319 resp . of @xmath41 $ ] , then @xmath320 where @xmath321 for @xmath322 .
this is the analogue of the celebrated result on the poisson structure of the eigenvalues for a random system ( see e.g. @xcite ) .
+ when considering the model for @xmath224 , theorem [ thr:7 ] is theorem [ thr:23 ] .
.1cmin @xcite , we proved decorrelation estimates that can be used in the present setting to prove [ thr:8 ] fix @xmath314 and @xmath323 such that @xmath324 , @xmath46 and @xmath325 .
then , the limits of the processes @xmath310 and @xmath326 are stochastically independent . due to the rescaling , the above results give only a picture of the resonances in a zone of the type @xmath327 -i\,\left[e^{-2\eta_\bullet(1+\varepsilon)\rho(e)l } , e^{-2\varepsilon\eta_\bullet\rho(e)l}\right]\ ] ] for @xmath328 arbitrarily small .
+ when @xmath9 gets large , this rectangle is of a very small width and located very close to the real axis .
theorems [ thr:6 ] , [ thr:7 ] and [ thr:8 ] describe the resonances lying closest to the real axis .
as a comparison between points ( 1 ) and ( 2 ) in theorem [ thr:6 ] shows , they are the most numerous .
+ one can get a number of other statistics ( e.g. the distribution of the spacings between the resonances ) using the techniques developed for the study of the spectral statistics of the eigenvalues of a random system in the localized phase ( see @xcite ) combined with the analysis developed in section [ sec : random - case-2 ] .
it is natural to question what happens deeper in the complex plane . to answer this question , fix an increasing sequence of scales @xmath329 such
that @xmath330 we first show that there are only few resonances below the line @xmath144im@xmath331 , namely [ thr:27 ] pick @xmath332 a sequence of scales satisfying and @xmath31 as above .
+ @xmath51 almost surely , for @xmath9 large , one has @xmath333 as we shall show now , after proper rescaling , the structure of theses resonances is the same as that of the resonances closer to the real axis .
+ fix @xmath293 so that @xmath46 .
recall that @xmath334 be the resonances of @xmath53 .
we now rescale the resonances using the sequence @xmath332 ; this rescaling will select resonances that are further away from the real axis .
define @xmath335 consider now the two - dimensional point process @xmath336 we prove the following analogue of the results of theorems [ thr:6 ] , [ thr:7 ] and [ thr:8 ] for resonances lying further away from the real axis .
[ thr:9 ] fix @xmath314 and @xmath323 such that @xmath324 , @xmath46 and @xmath325 .
fix a sequence of scales @xmath332 satisfying .
then , one has 1 . for @xmath298 $ ]
, @xmath51-almost surely @xmath337 where @xmath301 is defined in theorem [ thr:6 ] ; 2 .
the point processes @xmath338 and @xmath339 converge weakly to poisson processes in @xmath340 of intensity @xmath63 ; 3 .
the limits of the processes @xmath338 and @xmath339 are stochastically independent .
point ( 1 ) shows that , in , one actually has @xmath341 notice also that the effect of the scaling is to select resonances that live in the rectangle @xmath342 -i\,\left[e^{-2\eta_\bullet(1+\varepsilon)\rho(e)\ell_l } , e^{-2\varepsilon\eta_\bullet\rho(e)\ell_l}\right]\ ] ] this rectangle is now much farther away from the real axis than the one considered in section [ sec : descr - reson1 ] .
+ modulo rescaling , the picture one gets for resonances in such rectangles is the same one got before .
this description is valid almost all the way from distances to the real axis that are exponentially small in @xmath9 up to distances that are polynomially small ( see ) .
one can also study the resonances that are even further away from the real axis in a way similar to what was done in the periodic case in section [ sec : descr - reson - ll ] .
define the following random potentials on @xmath343 and @xmath344 @xmath345\\ \omega_{l - n}\text { for } [ l/2]+1\leq n\leq l\\ 0\text { for } l+1\leq n \end{cases } \end{aligned}\ ] ] where @xmath346 and @xmath347 are i.i.d . and satisfy the assumptions of the beginning of section [ sec : random - case ] .
+ consider the operators * @xmath348 on @xmath69 with dirichlet boundary condition at @xmath14 , * @xmath349 on @xmath0 . clearly , the random operator @xmath350 ( resp .
@xmath351 ) has the same distribution as @xmath352 ( resp .
@xmath353 ) . thus , for the low lying resonances
, we are now going to describe those of @xmath350 ( resp .
@xmath351 ) instead of those of @xmath352 ( resp .
@xmath353 ) . [ rem:4 ]
the reason for this change of operators is the reason why , in the case of the periodic potential , we had to distinguish various auxiliary operators depending on the congruence of @xmath9 modulo @xmath20 , the period : this gives a meaning to the limiting operators when @xmath19 .
define the probability measure @xmath354 using its borel transform by , for im@xmath355 , @xmath356 consider the function @xmath357 where the determinations of @xmath147 and @xmath150 are those described after .
+ this random function @xmath358 is the analogue of @xmath163 in the periodic case .
one proves the analogue of proposition [ pro:4 ] [ pro:6 ] if @xmath359 , one has @xmath360 .
thus , @xmath51 almost surely , @xmath358 does not vanish identically in @xmath144im@xmath145 .
+ pick @xmath361 compact . then , @xmath51 almost surely , the number of zeros of @xmath358 ( counted with multiplicity ) in @xmath362 $ ] is asymptotic to @xmath363 as @xmath364 ; moreover , @xmath51 almost surely , there exists @xmath365 such that all the zeroes of @xmath358 in @xmath366 are simple .
it seems reasonable to believe that except for the zero at @xmath227 , @xmath51 almost surely all the zeros of @xmath358 are simple ; we do not prove it + for the `` deep '' resonances , we then prove [ thr:21 ] fix @xmath361 a compact interval .
there exists @xmath50 such that , with probability 1 , for @xmath9 sufficiently large , there exists @xmath367 such that 1 . for each resonance of @xmath350 ( resp .
@xmath368 ) in @xmath369 $ ] , say @xmath268 , there exists a unique zero of @xmath358 ( resp .
@xmath370 ) , say @xmath371 , such that @xmath372 ; 2 .
reciprocally , to each zero of @xmath358 ( resp .
@xmath370 ) in @xmath373 $ ] , say @xmath374 , one can associate a unique resonance of @xmath350 ( resp .
@xmath375 ) , say @xmath268 , such that @xmath372 .
one can combine this result with the description of the asymptotic distribution of the resonances given by theorem [ thr:9 ] to obtain the asymptotic distributions of the zeros of the function @xmath358 near a point @xmath376 when @xmath377 .
indeed , let @xmath378 be the zeros of @xmath358 in @xmath144im@xmath145 .
rescale the zeros : @xmath379 and consider the two - dimensional point process @xmath380 defined by @xmath381 then , one has [ cor:3 ] fix @xmath293 such that @xmath46 . then , the point process @xmath380 converges weakly to a poisson process in @xmath382 with intensity @xmath63 .
the function @xmath358 has been studied in @xcite where the average density of its zeros was computed . here
, we obtain a more precise result .
fix @xmath21 . on @xmath0
, we now consider the operator @xmath383 where @xmath384 for @xmath239 and @xmath385 for @xmath241 and @xmath386 are i.i.d . and have the same distribution as above .
recall that @xmath43 is the almost sure spectrum of @xmath387 ( on @xmath0 ) .
we prove [ thr:26 ] first , @xmath51 almost surely , the resolvent of @xmath388 does not admit an analytic continuation from the upper half - plane through @xmath389 to any subset of the lower half plane . nevertheless , @xmath51-almost surely , the spectrum of @xmath390 in @xmath389 is purely absolutely continuous .
+ second , @xmath51 almost surely , the resolvent of @xmath388 does admit a meromorphic continuation from the upper half - plane through @xmath391 to the lower half plane ; the poles of this continuation are exactly the zeros of the function @xmath392 when continued from the upper half - plane through @xmath391 to the lower half - plane .
+ third , @xmath51 almost surely , the spectrum of @xmath388 in @xmath393 $ ] is pure point associated to exponentially decaying eigenfunctions ; hence , the resolvent of @xmath388 can not be be continued through @xmath393 $ ] .
let us also note here that if @xmath394supp@xmath249 ( where @xmath249 is the density of the random variables defining the random potential ) , then , by , one has @xmath16\subset\sigma$ ] . in this case , there is no possibility to continue the resolvent of @xmath388 to the lower half plane passing through @xmath16 $ ] .
+ comparing theorem [ thr:26 ] to theorem [ thr:25 ] , we see that , as the operator @xmath71 , when continued through @xmath389 , the operator @xmath390 does not have any resonances but for very different reasons .
+ when one does the continuation through @xmath391 , one sees that the number of resonances is finite ; `` near '' the real axis , the continuation of the function @xmath395 has non trivial imaginary part and near @xmath396 it does not vanish .
+ theorem [ thr:26 ] also shows that the equation studied in @xcite , i.e. the equation @xmath397 , does not describe the resonances of @xmath388 as is claimed in these papers : these resonances do not exist as there is no analytic continuation of the resolvent of @xmath388 through @xmath398 ! as is shown in theorem [ thr:21 ] , the solutions to the equation @xmath397 give an approximation to the resonances of @xmath352 ( see theorem [ thr:21 ] ) .
the proof of theorem [ thr:1 ] follows standard ideas that were first applied in the continuum setting i.e. for partial differential operators instead of finite difference operators ( see e.g. @xcite and references therein ) .
+ the proof relies on the fact that the resolvent of free laplace operator can be continued holomorphically from @xmath79 to @xmath80\cup[2,+\infty)\right)$ ] as an operator valued function from @xmath81 to @xmath82 .
this is an immediate consequence of the fact that , by discrete fourier transformation , @xmath399 is the fourier multiplier by the function @xmath400 .
+ indeed , for @xmath399 on @xmath0 and im@xmath401 , one thus has , for @xmath402 ( assume @xmath403 ) @xmath404 where @xmath405 where the determination @xmath406 is chosen so that im@xmath407 and re@xmath408 for im@xmath401 .
the determination satisfies @xmath409 .
+ the map @xmath410 can continued analytically from @xmath79 to the cut plane @xmath80\cup[2,+\infty)\right)$ ] as shown in figure [ fig:5 ] . ,
height=94 ] + the continuation is one - to - one and onto from @xmath80\cup[2,+\infty)\right)$ ] to @xmath411 .
it defines a determination of @xmath412 .
+ clearly , using , this continuation yields an analytic continuation of @xmath413 from @xmath144im@xmath414 to @xmath80\cup[2,+\infty)\right)$ ] as an operator from @xmath81 to @xmath82 .
.1cmlet us now turn to the half - line operator i.e. @xmath399 on @xmath343 with dirichlet condition at @xmath14 .
pick @xmath268 such that im@xmath401 and set @xmath415 where the determination @xmath406 is chosen as above . if for @xmath416 bounded and @xmath417 , one sets @xmath418 and @xmath419_n&=\frac1{\sin\theta(e)}\sum_{j=-1}^n v_j\cdot\sin((n - j)\theta(e ) ) \\&\hskip1cm - e^{i\theta(e)}\frac{\sin((n+1)\theta(e))}{\sin(\theta(e ) ) } \sum_{j\geq0}e^{ij\theta(e)}v_j . \end{split}\ ] ] then , for im@xmath401 , a direct computations shows that 1 . for @xmath420 ,
the vector @xmath421 is in the domain of the dirichlet laplacian on @xmath69 i.e. @xmath422_{-1}=0 $ ] ; 2 . for @xmath239
, one has @xmath423_{n+1}+[r_0^{\mathbb{n}}(e)(v)]_{n-1}-e[r_0^{\mathbb{n}}(e)(v)]_n = v_n.\ ] ] to prove this , let us just note that @xmath424_n&=\sum_{j=-1}^n v_j\cdot e^{i(n - j)\theta(e)}+ \sum_{j\geq n+1}v_j\cdot e^{-i(n - j)\theta(e ) } \\ & \hskip1cm - e^{i(n+2)\theta(e)}\sum_{j\geq-1 } v_j\cdot e^{ij\theta(e)}. \end{split}\ ] ] 3 .
@xmath425 defines a bounded map from @xmath69 to itself ; thus , @xmath425 is the resolvent of the dirichlet laplacian on @xmath343 at energy @xmath268 for im@xmath401 . using the continuation of @xmath410 , formula yields an analytic continuation of the resolvent @xmath425 as an operator from @xmath81 to @xmath82 .
[ rem:5 ] note that the resolvent @xmath425 at an energy @xmath268 s.t .
im@xmath426 is given by the formulae or where @xmath427 is replaced by @xmath428 . to deal with the perturbation @xmath1
, we proceed in the same way on @xmath344 and on @xmath343 .
set @xmath429 ( seen as a function on @xmath343 or @xmath344 depending on the case ) .
letting @xmath430 be either @xmath431 or @xmath425 , we compute @xmath432 thus , it suffices to check that the operator @xmath433 ( resp .
@xmath434 ) can be analytically continued as an operator from @xmath81 to @xmath81 ( resp .
@xmath82 to @xmath82 ) .
this follows directly from the fact @xmath435 has finite rank .
+ to complete the proof of theorem [ thr:1 ] , we just note that @xmath433 ( resp .
@xmath434 ) being finite rank , @xmath436 ( resp .
@xmath437 ) ) is not invertible if and only if @xmath438 if an eigenvalue of @xmath433 ( resp .
@xmath434 ) . as @xmath439 ( resp .
@xmath440 ) is analytic , this happens at a discrete set of energies or at all energies in @xmath80\cup[2,+\infty)\right)$ ] ( which is connected ) .
as we know that for im@xmath401 , @xmath438 is not an eigenvalue of @xmath433 ( resp .
@xmath434 ) , the set of energies @xmath268 for which @xmath438 is not an eigenvalue of @xmath433 ( resp .
@xmath434 ) is discrete .
+ this completes the proof of the first part of theorem [ thr:1 ] . to prove the second part ,
we will first write a characteristic equation for resonances . the bound on the number of resonances will then be obtained through a bound on the number of solutions to this equation .
the characteristic equation will take different forms depending on whether we deal with the half - line or the full line operator .
but in both cases , the coefficients of the characteristic equation will be constructed from the spectral data ( i.e. the eigenvalues and eigenfunctions ) of the operator @xmath15 ( see remark [ rem:6 ] ) we first consider @xmath76 on @xmath69 and prove [ thr:11 ] consider the operator @xmath15 defined as @xmath76 restricted to @xmath121 with dirichlet boundary conditions at @xmath9 and define * @xmath441 are the dirichlet eigenvalues of @xmath76 ordered so that @xmath442 ; * @xmath443 where @xmath444 is a normalized eigenvector associated to @xmath445 .
then , an energy @xmath268 is a resonance of @xmath195 if and only if @xmath446 the determination of @xmath427 being chosen so that im@xmath447 and re@xmath448 when im@xmath401 .
let us note that @xmath449 by the proof of the first statement of theorem [ thr:1 ] ( see the beginning of section [ sec : proof - theorem ] ) , we know that an energy @xmath268 is a resonance if and only if @xmath438 if an eigenvalue of @xmath433 where @xmath430 is defined by .
pick @xmath268 an resonance and let @xmath450 be a resonant state that is an eigenvector of @xmath433 associated to the eigenvalue @xmath438 . as @xmath451 for @xmath452 , equation yields that , for @xmath452 , @xmath453 for some fixed @xmath454 . as @xmath455 , for @xmath452
, it satisfies @xmath456 .
thus , @xmath457 and by , @xmath458 is a solution to the eigenvalues problem @xmath459 this can be equivalently be rewritten as @xmath460 the matrix in is the dirichlet restriction of @xmath76 to @xmath121 perturbed by the rank one operator @xmath461 .
thus , by rank one perturbation theory ( see e.g. @xcite ) , an energy @xmath268 is a resonance if and only if if satisfies .
+ this completes the proof of theorem [ thr:11 ] .
let us now complete the proof of theorem [ thr:1 ] for the operator on the half - line .
let us first note that , for im@xmath401 , the imaginary part of the left hand side of is positive by .
on the other hand , the imaginary part of the right hand side of is equal to @xmath462re@xmath463 and , thus , is negative ( recall that re@xmath448 ( see fig .
[ fig:1 ] ) .
thus , as already underlined , equation has no solution in the upper half - plane or on the interval @xmath17 .
+ clearly , equation is equivalent to the following polynomial equation of degree @xmath464 in the variable @xmath465 @xmath466 we are looking for the solutions to in the upper half - plane .
as the polynomial in the right hand side of has real coefficients , its zeros are symmetric with respect to the real axis .
moreover , one notices that , by , @xmath14 is a solution to .
hence , the number of solutions to in the upper half - plane is bounded by @xmath9 .
this completes the proof of theorem [ thr:1 ] .
now , consider @xmath73 on @xmath0 .
we prove [ thr:24 ] using the notations of theorem [ thr:11 ] , an energy @xmath268 is a resonance of @xmath73 if and only if @xmath467 where @xmath468 denotes the determinant of a square matrix , @xmath469 and the determination of @xmath427 is chosen as in theorem [ thr:11 ] .
so , an energy @xmath268 is a resonance of @xmath73 if and only if @xmath470 belongs to the spectrum of the @xmath471 matrix @xmath472 the proof is the same as that of theorem [ thr:11 ] except that now @xmath268 is a resonance is there exists @xmath458 a non trivial solution to the eigenvalues problem @xmath473 this can be equivalently be rewritten as @xmath474 thus , using rank one perturbations twice , we find that an energy @xmath268 is a resonance if and only if @xmath475 that is , if and only is holds .
this completes the proof of theorem [ thr:24 ] .
let us now complete the proof of theorem [ thr:1 ] for the operator on the full - line .
let us first show that has no solution in the upper half - plane .
therefore , if @xmath470 belongs to the spectrum of the matrix defined by and if @xmath476 is a normalized eigenvector associated to @xmath470 , one has @xmath477 this is impossible in the upper half - plane and on @xmath17 as the two sides of the equation have imaginary parts of opposite signs .
+ note that @xmath478 note also that @xmath470 is an eigenvalue of if and only if it satisfies @xmath479 as the eigenvalues of @xmath15 are simple , one computes @xmath480 thus , equation is equivalent to the following polynomial equation of degree @xmath481 in the variable @xmath465 @xmath482 where we have defined @xmath483 and @xmath484 the sequence @xmath485 also satisfies . taking @xmath486 to @xmath487 in , one notes that @xmath488 we are looking for the solutions to in the upper half - plane .
as the polynomial in the right hand side of has real coefficients , its zeros are symmetric with respect to the real axis .
moreover , one notices that , by , @xmath14 is a root of order two of the polunomial in .
hence , as the polynomial has degree @xmath489 , the number of solutions to in the upper half - plane is bounded by @xmath9 .
this completes the proof of theorem [ thr:1 ] .
by theorems [ thr:11 ] and [ thr:24 ] , we want to solve equations and in the lower half - plane .
we first derive some general estimates for zones in the lower half - plane free of solutions to equations and ( i.e. resonant free zones for the operators @xmath76 and @xmath73 ) and later a result on the existence of solutions to equations and ( i.e. resonances for the operators @xmath76 and @xmath73 ) .
we keep the notations of theorems [ thr:11 ] and [ thr:24 ] . to simplify the notations in the theorems of this section
, we will write @xmath490 for either @xmath491 when solving or @xmath492 when solving
. we will specify the superscript only when there is risk of confusion .
+ we first prove [ thr:13 ] fix @xmath493
. then , there exists @xmath177 ( independent of @xmath1 and @xmath9 ) such that , for any @xmath9 and @xmath494 such that @xmath495 , equations and have no solution in the set @xmath496\\
\quad 0\geq c\cdot\theta'_\delta\,&\text{im}\,e>-a_j\,d^2_j\ , |\sin\,\text{re}\;\theta(e)| \end{split } \right\}\ ] ] where the map @xmath410 is defined in section [ sec : proof - theorem ] and we have set @xmath497 in theorem [ thr:13 ] there are no conditions on the numbers @xmath498 or @xmath499 except their being positive . in our application to resonances , this property holds as we are dealing with the eigenvalues and the values of the eigenfunctions at the boundary of some dirichlet problem .
+ when @xmath500 , one proves the existence of another resonant free region near a energy @xmath445 , namely , [ thr:14 ] fix @xmath493 .
pick @xmath494 such that @xmath501 .
there exists @xmath177 ( depending only on @xmath309 ) such that , for any @xmath9 , if @xmath502 , equations and have no solution in the set @xmath503\cup
\left[\lambda_j+ca_j,\frac{\lambda_j+\lambda_{j+1}}2\right]\\ \quad & -ca_j\leq \text{im}\,e\leq -a_j\,d^2_j / c \end{split } \right\}\\&\hskip3cm\bigcup\left\{e\in{\mathbb{c}};\ \begin{split } \text{re}\,e&\in\left[\frac{\lambda_j+ \lambda_{j-1}}2,\frac{\lambda_j+\lambda_{j+1}}2\right]\\ \quad & -d^2_j / c\leq\text{im}\,e\leq -ca_j \end{split } \right\ } \end{split}\ ] ] theorem [ thr:13 ] becomes optimal when @xmath504 . in our application to resonances , for periodic operators , one has @xmath505 and @xmath506 ( see theorem [ thr:19 ] ) and for random operators , one has @xmath507 and @xmath508 ( see theorem [ thr:10 ] and ) .
thus , in the random case , theorem [ thr:13 ] will provide an optimal strip free of resonances whereas in the periodic case we will use a much more precise computation ( see theorem [ thr:17 ] ) to obtain sharp results .
+ .45 theorem [ thr:14 ] becomes optimal when @xmath490 is small and @xmath509 is of order one .
this will be sufficient to deal with the isolated eigenvalues .
it will also be sufficient to give a sharp description of the resonant free region for random potentials .
for the periodic potential , we will rely a much more precise computations ( see theorem [ thr:17 ] ) .
+ note that theorem [ thr:14 ] guarantees that , if @xmath509 is not too small , outside @xmath510 , resonances are quite far below the real axis . the basic idea of the proof is that , for @xmath268 close to @xmath445 , @xmath511 and the matrix @xmath512 are either large or have a very small imaginary part while , as @xmath513 , @xmath514 has a large imaginary part .
thus , and have no solution in this region . + we start with equation .
pick @xmath515 for some @xmath516 large to be chosen later on .
assume first that @xmath517 for @xmath518 .
recall that @xmath519 .
note that , for @xmath516 sufficiently large , for @xmath520 , one has @xmath521 } |\sin\text{re}\,\theta(e)|\\&\geq e^{\theta'_\delta\text{im } e } |\sin\text{re}\,\theta(e)|\geq e^{-1/c } |\sin\text{re}\,\theta(e)| \end{split}\ ] ] and @xmath522 one estimates @xmath523 thus , comparing and , we see that equation has no solution in the set @xmath524 .
+ assume now that @xmath525 .
then , for @xmath515 , one has @xmath526 thus , for @xmath527 , one computes @xmath528 provided @xmath516 satisfies @xmath529 .
+ hence , the comparison of with shows that has no solution in @xmath530 if we choose @xmath516 large enough ( independent of @xmath498 and @xmath531 ) .
thus , we have proved that for some @xmath177 large enough ( independent of @xmath498 and @xmath531 ) , has no solution in @xmath532 .
.1 cm let us now turn to the case of equation .
the basic ideas are the same as for equation .
consider the matrix @xmath512 defined by .
the summands in are hermitian , of rank @xmath63 and their norm is given by .
+ assume that @xmath515 is a solution to .
define the vectors @xmath533 here @xmath534 .
+ note that by defintion of @xmath490 , one has @xmath535 .
pick @xmath458 in @xmath536 , a normalized eigenvector of @xmath512 associated to the eigenvalue @xmath470 .
thus , @xmath458 satisfies @xmath537 note that , by assumption , one has @xmath538 where the constants are independent of the one defining @xmath532 .
+ taking the ( real ) scalar product of equation by @xmath539 , and then the imaginary part , we obtain @xmath540 thus , for @xmath515 , as @xmath541 , for @xmath516 in sufficiently large ( depending only on @xmath309 ) , @xmath542 hence , for a solution to in @xmath532 and @xmath458 as above , one has @xmath543 hence , by the definition of @xmath532 , for @xmath516 large , we get @xmath544 by , the operator @xmath512 can be written as @xmath545 where @xmath546 and @xmath547 are self - adjoint ( @xmath548 is non negative ) and satisfy @xmath549 an explicit computation shows that the eigenvalues of the two - by - two matrix @xmath550 satify @xmath551 where the implicit constants are independent of the one defining @xmath532 .
+ thus , by , using and the second estimate in , we see that the eigenvalues of the matrix @xmath552 satify @xmath553 clearly , for @xmath516 large , no such value can be equal to @xmath470 being to large by in the first case and having too small imaginary part in the second .
the proof of theorem [ thr:13 ] is complete .
again , we start with the solutions to . for @xmath554 , we compute @xmath555 when @xmath556 , the second equality above and yield , for @xmath516 sufficiently large , @xmath557 on the other hand , for some @xmath558 , one has @xmath559 now , as , under the assumptions of theorem [ thr:14 ] , one has @xmath560 } \left|\text{im}\,e^{-i\theta(e)}\right|\\\geq\frac14\min \left(\sqrt{16-(\lambda_j+\lambda_{j-1})^2},\sqrt{16-(\lambda_j+ \lambda_{j+1})^2}\right ) , \end{gathered}\ ] ] we obtain that has no solution in @xmath561 .
+ pick now @xmath562 such that @xmath563 . then , and yield , for @xmath516 sufficiently large , @xmath564 the imaginary part of @xmath514 is estimated as above .
thus , for @xmath516 sufficiently large , has no solution in @xmath565 .. 1 cm the case of equation is studied in exactly the same way except that , as in the proof of theorem [ thr:13 ] , one has to replace the study of @xmath511 by that of @xmath566 for @xmath458 a normalized eigenvector of @xmath512 associated to @xmath470 and , thus , the coefficient @xmath567 in gets multiplied by a factor @xmath568 that is bounded by @xmath569 . +
this completes the proof of theorem [ thr:14 ] .
we will now solve equation near a given @xmath445 under the additional assumptions that @xmath504 . by theorems [ thr:13 ] and [ thr:14 ]
, we will do so in the rectangle @xmath510 ( see fig .
[ fig:6 ] ) . actually , we prove that , in @xmath510 , there is exactly one resonance and give an asymptotic for this resonance in terms of @xmath490 , @xmath509 and @xmath445 .
this result is going to be applied to the case of random @xmath1 and to that of isolated eigenvalues ( for any @xmath1 ) .
+ using the notations of section [ sec : char - reson ] , for @xmath494 , we define @xmath570 we prove [ thr:12 ] pick @xmath494 such that @xmath513 .
there exists @xmath571 ( depending only on @xmath572 and @xmath573 ) such that , for any @xmath9 , if @xmath574 , equation and has exactly one solution in the set @xmath575\\
\quad -ca_j\leq & \text{im}\,e\leq -a_j\,d^2_j / c \end{split } \right\}.\ ] ] moreover , the solution to , say @xmath576 , satisfies @xmath577 and the solution to , say @xmath578 , satisfies @xmath579 note that , if @xmath580 is small , formula gives the asymptotic of the width of the solution @xmath576 , namely , @xmath581 ^ 2 + \text{im}^2\left(e^{-i\theta(\lambda_j)}\right ) } ( 1+o(1)).\ ] ] for @xmath75 , using the bounds and , we see that the asymptotic of the imaginary part of the solution @xmath578 satisfies @xmath582 this and will be useful when @xmath504 as will be the case for random potentials .
the case when @xmath490 and @xmath509 are of the same order of magnitude requires more information .
this is the case that we meet in the next section when dealing with periodic potentials .
+ the proof of theorem [ thr:12 ] also yields the behavior of @xmath583 and @xmath584 near their zeros in @xmath510 and , in particular shows the following [ pro:7 ] fix @xmath493 . under the assumptions of theorem [ thr:12 ] ,
there exists @xmath50 such that , for @xmath501 , one has @xmath585 proposition [ pro:7 ] is a consequence of the analogues of estimates and on rectangles @xmath586\times[-1,1]$ ] for @xmath95 and @xmath587 sufficiently small .
let us start with equation . to prove the statement on equation , in @xmath510 , we compare the function @xmath588 to the function @xmath589 clearly , in @xmath590 , the equation @xmath591 admits a unique solution given by @xmath592 for @xmath593 , the boundary of @xmath510 , one has @xmath594 hence , as @xmath595
, one gets @xmath596 thus , by rouch s theorem , equation has a unique solution in @xmath510 .
+ to obtain the asymptotics of the solution , it suffices to use rouch s theorem again with the functions @xmath597 and @xmath598 on the smaller rectangle @xmath599\times[-1,1]$ ] .
one then estimates @xmath600 thus , for @xmath601 sufficiently large , this completes the proof of the statements on the solutions to equation contained in theorem [ thr:12] .. 1 cm let us turn to equation . on @xmath510 ,
we now compare @xmath602 to the matrix valued function @xmath603 the matrix @xmath604 is rank @xmath63 and can be diagonalized as @xmath605 where and @xmath492 is given by and @xmath606 thus , @xmath607 is unitarily equivalent to @xmath608 as @xmath609 is real and the imaginary part of @xmath610 does not vanish , the matrix @xmath611 is invertible . by rank @xmath63 perturbation theory ( see e.g. @xcite ) , we know that @xmath612 is invertible if and only if @xmath613_{11}+\lambda_j\not = e$ ] ( where @xmath614_{11}$ ] is the upper right coefficient of the @xmath615 matrix @xmath612 ) . in this case
, one has @xmath616_{11}+\lambda_j - e } m_0^{-1}\begin{pmatrix } 1 & 0\\0 & 0 \end{pmatrix } m_0^{-1}.\ ] ] hence , @xmath14 is an eigenvalue of @xmath612 if and only if @xmath617_{11}\\&= \lambda_j+\left\langle \begin{pmatrix}\overline{\varphi_j(l ) } \\
\varphi_j(0)\end{pmatrix } , \left(\gamma_{l , j}(\lambda_j ) + e^{-i\theta(\lambda_j)}\right)^{-1}\begin{pmatrix}\overline{\varphi_j(l ) } \\ \varphi_j(0)\end{pmatrix}\right\rangle .
\end{split}\ ] ] note that , as @xmath618 is real symmetric and @xmath619 , one has @xmath620 and @xmath621 using , , and , we see that , for @xmath593 , the boundary of @xmath510 , @xmath622 is invertible and that one has @xmath623^{-1}\right\|\leq
2c\quad\text{and}\quad \left\| \gamma_{l , j}(e)-\gamma_{l , j}(\lambda_j)\right\|\leq ca^{\mathbb{z}}_jd^{-2}_j .
\end{gathered}\ ] ] hence , as @xmath595 , taking into account , one gets @xmath624^{-1 } \left(\gamma_l(e)+e^{-i\theta(e)}\right)\right\|\leq 4c^2a^{\mathbb{z}}_jd_j^{-2}\ ] ] in the same way , one proves @xmath625^{-1}\left(\gamma_l(e)+e^{-i\theta(e)}\right)\right\|\lesssim k^{-1}\ ] ] where we recall that @xmath599\times[-1,1]$ ] .
+ thus , we can apply rouch s theorem to compare the following two functions on @xmath626 and @xmath627 ( for @xmath601 sufficiently large ) , @xmath628 as @xmath629^{-1 } \left(\gamma_l(e)+e^{-i\theta(e)}\right)\right]\right)\right| .
\end{gathered}\ ] ] we then conclude as in the case of equation .
this completes the proof of theorem [ thr:12 ] .
combining theorems [ thr:12 ] , [ thr:13 ] and [ thr:14 ] , we get a pretty clear picture of the resonances near the dirichlet eigenvalues in @xmath17 as long as the associated @xmath490 and @xmath509 behave correctly .
as said , this and the knowledge of the spectral statistics for random operators will enable us to prove the results described in section [ sec : random - case ] .
for the periodic case , theorems [ thr:12 ] , [ thr:13 ] and [ thr:14 ] will prove not too be sufficient .
as we shall see , in this case , @xmath490 and @xmath509 are of the same order of magnitude .
thus , neighboring dirichlet eigenvalues have a sizable effect on the location of resonances .
therefore , in the next section , we compute the dirichlet spectral data for the periodic potential .
as we did not find any suitable reference for this , we first derive some estimates for the spectral data ( i.e. the @xmath498 and @xmath531 ) for the dirichlet restriction of a periodic operator to the interval @xmath121 when @xmath9 becomes large . + consider a potential @xmath630 such that , for some @xmath90 , one has @xmath631 for all @xmath632 .
we assume @xmath20 to be minimal i.e. to be the period of @xmath1 . in our first result
, we describe the spectrum of @xmath87 on @xmath0 and @xmath86 on @xmath69 ( with dirichlet boundary conditions at @xmath14 ) . in the second result we turn to @xmath15 , the dirichlet restriction @xmath99 to @xmath633 and described its spectral data .
+ we prove [ thr:15 ] the spectrum of @xmath91 , say @xmath22 , is a union of intervals that consist in purely absolutely continuous spectrum .
+ the spectrum of @xmath99 is the union of @xmath22 with at most finitely many simple eigenvalues outside @xmath22 , say , @xmath93 .
@xmath22 consists of purely absolutely continuous spectrum of @xmath99 and the eigenfunctions associated to @xmath93 , say @xmath634 , are exponentially decaying at infinity . except for the exponential decay of the eigenfunctions , the proof of the statement for the periodic operator on @xmath344 and @xmath343 is classical and can e.g. be found in a more general setting in ( * ? ? ?
* and 7 ) ( see also @xcite ) . the exponential decay is an immediate consequence of floquet theory for the periodic hamiltonian on @xmath344 and the fact that the eigenvalues lie in gaps of @xmath22 . + for @xmath91 one can define its bloch quasi - momentum ( see the beginning of section [ sec : proof - theorem-2 ] ) that we denote by @xmath635 ; it is real analytic on @xmath100 , , the interior of @xmath22 .
moreover , on this set , its derivative can be expressed in terms of the density of states defined in as @xmath636 we prove [ thr:16 ] there exits @xmath104 real analytic functions in a neighborhood of @xmath100 , say @xmath637 such that , for @xmath9 s.t . @xmath186
sufficiently large , the eigenvalues of @xmath15 within @xmath100 , say @xmath638 , are the solutions ( in @xmath100 ) to the quantization conditions @xmath639 moreover , there exits @xmath640 real analytic functions in a neighborhood of @xmath100 , say @xmath641 and @xmath642 , such that @xmath642 are positive on @xmath100 and such that the ( normalized ) eigenfunctions , say @xmath643 , associated to @xmath644 satisfy , for @xmath9 sufficiently large s.t .
@xmath186 , @xmath181 , @xmath645 here the eigenfunctions are enumerated according to the quantization condition .
+ recall that @xmath646 ( resp .
@xmath647 ) denotes the spectrum of @xmath122 ( resp .
@xmath117 ) .
if @xmath648 is an eigenvalue of @xmath15 outside @xmath22 , then , there exists @xmath50 such that , for @xmath186 sufficiently large , there exists @xmath649 s.t .
, one has @xmath650 ; if @xmath651 is a normalized eigenfunction associated to @xmath648 and @xmath15 , one has one of the following alternatives for @xmath9 large 1 .
if @xmath652 , one has @xmath653 2 . if @xmath654 , one has @xmath655 3 .
if @xmath656 , one has @xmath657 for later use , let us define @xmath658 , @xmath659 and @xmath660 by @xmath661 where @xmath635 , @xmath662 , @xmath663 , @xmath664 and @xmath641 are defined in theorem [ thr:16 ] .
+ as a consequence of theorem [ thr:16 ] , we obtain [ cor:1 ] for @xmath665 , for @xmath128 sufficiently large , one has @xmath666 here , @xmath667 , @xmath668 and @xmath669 are defined the functions defined in theorem [ thr:16 ] .
let us note here that proposition [ pro:1 ] is an immediate consequence of theorems [ thr:17 ] and [ thr:19 ] and corollary [ cor:1 ] . to prove the first equalities in and
, it suffices to prove that , for any @xmath670 , one computes @xmath671 the full statement then following by standard density argument .
the operator @xmath15 converges to @xmath122 in norm resolvent sense .
thus , we know that @xmath672 . now , by theorem [ thr:16 ] , as @xmath673 is supported in @xmath674 , using the poisson formula , one computes @xmath675 thus , using the non stationary phase i.e. integrating by parts , one gets , for any @xmath676 , @xmath677 here , we have used the analyticity of the functions @xmath658 and @xmath659 .
+ to deal with @xmath120 , we recall the operator @xmath678 ( that is unitarily equivalent to @xmath15 ) defined in remark [ rem:6 ] .
one has @xmath679 , thus , as @xmath120 is the strong resolvent sense limit of @xmath678 , one gets @xmath680 .
+ then , and and , thus , the first equalities in and , follow as @xmath681 , @xmath659 and @xmath660 converge ( locally uniformly on @xmath100 ) respectively to @xmath682 , @xmath668 and @xmath669 ( see and theorem [ thr:16 ] ) .
+ let us now prove the second equalities in and .
therefore , we use an _ almost analytic extension _ ( see @xcite ) of @xmath673 , say , @xmath683 , that is , a function @xmath684 satisfying ( 1 . for @xmath685 , @xmath686 , 2 .
supp@xmath687 , 3 .
@xmath688 , 4 .
the family of functions @xmath689 ( for @xmath690 ) is bounded in @xmath691 for any @xmath692 .
moreover , one has the following estimates : for @xmath693 , @xmath694 , @xmath695 , there exists @xmath696 such that @xmath697 by the definition of @xmath673 , one has that the right hand side of is bounded uniformly in @xmath268 complex .
+ let @xmath698 and @xmath699 be an almost analytic extension of @xmath700 .
then , by @xcite and @xcite , we know that , for any @xmath101 and @xmath701 , the following formula hold , @xmath702 where @xmath703 or @xmath120 .
+ using the geometric resolvent equation ( see e.g. ( * ? ? ?
* theorem 5.20 ) ) and the combes - thomas estimate ( see e.g. ( * ? ? ?
* theorem 11.2 ) ) , we know that for some @xmath177 , for im@xmath704 , @xmath705\delta_0\right\rangle\right|\\+ \left|\left\langle\delta_0,\left[(h_l - z)^{-1}-(h_0^+-z)^{-1 } \right]\delta_0\right\rangle\right|\\\leq \frac{c}{|\text{im}z|}e^{-l|\text{im}z|/c}. \end{gathered}\ ] ] plugging into and using , we get @xmath706 thus , by and , we obtain that , for @xmath670 and any @xmath707 , there exists @xmath708 such that @xmath709d\lambda\right|\\&= \left|\int_{{\mathbb{r } } } \chi\left ( \lambda\right ) f_{0,l}\left(\lambda\right)\theta'_{p , l}(\lambda)d\lambda- \int_{{\mathbb{r}}}\chi(\lambda)dn^+_0(\lambda ) \right|\leq c_n l^{-n}. \end{split}\ ] ] now , by and , the function @xmath710 admits an expansion in inverse powers of @xmath9 that is converging absolutely on compact subsets of @xmath100 , namely , @xmath711 thus , immediately yields that , for any @xmath712 , one has @xmath713 on @xmath100 .
hence , @xmath714 on @xmath100 .
this completes the proof of corollary [ cor:1 ] .
we will first describe some objects from the spectral theory of @xmath91 , then use them to describe the spectral theory of @xmath99 and finally prove theorem [ thr:16 ] .
let us now turn to the spectrum of the operator on the half - lattice .
[ [ sec : operator ] ] _ the operator @xmath122 _ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + for the operator @xmath750 ( that is @xmath97 on @xmath69 with dirichlet boundary conditions at @xmath14 ) , @xmath268 is in the spectrum if and only if either @xmath742 or @xmath744 and @xmath751^n\begin{pmatrix}1\\0\end{pmatrix}$ ] stays bounded as @xmath752 ( or equivalently @xmath753^nt_{k-1}(e)\cdots t_0(e)\begin{pmatrix}1\\0\end{pmatrix}$ ] stays bounded as @xmath752 ) .
+ when @xmath754 and @xmath755 , one can diagonalize @xmath734 in the following way @xmath756 thus , using @xmath757 for @xmath739 , one computes @xmath758 where @xmath759 clearly , the formulae , and stay valid even if @xmath760 . + thus , simple computations show that @xmath268 is in the spectrum @xmath97 on @xmath69 with dirichlet boundary conditions at @xmath14 if and only if one of the following conditions is satisfied : 1 .
@xmath742 : in this case , the set @xmath761 is contained in the absolutely continuous spectrum .
@xmath744 and @xmath762 note that , in case ( 2 ) , @xmath751^n\begin{pmatrix}1\\0\end{pmatrix}$ ] actually decays exponentially fast . in this case
, @xmath268 is an eigenvalue associated to the ( non normalized ) eigenfunction @xmath763 where , for @xmath239 and @xmath134 , @xmath764^n\\&=a_k(e)\left[a^0_p(e)\right]^n \end{split}\ ] ] writing @xmath765 it is well know that , for any @xmath20 , the zeros of @xmath766 are simple ( see e.g. @xcite ) .
let @xmath767 be a root of @xmath768 satisfying @xmath769 and @xmath770 .
differentiating , we get @xmath771 on @xmath22 , the spectrum of @xmath122 is purely absolutely continuous ; it does not contain any embedded eigenvalues .
+ let us now turn to @xmath117 . recalling and using the representation , we obtain that the eigenvalues of @xmath117 outside @xmath22 satisfy @xmath772 [
[ sec : operator2 ] ] _ the eigenvalues of the operator @xmath120 _ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + as for @xmath122 , the eigenfunction associated to @xmath268 and @xmath120 decays exponentially fast .
indeed , the eigenvalues of @xmath120 in the region @xmath744 can be analyzed as we analyzed those of @xmath122 i.e. they are the energies such that @xmath773^{-n}\begin{pmatrix}0\\1\end{pmatrix}$ ] stays bounded ; this yields the quantization conditions @xmath774 . in this case
, @xmath268 is an eigenvalue associated to the ( non normalized ) eigenfunction @xmath775 where , for @xmath239 and @xmath134 , @xmath776^{-n}.\ ] ] [ [ sec : operator3 ] ] _ common eigenvalues to @xmath122 and @xmath120 _
+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + assume now that @xmath767 is simultaneously an eigenvalue of @xmath120 and @xmath122 . in this case , one has @xmath777 , @xmath778 and @xmath779 .
so ( see also ) becomes @xmath780 hence , the analytic function @xmath781 has a root of order at least 2 at @xmath767 .
it also implies that @xmath782 .
indeed , if @xmath783 , implies @xmath784 as @xmath785 .
+ conversely , if @xmath786 such that @xmath770 and @xmath787 has a root of order at least 2 at @xmath767 , then holds and @xmath767 is an eigenvalue of @xmath120 .
let us now turn to the study of the eigenvalues and eigenvectors of @xmath15 i.e. to the proof of theorem [ thr:16 ] .
we first prove the statements for the eigenvalues and then , in the next section , turn to the eigenvectors .
+ recall that @xmath186 . by definition
, @xmath268 is an eigenvalue of @xmath97 on @xmath121 with dirichlet boundary conditions if and only if @xmath788^n \begin{pmatrix}1\\0\end{pmatrix } , \begin{pmatrix}0\\1\end{pmatrix}\right ) \end{split}\ ] ] where @xmath186 and @xmath718 is the monodromy matrix defined above .
+ using the notations of sections [ sec : spectral - theory ] and [ sec : spectral - theory - z ] , let first assume that @xmath789 . then , @xmath718 is diagonalizable and either of two things happen ( see e.g. @xcite ) : 1 .
either @xmath754 and @xmath790 , 2 . or @xmath733 and @xmath791id . in case
( 2 ) , the quantization condition becomes @xmath792 for case ( 1 ) i.e. when @xmath754 , diagonalizing @xmath793 using yields the following characteristic equations for the eigenvalues @xmath794
\rho^{-n}(e)-\left[b^0_{p-1}(e)-\rho(e)\right]\rho^n(e ) & -b_{k+1}(e ) \\a^0_{p-1}(e)(\rho^n(e)-\rho^{-n}(e ) ) & a_{k+1}(e)\end{matrix } \right|=0.\ ] ] as @xmath795 if we define @xmath796 equation can be rewritten as @xmath797 , thus , as @xmath798 , also as @xmath799 if @xmath744 , we note that the left hand side in is the first equation in and that the right hand side in is exponentially small for large @xmath262 . by and , the product @xmath800 is a monic polynomial of the variable @xmath268 .
thus , the solutions to are exponentially close to either an eigenvalue of @xmath122 or one of @xmath120 .
one distinguishes between the following cases : 1 .
if @xmath767 is an eigenvalue of @xmath122 but not of @xmath120 , then @xmath767 is a simple root of the function @xmath801 ( see section [ sec : operator3 ] ) ; one has to distinguish two cases depending on whether @xmath802 vanishes or not .
assume first @xmath803 ; then , by , we know that the eigenvector of @xmath122 actually satisfies the dirichlet boundary conditions at @xmath9 ; thus @xmath804 and gives a ( non normalized ) eigenvector .
+ assume now that @xmath805 ; then , the unique solution to close to @xmath767 satisfies @xmath806 2 .
if @xmath767 is an eigenvalue of @xmath120 but not of @xmath122 , mutandi mutandis , the analysis is the same .
if @xmath767 is an eigenvalue of both @xmath122 and @xmath120 , then , we are in a resonant tunneling situation . the analysis done in the appendix
, section [ sec : appendix ] shows that near @xmath767 , @xmath15 has two eigenvalues , say @xmath807 satisfying , for some constant @xmath221 , @xmath808 this completes the proof of the statements of theorem [ thr:16 ] for the eigenvalues in the region @xmath744 .
.1cmwe now study the eigenvalues in the region @xmath809 .
one can express @xmath265 in terms of the bloch quasi - momentum @xmath810 and use @xmath743 .
thus , as @xmath186 , the characteristic equation becomes @xmath811}{\overline { a^0_{p-1}(e)b_{k+1}(e)+a_{k+1}(e)\left[a^0_p(e)- \rho(e)\right]}}=:e^{2i h_k(e)}. \end{split}\ ] ] notice that as im@xmath812 does not vanish on @xmath100 , * the function @xmath813 is analytic , * the functions @xmath814 , @xmath815 , @xmath816 and @xmath817 are real analytic for any @xmath104 , * the function @xmath818 $ ] does not vanish .
thus , the function @xmath819 is real analytic on @xmath100 .
the quantization condition is thus equivalent to .
+ this completes the proof of the statement of theorem [ thr:15 ] about the eigenvalues of @xmath15 .
let us now turn to the associated eigenfunctions .
recall that we assume @xmath186 .
first , if @xmath820 is an eigenfunction associated to the eigenvalue @xmath445 , the eigenvalue equation reads @xmath821 to normalize the solution , we assume that @xmath822 . the coefficients we want to compute are @xmath823 fix @xmath824 .
thus , using the notations of section [ sec : dirichl - eigenv ] and the diagonalization , one has @xmath825 where @xmath826 and @xmath827 are defined in . by definition
, one has @xmath828 we start with the case ( 1 ) i.e. the case when @xmath829 . as @xmath830 , we obtain that @xmath831 \right)}{\left|\rho(\lambda_j)-\rho^{-1}(\lambda_j)\right|^2}.\ ] ] as @xmath186 , using the quantization condition , we compute @xmath832\right ) \\&\hskip1.5cm+ 2\sum_{m=0}^{p-1}|\alpha_m(\lambda_j)|^2\sum_{n=0}^n\left ( 1-\text{re}\left[\frac{\alpha_m(\lambda_j ) } { \overline{\alpha_m(\lambda_j ) } } \rho^{2n}(\lambda_j)\right]\right)\\ & = n\,p\,f(\lambda_j)\left(1-\frac1{np}\tilde f(\lambda_j ) \right ) \end{split}\ ] ] where we have defined @xmath833 and @xmath834 \frac{\text{im}\,(\overline{\alpha_{k+1}(e)}\rho(e))}{\text{im}\,(\rho(e ) ) } \\&\hskip2 cm -\frac{1}{f(e)}\sum_{m=0}^k|\alpha_m(e)|^2\left ( \text{re}\left[\frac{\alpha_m(e)\,\overline{\alpha_{k+1}(e ) } } { \overline{\alpha_m(e)}\,\alpha_{k+1}(e)}\right]-1\right ) .
\end{split}\ ] ] in the computation of @xmath641 , we have used the quantization condition . + for @xmath835 , as the matrix @xmath836 is invertible on @xmath100 , the function @xmath837 does not vanish on this set .
thus , the function @xmath838 does not vanish on @xmath100 .
note also that @xmath839 and @xmath641 are real analytic on @xmath100 . +
as @xmath186 , the characteristic equation for @xmath445 , that is , reads @xmath840 hence , one computes @xmath841 } \frac{a_{p-1}^0(\lambda_j)}{|\alpha_{k+1}(\lambda_j)| } \end{split}\ ] ] using . plugging this and into , recalling that @xmath842
, we obtain if , in addition to and we set @xmath843 and @xmath844 this completes the proof of the statements in theorem [ thr:16 ] on the eigenfunctions of @xmath15 associated to eigenvalues in @xmath100 .. 1cmlet us now turn to the eigenfunctions associated to eigenvalues @xmath15 in the gaps of @xmath22 i.e. is the region @xmath744 . on @xmath118 ,
the eigenvalue @xmath813 is real valued and so are all the functions @xmath845 and @xmath846 . as @xmath747 , for @xmath847 , becomes @xmath848
let us now distinguish the same cases as for the eigenvalues i.e. when @xmath268 is close to an eigenvalue of @xmath122 or @xmath120 : 1 .
pick @xmath767 an eigenvalue of @xmath122 but not of @xmath120and let @xmath268 be close to @xmath767 . then , by definition @xmath849 .
thus , for @xmath835 , one has @xmath850 . as @xmath268 satisfies , using , becomes @xmath851\rho^{2(n - n)}(e')+ o(\rho^{2n}(e))\biggl|^2 .
\end{split}\ ] ] where @xmath835 if @xmath852 and @xmath853 if @xmath854 . + as @xmath855 , one computes @xmath856 this yields @xmath857 thus , by and , as @xmath849 , we obtain @xmath858 ^ 2\sum_{m=0}^{p-1}a^2_m(e ' ) } \left|\begin{matrix}\beta'_k(e ' ) & a_k(e')\\\beta'_{k+1}(e')&a_{k+1}(e ' ) \end{matrix}\right|^2 + o(n\rho^{4n}(e))\\ & = \gamma\rho^{2n}(e ' ) + o(n\rho^{4n}(e ) ) .
\end{split}\ ] ] where @xmath859 ^ 2\sum_{m=0}^{p-1}a^2_m(e ' ) } \left(\frac{d a_{p-1}^0}{de}(e')\right)^2>0\ ] ] 2 .
if @xmath767 is an eigenvalue of @xmath120 but not of @xmath122 , then inverting the parts of @xmath120 and @xmath122 , we see that @xmath860 is of order @xmath63 .
a precise asymptotic can be computed but it wo nt be needed .
if @xmath767 is an eigenvalue of @xmath122 and of @xmath120 , the double well analysis done in section [ sec : appendix ] shows that for normalized eigenvectors , say , @xmath861 associated to the two eigenvalues of @xmath15 close to @xmath767 , the four coefficients @xmath862 and @xmath863 are of order @xmath63 . again precise asymptotics can be computed but wo nt be needed .
this completes the description of the eigenfunctions given by theorem [ thr:16 ] and completes the proof of this result .
we are now in the state to prove the results stated in section [ sec : periodic - case ] .
therefore , we first study the function @xmath864 and @xmath865 in the complex strip @xmath228 for @xmath866 . using theorem [ thr:16 ] , we first prove [ thr:17 ] fix @xmath866 a compact interval .
there exists @xmath868 such that , for any @xmath707 , there exists @xmath708 such that , for @xmath9 sufficiently large s.t @xmath869 , one has @xmath870 where @xmath871 and @xmath872 ( see ) , deeper into the lower complex plane we obtain the following simpler estimate [ thr:19 ] there exists @xmath177 such that , for any @xmath328 and for @xmath873 sufficiently large s.t .
@xmath186 , one has @xmath874 in sections [ sec : proof - theorem-3 ] , the approximations and theorems will be used to prove theorems [ thr:5 ] , [ thr:2 ] and [ thr:3 ] .
+ let us note that , as @xmath875 , for @xmath876 , the asymptotics given by theorems [ thr:17 ] and [ thr:19 ] coincide in the region @xmath144re@xmath877im@xmath878 : indeed one has , @xmath879 let us now turn to the proofs of theorems [ thr:17 ] and [ thr:19 ] . to prove theorem [ thr:17 ] , we split the sum @xmath511 into two parts , one containing the dirichlet eigenvalues `` close '' to re@xmath880 , the second one containing those `` far '' from re@xmath880 . by `` far '' , we mean at that the distance to re@xmath880 is lower bounded by a small constant independent of @xmath9 .
the `` close '' eigenvalues are then described by theorem [ thr:16 ] .
for the `` far '' eigenvalues , the strong resolvent convergence of @xmath15 to @xmath122 , that of @xmath678 to @xmath120 ( see remark [ rem:6 ] ) and combes - thomas estimates enable us to compute the limit and to show that the prelimit and the limit are @xmath881 close to each other .
for the `` close '' eigenvalues , the sum coming up in , the definition of @xmath867 , is a riemann sum .
we use the poisson formula to obtain a precise approximation .. 1cmas @xmath31 is a compact interval in @xmath100 , we pick @xmath328 such that , for @xmath293 , one has @xmath882\subset\overset{\circ}{\sigma}_{\mathbb{z}}$ ] .
let @xmath698 be a non - negative cut - off function such that @xmath883 on @xmath884 $ ] and @xmath885 outside @xmath886 $ ] . for @xmath293 ,
define @xmath887 ..
1cmwe first give the asymptotic for the sum over the dirichlet eigenvalues far from re@xmath880 .
we prove [ le:1 ] for any @xmath888 , there exists @xmath708 such that , for @xmath9 sufficiently large such that @xmath128 , one has @xmath889 where @xmath890 recall ( see theorems [ thr:11 ] ) that @xmath15 is the operator @xmath122 restricted to @xmath121 with dirichlet boundary condition at @xmath9 ; as @xmath128 , it is unitarily equivalent to the operator @xmath120 restricted to @xmath891 with dirichlet boundary condition at @xmath124 ( see remark [ rem:6 ] ) .
+ pick @xmath892 such that @xmath893 on @xmath894 .
first , we note that @xmath895(h_l)(h_l - e)^{-1 } \delta_0\rangle\\&\hskip5 cm -\langle\delta_0,\left[\tilde\chi(1-\chi_{\text{re}\,e})\right](h_0^+)(h_0^+-e)^{-1 } \delta_0\rangle , \end{split}\ ] ] @xmath896(h_l)(h_l - e)^{-1 } \delta_l\rangle\\&\hskip5cm-\langle\delta_l,\left[\tilde\chi(1-\chi_{\text{re}\,e})\right ] ( h_k^-)(h_k^--e)^{-1}\delta_l\rangle , \end{split}\ ] ] and @xmath897(h_l)(h_l - e)^{-1 } \delta_0\rangle.\ ] ] by the definition of @xmath898 , the function @xmath899 is @xmath900 on @xmath319 ; moreover , its semi - norms ( see ) are bounded uniformly in @xmath901 .
thus , there exists an almost analytic extension of @xmath902(\cdot)(\cdot - e)^{-1}$ ] such that , uniformly in @xmath268 , one has .
+ in the same way as we obtained , we obtain @xmath903\delta_l\right\rangle\right|\\+ \left|\left\langle\delta_0,\left[(h_l - z)^{-1}-(h_0^+-z)^{-1 } \right]\delta_0\right\rangle\right|\\+ \left|\left\langle\delta_0,(h_l - z)^{-1}\delta_l\right\rangle\right|\leq \frac{c}{|\text{im}z|^2}e^{-l|\text{im}z|/c } \end{gathered}\ ] ] plugging into and using for @xmath902(\cdot)(\cdot - e)^{-1}$ ] , we get @xmath904 this entails and completes the proof of lemma [ le:1 ] .
let us now estimate the part of @xmath512 associate to the dirichlet eigenvalues close to re@xmath880 .
therefore , define @xmath905 we prove [ le:13 ] there exists @xmath328 such that , for @xmath906 , there exists @xmath907 such that , for @xmath9 sufficiently large such that @xmath869 , one has @xmath908 where @xmath909 is defined in .
clearly lemmas [ le:1 ] and [ le:13 ] immediately yield theorem [ thr:17 ] .
the quasi - momentum @xmath635 defines a real analytic one - to - one monotonic map from the interior of each band of spectrum onto the set @xmath910 , @xmath911 or @xmath912 ( depending on the spectral band containing @xmath913 $ ] where @xmath328 has been fixed above ) ( see e.g. @xcite ) .
moreover , the derivative @xmath682 does not vanish in the interior of a spectral band .
thus , for @xmath9 sufficiently large , the derivative @xmath681 does not vanish on @xmath914 $ ] and @xmath658 is real analytic one - to - one on a complex neighborhood of @xmath915)+i[-3\varepsilon,3\varepsilon]$ ] ( possibly at the expense of reducing @xmath308 somewhat ) .
+ by , and theorem [ thr:16 ] , one may write @xmath916 where @xmath917 the matrix @xmath612 can also be assumed to be analytic in the rectangle @xmath915)+i[-3\varepsilon,3\varepsilon]$ ] .
thus , the poisson formula tells us that @xmath918 by the definition of @xmath898 ; here , we have set @xmath919 let us now study the individual terms in the last sum in .
therefore , recall that in @xmath884 $ ] , @xmath673 is identically @xmath63 and that @xmath920 and @xmath921 are analytic in @xmath915)+i[-3\varepsilon,3\varepsilon]$ ] ; moreover , by , for some @xmath493 , one has @xmath922 } \theta'_{p , l}(\lambda+\text{re}\,e)\geq \liminf_{l\to+\infty}\inf_{e\in i } \theta'_{p , l}(e)\geq \delta.\ ] ] recall also that im@xmath426 . consider @xmath923 $ ] smooth such that @xmath924 on @xmath925 $ ] and @xmath926 outside @xmath927 $ ] .
+ in the complex plane , consider the paths @xmath928 defined by @xmath929 as @xmath930im@xmath931 , by contour deformation , we have @xmath932 we then estimate * for @xmath933 , using a non - stationary phase argument as the integrand is the product of a smooth function with an rapidly oscillatory function ( using @xmath934 as the large parameter ) , one then estimates @xmath935 the phase function is complex but its real part is non positive as im@xmath936 on the support of @xmath673 ( by ) .
note that the off - diagonal terms of @xmath937 also carry a rapidly oscillating exponential ( see ) but it clearly does not suffice to counter the main one . * in the same way , for @xmath938 , one has @xmath939 thus , we compute @xmath940 finally , for @xmath941 , the contour deformation along @xmath942 yields @xmath943 by corollary [ cor:1 ] . + plugging this , and into and computing the geometric sum immediately yields the following asymptotic expansion ( where the remainder term is uniform on the rectangle @xmath944 ) @xmath945 this completes the proof of lemma [ eq:65 ] . to prove , for im@xmath946
, it suffices to write @xmath947\delta_0\rangle \\&=\langle\delta_0,(h_l - e)^{-1}\delta_l\rangle \langle\delta_{l+1},(h_0^+-e)^{-1 } \delta_0\rangle \end{split}\ ] ] and @xmath948 and use the combes - thomas estimate .
this completes the proof of theorem [ thr:19 ] .
we will now use theorems [ thr:17 ] and [ thr:19 ] to prove theorems [ thr:5 ] , [ thr:2 ] and [ thr:3 ] .
the first statement of theorem [ thr:5 ] is an immediate consequence of the characteristic equations for the resonances and and the description of the eigenvalues of @xmath15 given in theorem [ thr:16 ] .
when @xmath225 i.e. for the operator on the half - line , if @xmath949 does not meet @xmath85 , there exists @xmath177 s.t . for @xmath9 sufficiently large dist@xmath950
thus , on the set @xmath951 , one has im@xmath952im@xmath953 . as on @xmath31 ,
one has im@xmath954 ( see section [ sec : proof - theorem ] ) , the characteristic equation admits a solution @xmath268 such that re@xmath955 only if im@xmath956 . this completes the proof of point ( 1 ) of theorem [ thr:5 ] for @xmath225 . for @xmath224 i.e. to study equation , one reasons in the same way except that one replaces the study of @xmath511 by that of @xmath957 for @xmath458 an arbitrary vector in @xmath958 of unit length
this completes the proof of point ( 1 ) of theorem [ thr:5 ] + point ( 3a ) is an immediate consequence of theorems [ thr:12 ] and [ thr:14 ] and the description of the eigenvalues of @xmath15 outside @xmath22 .
notice that in the present case @xmath509 in theorems [ thr:12 ] and [ thr:14 ] is bounded from below by a constant independent of @xmath9 and @xmath959 is exponentially small and described by theorem [ thr:16 ] .
+ point ( 3b ) is an immediate consequence of the description of the eigenvalues of @xmath15 outside @xmath22 in case ( 3 ) of theorem [ thr:19 ] and theorem [ thr:13 ] .
indeed , in the present case @xmath509 and and @xmath959 are both of order @xmath63 ; thus , theorem [ thr:13 ] guarantees , around the common eigenvalue for @xmath120 and @xmath122 , a rectangle of width of order @xmath63 free of resonances .
+ let us now turn to the proof of point ( 2 ) .
therefore , we first prove the following corollary of theorem [ thr:17 ] [ cor:2 ] fix @xmath866 compact .
there exists @xmath210 such that , for @xmath9 sufficiently large , one has @xmath960 clearly , corollary [ cor:2 ] implies that neither equation nor equation can have a solution in @xmath961-\eta_0/l,0]$ ] .
this proves point ( 2 ) of theorem [ thr:5] .. 2cmbefore proving corollary [ cor:2 ] , we first prove propositions [ pro:4 ] and [ pro:5 ] as these will be of use in the proof of corollary [ cor:2 ]
. we will do the proofs for the function @xmath163 .
proposition [ pro:4 ] is an immediate consequence of the fact that , in the lower half - plane , the function @xmath962 ( i.e. the determination of it defined above ) is equal to the stieljes ( or borel ) transform of the spectral measure associated to the dirichlet laplacian on @xmath343 and the vector @xmath963 ; this follows from a direct computation ( see remark [ rem:5 ] and for @xmath964 ) .
now , the spectral measure @xmath965 is also the spectral measure of a schrdinger operator , say , @xmath966 on @xmath343 associated to @xmath963 : it suffices to take the symmetric of @xmath967 with respect to @xmath14 .
the equality of the borel transforms implies the equality of the measures but @xmath963 is cyclic for both operators so the operators have equal spectral measures .
this implies that the two operators are equal and thus the symmetric of @xmath967 has to vanish identically on @xmath343 . as @xmath1 is periodic ,
@xmath1 must vanish identically .
+ as for the second point , if the function @xmath163 were to vanish to infinite order at @xmath968 , as each of the terms @xmath969 and @xmath970 admits an infinite asymptotic expansion in powers of @xmath971 , these two expansions would be equal .
the @xmath101-th coefficient of these expansion are respectively the @xmath101-th moments of the spectral measures of @xmath966 and @xmath972 ( associated to the cyclic vector @xmath963 ) .
so these moments would coincide and thus the spectral measures would coincide .
one then concludes as above . on @xmath144im@xmath145 , define the functions @xmath973 first , the analyticity of @xmath974 and @xmath975 is clear ; indeed , all the functions involved are analytic and the functions @xmath164 and @xmath161 stay positive on @xmath100 .
moreover , these functions can be analytically continued through @xmath242 . by , for @xmath268 real , one has im@xmath976im@xmath977 im@xmath978 which is positive ( see section [ sec : proof - theorem ] ) .
thus , the functions @xmath979 and @xmath980 do not vanish on @xmath31 .
moreover , as @xmath981 this function has a positive imaginary part on @xmath31 .
+ this proves the first two properties of @xmath982 stated in proposition [ pro:5 ] . by the very definition of @xmath982 and @xmath974 ,
the last property is obvious in the case of the half - line ; for the full line i.e. if @xmath224 , the last property is a consequence of the following computation @xmath983 this completes the proof of proposition [ pro:5 ] . in view of theorem [ thr:17 ] , to obtain , it suffices to prove that there exists @xmath210 such that , for @xmath9 sufficiently large , one has @xmath984 where @xmath872 .
+ we compute @xmath985 where @xmath974 is defined in .
thus , @xmath986 as , for @xmath987 sufficiently small and @xmath873 , one has @xmath988 now , notice that , by corollary [ cor:1 ] , for @xmath293 , one has @xmath989 thus , as @xmath990im@xmath978 is positive on @xmath31 , the analytic function @xmath991 has positive imaginary part larger than , say , @xmath992 on @xmath31 ; hence , it has imaginary part larger than , say , @xmath993 in some neighborhood of @xmath994 ( for sufficiently small @xmath210 ) .
let @xmath612 be the maximum modulus of this function on @xmath994 .
thus , as @xmath995 , one has @xmath996 possibly reducing @xmath997 , this guarantees that , for @xmath998 and @xmath999 , one has @xmath1000 this completes the proof of the first lower bound in in corollary [ cor:2 ] .
+ to prove the second bound in , using , we compute @xmath1001 where @xmath1002 and @xmath975 are defined by and .
+ using proposition [ pro:5 ] , one then concludes the non - vanishing of the function @xmath1003 in the complex rectangle @xmath1004 ( for @xmath997 sufficiently small ) in the same way as above .
this completes the proof of corollary [ cor:2 ] . to solve and , by theorem [ thr:17 ] , we respectively first solve the equations @xmath1005 in a rectangle @xmath1006 $ ] .
indeed , in such a rectangle , by theorem [ thr:17 ] , equations and are respectively equivalent to @xmath1007 where the terms @xmath1008 are analytic in a rectangle @xmath1009 ( where @xmath1010 ) and the bound @xmath1008 holds in the supremum norm .
+ thanks to and , for @xmath95 , to solve the equations , it suffices to solve @xmath1011 where we recall @xmath872 and , @xmath975 and @xmath974 being respectively defined in and , and , as in section [ sec : descr - reson ] , one has set * @xmath1012 in the case of the half - line , * @xmath1013 in the case of the line .
we want to solve is a rectangle @xmath944 for some @xmath308 small but fixed . using proposition [ pro:5 ] ,
we pick @xmath308 so small that , in the rectangle @xmath1014 $ ] , the only zeros of @xmath174 are those on the real line and im@xmath1015 is positive in @xmath944 .
+ to solve , we change variables @xmath1016 that is , we write @xmath1017 as , for @xmath1018 sufficiently large , @xmath1019 , at the cost of possibly reducing @xmath308 , this real analytic change of variables maps @xmath1020+i[-\varepsilon,0)$ ] into , say , @xmath1021 such that @xmath1022\subset d_l$ ] ( for some @xmath230 ) where @xmath1023)$ ] ; the inverse change of variable maps @xmath1022 $ ] into some domain , say , @xmath1024 such that @xmath1025+i[-\varepsilon',0]\subset\tilde d_l$ ] ( for some @xmath1026 ) .
now , to find all the solutions to in @xmath1027 , we solve the following equation in @xmath1022 $ ] @xmath1028 as @xmath1029 is @xmath1030 periodic , we split @xmath1022 $ ] into vertical strips of the type @xmath1031+i[-\eta(l - k),0]$ ] , @xmath1032 , @xmath1033 . without loss of generality , we may assume that @xmath1034\pi$ ] . to solve on the rectangle @xmath1031+i[-\eta(l - k),0]$ ] , we shift @xmath458 by @xmath1035 and solve the following equation on @xmath1036+i[-\eta(l - k),0]$ ] @xmath1037 in proving theorem [ thr:5 ] , we have already shown that for some @xmath1038 ( independent of @xmath9 sufficiently large and @xmath1039 ) , does not have a solution in @xmath1036+i[-\tilde\eta,0]$ ] .
the cotangent is an analytic one - to - one mapping from @xmath1040 $ ] to @xmath203 .
thus , for @xmath9 sufficiently large and @xmath993 sufficiently small , the cotangent defines a one - to - one mapping from @xmath1041 $ ] onto @xmath1042 , analytic in the interior of @xmath1041 $ ] and continuous up to the boundary where we have defined @xmath1043 moreover , the boundaries @xmath1044 $ ] and @xmath1045 $ ] are mapped onto the interval @xmath1046 $ ] .
+ let @xmath1047 denote the finite set of zeros of @xmath1048 in @xmath31 .
then , by a taylor expansion near the zeros of @xmath1049 , we know that , for @xmath987 sufficiently small , there exists @xmath1050 and @xmath1051 such that , for @xmath9 sufficiently large , * for @xmath1052 , there exists @xmath1053 such that , for @xmath1032 , if @xmath1054 , @xmath1055 then @xmath1056 + i[-\eta(l - k),0]$ ] , one has @xmath1057 ; * for @xmath1058+i[-\eta(l - k),0]$ ] and @xmath374 the point in @xmath1047 closest to @xmath1059 , one has @xmath1060^{-\tilde k}\leq\frac1{\varepsilon_0}\ ] ] where @xmath1061 is the order of @xmath374 as a zero of @xmath1062 . as a consequence of the above description of @xmath1063
, we obtain [ le:12 ] there exists @xmath993 and @xmath987 mall such that , for @xmath9 sufficiently large , for all @xmath1032 , @xmath1064 maps the rectangle @xmath1036+i[-\eta(l - k),-\tilde\eta]$ ] into a compact subset of @xmath1065 in such a way that @xmath1066+i[-\eta(l - k),-\tilde\eta ] ) } \left|\cot u- c^\bullet_{l , l}(u)\right|\gtrsim \left(\left|\tilde e-\theta_{p , l}^{-1}\left ( \frac{l\pi}{l - k}\right)\right|+\frac{\tilde\eta}{l - k } \right)^{\tilde k}\ ] ] where @xmath374 is the zero of @xmath208 closest to @xmath1059 and @xmath1061 is its order .
note that , under the assumptions of lemma [ le:12 ] , implies that @xmath1067+i[-\eta(l - k),-\tilde\eta ] ) } \left|\cot u- c^\bullet_{l ,
l}(u)\right|\gtrsim l^{-\tilde k}\ ] ] thus , we can define the analytic mapping @xmath1068 on @xmath1036+i[-\eta(l - k),-\tilde\eta]$ ] ; it maps the rectangle @xmath1036+i[-\eta(l - k),-\tilde\eta]$ ] into a compact subset of @xmath1069 .
the equation on @xmath1036+i[-\eta(l - k),-\tilde\eta]$ ] is , thus , equivalent to the following fixed point equation on the same rectangle @xmath1070 we note that , for @xmath1071 , for @xmath9 sufficiently large , if for some @xmath1072 of multiplicity @xmath1061 , one has @xmath1073 then , equation has no solution in @xmath1036+i[-\eta(l - k),-\tilde\eta]$ ] outside of the set @xmath1074+i\left[-\eta(l - k),\frac{\alpha\tilde k}4\log\left [ \left|\theta_{p , l}^{-1}\left(\frac{l\pi}{l - k}\right)- \tilde e\right|+\frac1l \right]\right].\ ] ] indeed , for @xmath1075+i[-\eta(l - k),-\tilde\eta])\setminus r_{l , l}$ ] , by , that is , for @xmath1076 and @xmath1077\leq\text{im}\,u\leq -\tilde\eta\ ] ] one has @xmath1078 and @xmath1079 .
+ so , if for some @xmath1072 , one has @xmath1073 , it suffices to solve on @xmath1080 .
we compute the derivative of @xmath1063 in the interior of @xmath1080 @xmath1081 thus , fixing @xmath1071 , * if @xmath204 is such that , for some @xmath1072 , one has @xmath1082 , for @xmath1083 , we estimate @xmath1084^{-1 } \\&\lesssim\frac1{(l - k)\left| \theta_{p , l}^{-1}\left(\frac{l\pi}{l - k}\right)- \tilde e\right|+\left|\log\left [ \left|\theta_{p , l}^{-1}\left(\frac{l\pi}{l - k}\right)- \tilde e\right|+\frac{\tilde\eta}{l - k } \right]\right| } \\&\lesssim\frac1{\log l } ; \end{split}\ ] ] * if @xmath204 is such that , for all @xmath1072 , one has @xmath1085 , for @xmath1086+i[-\eta(l - k),-\tilde\eta]$ ] , we estimate @xmath1087^{-1 } \\&\lesssim\frac1{(l - k)\left| \theta_{p , l}^{-1}\left(\frac{l\pi}{l - k}\right)- \tilde e\right|}\lesssim\frac1{l^{1-\alpha}}. \end{split}\ ] ] hence , for @xmath9 sufficiently large , @xmath1068 is a contraction on @xmath1080 .
equation thus admits a unique solution , say , @xmath1088 in the rectangle @xmath1036+i[-\eta(l - k),-\tilde\eta]$ ] .
this solution is a simple root of @xmath1089 .
hence , @xmath1090 is the only solution to equation in @xmath1036+i[-\eta(l - k),-\tilde\eta]$ ] .
+ by , for @xmath9 sufficiently large , for @xmath1032 , both the equations @xmath1091 can be rewritten as @xmath1092 in @xmath1036+i[-\eta(l - k),-\tilde\eta]$ ] .
+ thus , each of the equations in admits a single solution in @xmath1036+i[-\eta(l - k),-\tilde\eta]$ ] and this root is simple ; moreover , this solution , say , @xmath1093 satisfies @xmath1094 ; indeed , the bounds and guarantee that one can apply rouch s theorem on the disk @xmath1095 for any @xmath632 . + thus , we have proved the [ le:9 ] pick @xmath31 as above .
then , there exists @xmath230 such that , for @xmath9 sufficiently large s.t .
@xmath186 , the resonances in @xmath231 $ ] are the energies @xmath1096 defined by @xmath1097 belonging to @xmath231 $ ] .
let us complete the proof of theorem [ thr:6 ] that is , prove that , for @xmath987 sufficiently small , for @xmath9 sufficiently large such that @xmath128 , is the unique resonance in @xmath1098 + i\left[-\eta,0\right]$ ] ; recall that @xmath207 is defined in .
+ therefore , we first note that the taylor expansion of @xmath1099 , and the quantization condition imply that @xmath1100 as re@xmath1101 and @xmath1102im@xmath1103 .
+ moreover , as @xmath1104 $ ] using and , we compute @xmath1105 \right).\end{gathered}\ ] ] thus , one has @xmath1106
\right)\\ + o\left(\left(\frac{\log l}l\right)^2\right).\end{gathered}\ ] ] as @xmath1093 solves , using and , we thus obtain that @xmath1107 using again re@xmath1101 and @xmath1102im@xmath1103 .
+ taking into account , this complete the proof of theorem [ thr:2 ] . proposition [ pro:3 ] is an immediate consequence of theorem [ thr:2 ] , the definition of @xmath1108 and the standard asymptotics of @xmath1109 near @xmath227 i.e. @xmath1110 .
.1cmto prove proposition [ pro:2 ] , it suffices to notice that , under the assumptions of proposition [ pro:2 ] , the bound on the derivative of @xmath1111 on the the rectangle @xmath1080 becomes @xmath1112 thus , as a solution to , @xmath1113 admits an asymptotic expansion in inverse powers of @xmath9 . plugging this into yields the asymptotic expansion for the resonance .
then , follows from the computation of the first terms .
theorem [ thr:3 ] is an immediate consequence of theorem [ thr:19 ] , the fact that the functions are analytic in the lower complex half - plane and have only finitely many zeros there and the argument principle . using the same notations as above , we can write @xmath1114 where @xmath1115 is the dirichlet laplacian on @xmath69 .
+ define the operators @xmath1116 for im@xmath1117 , @xmath1118 and @xmath1119 have a non vanishing imaginary part of the same sign ; hence , the complex number @xmath1120 does not vanish .
thus , by rank one perturbation theory , ( see e.g. @xcite ) , we thus know that @xmath1121 and @xmath1122 are invertible and that @xmath1123 and @xmath1124 thus , for im@xmath1117 , using schur s complement formula , we compute @xmath1125 where @xmath1126 is the adjoint of @xmath1127 and @xmath1128 now , when coming from im@xmath401 and passing through @xmath25 , the complex numbers @xmath1118 and @xmath1119 keep imaginary parts of the same positive sign ; thus , the two operator - valued functions @xmath1129 and @xmath1130 can be analytically continued through @xmath25 from the upper to the lower complex half - plane ( as operators respectively from @xmath1131 to @xmath1132 and from @xmath1133 to @xmath1134 ) .
+ when coming from the upper half - plane and passing through @xmath1135 and @xmath1136 $ ] , also provides an analytic continuation of @xmath1137 . definition and formula immediately show that the poles of these continuations only occur at the zeros of the function @xmath1138 when continued from the upper half - plane through the sets @xmath243 and @xmath244 $ ] ( these sets are finite unions of intervals ) .
+ this completes the proof of theorem [ thr:25 ] .
as for the periodic potential , for the random potential , we start with a description of the function @xmath865 ( see ) , that is , with a description of the spectral data for the dirichlet operator @xmath53 .
we recall a number of results on the dirichlet eigenvalues of @xmath53 that will be used in our analysis .
+ it is well known that , under our assumptions , in dimension one , the whole spectrum of @xmath1139 is in the localization region ( see e.g. @xcite ) that is [ thr:18 ] there exists @xmath1140 and @xmath1071 such that , one has @xmath1141 and @xmath1142 where @xmath1143 and @xmath1144 . the supremum is taken over the functions @xmath839 that are borelian and compactly supported . as a consequence
, one can define localization centers e.g. by means of the following results [ le:2 ] fix @xmath1145 a sequence of scales i.e. @xmath107 as @xmath19 .
there exists @xmath1140 such that , for @xmath9 sufficiently large , with probability larger than @xmath1146 , if 1 .
@xmath1147 is a normalized eigenvector of @xmath53 associated to @xmath1148 in @xmath43 , 2 .
@xmath1149 is a maximum of @xmath1150 in @xmath121 , then , for @xmath1151 , one has @xmath1152 this result can e.g. be applied for the scales @xmath1153 . in this case , the probability estimate of the bad sets ( i.e. when the conclusions of lemma [ le:3 ] does not hold ) is summable .
the point @xmath1154 is a localization center for @xmath1148 or @xmath1147 .
it is not defined uniquely , but , one easily shows that there exists @xmath177 such that for any two localization centers , say , @xmath1155 and @xmath1156 , one has @xmath1157 ( see @xcite ) . to fix ideas
, we set the localization center associated to the eigenvalue @xmath1148 to be the left most maximum of @xmath1158 .
+ we show [ le:3 ] for any @xmath83 , there exists @xmath177 and @xmath29 ( depending on @xmath1159 and @xmath20 ) such that , for @xmath1160 , for any sequence satisfying , with probability at least @xmath1161 , there exists at most @xmath1162 eigenvalues having a localization center in @xmath1163 .
we will use the fact that we are dealing with one - dimensional systems to improve upon the estimate .
we prove [ thr:10 ] for any @xmath493 and @xmath316 , there exists @xmath177 and @xmath29 ( depending on @xmath20 and @xmath309 ) such that , for @xmath1160 , with probability at least @xmath1161 , if @xmath1148 is an eigenvalue in @xmath43 associated to the eigenfunction @xmath1147 and the localization center @xmath1164 then , * if @xmath1165 , one has @xmath1166 * if @xmath1167 , one has @xmath1168 to analyze the resonances of @xmath352 ( resp .
@xmath353 ) , we shall use ( resp . and ) .
+ we now use these estimates as the starting point of a short digression from the main theme of this paper .
let us first state a corollary to theorem [ thr:10 ] , we prove [ thr:20 ] for any @xmath493 and @xmath316 , there exists @xmath328 such that , for @xmath9 sufficiently large ( depending on @xmath20 and @xmath309 ) , with probability at least @xmath1161 , if @xmath1148 is an eigenvalue in @xmath43 associated to the eigenfunction @xmath1147 and the localization center @xmath1164 then , for @xmath1169 and @xmath1170 , one has @xmath1171 compare to .
there are two improvements .
first , the unknown rate of decay @xmath279 is replaced by the lyapunov exponent @xmath1172 which was expected to be correct decay rate . indeed , for the one - dimensional discrete anderson model on the half - axis , it is well known ( see e.g. @xcite ) that , @xmath51-almost surely , the spectrum is localized and the eigenfunctions decay exponentially at infinity at a rate given by the lyapunov exponent . in theorem
[ thr:20 ] , we state that , with a good probability , this is true for finite volume restrictions .
+ second , in , we get both an upper and lower bound on the eigenfunction .
this is more precise than .
+ to our knowledge , such a result was not known up to now .
the strategy that we use to prove this result can be applied in a more general one - dimensional setting to obtain analogues of ( see @xcite ) .
+ we complement this with the much simpler [ le:8 ] for any @xmath177 and @xmath316 , there exists @xmath558 and @xmath29 ( depending on @xmath31 , @xmath20 and @xmath309 ) such that , for @xmath1160 , with probability at least @xmath1161 , if @xmath1148 is an eigenvalue in @xmath43 associated to the eigenfunction @xmath1147 and the localization center @xmath1164 then , * if @xmath1173 , one has @xmath1174 ; * if @xmath1175 , one has @xmath1176 .
the proof of this result is obvious and only uses the fact that the matrices in the cocycle defining the operator ( see section [ sec : estim - growth - eigenf ] ) are bounded that is , equivalently , that the solutions to the schrdinger equation grow at most exponentially at a rate controlled by the potential .
.2cmlet us return to the resonances in the random case and the description of the function @xmath1177 .
recall that in , the values @xmath531 are the eigenvalues @xmath1178 of @xmath53 and the coefficients @xmath1179 are defined in theorem [ thr:11 ] and by .
thus , theorem [ thr:10 ] describes the coefficients @xmath1180 coming into @xmath1177 and @xmath867 ( see and ) .
let us now state a few consequences of theorem [ thr:10 ] . + fix @xmath31 a compact interval in @xmath43 the almost sure spectrum of @xmath1139 . for @xmath95
, define @xmath1181 taking @xmath1182 in theorem [ thr:10 ] and using borel - cantelli argument , we obtain that @xmath1183 this and the continuity of the lyapunov exponent ( see e.g. @xcite ) guarantees that @xmath1184 where @xmath1185 is defined in theorem [ thr:4 ] .
+ to use the analysis performed in section [ sec : char - reson ] , we also need a description for the @xmath531 i.e. the dirichlet eigenvalues of @xmath53 .
therefore , we will use the results of @xcite , @xcite and @xcite ( see also @xcite ) .
+ we first recall the minami estimate satisfied by @xmath53 ( see e.g. @xcite and references therein ) : there exists @xmath177 such that , for @xmath1186 , one has @xmath1187)\right ) \leq c|i|^2(l+1)^2.\ ] ] by a simple covering argument , this entails the following estimate @xmath1188 thus , for @xmath1189 , a borel - cantelli argument yields , that @xmath1190 we are now going to prove the results stated in section [ sec : random - case ] . as for theorem [ thr:5 ] , this result follows from theorem [ thr:13 ] .
the point ( 1 ) is proved exactly as the point ( 1 ) in theorem [ thr:5 ] .
point ( 2 ) follows immediately from theorem [ thr:13 ] and .
this completes the proof of theorem [ thr:4 ] .
recall that @xmath291 . to prove ( 1 ) we proceed as follows .
the standard result guaranteeing the existence of the density of states @xmath262 ( see e.g. @xcite ) imply that , @xmath51 almost surely , one has @xmath1191 this , in particular , shows that , if @xmath275 is a compact interval , then , @xmath51 almost surely , for @xmath9 sufficiently large @xmath31 is covered by intervals of the form @xmath1192 $ ] and their number is of size @xmath1193 ( actually this holds for @xmath1194 $ ] if @xmath328 is chosen small enough ) .
moreover , the estimate guarantees that @xmath1195 ( for any @xmath1189 fixed ) for all @xmath1196 .
thus , theorems [ thr:13 ] , [ thr:14 ] and [ thr:12 ] and the estimate guarantee that , @xmath51 almost surely , all the resonances in the strip @xmath1197 are described by theorem [ thr:12 ] . indeed , for such a resonance the imaginary part must be larger than @xmath1198 ; thus , by theorem [ thr:13 ] , for every rectangle @xmath1199-i[e^{-l^\kappa},0)$ ] containing a resonance , one has @xmath1200 thus , @xmath500 and one can apply theorem [ thr:12 ] to compute the resonance .
+ let count the number of those resonances . therefore ,
let @xmath1201 where @xmath1202 is to be chosen . by and
, @xmath51 almost surely , one has @xmath1203 for all @xmath1204 such that @xmath1196 as long as this dirichlet eigenvalue is associated to a localization center in @xmath1205 ( actually it holds for @xmath1206 $ ] if @xmath328 is chosen small enough ) ; thus , we can apply theorems [ thr:12 ] and [ thr:14 ] to each of the @xmath531 that are associated to a localization center in @xmath1207 . by formula ,
each of these eigenvalues gives rise to a single simple resonance the imaginary part of which is of size @xmath1208 .
thus , it lies above the line im@xmath1209 for @xmath1210 .
actually , the estimate guarantees that @xmath1195 ( for any @xmath1189 fixed ) and theorem [ thr:14 ] shows that these resonances are the only ones up to a line im@xmath1211 .
moreover , by lemma [ le:3 ] , we know there at most @xmath1162 eigenvalues @xmath445 that do not have their localization center in @xmath1212 .
thus , we obtain , @xmath51 almost surely , @xmath1213 point ( 2 ) is proved in the same way .
pick @xmath294 .
in addition to what was used above , one uses the continuity of the density of states @xmath44 and lyapunov exponent @xmath813 .
assume @xmath268 is as in point ( 2 ) .
then , @xmath51 almost surely , the reasoning done above shows that , for any @xmath230 , there exists @xmath1050 such that , for @xmath1052 and @xmath1214 , for @xmath9 sufficiently large one has , @xmath1215\text
{ such } \\\text { that } - e^{\eta_\bullet\rho(e)\delta l } \lesssim e^{2\eta_\bullet\rho(e)\lambda\,l }
a_l\lesssim -e^{-\eta_\bullet\rho(e)\delta l } \end{aligned } \right\}\\ \leq \#\left\{z\text { resonance of } h^\bullet_{\omega , l}\text { in } r^\bullet(e,\lambda , l,\varepsilon,\delta ) \right\}\\\leq \#\left\ { \begin{aligned } \lambda_l
\text { e.v of } h^{\mathbb{n}}_{\omega , l}\text { in } e+\frac{\varepsilon}{2\,n(e)}\left[-1-\eta , 1+\eta\right]\text { such } \\\text { that } -e^{\eta_\bullet\rho(e)\delta l } \lesssim e^{2\eta_\bullet\rho(e)\lambda\,l } a_l\lesssim -e^{-\eta_\bullet\rho(e)\delta l } \end{aligned } \right\}\end{gathered}\ ] ] using theorem [ thr:10 ] and the continuity of the lyapunov exponent in conjunction with the definition of @xmath490 ( see and ) , we obtain that , @xmath51 almost surely , for any @xmath230 , there exists @xmath1050 such that , for @xmath1052 and @xmath1214 , for @xmath9 sufficiently large one has , @xmath1216 \\\text { with localization center in } i^\bullet(l,\delta,-\eta ) \end{aligned } \right\}\\ \leq \#\left\{z\text { resonance of } h^\bullet_{\omega , l}\text { in } r^\bullet(e,\lambda , l,\varepsilon,\delta ) \right\}\\\leq \#\left\ { \begin{aligned } \text{e.v of } h^{\mathbb{n}}_{\omega ,
l}\text { in } e+\frac{\varepsilon}{2\,n(e)}\left[-1-\eta , 1+\eta\right]\\\text { with localization center in } i^\bullet(l,\delta,\eta ) \end{aligned } \right\}\end{gathered}\ ] ] where @xmath1217+\llbracket -l\delta(1+\eta ) , l\delta(1+\eta)\rrbracket\\\intertext{and } \begin{aligned } i^{\mathbb{z}}(l,\lambda,\delta,\eta)&= \left(\left[\frac{l\lambda}2\right]+\llbracket -l\delta(1+\eta),l\delta(1+\eta))\rrbracket\right)\\ & \hskip1.5cm\cup \left ( \left[l\left(1-\frac{\lambda}2\right)\right]+ \llbracket - l\delta(1+\eta)),l\delta(1+\eta))\rrbracket\right ) .
\end{aligned}\end{gathered}\ ] ] now , using the exponential localization of the eigenfunctions , one has that , @xmath51 almost surely , for any @xmath230 , there exists @xmath1050 such that , for @xmath1052 and @xmath1214 , for @xmath9 sufficiently large one has , @xmath1218 \right\}\\ \leq \#\left\{z\text { resonance of } h^\bullet_{\omega , l}\text { in } r^\bullet(e,\lambda , l,\varepsilon,\delta ) \right\}\\\leq \#\left\ { \text{e.v of } h^{\mathbb{n}}_{\omega , l,\lambda,\delta,2\eta,\bullet}\text { in } e+\frac{\varepsilon}{2\,n(e)}\left[-1 - 2\eta , 1 + 2\eta\right ] \right\}\end{gathered}\ ] ] where @xmath1219 with dirichlet boundary conditions at the edges . +
this immediately yields point ( 2 ) for @xmath294 using the convergence of the density of states measure for our random model . the case
@xmath307 is dealt with in the same way .
+ as already said , point ( 3 ) is an `` integrated '' version of point ( 2 ) . using the same ideas as above , partitioning @xmath1220 s.t .
@xmath1221 centered in @xmath1222 , one proves @xmath1223
\right\}\\ \leq \#\left\{z\text { resonance of } h^\bullet_{\omega , l}\text { in } i+\left[-e^{-l^\kappa},-e^{-cl } \right]\right\}\\\leq \sum_{p=0}^p \#\left\ { \text{e.v of } h^+_{\omega , p , l,\bullet}\text { in } e_p+\frac{\varepsilon}{2\,n(e_p)}\left[-1 - 2\eta , 1 + 2\eta\right ] \right\}\end{gathered}\ ] ] where * @xmath1224 is the operator @xmath252 restricted to * * @xmath1225 if @xmath225 , * * to @xmath1226 if @xmath224 ; * @xmath1227 is the operator @xmath252 restricted to * * @xmath1228 if @xmath225 , * * to @xmath1229 if @xmath224 ; in the computation above , we used the continuity of both , the density of states @xmath44 and lyapunov exponent @xmath813 . thus , we obtain @xmath1230\right\}\\= l\left(\sum_{p=0}^p\inf(c\rho^{-1}(e_p),1)n(e_p)|i_p|+ o(1)\right)\\+\#\left\{z\text { resonance of } h^\bullet_{\omega , l}\text { in } i+\left(-\infty , e^{-l^\kappa}\right]\right\}.\end{gathered}\ ] ] the last term being controlled by theorem [ thr:27 ] , one obtains point ( 3 ) as the riemann sum in the right hand side above converges to the integral in the right hand side of as @xmath1231 .
this completes the proof of theorem [ thr:6 ] .
the proof of theorem [ thr:7 ] relies on ( * ? ? ?
* theorem 1.13 ) which describes the local distribution of the eigenvalues and localization centers @xmath1232 : namely , one has @xmath1233 where @xmath1234 for @xmath322 .
+ recall that @xmath1235 are the resonances of @xmath53 . by the argument used in the proof of theorem [ thr:6 ] , we know that , @xmath51 almost surely , all the resonances in @xmath1236 + i\left[-e^{-l^\kappa},0\right]$ ] are constructed from the @xmath1237 by formula . thus , up to renumbering ,
the rescaled real and imaginary parts ( see ) become @xmath1238 where @xmath1239 and @xmath1164 is the associated localization center ; here we used the continuity of @xmath813 .
+ on the other hand , for the resonances below the line in @xmath144im@xmath1240 , one has @xmath1241 .
so all these resonances are `` pushed upwards '' towards the upper half - plane .
hence , the statement of theorem [ thr:7 ] is an immediate consequence of .
using the computations of the previous section , as @xmath324 , theorem [ thr:8 ] is a direct consequence of ( * ? ? ?
* theorem 1.2 ) ( see also ( * ? ? ? * theorem 1.11 ) ) .
consider equations and . by theorem [ thr:10 ] and lemma [ le:3 ] , @xmath51 almost surely , for @xmath9 large , the number of @xmath1179 larger than @xmath1242
is bounded by @xmath1162 .
solving and in the strip @xmath144re@xmath877im@xmath1243 , we can write @xmath1244 and similarly @xmath1245 .
one then has @xmath1246 the count of the number of resonances given by the proof of theorems [ thr:11 ] and [ thr:24 ] then shows that the equations and where @xmath1177 and @xmath867 are respectively replaced by @xmath1247 and @xmath1248 have at most @xmath1162 solutions in the lower half plane .
the equations where @xmath1177 and @xmath867 are replaced by @xmath1247 and @xmath1248 we will call the @xmath1249-equations .
the analogue of theorems [ thr:13 ] , [ thr:14 ] and [ thr:12 ] for the @xmath1249-equations and theorem [ thr:10 ] show that the only solutions to the @xmath1249-equations in the strip @xmath144re@xmath1250im@xmath1251 are given by formulas and for the eigenvalues of the dirichlet problem associated to a localization center in @xmath1252 if @xmath225 and in @xmath1253 if @xmath224 .
thus , these zeros are simple and separated by a distance at least @xmath1254 from each other ( recall ) .
moreover , we can cover the interval @xmath31 by intervals of the type @xmath1255 $ ] , that is , one can write @xmath1256\ ] ] where @xmath1257 , @xmath1258 , @xmath1259 and @xmath1260
. + consider now the line @xmath144im@xmath1261 and its intersection with the vertical strip @xmath1255-i{\mathbb{r}}^+$ ] .
three thing may occur : + ( 1 ) either @xmath1262 ( the constant @xmath516 is defined in theorem [ thr:13 ] ) , then , on the interval @xmath1263-ie^{-\ell_l}$ ] , one has @xmath1264 this follows from the proof of theorem [ thr:13 ] ( see in particular , , and ) for some fixed @xmath50 ; recall that , on the interval @xmath1265 , one has @xmath1266 ; + ( 2 ) either @xmath1267 ( the constant @xmath516 is defined in theorem [ thr:14 ] ) , then , on the interval @xmath1263-ie^{-\ell_l}$ ] , one has again for a possibly different constant ; this follows from he proof of theorem [ thr:14 ] ( see in particular and ) ; + .45 \(3 ) if we are not in cases ( 1 ) or ( 2 ) , then the line @xmath144im@xmath1261 may cross @xmath510 ( defined in theorem [ thr:12 ] ; see also fig .
[ fig:6 ] ) ; we change the line @xmath144im@xmath1261
so as to enter @xmath1268 until we reach the boundary of @xmath510 and then follow this boundary getting closer to the real axis , turning around @xmath510 and finally reaching the line @xmath144im@xmath1261 again on the other side of @xmath510 and following it up to the boundary of @xmath1268 ( see figure [ fig:7 ] ) ; on this new line , the bound again holds ; moreover , this new line is closer to the real axis than the line @xmath144im@xmath1261 .. 1cmlet us call @xmath1269 the path obtained by gluing together the paths constructed in points ( 1)-(3 ) for @xmath1270 and the half - lines @xmath1271 -i[e^{-\ell_l},+\infty)$ ] and @xmath1272 -i[e^{-\ell_l},+\infty)$ ] ( see ) .
one can then apply rouch s theorem to compare the @xmath1249 equations to the equations and : by and , on the line @xmath1269 , one has @xmath1273 and @xmath1274 thus , the number of solutions to equations and below the line @xmath1269 is bounded by @xmath1162 ; as @xmath1269 lies above @xmath144im@xmath1261 , in the half - plane @xmath144im@xmath1243 , the equations and have at most @xmath1162 solutions .
we have proved theorem [ thr:27 ] . the first point in theorem [ thr:9 ]
is proved in the same way as point ( 2 ) in theorem [ thr:6 ] up to the change of scales , @xmath9 being replaced by @xmath1275 .
pick scales @xmath1276 satisfying such that @xmath1277 .
one has [ le:4 ] fix two sequences @xmath1278 and @xmath1279 such that @xmath1280 . with probability one , for @xmath9
sufficiently large , @xmath1281\right\}\\\leq\#\left\ { \text{e.v . of } h_{\omega
, l}\text { in } [ a_l , b_l ] \text { with loc . cent . in } \llbracket0,\ell_l\rrbracket \right\}\\\leq \#\left\{\text{e.v . of }
h_{\omega , \ell_l+2\ell'_l/\rho}\text { in } \left[a_l - e^{-\ell'_l } , b_l+e^{-\ell'_l}\right]\right\ } \end{gathered}\ ] ] where @xmath279 is given by lemma [ le:2 ] .
to prove lemma [ le:4 ] , we apply lemma [ le:2 ] to @xmath1282 ( i.e. for the operator @xmath1139 restricted to the interval @xmath1283 ) and @xmath1284 .
the probability of the bad set is the @xmath1008 , thus , summable in @xmath9 . using the localization estimate ,
one proves that * each eigenvalue of @xmath1285 is at a distance of at most @xmath1286 of an eigenvalue of @xmath53 with loc .
cent . in @xmath1287 ; * each eigenvalue of @xmath53 with loc .
. in @xmath1287 is at a distance of at most @xmath1286 of an eigenvalue of @xmath1288 .
lemma [ le:4 ] follows .
the first point in theorem [ thr:9 ] is then point ( 2 ) of theorem [ thr:6 ] for the operator @xmath1289 and @xmath1290 and the fact that @xmath1277 .. 1cmthe proof of the second statement in theorem [ thr:9 ] is very similar to that of theorem [ thr:7 ] .
fix @xmath31 a compact interval in @xmath1291 .
as @xmath1275 satisfies , one can find @xmath1292 also satisfying such that @xmath1293 .
for the same reasons as in the proof of theorem [ thr:7 ] , after rescaling , all the resonances in @xmath1294 outside the strip @xmath1295 are then pushed to either @xmath14 or @xmath1296 as @xmath19 .
+ on the other hand , the resonances in the strip @xmath1297 are described by .
the rescaled real and imaginary parts of the resonances ( see ) now become @xmath1298 and @xmath1299 . +
now , to compute the limit of @xmath1300 , using the exponential decay property , it suffices to use ( * ? ? ?
* theorem 1.14 ) .
let us note here that ( * ? ? ?
* condition ( 1.50 ) ) on the scales @xmath332is slightly stronger than .
that condition suffices is a consequence of the stronger localization property known in the present case ( compare theorem [ thr:10 ] to ( * ? ? ?
* assumption ( loc ) ) ) .
this completes the proof of the second point in theorem [ thr:9 ] . the final statement in [ thr:9 ] is proved in exactly the same way as theorem [ thr:8 ] . + the proof of theorem [ thr:9 ] is complete .
localization for the operator @xmath252 can be described by the following [ le:10 ] there exists @xmath1140 and @xmath1301 such that , @xmath51 almost surely , there exists @xmath1302 s.t . for @xmath9
sufficiently large , if 1 .
@xmath1147 is a normalized eigenvector of @xmath53 associated to @xmath1148 in @xmath43 , 2 .
@xmath1303 is a maximum of @xmath1150 in @xmath343 , then , for @xmath1304 , one has @xmath1305 moreover , the mapping @xmath1306 is measurable and @xmath1307 .
this result for our model is a consequence of theorem [ thr:18 ] ( see e.g. @xcite ) and ( * ? ? ?
* theorem 6.1 ) .
+ we thus obtain the following representation for the function @xmath358 @xmath1308 as @xmath1309 satisfies a dirichlet boundary condition at @xmath438 , one has @xmath1310 as @xmath1311 , the representation yields @xmath1312 this proves the first point in proposition [ pro:6 ] .
+ as a direct consequence of theorem [ thr:18 ] and the computation leading to theorem [ thr:19 ] ( see section [ sec : proof - theorem-7 ] ) , we obtain that there exists @xmath1313 s.t . for @xmath9
sufficiently large , with probability at least @xmath1314 , one has @xmath1315 taking @xmath1316 for some sufficiently small @xmath50 , this and rouch s theorem implies that , with probability @xmath1317 , the number of zeros of @xmath358 ( counted with multiplicity ) in @xmath1318 $ ] is bounded * from above by the number of resonances of @xmath1319 in @xmath1320 $ ] ; * from below by the number of resonances of @xmath1319 in @xmath1321 $ ] . where @xmath1322 $ ] and @xmath1323 $ ] if @xmath1324 $ ] .
+ here , to apply rouch s theorem , we apply the same strategy as in the proof of theorem [ thr:27 ] and construct a path bounding a region larger ( resp .
smaller ) than @xmath1325 $ ] ( resp .
@xmath1326 $ ] ) on which one can guarantee @xmath1327 .
+ now , we choose the constant @xmath587 ( see ) to be so small that @xmath1328 . applying point ( 3 ) of theorem [ thr:6 ] to @xmath1319 with this constant @xmath587
, we obtain that the number of resonances of @xmath1319 in @xmath1320 $ ] ( resp .
@xmath1321 $ ] ) is upper bounded ( resp lower bounded ) by @xmath1329 hence , we obtain the second point of proposition [ pro:6 ] .
the last point of this proposition is then an immediate consequence of the arguments developed to obtain the second point if one takes into account the following facts : * the minimal distance between the dirichlet eigenvalues of @xmath352 is bounded from below by @xmath1254 ( see ) , * the growth of the function @xmath583 near its zeros ( i.e. the resonances ) is well controlled by proposition [ pro:7 ] . indeed
, this implies that the resonances of @xmath352 are simple in @xmath1330 ( one can choose larger rectangles ) and that near each resonance one can apply rouch s theorem to control the zero of @xmath358 . note that this also yields @xmath51-almost surely , there exists @xmath1331 such that @xmath1332 this completes the proof of proposition [ pro:6] .. 2cmtheorem [ thr:21 ] is a consequence of the following [ thr:28 ] there exists @xmath1313 such that , @xmath51 almost surely , for @xmath873 sufficiently large one has @xmath1333 where * @xmath1334 is the matrix @xmath512 ( see ) constructed from the dirichlet data of @xmath1335 ( see ) on @xmath1336 ; * @xmath1337 is the function @xmath511 ( see ) constructed from the dirichlet data of @xmath1338 ( see ) on @xmath1336 ; theorem [ thr:28 ] is proved exactly as theorem [ thr:19 ] except that one uses the localization estimates instead of the combes - thomas estimates .
+ theorem [ thr:21 ] is then an immediate consequence of the estimate .
indeed , this implies that if @xmath1339 is a resonance for e.g. @xmath1340 in @xmath1341 $ ] , then @xmath1342 . by the last point of proposition [ pro:6 ] , @xmath51 almost surely , we know that the multiplicity of the zeros of @xmath358 is bounded by @xmath1343 . moreover , for the zeros of @xmath358 in @xmath1344 , we know the bound .
this bound and imply that @xmath1345 this yields point ( 2 ) in theorem [ thr:21 ] by an application of rouch s theorem .
point ( 1 ) is obtained in the same way using proposition [ pro:7 ] that gives @xmath1346 the case of @xmath1347 is dealt with in the same way .
+ this completes the proof of theorem [ thr:21 ] . in the present section
we are going to prove theorems [ thr:10 ] and [ thr:20 ] . at the end of the section
, we also prove the simpler lemma [ le:3 ] . + the proof of theorem [ thr:10 ] relies on locally uniform estimates on the rate of growth of the cocycle attached to the schrdinger operator that we present now .
define @xmath1348 where @xmath1349 we start with an upper bound on the large deviation of the growth rate of the cocycle that is uniform in energy .
fix @xmath1350 and @xmath1351 . for one part ,
the proof of theorem [ thr:10 ] relies on the following [ le:6 ] let @xmath1186 be a compact interval . for any @xmath493 , there exists @xmath1352 and @xmath230 such that , for @xmath1353 and any @xmath558 , one has @xmath1354 where we recall that @xmath1355 denotes the left shift ( i.e. if @xmath1356 then @xmath1357_n= \omega_{n+1}$ ] for @xmath239 ) and @xmath1358 @xmath101 times . at the heart of this result
is a large deviation principle for the growth rate of the cocycle ( see ( * ? ? ?
* and theorem 6.1 ) ) .
as it also serves in the proof of theorem [ thr:10 ] , we recall it now
. one has [ le:7 ] let @xmath1186 be a compact interval .
for any @xmath493 , there exists @xmath1352 and @xmath230 such that , for @xmath1353 , one has @xmath1359 while this result is not stated as is in @xcite , it can be obtained from ( * ? ? ?
* lemma 6.2 and theorem 6.1 ) . indeed , by inspecting the proof of (
* lemma 6.2 and theorem 6.1 ) , it is clear that the quantities involved ( in particular , the principal eigenvalue of @xmath1360 in ( * ? ? ?
* theorem 4.3 ) ) are continuous functions of the energy @xmath268 .
thus , taking this into account , the proof of ( * ? ? ?
* theorem 6.1 ) yields , for our cocycle , a convergence that is locally uniform in energy , that is , . .2cmto prove theorem [ thr:10 ] , in addition to lemma [ le:6 ] , we also need to guarantee that the uniform lower bound on the growth rate of the cocycle .
we need this bound at least on the spectrum of @xmath53 with a good probability .
actually , this is the best one can hope for : a uniform bound in the style of will not hold ; it is the same phenomenon that bars the lyapunov exponent from being continuous versus to upper semi - continuous . + we prove [ le:5 ] fix @xmath31 a compact interval and @xmath493 .
pick @xmath476 s.t .
@xmath1361 . for @xmath1362 ,
if @xmath1363 , define @xmath1364 finally , define @xmath1365 .
+ recall that @xmath1178 are the eigenvalues of @xmath53 and let @xmath1164 be the associated localization centers . + for @xmath1366 , define the sets @xmath1367 and @xmath1368 then , the sets @xmath1369 are measurable and , for any @xmath493 , there exists @xmath230 and @xmath1370 such that , for @xmath1371 , one has @xmath1372 here , the constant @xmath987 is the one given by .
first , let us explain the meaning of lemma [ le:5 ] .
as , by lemma [ le:6 ] , we already control the growth of the cocycle from above , we see that in the definitions of the set @xmath1373 resp .
@xmath1374 , it would have sufficed to require @xmath1375 resp .
@xmath1376 hence , what lemma [ le:5 ] measures is that the probability that the cocycle at energy @xmath1377 leading from a localization center @xmath1378 to either @xmath14 or @xmath9 decays at a rate smaller than the rate predicted by the lyapunov exponent .
+ the sets @xmath1369 are the sets of bad configurations i.e. the events when the rate of decay of the solution is far from the lyapunov exponent . indeed , for @xmath51 outside
@xmath1379 i.e. if the reverse of the inequalities defining @xmath1380 hold , when @xmath1381 and @xmath1382 , then , we know that the eigenfunction @xmath1383 has to decay from the center of localization @xmath1378 ( which is a local maximum of its modulus ) towards the edges at a rate approximately @xmath1384 ( where @xmath309 measures the error ) . the eigenfunction being normalized , at the localization center it is of size at least @xmath1385
. this will entail the estimates and with a good probability .
.1cmthere is a major difference in the uniformity in energy obtained in lemmas [ le:5 ] and [ le:6 ] . in lemma [ le:5 ] , we do not get a lower bound on the decay rate that is uniform all over @xmath31 : it is merely uniform over the spectrum inside @xmath31 ( which is sufficient for our purpose as we shall see ) .
the reason for this difference in the uniformity between lemma [ le:6 ] and [ le:5 ] is the same that yields that the lyapunov exponent @xmath1386 is in general only upper semi - continuous and not lower semi - continuous ( in the present situation , it actually is continuous ) . .2cmwe
postpone the proofs of lemmas [ le:6 ] and [ le:5 ] to the end of this section and turn to the proofs of theorems [ thr:10 ] and [ thr:20 ] . by lemma [ le:6 ] , as @xmath1387 , with probability at least @xmath1388 , for @xmath1353 and any @xmath558 ,
one also has @xmath1389 now pick @xmath1390 where @xmath177 is to be chosen later on . we know that
, with probability @xmath1391 satisfying @xmath1392 for @xmath1353 and any @xmath1393 $ ] and any @xmath1394 $ ] , one also has @xmath1395 let @xmath1147 be a normalized eigenfunction associated to the eigenvalue @xmath1396 with localization center @xmath1164 . by the definition of the localization center
, one has @xmath1397 by the eigenvalue equation , for @xmath1151 , one has @xmath1398 hence , by and , with probability at least @xmath1399 , if @xmath1400 , for @xmath1401 , one has @xmath1402 and , for @xmath1403 , one has @xmath1404 on the other hand , by the definition of the dirichlet boundary conditions , we know that @xmath1405 thus , @xmath1406 and @xmath1407 thus , for @xmath1408 where we have set @xmath1409 and @xmath1410 , we know that @xmath1411 and @xmath1412 thus , we obtain that , for @xmath1413 , one has @xmath1414 and @xmath1415 the estimates given by lemma [ le:5 ] on the probability of @xmath1416 and @xmath1417 for @xmath1390 and the estimate then imply that , with a probability at least @xmath1418 , the bounds , , and hold .
thus , picking @xmath1390 for @xmath177 sufficiently large ( depending only on @xmath987 , thus , on @xmath309 and @xmath20 ) , these bounds hold with a probability at least @xmath1161 .
this complete the proof of theorem [ thr:10 ] .
[ rem:1 ] one may wonder whether the uniform growth estimate given by lemmas [ le:6 ] and [ le:5 ] is actually necessary in the proof of theorem [ thr:10 ] . that it is is due to the fact that both the eigenvalue @xmath1148 and the localization center @xmath1164 ( and , thus , the vector @xmath1419 ) depend on @xmath51 .
thus , is not sufficient to estimate the second term in the left hand sides of and . to prove theorem [ thr:20 ]
, we follow the strategy that led to the proof of theorem [ thr:10 ] .
first , note that and provide the expected lower bounds on the eigenfunction with the right probability .
as for the upper bound , by , using the conclusions of theorem [ thr:10 ] and the bounds given by lemma [ le:6 ] , we know that , e.g. for @xmath1420 @xmath1421 if @xmath1422 i.e. @xmath1423 . + for @xmath1424 one reasons similarly and , thus , completes the proof of theorem [ thr:20 ] .
[ rem:2 ] actually , as the proof shows , the results one obtains are more precise than the claims made in theorem [ thr:20 ] ( see @xcite ) . the proofs for the two sets @xmath1369 are the same .
we will only write out the one for @xmath1425 .
let us first address the measurability of @xmath1425 .
the functions @xmath1426 and @xmath1427 are continuous ( as the eigenvalues and eigenvectors of finite dimensional matrices depending continuously on the parameter @xmath1428 ) .
thus , for fixed @xmath1204 , the sets @xmath1429 and @xmath1430 are open ( we used the definition of @xmath1164 as the left most localization center ( see theorem [ thr:10 ] ) ) .
this yields the measurability of @xmath1425 . .2cmwe
claim that @xmath1431 where @xmath1432 denotes the spectral projector associated to @xmath53 on the open set @xmath1374 . indeed ,
if one has @xmath1433 for all @xmath1204 then the left hand side of vanishes and the right hand side is non negative . on the other hand ,
if , for some @xmath1204 , one has @xmath1434 and @xmath1435 then , we compute @xmath1436 by the definition of @xmath1164 .
+ an important fact is that , by construction ( see lemma [ le:5 ] ) , the set of energies @xmath1374 does not depend on @xmath1437 . hence , denoting by @xmath1438 the expectation with respect to @xmath1437 and @xmath1439 the expectation with respect to @xmath1440 , we compute @xmath1441 as @xmath1437 is assumed to have a bounded compactly supported distribution and as @xmath1374 does not depend on @xmath1437 , a standard spectral averaging lemma ( see e.g. ( * ? ? ?
* theorem 11.8 ) ) yields @xmath1442 where @xmath1443 denotes the lebesgue measure .
thus , we obtain @xmath1444 by lemma [ le:7 ] and the fubini - tonelli theorem , we know that @xmath1445 taking the expectation of both sides of and plugging this into , we obtain @xmath1446 in the same way , one obtains @xmath1447 this completes the proof of lemma [ le:5 ] .
this proof can be seen as the analogue of the so - called kotani trick for products of finitely random matrices ( see e.g. @xcite ) . the basic idea of this proof is to use the estimate , in particular , the exponentially small probability and some perturbation theory for the cocycles so as to obtain a uniform estimate .
+ let @xmath987 be given by .
fix @xmath1448 and write @xmath1449\text { where } e^{-\eta'(l+1)}/2\leq e_{j+1}-e_j\leq 2e^{-\eta'(l+1)};\ ] ] thus , @xmath1450 .
+ we now want to estimate what happens for @xmath1451 $ ] . therefore ,
using and @xmath1452 where @xmath1453 we compute @xmath1454 where @xmath1455 clearly , as the random variables have compact support , one has the uniform bound @xmath1456 thus one has @xmath1457 hence , for @xmath1458 fixed , one computes @xmath1459 if @xmath1460 and @xmath9 is sufficiently large ( depending only on @xmath1461 and @xmath516 ) .
+ we now assume that @xmath1458 satisfies @xmath1460 and pick @xmath1462 .
pick @xmath1463 small to be fixed later .
assume moreover that @xmath9 is so that @xmath1464 where @xmath1465 is defined in lemma [ le:7 ]
. then , by lemma [ le:7 ] , for @xmath1466 , one has 1 .
either @xmath1467 ; then , one has @xmath1468 2 . or @xmath1469 ; then , by , with probability at least equal to @xmath1470 , one has @xmath1471 define @xmath1472 by definition , one has @xmath1473 clearly , for any fixed sequence @xmath1474 , the random variables @xmath1475 are independent . hence , by , for a fixed @xmath1476 , one has @xmath1477 thus , for @xmath276 , one has @xmath1478 hence , with probability at least @xmath1479 , we know that @xmath1480 using estimates and for the remaining terms in the product below , for any given @xmath1481-uple @xmath1482 , one obtains @xmath1483 hence , with probability at least @xmath1484 , for @xmath1485 , we estimate @xmath1486l + [ c-(\rho(e_j)+\delta)(1-\varepsilon)]l_\delta
l}\\&\leq l^{l_0}e^{[\rho(e_j)+\delta+(c-\rho(e_j)-\delta)\varepsilon]l + [ c-(\rho(e_j)+\delta)(1-\varepsilon)]l_\delta l_0}. \end{split}\ ] ] it remains now to choose the quantities @xmath1461 , @xmath1458 and @xmath308 so that the following requirements be satisfied @xmath1487l_\delta l_0}{l+1}&\leq\frac{\delta}{2(\rho(e_j)+\delta)}. \end{split}\ ] ] fixing @xmath308 small , picking @xmath1488 and setting @xmath1489 where @xmath1071 , we see that all the conditions in are satisfied for @xmath9 sufficiently large . moreover , one has @xmath1490 .
+ plugging this and the last estimate in into , we obtain that , with probability at least @xmath1491 , for any @xmath1492 ( see ) , for @xmath1451 $ ] , one has @xmath1493 as @xmath279 is continuous ( see e.g. @xcite ) , one gets that , for any @xmath493 , for @xmath9 sufficiently large , with probability at least @xmath1491 , one has , for any @xmath293 , @xmath1494 hence , as @xmath1387 , one has @xmath1495 .
+ using the fact that the probability measure on @xmath1496 is invariant under the shift ( it is a product measure ) , we obtain .
this completes the proof of lemma [ le:6 ] .
assume the realization @xmath51 is such that the conclusions of lemma [ le:2 ] hold in @xmath31 for the scales @xmath1153 .
fix @xmath221 and let @xmath1497 be the set of indices of the eigenvalues @xmath1178 of @xmath53 having a localization center in @xmath1498 .
fix @xmath1499 and consider the projector on the sites in @xmath1500 i.e. @xmath1501 .
+ consider the following gram matrices @xmath1502 where @xmath1503 and @xmath1504 by definition , the rank of @xmath1505 is bounded by @xmath1162 .
moreover , as by one has @xmath1506 , one has @xmath1507 thus , picking @xmath1508 yields that , for @xmath9 sufficiently large , @xmath1509 is invertible and its rank , @xmath262 , is bounded by that of @xmath1510 i.e. by @xmath1162 .
this yields @xmath1511 and the proof of lemma [ le:3 ] is complete . using the same notations as above , we can write @xmath1512 where * @xmath1115 is the dirichlet laplacian on @xmath69 , * @xmath1513 on @xmath1514 with dirichlet boundary conditions at @xmath14 .
define the operators @xmath1515 for im@xmath1117 , the numbers @xmath1516 and @xmath1119 have non vanishing imaginary parts of the same sign ; hence , the complex number @xmath1517 does not vanish .
thus , by rank one perturbation theory , ( see e.g. @xcite ) , we thus know that @xmath1518 and @xmath1519 are invertible and that @xmath1520 thus , for im@xmath1117 , using schur s complement formula , we compute @xmath1521 where @xmath1126 is the adjoint of @xmath1127 and @xmath1522 thus , considering the lower right coefficient of this matrix , we see that , when coming from upper half - plane through @xmath389 , @xmath1523 can be continued meromorphically to the lower half plane ( as an operator from @xmath1133 to @xmath1134 ) only if @xmath1524 can be meromorphically ( as an operator from @xmath1131 to @xmath1132 ) . + as @xmath1525 can be analytically continued ( see section [ sec : proof - theorem ] ) , by , the meromorphic continuation of @xmath1524 will exist if and only if the complex valued map @xmath1526 can be meromorphically continued from the upper half - plane through @xmath389 .
fix @xmath51 s.t .
the spectrum of @xmath1527 be equal to @xmath43 and pure point ( this set of @xmath51 is of probability @xmath63 ) . as @xmath1528 is a cyclic vector for @xmath1527 , for @xmath268 an eigenvalue of @xmath1527 , one then has @xmath1529 hence
, if the analytic continuation of @xmath1530 would exist , on @xmath389 , it would be equal to @xmath1531 this in turn would imply that the same equality holds on the whole upper half - plane , thus , in view of the definition of @xmath1530 , that holds on the whole upper half plane : this is absurd !
thus , we have proved that , @xmath51 almost surely , @xmath1523 does not admit a meromorphic continuation through @xmath389 .
.1cmlet us now prove that the spectral measure of @xmath388 in @xmath389 is purely absolutely continuous .
therefore , it suffices ( see e.g. @xcite and ( * ? ? ?
* theorem 11.6 ) ) to prove that , for all @xmath314 , one has @xmath1532 using , and , for im@xmath1117 , we compute @xmath1533 for @xmath1534 , @xmath1535 , @xmath1536 and @xmath1537 thus , to prove the absolute continuity of the spectral measure of @xmath388 in @xmath389 , it suffices to prove that , for @xmath314 , one has @xmath1538 this is the case as * the signs of the imaginary parts of @xmath1539 and @xmath1540 are the same ( negative if im@xmath426 and positive if im@xmath401 ) , * for @xmath1541 , @xmath1540 has a finite limit when @xmath377 , * for re@xmath1542 , the imaginary part of @xmath1540 does not vanish in the limit @xmath377 .
so , we have proved the part of theorem [ thr:26 ] concerning the absence of analytic continuation of the resolvent of @xmath388 through @xmath389 and the nature of its spectrum in this set .
+ let us now turn to the continuation through @xmath391 .
as the spectrum of @xmath1527 outside @xmath43 is discrete , the representation immediately shows that the resolvent @xmath1543 can be continued through @xmath391 , the poles of the continuation being given by the zeros of the function @xmath1544 let us now prove the last part of theorem [ thr:26 ] .
the proof relies again on .
we pick @xmath1545 where @xmath1159 is determined by theorem [ thr:18 ] for @xmath1527 . then , for @xmath1534 and @xmath1535 , using the cauchy - schwartz inequality , for im@xmath1117 , we compute @xmath1546 for @xmath1547 a compact interval , we know that , for @xmath1534 and @xmath1535 , * @xmath1548 by the combes - thomas estimates ; * @xmath1549 by the characterization of localization in @xmath43 for @xmath1527 .
it suffices now to estimate the last term in using a standard decomposition of rank one perturbations ( see e.g. @xcite ) , one write @xmath1550 where @xmath1551 and @xmath1552 only depend on @xmath1553 .
thus , as @xmath1554 have a bounded density , for im@xmath1117 , one has @xmath1555 thus , we have proved that , for @xmath1556 $ ] a compact interval , for @xmath1545 and some @xmath1557 , for @xmath1534 and @xmath1535 , one has @xmath1558 in the same way , using and , one proves that @xmath1559 thus , we have proved that , for some @xmath1557 , one has @xmath1560 hence , we know that the spectrum of @xmath388 in @xmath1561 $ ] ( as @xmath1562 can be taken arbitrary contained in this set ) is pure point associated to exponentially decaying eigenfunctions ( see e.g. @xcite ) .
this completes the proof of theorem [ thr:26 ] .
in this section we study the eigenvalues and eigenvectors of @xmath15 ( see remark [ rem:6 ] ) near an energy @xmath767 that is an eigenvalue of both @xmath122 and @xmath120 ( see the ends of sections [ sec : dirichl - eigenv ] and [ sec : dirichl - eigenf ] ) .
we keep the notations of sections [ sec : dirichl - eigenv ] and [ sec : dirichl - eigenf ] .
+ let @xmath1563 ( resp .
@xmath1564 ) be normalized eigenvectors of @xmath122 ( resp .
@xmath117 ) associated to @xmath1565 .
thus , by and , we can pick , for @xmath239 and @xmath1566 , @xmath1567 assume @xmath186 and , for @xmath1568 , define @xmath1569 by @xmath1570 thus , one has @xmath1571 recall that @xmath1572 ( see sections [ sec : dirichl - eigenv ] and [ sec : dirichl - eigenf ] ) ; thus , by , one has @xmath1573 moreover , as @xmath15 converges to @xmath122 in strong resolvent sense , for @xmath328 sufficiently small , for @xmath9 sufficiently large , @xmath1574 has no spectrum in the compact @xmath1575\cup[\varepsilon/2,2\varepsilon]$ ] .
let @xmath1576 be the spectral projector onto the interval @xmath1577 $ ] that is @xmath1578 . by
, one computes @xmath1579 thus , one gets @xmath1580 define @xmath1581 the gram matrix of @xmath1582 then reads id@xmath1583 .
orthonormalizing @xmath1582 into @xmath1584 and , computing the matrix elements of @xmath1585 in this basis , we obtain @xmath1586 thus , we obtain that the eigenvalues of @xmath15 near @xmath767 are given by @xmath1587 and the eigenvectors by @xmath1588 .
in particular , their components at @xmath14 and @xmath9 are asymptotic to non vanishing constants .
michael aizenman , jeffrey h. schenker , roland m. friedrich , and dirk hundertmark .
finite - volume fractional - moment criteria for anderson localization . , 224(1):219253 , 2001 .
dedicated to joel l. lebowitz .
franois germinet and frdric klopp .
spectral statistics for the discrete anderson model in the localized regime . in _ spectra of random operators and related topics
_ , rims kkyroku bessatsu , b27 , pages 1124 .
inst . math .
( rims ) , kyoto , 2011 .
werner kirsch .
an invitation to random schrdinger operators . in _
random schrdinger operators _ , volume 25 of _ panor .
synthses _ , pages 1119 .
france , paris , 2008 . with an appendix by frdric klopp .
frdric klopp .
resonances for `` large '' ergodic systems in one dimension : a review . in _ proceedings of the conference ``
spectra days 2010 '' , santiago ( chile ) , september 20 - 24 , 2010 . _ ,
arxiv : http://fr.arxiv.org/abs/1107.0534 .
herv kunz and bernard souillard .
random schrdinger operators . some rigorous results . in _ mathematical problems in theoretical physics
phys . , lausanne , 1979 ) _ , volume 116 of _ lecture notes in phys .
_ , pages 8387 .
springer , berlin , 1980 . j. n. mather .
on nirenberg s proof of malgrange s preparation theorem . in _ proceedings of liverpool singularities -
symposium i _ , number 192 in lecture notes in mathematics , berlin , 1971 .
springer verlag .
barry simon .
spectral analysis of rank one perturbations and applications . in _ mathematical quantum theory .
schrdinger operators ( vancouver , bc , 1993 ) _ , pages 109149 .
soc . , providence , ri , 1995 .
m. zworski . quantum resonances and
partial differential equations . in _ proceedings of the international congress of mathematicians ,
iii ( beijing , 2002 ) _ , pages 243252 , beijing , 2002 .
higher ed . press . | the present paper is devoted to the study of resonances for one - dimensional quantum systems with a potential that is the restriction to some large box of an ergodic potential . for discrete models
both on a half - line and on the whole line , we study the distributions of the resonances in the limit when the size of the box where the potential does not vanish goes to infinity . for periodic and random potentials ,
we analyze how the spectral theory of the limit operator influences the distribution of the resonances .
.5 cm rsum .
dans cet article , nous tudions les rsonances dun systme unidimensionnel plong dans un potentiel qui est la restriction un grand intervalle dun potentiel ergodique .
pour des modles discrets sur la droite et la demie droite , nous tudions la distribution des rsonances dans la limite de la taille de bote infinie .
pour des potentiels priodiques et alatoires , nous analysons linfluence de la thorie spectrale de loprateur limite sur la distribution des rsonances . |
sample - to - sample fluctuations of the conductance of disordered systems have a universal regime , in which they are independent of the mean conductance . the requirement for these universal conductance fluctuations
@xcite is that the sample size should be small compared to the localization length .
the mean conductance is then much larger than the conductance quantum @xmath5 .
the same condition applies to the universality of conductance fluctuations in ballistic chaotic quantum dots @xcite , although there is no localization in these systems .
random - matrix - theory ( rmt ) has the universal limit @xmath6 for the variance of the conductance @xmath7 in units of @xmath5 . here
@xmath1 is the number of modes transmitted through each of the two ballistic point contacts that connect the quantum dot to electron reservoirs . since the mean conductance @xmath8
, the condition for universality remains that the mean conductance should be large compared to the conductance quantum . in the present paper
we will show that there is actually an upper limit on @xmath1 , beyond which rmt breaks down in a quantum dot and the universality of the conductance fluctuations is lost . since the width @xmath9 of a point contact should be small compared to the size @xmath10 of the quantum dot , in order to have chaotic scattering , a trivial requirement is @xmath11 , where @xmath0 is the number of transverse modes in a cross - section of the quantum dot .
( in two dimensions , @xmath12 and @xmath13 , with @xmath14 the fermi wavelength . )
by considering the quantum - to - classical crossover , we arrive at the more stringent requirement @xmath15 with @xmath16 the lyapunov exponent and @xmath17 the ergodic time of the classical chaotic dynamics .
the requirement is more stringent than @xmath11 because , typically , @xmath18 and @xmath17 are both equal to the time of flight @xmath19 across the system , so the exponential factor in eq .
( [ nllsqrtm ] ) is not far from unity .
expressed in terms of time scales , the upper limit in eq .
( [ nllsqrtm ] ) says that @xmath17 should be larger than the ehrenfest time @xcite @xmath20.\ ] ] the condition @xmath21 which we find for the universality of conductance fluctuations is much more stringent than the condition @xmath22 for the validity of rmt found in other contexts .
@xcite here @xmath23 is the mean dwell time in the quantum dot , which is @xmath24 in any chaotic system .
the outline of this paper is as follows . in sec .
ii we describe the quantum mechanical model that we use to calculate @xmath25 numerically , which is the same stroboscopic model used in previous investigations of the ehrenfest time @xcite .
the data is interpreted semiclassically in sec .
iii , leading to the crossover criterion ( [ nllsqrtm ] ) .
we conclude in sec . iv .
) . inside each lead
we plot the initial and final coordinates of trajectories which are transmitted from the left to the right lead after at most 3 iterations ( with @xmath26 ) . the points cluster along narrow `` transmission bands '' .
[ fig_lead ] , width=283,height=283 ] the physical system we have in mind is a ballistic ( clean ) quantum dot in a two - dimensional electron gas , connected by two ballistic leads to electron reservoirs .
while the phase space of this system is four - dimensional , it can be reduced to two dimensions on a poincar surface of section.@xcite the open kicked rotator @xcite is a stroboscopic model with a two - dimensional phase space .
we summarize how this model is constructed , following ref . .
one starts from the closed system ( without the leads ) .
the kicked rotator describes a particle moving along a circle , kicked periodically at time intervals @xmath27 .
we set to unity the stroboscopic time @xmath27 and the plank constant @xmath28 .
the stroboscopic time evolution of a wave function is given by the floquet operator @xmath29 , which can be represented by an @xmath30 unitary symmetric matrix .
the even integer @xmath0 defines the effective planck constant @xmath31 . in the discrete coordinate representation ( @xmath32 , @xmath33 )
the matrix elements of @xmath29 are given by @xmath34 where @xmath35 is the map generating function , @xmath36 and @xmath37 is the kicking strength .
the eigenvalues @xmath38 of @xmath29 define the quasi - energies @xmath39 .
the mean spacing @xmath40 of the quasi - energies plays the role of the mean level spacing @xmath41 in the quantum dot . to model a pair of @xmath1-mode ballistic leads
, we impose open boundary conditions in a subspace of hilbert space represented by the indices @xmath42 .
the subscript @xmath43 labels the modes and the superscript @xmath44 labels the leads .
a @xmath45 projection matrix @xmath46 describes the coupling to the ballistic leads .
its elements are @xmath47 the mean dwell time is @xmath48 ( in units of @xmath19 ) . the matrices @xmath46 and @xmath29 together determine the quasi - energy dependent scattering matrix @xmath49^{-1}{\cal f}p^{\rm t}.\label{sdef}\ ] ] the symmetry of @xmath29 ensures that @xmath35 is also symmetric , as it should be in the presence of time - reversal symmetry . by grouping together the @xmath1 indices belonging to the same lead
, the @xmath50 matrix @xmath35 can be decomposed into 4 sub - blocks containing the @xmath51 transmission and reflection matrices , @xmath52 the conductance @xmath7 ( in units of @xmath5 ) follows from the landauer formula @xmath53 the open quantum kicked rotator has a classical limit , described by a map on the torus @xmath54 .
the classical phase space , including the leads , is shown in fig .
[ fig_lead ] .
the map relates @xmath55 at time @xmath56 to @xmath57 at time @xmath58 : @xmath59 the classical mechanics becomes fully chaotic for @xmath60 , with lyapunov exponent @xmath61 . for smaller @xmath37 the phase space
is mixed , containing both regions of chaotic and of regular motion .
we will restrict ourselves to the fully chaotic regime in this paper .
variance of the conductance fluctuations obtained numerically by varying @xmath62 with fixed lead positions .
error bars indicate the scatter of values obtained for different lead positions .
results are shown as a function of @xmath63 , for two values of the dwell time @xmath64 .
the dashed line is the rmt prediction @xmath65 .
, width=302 ] to calculate the conductance ( [ gtt ] ) we need to invert the @xmath30 matrix between square brackets in eq .
( [ sdef ] ) .
we do this numerically using an iterative procedure .
@xcite the iteration can be done efficiently using the fast - fourier - transform algorithm to calculate the application of @xmath66 to a vector .
the time required to calculate @xmath35 scales as @xmath67 , which for large @xmath0 is quicker than the @xmath68 scaling of a direct inversion .
the memory requirements scale as @xmath0 , because we need not store the full scattering matrix to obtain the conductance .
we distinguish two types of mesoscopic fluctuations in the conductance .
the first type appears upon varying the quasi - energy @xmath62 for a given scattering matrix @xmath69 .
since these fluctuations have no classical analogue ( the classical map ( [ mapdef ] ) being @xmath62-independent ) , we refer to them as quantum fluctuations .
the second type appears upon varying the position of the leads , so these involve variation of the scattering matrix at fixed @xmath62 .
we refer to them as sample - to - sample fluctuations .
they have both a quantum mechanical component and a classical analogue .
one could introduce a third type of fluctuations , involving both variation of @xmath62 and of the lead positions .
we have found ( as expected ) that these are statistically equivalent to the sample - to - sample fluctuations at fixed @xmath62 , so we need not distinguish between fluctuations of type two and three .
, but now for an ensemble in which the lead positions and the quasi - energy are both varied .
the dashed lines are the sum of the rmt value ( [ vrmt ] ) and the classical result ( [ vcl ] ) .
results are shown for three values of the kicking strength @xmath37 .
open symbols are for the dwell time @xmath70 and closed ones for @xmath71 .
[ fig2_a ] , width=302 ] we have calculated the variance @xmath72 of the conductance , either by varying @xmath62 at fixed lead positions ( quantum fluctuations ) or by varying both @xmath62 and lead positions ( sample - to - sample fluctuations ) . since the quantum interference pattern is completely different only for energy variations of order of the thouless energy @xmath73 , we choose a number @xmath74 of equally spaced values of @xmath62 in the interval @xmath75 .
we take @xmath76 different lead positions , randomly located at the @xmath77-axis in fig .
[ fig_lead ] . to investigate the quantum - to - classical crossover , we change @xmath31 while keeping the dwell time @xmath48 constant .
the results are plotted in figs .
[ fig1 ] and [ fig2_a ] .
we interpret the numerical data by assuming that the variance of the conductance is the sum of two contributions : a universal quantum mechanical contribution @xmath78 given by random - matrix theory and a nonuniversal quasiclassical contribution @xmath79 determined by sample - to - sample fluctuations in the classical transmission probabilities .
the rmt contribution equals @xcite @xmath80 in the presence of time - reversal symmetry .
the classical contribution is calculated from the classical map ( [ mapdef ] ) , by determining the probability @xmath81 of a particle injected randomly through lead @xmath82 to escape via lead @xmath83 .
since the conductance is given semiclassically by @xmath84 , we obtain @xmath85 we plot @xmath86 in fig .
[ fig2_a ] ( dashed curves ) , for comparison with the results of our full quantum mechanical calculation .
the agreement is excellent . upon changes of lead positions , calculated numerically from the map ( [ mapdef ] ) .
the data is shown for four values of the dwell time @xmath74 , as a function of the lyapunov exponent @xmath87 .
the dotted lines are the analytical prediction ( [ varpestimate ] ) , with fit parameters @xmath88 and @xmath89 ( the same for all data sets ) .
[ fig3 ] , width=302 ] we now would like to investigate what ratio of time scales governs the crossover from quantum to classical fluctuations . to estimate the magnitude of the sample - to - sample fluctuations in the classical transmission probability , we use results from ref . .
there it was found that the starting points ( and end points ) of transmitted trajectories are not homogeneously distributed in phase space .
instead , they cluster together in nearly parallel , narrow bands . these transmission bands are clearly visible in fig .
[ fig_lead ] .
the largest band has an area @xmath90 determined by the ergodic time @xmath17 .
this is the time required for a trajectory to explore the whole accessible phase space .
the values of @xmath17 and @xmath91 depend on the degree of collimation of the beam of trajectories injected into the system .
@xcite for our model , without collimation , one has @xmath17 of order unity ( one stroboscopic period ) and @xmath92 .
the typical transmission band has an area @xmath93 which is exponentially smaller than @xmath94 ( since @xmath95 ) . as the position of the lead
is moved around , transmission bands move into and out of the lead .
the resulting fluctuations in the transmission probability @xmath81 are dominated by the largest band . since there is an exponentially large number @xmath96 of typical bands , their fluctuations average out .
the total area in phase space of the lead is @xmath97 , so we estimate the mean squared fluctuations in @xmath98 at @xmath99 with @xmath100 and @xmath17 of order unity .
we have tested this functional dependence numerically for the map ( [ mapdef ] ) , and find a reasonable agreement ( see fig .
[ fig3 ] ) .
both the exponential dependence on @xmath16 and the quadratic dependence on @xmath64 are consistent with the data .
we find @xmath89 of order unity , as expected .
( [ vcl ] ) and ( [ varpestimate ] ) imply @xmath101 in fig .
[ fig2_b ] we plot the same data as in fig .
[ fig2_a ] , but now as a function of @xmath102 .
we see that the functional dependence ( [ varform ] ) is approached for large dwell times . rescaled to show the approach to a single limiting curve in the large dwell time limit .
the solid line is calculated from eq .
( [ varform ] ) , with the same parameters @xmath103 , @xmath104 as in fig .
[ fig2_b ] , width=302 ] the quantum fluctuations of rmt dominate over the classical fluctuations if @xmath105 . using the estimate ( [ varpestimate ] ) ,
this amounts to the condition @xmath106
\equiv \tau_{e}\label{crossovercondition}\ ] ] that the ergodic time exceeds the ehrenfest time .
notice that condition ( [ crossovercondition ] ) is always satisfied if @xmath107 .
this agrees with the findings of ref .
, that the breakdown of rmt starts when @xmath108 .
in summary , we have presented both a fully quantum mechanical and a semiclassical calculation of the quantum - to - classical crossover from universal to non - universal conductance fluctuations .
the two calculations are in very good agreement , without any adjustable parameter ( compare data points with curves in fig .
[ fig2_a ] ) .
we have also given an analytical approximation to the numerical data , which allows us to determine the parametric dependence of the crossover .
we have found that universality of the conductance fluctuations requires the ergodic time @xmath17 to be larger than the ehrenfest time @xmath109 .
this condition is much more stringent than the condition that the dwell time @xmath74 should be larger than @xmath109 , found previously for universality of the shot noise in a quantum dot .
@xcite the universality of the excitation gap in a quantum dot connected to a superconductor is also governed by the ratio @xmath110 rather than @xmath111 @xcite , as is the universality of the weak localization effect .
@xcite these two properties have in common that they represent ensemble averages , rather than sample - to - sample fluctuations .
we propose that what we have found here for the conductance is generic for other transport properties : that the breakdown of rmt with increasing @xmath109 occurs when @xmath112 for ensemble averages and when @xmath113 for the fluctuations .
this has immediate experimental consequences , because it is much easier to violate the condition @xmath113 than the condition @xmath112 . to test this proposal ,
an obvious next step would be to determine the ratio of time scales that govern the breakdown of universality of the fluctuations in the superconducting excitation gap .
the numerical data in refs . and was interpreted in terms of the ratio @xmath114 , but an alternative description in terms of the ratio @xmath115 was not considered . one final remark about the distinction between classical and quantum fluctuations , explained in sec .
it is possible to suppress the classical fluctuations entirely , by varying only the quasi - energy at fixed lead positions . in that case
we would expect the breakdown of universality to be governed by @xmath110 instead of @xmath111 .
our numerical data ( fig .
[ fig1 ] ) does not show any systematic deviation from rmt , probably because we could not reach sufficiently large systems in our simulation .
_ note added _ : our final remark above has been criticized by jacquod and sukhorukov [ ] .
they argue that the numerical data of fig .
[ fig1 ] ( and similar data of their own ) does not show any systematic deviation from rmt because quantum fluctuations remain universal if @xmath112 .
their argument relies on the assumption that the effective rmt of ref .
holds not only for the classical fluctuations ( as we assume here ) , but also for the quantum fluctuations .
the effective rmt says that quantum fluctuations are due to a number @xmath116 of transmission channels with an rmt distribution .
universality of the quantum fluctuations is then guaranteed even if @xmath117 , as long as @xmath118 is still large compared to unity .
this line of reasoning , if pursued further , contradicts the established theory @xcite of the @xmath109 dependence of weak localization .
rmt says that the weak localization correction @xmath119 is independent of the number of channels @xcite .
validity of the effective rmt at the quantum level would therefore imply that weak localization remains universal if @xmath112 , as long as @xmath120 .
this contradicts the result @xmath121 of refs . and .
this work was supported by the dutch science foundation nwo / fom .
j.t . acknowledges the financial support provided through the european community s human potential programme under contract hprn
ct2000 - 00144 , nanoscale dynamics .
b. l. altshuler , jetp lett .
* 41 * , 648 ( 1985 ) .
p. a. lee and a. d. stone , phys .
* 55 * , 1622 ( 1985 ) .
h. u. baranger and p. a. mello , phys .
* 73 * , 142 ( 1994 ) .
r. a. jalabert , j .- l .
pichard , and c. w. j. beenakker , europhys .
* 27 * , 255 ( 1994 ) .
m. g. vavilov and a. i. larkin , phys .
b * 67 * , 115335 ( 2003 ) .
p. g. silvestrov , m. c. goorden , and c. w. j. beenakker , phys .
b * 67 * , 241301 ( 2003 ) .
a. lodder and yu .
v. nazarov , phys .
b * 58 * , 5783 ( 1998 ) .
f. borgonovi , i. guarneri , and d. l. shepelyansky , phys .
a * 43 * , 4517 ( 1991 ) . f. borgonovi and i. guarneri , j. phys .
a * 25 * , 3239 ( 1992 ) .
y. v. fyodorov and h .- j .
sommers , jetp lett .
* 72 * , 422 ( 2000 ) .
a. ossipov , t. kottos , and t. geisel , europhys .
* 62 * , 719 ( 2003 ) . | we calculate the system - size - over - wave - length ( @xmath0 ) dependence of sample - to - sample conductance fluctuations , using the open kicked rotator to model chaotic scattering in a ballistic quantum dot coupled by two @xmath1-mode point contacts to electron reservoirs .
both a fully quantum mechanical and a semiclassical calculation are presented , and found to be in good agreement .
the mean squared conductance fluctuations reach the universal quantum limit of random - matrix - theory for small systems . for large systems
they increase @xmath2 at fixed mean dwell time @xmath3 .
the universal quantum fluctuations dominate over the nonuniversal classical fluctuations if @xmath4 .
when expressed as a ratio of time scales , the quantum - to - classical crossover is governed by the ratio of ehrenfest time and ergodic time . |
the physical properties of ( quasi-)one - dimensional electron systems are often very different from their 2d counterpart , for example , in ( spin-@xmath13 ) heisenberg antiferromagnets ( afm ) : the ( quasi-1d ) ladders exhibit in the groundstate an odd - even effect , i.e. , odd - leg ladders have one gapless spinon - mode , while even - leg ladders have a spin - gap @xcite .
the 2d heisenberg model however , has two gapless magnon - modes and long - range order @xcite .
similarly , ( quasi-)1d conductors are generally luttinger liquids ( ll ) @xcite , while 2d and 3d electron systems are ( usually ) fermi liquids ( fl ) . for the afm heisenberg model , the dimensional crossover from 1d to 2d has been studied in ref . @xcite . via the correspondence between number of legs @xmath0 and temperature @xmath14 , @xmath15
, the nonlinear sigma - model allows one in this case to calculate various physical properties ; in particular , the spin - gap present in even - leg ladders vanishes exponentially as a function of the number @xmath0 of coupled spin - chains .
coupled lls @xcite and the dimensional crossover to 2d have been investigated by many different authors @xcite .
one of the reasons was the search for a non - fl ( i.e. , possible ll - like ) phase in 2d @xcite .
however , no clear evidence for such a phase has been found so far @xcite . in the half - filled and lightly doped hubbard model
, there is a _ strong competition _ between cooper , umklapp , and afm processes . for an unbiased determination of the dominant phase , it is therefore advantageous to use the _ renormalization group _ ( rg ) technique @xcite . in the ( weakly interacting ) 2d case , such rg studies found afm at and close to half - filling and @xmath10-wave superconductivity for sufficient hole doping @xcite .
the dimensional crossover from 1d to 2d was only investigated _
away _ from half - filling ( without umklapp and afm processes ) @xcite .
the authors concluded that the system becomes a fl for @xmath16 . in a previous work ,
we have studied in detail the groundstate properties of the half - filled and lightly doped @xmath17 and 4-leg hubbard ladders in the case of weak interactions @xcite .
in contrast to the large @xmath18 case , where the half - filled hubbard model converges onto the heisenberg model of spin-1/2 @xcite , in the small @xmath18 case , the _ charge degrees _ of freedom are still _ present _ , which gives insight in the effect of doping away from half - filling . here
, we analyze the crossover from 1d to 2d for the half - filled ( and lightly doped ) ladders .
we will show that depending on the energy - scale ( respectively length - scale ) , the physical properties are very different and that the phases associated with these properties are separated by crossover energies which are a function of @xmath0 . in particular , the quasi-1d analog of the 2d phase is present at higher energies , implying that the crossover between _ different _ phases can be studied for @xmath0 _ finite_. an advantage in quasi-1d compared to the 2d case is the limited number of different interactions .
this allows us to obtain ( some ) _ analytical _ results and to disentangle the interplay between afms ( two gapless magnon - modes ) , isls ( insulating spin - liquid , i.e. , only short - range correlations ) , and superconductivity . the 2d limit ( @xmath19 )
can then be taken within the _ same _ phase . for weak interactions , the hubbard ladders are equivalent to a @xmath0-band model , characterized by fermi velocities @xmath20 . at half - filling ,
due to nesting of the fermi surface ( fs ) , the fermi velocities of the bands @xmath21 and @xmath22 are equal ( to be called `` band pairs '' below ) , such that afm and umklapp processes take place . using rg techniques , we find that the half - filled ladders exhibit ( in the spin - sector ) an odd - even effect only below a crossover energy @xmath1 $ ] ( @xmath2 and @xmath3 ) , but are 2d - like ( insulating ) afms above the scale @xmath4 .
the reason is , that below @xmath4 only interactions within band pairs @xmath5 [ and within the band @xmath6 for @xmath0 odd ] are present , while above @xmath4 all bands are interacting with each other .
we obtain an analytical expression for the hamiltonian which is similar to the 2d heisenberg afm . analyzing the charge - sector by bosonization techniques @xcite ,
we find , that the mott insulator is of the same type as in @xmath7 half - filled two - leg ladders .
the scale @xmath4 decreases fast to zero as the number of chains @xmath0 goes to infinity .
the 2d limit is thus an insulating afm , in agreement with what has been found in studies of the weakly interacting 2d hubbard model @xcite .
doping away from half - filling ( at energies above @xmath4 ) , we find that the effect of a doping @xmath23 is the same as of a finite @xmath24 : doped holes suppress the interactions between unpaired bands and thereby reduce the antiferromagnetic correlations .
the effect of the doping on the remaining interactions within a band pair @xmath5 is then the same as when doping @xmath7 uncoupled isls ; there is a spin - gap and @xmath10-wave - like phase coherence between bands @xmath21 and @xmath22 . at a higher doping value @xmath9 , phase coherence between all band pairs sets in and the @xmath0-leg ladders
become a 2d - like @xmath10-wave superconductor [ @xmath11 for @xmath12 .
the low - energy physics is given by a @xmath10-wave bcs hamiltonian .
the origin of the superconducting instability is the same as for the 2d case @xcite : a kohn - luttinger - type attraction @xcite is generated by afm processes @xcite .
furthermore , while at half - filling the antiferromagnetic correlations are commensurate , upon doping , they become incommensurate .
[ f : sphd ] gives a schematic overview .
summarizing , the @xmath0-leg hubbard model exhibits below the length - scale @xmath25 almost the same physical properties as the 2d model ( i.e. , the same correlation functions dominate ) , but in contrast to the 2d case , this quasi-1d analog can be treated ( partially ) _
we do all our calculations for _ even _
@xmath0 and only refer to the odd @xmath0 case where it is of interest . in sec .
ii , we introduce the @xmath0-leg ladder hamiltonian , discuss the various interactions , and specify along which `` path '' the dimensional crossover is studied . in sec .
iii , we briefly revisit the results obtained previously in ref .
@xcite and analyze the half - filled hamiltonian as a function of the energy - scale .
we show that there exist for finite @xmath0 two clearly different phases , separated by a crossover energy @xmath4 .
we derive an ( analytical ) expression for @xmath4 . in sec .
iv , the phase above the energy @xmath4 is investigated in detail and shown to be a 2d - like afm . finally , in sec .
v , we treat the effect of doping away from half - filling .
the @xmath0-leg ( @xmath0-chain ) hubbard model is given by @xmath26 , where the kinetic energy is @xmath27 and the interaction term is @xmath28 the hopping matrix elements along- and perpendicular to the chains are denoted by @xmath29 and @xmath30 respectively and @xmath31 is the on - site repulsion . for our approach , it is advantageous first to diagonalize @xmath32 , @xmath33 where @xmath34 and @xmath35 are the creation- and annihilation operators for the band @xmath21 . for open boundary conditions perpendicular to the legs , the dispersion relations @xmath36
are given by ( @xmath37 ) @xmath38 .
\label{eq : dispreln}\ ] ] the fermi momenta in each band @xmath39 are determined by the chemical potential @xmath40 and the filling @xmath41 .
since we are only interested in the low - energy physics , we linearize the dispersion relation at the fs , resulting in fermi velocities @xmath42 .
we introduce operators @xmath43 for right and left movers at the fs . at half - filling , @xmath44 , and @xmath45^{2}},\ ] ]
where @xmath46 .
for @xmath47 , @xmath48 , which leads to a singular behavior for large @xmath0 ( the quasi-1d analog of the van hove singularities in 2d ) . for the following ,
we take @xmath49 in order to avoid singularities .
note that the fs is then not flat @xcite , but that we still have nesting , i.e. , @xmath50 .
the crucial difference between quasi-1d and 2d are the interactions which control the low - energy physics .
the system is quasi-1d and only a finite number of different interactions play a role , provided the energy difference between two neighboring bands is larger than the largest energy - scale of the system , i.e. , @xmath51 .
therefore , we investigate the crossover between _ different phases _ ( isl and afm ) in quasi-1d , @xmath52 , as a function of the _ energy _
( this can partially be done analytically ) .
the 2d limit , @xmath53 , is then taken within the _ same _ phase .
in other words , along this particular `` path '' , the dimensional crossover from 1d to 2d can be treated in a controlled way .
since we are interested in the small-@xmath18 low - energy physics , we have to take into account only the processes , which are present down to arbitrarily small energy - scales .
for the half - filled case , this implies that only umklapp processes where the momenta add up exactly to @xmath54 , @xmath55 are relevant .
this condition is only fulfilled for the band pairs @xmath56 , @xmath50 .
when doping away from half - filling , the chemical potential ( measured from its value at half - filling ) introduces a low - energy cutoff for the various umklapp processes .
the interactions which have to be taken into account initially are then different . at half - filling ,
the interacting part of the hamiltonian consists of forward ( f ) , cooper ( c ) , and umklapp ( u ) processes between 2 different bands and ( non-)umklapp processes between 4 different bands ( in our notation , we follow refs .
@xcite ) , @xmath57 , where @xmath58\end{aligned}\ ] ] where the u(1 ) and su(2 ) currents are defined as @xmath59 and @xmath60 where again @xmath61 and the @xmath62 are the pauli matrices ( @xmath63 ) . due to symmetry , @xmath64 and @xmath65 .
the umklapp currents are @xmath66 and @xmath67 where @xmath68 .
the 4-band interactions involve processes between two different band pairs @xmath56 and @xmath69 , @xmath70 where @xmath71 in our case @xmath72 implying @xmath73 , @xmath74 .
note that the 4-band couplings describe _ afm processes _
, i.e. , the terms @xmath75 result in a momentum transfer of @xmath54 ( along the chains ) .
using the rg method , we study the half - filled @xmath0-leg ladders as a function of the _ energy - scale_. we show that there exist two clearly different phases , separated by a crossover energy @xmath76 .
the phase below @xmath4 is an isl , i.e. , 4-band - afm processes are suppressed and the physics is dominated by 2-band umklapp and cooper processes @xcite . in contrast , in the phase above @xmath4 ( but still below the bandwidth @xmath29 ) the 4-band - afm processes are large and dominating ( this phase is investigated in detail in sec .
first ( sec .
iii a ) , we revisit the results for @xmath0-leg ladders obtained previously in ref .
@xcite and give a criterion for the couplings allowing for a ( rather ) precise distinction between the two phases . in sec .
iii b , we calculate the crossover energy @xmath4 as a function of @xmath0 . in ref .
@xcite , we have shown , that at half - filling ( for @xmath0 even ) the band pair @xmath56 scales towards an isl and is frozen out at the energy - scale @xmath77 , where @xmath78 . the low - energy hamiltonian is then the same as of @xmath7 half - filled two - leg ladders .
for increasing @xmath0 , the initial - value @xmath79 ( and therefore the energy - scale ) , which leads to this decoupling ( in particular to a suppression of the 4-band - afm processes ) and an isl phase , decreases fast .
next , using bosonization techniques @xcite , we show that the rg flow of the _ luttinger liquid parameter _ for spin triplet pairing , @xmath80 , allows one to determine the energy - scale , where the spin - gap opens .
this energy - scale can then be interpreted as the crossover energy between the isl phase at low energies and the afm phase at higher energies ( which has no spin - gap ) .
bosonization is the method of rewriting fermionic creation ( annihilation ) operators in terms of field operators @xmath81 and @xmath82 satisfying the commutation relation @xmath83=i\delta(x - y)$ ] ( @xmath84 labels the charge and spin - modes for the different bands ) .
the dual field of @xmath81 is defined as @xmath85 .
the fields @xmath81 contain the ( physical ) charge respectively spin of the particle , i.e. , the charge / spin - density is @xmath86 , while the fields @xmath87 can be interpreted as the phase of the charge - part respectively the spin - part ( for ladders , see , e.g. , refs .
we introduce the fields @xmath88 and @xmath89 combining the fields of the bands @xmath21 and @xmath90 ( @xmath91 ) , e.g. , the spin - fields @xmath92 describe triplet / singlet pairs between the momenta @xmath39 and @xmath93 .
the bosonized hamiltonian takes the following form .
the noninteracting hamiltonian @xmath32 plus the contributions of the charge - currents @xmath94 and @xmath95 and the @xmath96-component of the spin - currents @xmath97 and @xmath98 can be written as @xmath99 , \label{eq : h0bos}\ ] ] where @xmath91 and @xmath100 and the luttinger liquid parameters ( llp ) are @xmath101 and @xmath102 the velocities are @xmath103 .
the interacting part of the @xmath0-leg ladder hamiltonian contains ( apart form other terms ) the expression @xcite @xmath104 fields which appear in a cosine become `` pinned '' in order to minimize the energy , e.g. , the term @xmath105 implies ( quasiclassicaly ) @xmath106 and the corresponding spin - mode acquires a _
gap_. the llps determine the exponents of correlation functions @xcite and together with cosine - terms the pinned fields . here ,
as we show in the following , the flow of the llp @xmath80 can be used to determine the energy - scale where the spin - gap in the @xmath107 sector opens .
note that initially ( i.e. , at high energies ) @xmath108 and @xmath109 ( and @xmath110 , @xmath111 ) . for the @xmath0-leg ladders ,
the rg flow of the couplings is such that @xmath112 and @xmath113 at low energies , implying @xmath114 .
hence , while @xmath115 is always @xmath116 , there is an energy , where @xmath80 becomes smaller than 1 , i.e. , @xmath117 .
the canonical transformation @xmath118 eliminates @xmath80 from @xmath119 ( it just renormalizes the velocities ) and results in @xmath120 a large @xmath80 therefore tends to depin the field @xmath121 , while a small @xmath80 results in a strong pinning of the field @xmath121 ( for the dual field @xmath122 , it is vice versa ) .
the crossover between the two regimes takes place at @xmath123 ( since we are at finite energy , it is a crossover and not a transition ) .
the isl phase therefore requires @xmath124 ( @xmath125 ) and @xmath126 can be used in the rg flow to determine a crossover energy .
we note that for the pure sine - gordon model ( at zero temperature ) , the value @xmath127 rigorously separates the massive from the massless phase @xcite . performing a numerical integration of the full set of rg equations ( rges ) @xcite
, we calculate the crossover energy @xmath128 , which separates ladder - like from 2d - afm - like behavior ( the rges are a generalization of the 3-leg rges given in ref .
@xcite ; the rges relevant for the afm phase are given below ) . in fig .
[ f : cflow ] , we have plotted the strong - coupling value of the ratio of a 4-band afm and a 2-band cooper coupling , @xmath129 , and of @xmath130 for @xmath131 ( we fix the strong - coupling value of @xmath132 at the bandwidth @xmath29 ) .
we take @xmath133 , because for @xmath134 , the band pair @xmath135 is close to the van hove singularities , such that its behavior would no more be typical for the system .
when the number of chains , @xmath0 , increases , the initial - value @xmath79 which leads to a suppression of the 4-band - afm couplings decreases fast .
similarly , for large @xmath0 , the @xmath136 becomes only negative when @xmath79 is very small .
note that for the other band pairs @xmath56 , the corresponding energy - scales are _ lower _ , since the velocities @xmath20 are larger and the gaps are of the order of @xmath137 .
hence , both the 4-band - afm processes and the @xmath126 have a strong and similar dependence on the energy - scale .
for small enough initial - values @xmath79 , the _ differences _ between the fermi velocities lead to a decoupling into band pairs .
however , since @xmath138 , for increasing @xmath0 , this decoupling becomes suppressed and the physics is dominated by 4-band - afm processes .
we calculate the crossover energy @xmath4 as follows .
we determine initial - values @xmath139 such that @xmath140 at the scale where the first coupling of the band pair @xmath135 has grown up to the bandwidth @xmath29 . for initial - values @xmath141 , a spin - gap then opens , starting in the band pair @xmath135 , and the 4-band - afm processes become suppressed .
we find that @xmath142 can be fitted by @xmath143 , where @xmath3 ( for @xmath144 , we obtain @xmath145 ) and @xmath146 , see fig .
[ f : ucross ] . on the other hand
, an initial - value @xmath18 corresponds to an energy - scale ( gap ) @xmath147 . the crossover energy @xmath4 is therefore given by @xmath148,\ ] ] where @xmath149 .
note that the energy @xmath4 is also the upper limit for the spin - gap in the isl phase . for large @xmath18 ( heisenberg afm ) , the corresponding scale ( spin - gap )
is @xmath150 , where @xmath151 ( see also table i ) @xcite . while for large @xmath18 the spin - gap decreases exponentially , in the small-@xmath18 case the decrease is _
double - exponentially_. as a result , the `` phase '' of the 2d system is for finite @xmath0 _ present _ above an energy - scale @xmath128 , where @xmath152 for @xmath16 .
equivalently , below the length - scale @xmath25 , the dominating correlation functions are the same as in the 2d system .
first ( sec .
iv a ) , using the rg method , we calculate the asymptotic ratios of the couplings and the charge - gap of the phase present above @xmath4 . comparing our results with the flow well away from half - filling @xcite ,
we conclude , that afm processes ( at high energies ) are necessary for an instability in the rg flow of the @xmath0-leg hubbard model . in sec .
iv b , we show that the phase above @xmath4 is the quasi-1d counterpart of a 2d insulating afm . in particular , there is no odd - even effect and the interacting part of the hamiltonian is similar to the 2d heisenberg afm . using bosonization techniques , we obtain , that the system is mott insulating and that the insulator is of the same type as in the half - filled two - leg ladder , see sec .
iv c. for the spin - sector , we find the same physical properties as for the heisenberg afm , i.e. , two gapless magnon modes , spinon confinement , and long - range order in the 2d limit .
we give the rges of the 4-band - afm couplings for large @xmath0 , i.e. , keeping in the rges only sums over @xmath0 products of couplings and dropping the ( 2-band ) contributions of the order of 1 .
we then determine ( analytically ) the asymptotic ratios of the couplings for hubbard initial - values and the size of the charge - gap .
the rges of the 4-band interactions are given by a sum over @xmath7 products of 4-band couplings @xmath153 where the energy - scale is related to @xmath154 by @xmath155 .
defining @xmath156 we find from the above rges , that @xmath157 given the hubbard initial - values @xmath158 and @xmath159 , we obtain @xmath160 for all @xmath154 ( this fixed point is stabilized by the 2-band interactions ) .
since @xmath161\ ] ] and @xmath162 , the @xmath163 flow to 0 , @xmath164 for increasing @xmath154 . from the flow of the @xmath165 , we then calculate the flow of the 4-band couplings and obtain the asymptotic ratios @xmath166 .the table shows the correspondence between crossover energy ( spin - gap ) and energy - scale of the 2d system for small and large @xmath18 .
the exponentially small charge - gap for small @xmath18 leads to a double exponentially suppressed 1d-2d crossover energy ( the result for the spin - gap in the large @xmath18 case is from ref .
[ 6 ] ) . [ cols="<,^,^,^,^",options="header " , ] for the charge - sector
, we then find , that the same fields are pinned as in the half - filled two - leg ( respectively @xmath0-leg ) hubbard ladder @xcite , i.e. , @xmath167 and @xmath168 .
the type of mott insulator is thus the same in 1d and 2d .
note that pinning of @xmath169 instead of @xmath170 leads to another type of insulator .
the difference between these two types of insulators becomes physically relevant upon ( hole ) doping . while the second type most likely
becomes a fl , the first type becomes ( in case we have a spin - gap ) a superconductor , since doping only depins the @xmath171-fields but does not destroy the phase coherence , i.e. , the @xmath172-fields remain pinned , see ref .
@xcite .
in contrast , in the spin - sector , both the fields @xmath92 and the dual fields @xmath173 appear in a cosine , resulting in a competition between different `` phases '' , i.e. , pinning of the field vs. its dual field . however , since @xmath108 and @xmath174 ( for a comparison , see sec . iii a ) , it is more favorable to pin @xmath175 and @xmath176 than the corresponding dual fields .
the physical interpretation is then the following .
the pinning of @xmath176 leads to a spinon confinement , leaving as physical particles spin 1 magnons .
note that only the _ differences _ @xmath177 appear in @xmath178 , such that the total magnon mode(s ) , given by @xmath179 and @xmath180 , remains gapless .
the @xmath54 and @xmath181 mode then result from a superposition of left / right going modes along the chains ( since we use open boundary conditions , the transvers momentum is always positive ) . for the calculation of the spin - spin correlation function , we first rewrite the real - space operators @xmath182 in terms of the band operators @xmath183 and then the band operators in terms of the bose - operators .
using that only products of fields , which contain the pinned charge - fields @xmath184 and @xmath170 give non - vanishing contributions to the spin - spin correlation function , the real - space spin - operator at the position @xmath185 takes the form @xmath186 where @xmath187 $ ] and at half - filling , @xmath50 . the abelian bosonization scheme used here , _
breaks _ non - abelian symmetries , i.e. , the su(2 ) spin - symmetry is broken down to u(1 ) , see , e.g. , refs . @xcite . in our case ,
only the @xmath188 and @xmath189 components of the spin - spin correlation function give then straightforwardly the correct physical result ( this `` problem '' occurs also in a single chain with a spin - gap @xcite ) .
the products of fields appearing in eq .
( [ eq : scf ] ) are then rewritten in terms of bose - operators according to @xmath190 the charge - fields and @xmath176 are pinned and can be set to zero . using that @xmath191
, we express @xmath175 in terms of the total spin - mode @xmath192 , @xmath193 .
since @xmath32 is gaussian in the fields @xmath194 and @xmath192 , the spin - spin correlation function takes the form ( for details about such calculations , consult , e.g. , ref .
@xcite ) @xmath195 where @xmath196 ( for @xmath197 we are in the isl phase , where the decay is exponentially with coherence length @xmath198 , respectively , for @xmath0 odd , @xmath199 ) .
the energy - scales for charge and spin - excitations are _
the llps in the charge - sector , see eq .
( [ eq : afkr ] ) , deviate more from their noninteracting value ( @xmath200 ) than the llps in the spin - sector , eq . ( [ eq : afks ] ) , i.e. , @xmath201 this implies that the charge - fields are pinned more strongly than the spin - fields and therefore the mott - gap ( we may interpret the rg - scale @xmath202 as the charge - gap ) is larger than the gap to spin-1/2 ( spinon ) excitations .
the coupling between the different bands makes it difficult to investigate the detailed excitation spectrum , such that we do not go beyond this qualitative result ( for the pure sine - gordon model , there exist rigorous results about the gap dependence on @xmath203 , see ref .
next , we study the effect of doping away from half - filling . there are two _ different _ doping regimes .
the _ lightly doped _ case , where the influence of the umklapp processes has to be taken into account , can be treated as a perturbation of the half - filled low - energy hamiltonian ( limit @xmath204 and @xmath18 finite ) . for _ increasing _ doping , one has then again to investigate the rg flow . for the lightly doped case ( sec . v a ) , we find above @xmath4 the same phase as below @xmath4 ( see ref .
@xcite ) : a conducting spin - liquid ( bound hole - pairs ) , see also fig .
[ f : sphd ] . furthermore , while at half - filling the magnetic correlations are only at @xmath205 , they become _
_ in__commensurate upon doping . in sec .
v b , we briefly show how and why for increasing doping phase - coherence between all band pairs sets in and the system becomes a 2d - like @xmath10-wave superconductor .
the low - energy hamiltonian takes the form of a @xmath10-wave bcs model , where the pairing is between electrons . note
that the superconducting instability is for @xmath206 generated in the same way as for the 2d case @xcite : a kohn - luttinger - type attraction @xcite is mediated by afm processes @xcite . in contrast , the superconducting instability in the two - leg ladder is generated by forward - scattering processes @xcite .
the isl phase of the half - filled two - leg hubbard ladder becomes superconducting when doped , because of the phase - coherence ( see discussion of the charge - sector in the previous section ) and the spin - gap present at half - filling @xcite .
here , we show that the doped hubbard afm becomes a conducting spin - liquid , formed by bound hole - pairs ( this phase may becomes superconducting at very low energies ; for the 4-leg ladder , see discussion in ref .
@xcite ) .
the _ 2-band _ interactions alone , see eqs .
( [ eq : tb1 ] ) and ( [ eq : tb2 ] ) , would result in an isl phase , i.e. , @xmath207 .
it is the presence of interactions between bands @xmath208 and @xmath21 which are _ not _ `` paired '' ( @xmath209 ) , which renders the system a 2d - like afm .
the effect of a decreasing @xmath24 is therefore that interactions between unpaired bands become suppressed and the afm correlations are reduced .
next , we show that the effect of doping away from half - filling is a similar one , i.e. , the doping @xmath8 corresponds to @xmath210 . at low dopings and at finite @xmath0 ,
we can dope the system perturbatively by introducing a chemical potential term @xmath211 ( @xmath212 is the total charge ) in the half - filled low - energy hamiltonian , see ref . @xcite . for the charge - sector , we can use bosonization , where the charge density is given by @xmath213 .
doping then introduces _ kinks _ in the fields @xmath184 , such that the expectation value of cosine terms , which contain this field gradually goes to zero ( the kink is a fermionic , strongly localized particle ) . here , the kinks are bound hole - pairs of charge 2 and zero spin .
all 4-band interactions contain the field @xmath184 , see eq .
( [ eq : cp4b ] ) , and therefore vanish .
the effect of the doping on the 2-band interactions is the same as when doping an isl , i.e. , in the bosonized version , the umklapp term ( [ eq : tb2 ] ) vanishes leaving only the term ( [ eq : tb1 ] ) , which opens a spin - gap , and leads to @xmath10-wave - like phase coherence between the bands @xmath21 and @xmath90 , corresponding to the fermi momenta @xmath214 and @xmath215 .
the charge - gaps close to @xmath216 are the smallest ones ( note that @xmath217 ) , such that the hole - pairs enter first there . in other words , it is the fact , that the ( @xmath218-wave ) afm processes @xmath219 ( partially ) coincide with the ( @xmath10-wave ) cooper processes with momentum transfer @xmath54 @xmath220 which makes a doped afm a superconductor : these particular afm / cooper processes flow at higher energies
driven by afm processes to strong coupling and are not suppressed at lower energies by doping ( see also fig . [
f : afp ] ) .
we therefore have afm mediated superconductivity @xcite .
the full phase - coherence around the fs then grows out of the phase - coherence between bands @xmath21 and @xmath90 , see below . upon doping ,
the afm correlations become _ _ in__commensurate . when doping away from half - filling , @xmath221 , and the afm peak shifts away from @xmath205 , see eq .
( [ eq : scf ] ) .
in particular , for small @xmath222 , all bands are ( almost ) equally doped and @xmath223 , i.e. , the incommensurability is equal the doping @xmath8 .
furthermore , doping leads to a cutoff of the afm processes at the energy - scale of the chemical potential @xmath224 and therefore to an exponential decay of the spin - spin correlation function , where the coherence - length is given by @xmath225 .
we argue , that for increasing doping , phase - coherence between the band pairs sets in , when the `` distance '' between neighboring bands is of the order of the `` distance '' of the fs to the umklapp surface , i.e. , for dopings @xmath226 .
the system then becomes a 2d - like @xmath10-wave superconductor .
since a finite chemical potential @xmath224 results in a low - energy cutoff for the umklapp and 4-band - afm interactions , one has to start initially with all interactions which are _ large _ at this cutoff energy - scale . in leading order , this implies that not only the 4-band interactions @xmath227 and @xmath228 have to be taken into account , but also `` neighbor '' processes of the form @xmath229 and @xmath230 .
these processes are only present down to an energy - scale @xmath231 . for dopings @xmath232 ,
the `` neighbor '' processes are cutoff _ before _ the main afm processes , which leads to a decoupling into band pairs @xmath56 , while for @xmath233 all afm processes are cutoff by the chemical potential . in this case ,
not only 2-band cooper interactions within band pairs @xmath56 are large at the cutoff energy - scale , but also the neighboring 2-band cooper interactions @xmath234 . integrating the rges without afm and umklapp processes , but with initial - values for the cooper and forward interactions given by the afm phase , see eq .
( [ eq : ratios2b ] ) , and similar initial - values for non - leading cooper processes @xmath235 ( @xmath236 ) , we obtain a @xmath237 superconducting instability at lower energies .
we always find that the scaling of the couplings is exactly such that ( in particular , @xmath238 , @xmath239 ) @xmath240 where the pairing operator is @xmath241 and @xmath242 for @xmath243 respectively @xmath244 and @xmath245 in all other cases ( @xmath237 symmetry ) .
the size ( but not the sign ) of @xmath246 depends on the initial - values ( i.e. , on the doping ) .
note that the energy - scale of the system is reduced by doping , i.e. , the superconductor has a lower scale than the afm . for large dopings ,
the 4-band interactions are cutoff _ before _ they could have flown sufficiently close to the afm fixed - point , such that there is no instability in the rg flow and the low - energy phase is a fl @xcite . as a result ,
for @xmath0 finite , there is between the insulating afm respectively isl phase at half - filling and the @xmath10-wave superconductor at higher doping , an intermediate , conducting phase with a spin - gap ( see also fig .
[ f : sphd ] ) . in the 2d - limit
, the intermediate phase seems to disappear .
we have studied the dimensional crossover from 1d to 2d in half - filled , weakly interacting , @xmath0-leg hubbard ladders for @xmath49 . using the rg method
, we found a crossover energy @xmath247 $ ] , where at energies below @xmath4 the behavior of the spin - sector is ladder - like and at energies above @xmath4 2d - afm - like .
we showed that the 4-band - afm processes are responsible for an instability in the rg flow . since without 4-band and umklapp processes
the hubbard ladders become a fl for large @xmath0 @xcite , afm processes ( at higher energies ) are necessary and sufficient for a nontrivial phase in the ( doped ) @xmath0-leg hubbard model . in the afm phase , the couplings flow towards universal ratios and we obtained an analytical expression for the hamiltonian which is similar to the heisenberg afm . using bosonization techniques , we further investigated this hamiltonian and found that the charge - sector is the same as in the half - filled two - leg ladder , i.e. , the type of mott insulator is the same in 1d and 2d .
for the spin - sector , bosonization techniques confirmed that the small-@xmath18 case has similar physical properties as the large-@xmath18 ( heisenberg ) case .
the effects of doping on the half - filled hubbard hamiltonian and of reducing the number of coupled chains @xmath0 is ( almost ) the same : in both cases , the interactions between unpaired bands are suppressed , the antiferromagnetic correlations are reduced and only interactions within band pairs @xmath56 remain . doping then opens a spin - gap , but phase coherence exists only between bands @xmath21 and @xmath90 . at higher doping levels @xmath248 , we can expect that phase coherence between all bands sets in , such that the system becomes a 2d - like @xmath10-wave superconductor . while the lightly doped case is conveniently described by bound hole - pairs , the 2d - like superconductor at higher dopings
is represented by a bcs - like @xmath10-wave hamiltonian with electron pairing . in the doped case
the antiferromagnetic correlations become incommensurate . as a result
, the hubbard model can not be understood in terms of an afm , isl or superconductor alone .
the phase ( correlation function ) depends always on the energy - scale ( respectively length - scale ) and is related to the value of the llps .
the afm exhibits the highest energy - scale . at half - filling ,
the groundstate ( for @xmath0 finite ) is the isl and for sufficient doping the @xmath10-wave superconductor .
we conjecture that in the doped case , the flow of the couplings from the afm to the @xmath10-wave takes place via an _ intermediate _ phase : consult table ii ; afm and superconductor differ in _ two _ llps .
this intermediate ( finite energy ) phase has then the same llps as the isl suggesting that the spin - gap opens _ before _
phase coherence sets in .
99 e. dagotto and t.m .
rice , science * 271 * , 618 ( 1996 ) .
s. chakravarty , b.i .
halperin , and d.r .
nelson , phys .
b * 39 * , 2344 ( 1989 ) . for a review , see e. manousakis , rev .
phys . * 63 * , 1 ( 1991 ) .
luttinger , j. math . phys . * 4 * , 1154 ( 1963 ) .
haldane , j. phys .
c * 14 * , 2585 ( 1981 ) .
s. chakravarty , phys .
rev . lett . * 77 * , 4446 ( 1996 ) . for an introduction , see h.j .
schulz , in _ correlated fermions and transport in mesoscopic systems _ , edited by t. martin _
frontiers , gif - sur - yvette , 1996 , p. 81 ) , and references therein . c. castellani , c. di castro , and w. metzner , phys .
. lett . * 72 * , 316 ( 1994 ) .
d. boies , c. bourbonnais , and a.m.s .
tremblay , phys .
* 74 * , 968 ( 1995 ) .
e. arrigoni , phys .
rev . lett . * 83 * , 128 ( 1999 ) .
anderson , phys .
* 64 * , 1839 ( 1990 ) .
an exception are strongly anisotropic systems , see a. vishwanath and d. carpentier , phys .
* 86 * , 676 ( 2001 ) .
r. shankar , rev .
66 * , 129 ( 1994 ) .
d. zanchi and h.j .
schulz , phys .
b * 54 * , 9509 ( 1996 ) ; europhys . lett . *
44 * , 235 ( 1998 ) ; phys . rev .
b * 61 * , 13609 ( 2000 ) . c.j .
halboth and w. metzner , phys .
b * 61 * , 7364 ( 2000 ) ; phys . rev . lett . * 85 * , 5162 ( 2000 ) . c. honerkamp , m. salmhofer , n. furukawa , and t.m .
rice , phys .
b * 63 * , 035109 ( 2001 ) .
b. binz , d. baeriswyl , and b. douot , cond - mat/0104424 .
h. lin , l. balents , and m. fisher , phys .
b * 56 * , 6569 ( 1997 ) .
u. ledermann , k. le hur , and t.m .
rice , phys .
b * 62 * , 16383 ( 2000 ) .
a. auerbach , in _ interacting electrons and quantum magnetism _
( springer , new york , 1994 ) .
schulz , in _ proceedings of les houches summer school lxi _ , edited by e. akkermans _ et al . _
( elsevier , amsterdam , 1995 ) .
j. voit , rep .
. phys . * 58 * , 977 ( 1995 ) .
w. kohn and j.m .
luttinger , phys .
* 15 * , 524 ( 1965 ) .
k. miyake , s. schmitt - rink , and c.m .
varma , phys .
b * 34 * , 6554 ( 1986 ) ; d.j .
scalapino , e. loh , and j.e .
hirsch , phys .
b * 34 * , 8190 ( 1986 ) .
for a related work with flat fermi surface , see s. dusuel , f. vistulo de abreu , and b. douot , cond - mat/0107548 .
h. lin , l. balents , and m. fisher , phys . rev .
b * 58 * , 1794 ( 1998 ) .
gogolin , a.a .
nersesyan , and a.m. tsvelik , in _ bosonization and strongly correlated systems _
( cambridge university press , cambridge , 1998 ) .
u. ledermann , ph.d .
thesis , eth zrich ( 2001 ) .
e. witten , commun .
phys . * 92 * , 455 ( 1984 ) . | we study the dimensional crossover from 1d to 2d in half - filled and lightly doped , weakly interacting @xmath0-leg hubbard ladders . in this case
, the hubbard ladders are equivalent to a @xmath0-band model . using renormalization group techniques , we find , that the half - filled ladders exhibit ( in the spin - sector ) an odd - even effect only below a crossover energy @xmath1 $ ] ( @xmath2 and @xmath3 depend on the interaction strength and on the hopping matrix elements ) : below @xmath4 , the dominant interactions take place within band pairs @xmath5 [ and within the band @xmath6 for @xmath0 odd ] , such that even - leg ladders are an insulating spin - liquid , while odd - leg ladders have one gapless spinon - mode .
in contrast , above the energy - scale @xmath4 , all bands are interacting with each other and the system is a 2d - like ( insulating ) antiferromagnet ; we obtain an analytical expression for the hamiltonian which is similar to the 2d heisenberg antiferromagnet .
bosonization techniques show , that in the charge - sector the mott insulator is as well below and above @xmath4 of the same type as in @xmath7 half - filled two - leg ladders .
doping away from half - filling , we find that the effect of an increasing doping @xmath8 is very similar to decreasing the number of legs @xmath0 : in both cases interactions between unpaired bands are suppressed and thereby the antiferromagnetic correlations reduced .
the resulting band pairs form then insulating spin - liquids and when doped , there is a spin - gap , but phase coherence exists only within the band pairs . at higher doping levels @xmath9 ,
phase coherence between all band pairs sets in and the system becomes a 2d - like @xmath10-wave superconductor [ @xmath11 for @xmath12 . 2 |
in this work we would like to illustrate and exemplify some general results from @xcite where the general framework of a morita theory which is covariant under a given hopf algebra was studied .
one main motivation to do so is coming from ( deformation ) quantization theory @xcite , see e.g @xcite for recent reviews . here
morita equivalence provides an important notion of equivalence of observable algebras @xcite .
in particular , on cotangent bundles the condition for star products to be morita equivalent is shown to coincide with dirac s integrality condition for magnetic charges of a background magnetic field @xcite leading to a natural interpretation of morita equivalence also in more general situations . from the differential geometric point of view , it is a natural question whether all these techniques as developed in @xcite can be made compatible with a certain given symmetry of the underlying manifold . on the purely algebraic level ,
a fairly general notion of ` symmetry ' is that of a hopf algebra action of a given hopf algebra . in @xcite
we studied this type of symmetry in the general situation . from this general framework we shall specialize now into two directions : on one hand
, the algebras on which the symmetry acts and whose morita theory shall be studied will now be commutative : we are interested in the algebra of functions @xmath1 on a manifold . on the other hand
, the symmetry in question will either be coming from a lie group action on @xmath2 or from a lie algebra action as its infinitesimal counterpart .
it is well - known that two commutative algebras are morita equivalent if and only if they are isomorphic , see e.g. the textbook @xcite , whence for commutative algebras morita equivalence seems to be a useless notion .
however , this is not true as things become interesting if one asks in addition in how many ways two algebras can be morita equivalent compared to the ways in which they can be isomorphic .
it turns out that in general there are new possibilities which makes morita theory interesting even in the commutative framework .
this phenomenon is precisely encoded in the so - called picard groupoid which we shall compute for the case of function algebras .
this way , we find an interesting and non - trivial class of examples illustrating the general ideas of @xcite .
moreover , it will also be of independent interest as we are now able to re - interpret several well - known problems and results in differential geometry from a morita theoretic point of view .
finally , the commutative situation with the algebras being function algebras @xmath1 is expected to be the starting point for a discussion of morita equivalence of star products as in @xcite but now being compatible with a symmetry of the classical phase space @xcite .
the article is organized as follows : in section [ sec : geometryalgebra ] we recall some well - known arguments why one should and how one can pass from a geometric to a more algebraic description of differential geometry . the next section is devoted to a general discussion on morita theory in different flavours , taking into account specific structures of the algebras in question .
here we are mainly interested in @xmath0-involutions and notions of positivity . in section [ sec : covariant ] we add one more structure to be preserved by morita theory , namely a symmetry which we model by a hopf algebra action . this can be specialized to group actions and lie algebra actions .
the last section contains some new material , namely the explicit computation of a certain part of the lie algebra covariant picard group .
* acknowledgments : * it is a pleasure for me to thank the organizers and in particular michel cahen and willy sarlet for their kind invitation to the 20th international workshop on differential geometric methods in theoretical mechanics in ghent where the content of this work was presented .
moreover , i would like to thank stefan jansen and nikolai neumaier for valuable discussions and comments on the manuscript .
in some sense , differential geometric methods in mathematical physics correspond mainly to classical theories : hamiltonian mechanics on a symplectic or poisson manifold @xmath2 is one prominent example . on the other hand ,
quantum theories require a more algebraic approach : here the uncertainty relations in physics are modelled mathematically by non - trivial commutation relations between observables in some noncommutative algebra , the observable algebra .
thus _ quantization _ in a very broad sense can be understood as the passage from geometric to noncommutative algebraic structures .
an intermediate step is of course to encode the geometric structures on @xmath2 in algebraic terms based on the commutative algebra of functions @xmath1 where , in view of applications to quantization , it is convenient to consider complex - valued functions .
then it is a folklore statement ( milnor s exercise ) that one can recover the smooth manifold @xmath2 from the @xmath0-algebra @xmath1 . more specifically : every @xmath0-homomorphism @xmath3 between function algebras is actually of the form @xmath4 with some smooth map @xmath5 between the underlying manifolds , see e.g. @xcite for a recent discussion .
thus the category of @xmath0-algebras with @xmath0-homomorphisms as morphisms becomes relevant to differential geometry .
the ` dictionary ' to translate geometric to algebraic terms , which is also one of the cornerstones of connes noncommutative geometry @xcite , can be extended in various directions .
we mention just one further example also relevant to morita theory : by the well - known serre - swan theorem , see e.g.@xcite , the ( complex ) vector bundles @xmath6 correspond to finitely generated projective modules over the algebra @xmath1 via @xmath7 . in more geometric terms
this means that for any vector bundle there exists another vector bundle @xmath8 such that @xmath9 is a trivial vector bundle .
this is of course the key to relate the algebraic @xmath10-theory of @xmath1 to the topological @xmath11-theory of @xmath2 .
let us now turn to morita theory .
its first motivation came from the question what one can say about two algebras @xmath12 and @xmath13 provided one knows that their categories of ( left ) modules are equivalent , see @xcite .
clearly , in view of applications to quantum mechanics a good understanding of the more specific modules given by @xmath0-representations of the observable algebras on ( pre ) hilbert spaces is crucial for any physical interpretation . as we shall not start to define what a reasonable category of modules over the @xmath0-algebra @xmath1 should be though this
can perfectly be done , see e.g. @xcite and references therein we take a different motivation which will lead essentially to the same structures .
the idea is to take the category of @xmath0-algebras , keep the objects and enhance the notion of morphisms .
this can be expected to be interesting as for function algebras we already know what the ` ordinary ' morphisms are : pull - backs by smooth maps . thus a generalization would lead to a generalization of smooth maps between manifolds , when we translate things back using our ` dictionary ' . in particular , it might happen that algebras become isomorphic in this new , enhanced category ( which will turn out to be not the case for function algebras ) and one might have more ` automorphisms ' of a given algebra ( which will indeed be the case for function algebras ) . in general , the invertible morphisms in a category form a ( large ) groupoid in the obvious sense which is called the picard groupoid of the category .
thus a major step in understanding the whole category is to consider its picard groupoid of invertible arrows first . in principle
, the whole idea should be familiar from geometric mechanics as one example of enhancing a category by allowing more general morphisms is given by the symplectic ` category ' : first one considers symplectic manifolds as objects and symplectomorphisms as morphisms .
though this is a reasonable choice to look at , it turns out to be rather boring as the choice for the morphisms is too restrictive .
more interesting is the ` category ' where one considers morphisms to be canonical relations , see e.g.@xcite .
however , this is no longer an honest category since the composition of morphisms is only defined when certain technical requirements like clean intesections of the canonical relations are fulfilled .
nevertheless this ` symplectic category ' is by far more interesting now .
other examples are the morita theory for ( integrable ) poisson manifolds by xu @xcite as well as the morita theory of lie groupoids , see e.g. @xcite .
after having outlined the general ideas in the previous section we should start being more concrete now . as warming up
we discuss the ` enhancing of the category ' for the category of unital algebras with usual algebra morphisms first , see e.g. @xcite for this classical approach . here
the generalized morphisms are the _ bimodules _ : for two algebras @xmath12 and @xmath13 a @xmath14-bimodule @xmath15 , which we shall frequently denote by @xmath16 to indicate that @xmath13 acts from the left while @xmath12 acts from the right , is considered as an arrow @xmath17
. why does this give a reasonable notion of morphisms ?
in particular , we have to define the composition of morphisms .
thus let @xmath16 and @xmath18 be bimodules then their tensor product @xmath19 over @xmath13 is a @xmath20-bimodule and hence an arrow @xmath21 .
however , this is not yet an associative composition law as for three bimodules @xmath22 , @xmath18 , @xmath16 we have a _
canonical isomorphism _ @xmath23 as @xmath24-bimodules but _ not equality_. the way out is to use isomorphism classes of bimodules as arrows instead of bimodules themselves .
then the tensor product becomes indeed associative and the isomorphism class of the canonical bimodule @xmath25 serves as the identity morphism of the object @xmath12 since we use unital algebras for simplicity .
the final restriction we have to impose is that in a category the morphism space between two objects has to be a set , which is a priori not clear in our enhanced category .
therefor one should pose additional constraints on the bimodules like finitely generatedness .
however , we shall ignore these subtleties in the following as the new notion of isomorphisms in this category will be unaffected anyway .
however , we still have to show that we really get an extension of our previous notion of morphisms .
thus let @xmath26 be an algebra homomorphism .
then on @xmath13 we define a right @xmath12-module structure by @xmath27 and obtain a bimodule @xmath28 .
its isomorphism class is denoted by @xmath29 .
it is easy to see that @xmath30 and @xmath31 is the class of @xmath25 whence our previous notion of morphisms is indeed contained in the new one .
if we denote this new category by @xmath32 then two unital algebras @xmath12 and @xmath13 are called _ morita equivalent _ iff they are isomorphic in @xmath32 . without going into the details this is equivalent to the existence of a certain bimodule which is ` invertible ' with respect to the composition @xmath33 .
in fact , such bimodules can be characterized rather explicitly , see e.g. @xcite .
the isomorphism classes of these invertible bimodules constitute now the _ picard groupoid _ of this category @xmath32 which we shall denote by @xmath34 .
the invertible arrows from @xmath12 to @xmath13 are denoted by @xmath35 while the isotropy group of this groupoid at the local unit @xmath12 is denoted by @xmath36 , the _ picard group _ of @xmath12 .
the map @xmath37 induces now a group homomorphism such that @xmath38 is exact , whence in the commutative case , the automorphism group of @xmath12 is a subgroup of the picard group @xmath36 .
finally , it can be shown that for commutative @xmath12 , the exact sequence is split whence @xmath39 where the subgroup @xmath40 consists of the symmetric invertible bimodules , i.e. those where @xmath41 for all @xmath42 and @xmath43 .
then @xmath40 is called the _ commutative _ or _ static picard group _ , see e.g. @xcite for a discussion and further references .
it is a well - known theorem in morita theory that for unital algebras @xmath12 , @xmath13 the equivalence bimodules @xmath16 are certain finitely generated projective right @xmath12-modules such that @xmath44 . coming back to our example
@xmath45 we see , using the serre - swan theorem , that the only candidates for the symmetric self - equivalence bimodules are the sections @xmath46 of a complex _
line bundle_. in fact , it turns out that @xmath46 is indeed invertible with inverse given by the class of @xmath47 since @xmath48 as @xmath1-bimodules .
this shows that the static picard group of @xmath1 is just the ` geometric ' picard group , i.e. the group of isomorphism classes of complex line bundles with the tensor product as multiplication . using the chern class to classify complex line bundles
then gives according to @xmath49 where the semidirect product structure comes from the usual action of diffeomorphisms on @xmath50 .
in general , all morita equivalence bimodules @xmath16 for @xmath51 are isomorphic to some @xmath52 with a vector bundle @xmath6 of non - zero fibre dimension .
moreover , @xmath13 has to be isomorphic to @xmath53 .
thus for function algebras @xmath1 we have a complete description of the picard groupoid .
we shall now specialize our notion of morita equivalence : we have already argued that the @xmath0-involution of @xmath1 should be taken into account when having applications to quantization in mind .
moreover , one can include notions of positivity into morita theory .
one defines a linear functional @xmath54 to be _ positive _ if @xmath55 for all @xmath56 .
then an element @xmath43 is called _ positive _ if @xmath57 for all positive linear functionals @xmath58 of @xmath12 , see @xcite for a detailed discussion .
it is clear that for applications to quantum theories such notions of positive functionals are crucial as they encode expectation value functionals and hence the physical _ states _ for the observable algebra . in particular
, for @xmath45 one finds that positive linear functionals are precisely the integrations with respect to compactly supported positive borel measures .
this follows essentially from riesz representation theorem , see ( * ? ? ?
b ) . from this
it immediately follows that @xmath59 is positive iff @xmath60 for all @xmath61 , whence the above , purely algebraic definition reproduces the usual notion .
we can now state the definition of @xmath0-morita equivalence @xcite and strong morita equivalence bimodules , see @xcite as well as @xcite for rieffel s original formulation in the context of @xmath62-algebras and @xcite for the general case of @xmath0-algebras .
instead of describing the ` enhanced category ' way , we directly give the definition in terms of bimodules which is entirely equivalent , see @xcite .
[ definition : moritabimodules ] a @xmath0-morita equivalence bimodule @xmath16 is a @xmath63-bimodule together with inner products @xmath64 and @xmath65 such that for all @xmath66 , @xmath43 and @xmath67 we have : 1 .
@xmath68 ( resp .
@xmath69 ) is linear in the right ( resp
. left ) argument .
2 . @xmath70 and @xmath71 .
@xmath72 and @xmath73 .
4 . @xmath74 and @xmath75 are non - degenerate . 5 .
@xmath74 and @xmath75 are full .
@xmath76 and @xmath77 .
@xmath78 .
if in addition the inner products are completely positive then @xmath16 is called a strong morita equivalence bimodule . here
@xmath74 is called _ full _ if the @xmath79-span of all elements @xmath80 is the whole algebra @xmath12 ; in general these elements constitute a @xmath0-ideal .
moreover , @xmath74 is called _ completely positive _ if for all @xmath81 and all @xmath82 the matrix @xmath83 is positive in the @xmath0-algebra @xmath84 .
the composition of bimodules is again the tensor product where on @xmath19 the @xmath12-valued inner product is now defined by rieffel s formula @xmath85 and analogously for the @xmath86-valued inner product .
it is then a non - trivial theorem that this is indeed completely positive again , if the inner products on @xmath15 and @xmath87 have been completely positive @xcite . passing to isometric isomorphism classes one can show that this gives a groupoid : the @xmath0-picard groupoid @xmath88 and the strong picard groupoid @xmath89 , respectively . in particular , the local unit at @xmath12
is given by the isometric isomorphism class of @xmath25 equipped with the inner products @xmath90 more generally , @xmath91 , viewed as @xmath92)-bimodule equipped with the canonical inner products @xmath93 implements the strong morita equivalence between @xmath12 and @xmath84 .
since we simply can forget the additional structures we obtain canonical groupoid morphisms @xmath94 \efig,\ ] ] which have been studied in @xcite : in general , none of them is surjective nor injective , even on the level of the picard groups .
the geometric interpretation of the inner products is that they correspond to _ hermitian fiber metrics _ on the corresponding line bundles or vector bundles , respectively : indeed , this can be seen easily from the very definitions . since up to isometry there is only one positive hermitian fiber metric on a given line bundle we have in the case of @xmath45 @xmath95 [ remark : weirdring ] in the approach of @xcite one main point was to replace the real numbers @xmath96 by an arbitrary _ ordered ring _ @xmath97 and @xmath79 by the ring extension @xmath98 with @xmath99 .
this allows to include also the formal star product algebras from deformation quantization into the game .
they are defined as algebras over the formal power series @xmath100 $ ] .
surprisingly , essentially all of the constructions involving positivity go through without problems .
let us now pass to the covariant situation : we want to incorporate some given symmetry of the @xmath0-algebras in question .
here we have two main motivations and examples from differential geometry : first , a smooth action @xmath101 of a lie group @xmath102 on @xmath2 , where by convention we choose a right action in order to have a left action @xmath103 on @xmath1 by @xmath0-automorphisms .
second , as infinitesimal version of @xmath104 , a lie algebra action , i.e. a lie algebra homomorphism @xmath105 from a real finite dimensional lie algebra @xmath106 into the lie algebra of real vector fields , which correspond to the @xmath0-derivations of @xmath1 . in order to formalize and unify both situations
it is advantageous to consider hopf @xmath0-algebras and their actions on algebras .
thus let @xmath107 be a _
hopf @xmath0-algebra _ ,
i.e. a unital @xmath0-algebra with a coassociative coproduct @xmath108 , a counit @xmath109 and an antipode @xmath110 such that @xmath111 as well as @xmath112 are @xmath0-homomorphisms and @xmath113 for all @xmath114 , see e.g. ( * ? ? ?
iv.8 ) . for the coproduct we shall use sweedler s notation @xmath115 .
the two geometric examples we want to discuss are now encoded in the following hopf @xmath0-algebras : first , recall that any group @xmath102 defines its group algebra @xmath116 $ ] which becomes a hopf @xmath0-algebra by setting @xmath117 , @xmath118 and @xmath119 for @xmath120 $ ] . in this case
@xmath121 $ ] is even _ cocommutative _ , i.e. @xmath122 where the opposite coproduct is defined by @xmath123 .
second , for any real lie algebra the complexified universal enveloping algebra @xmath124 becomes a hopf @xmath0-algebra by setting @xmath125 , @xmath126 and @xmath127 together with the resulting extensions to all of @xmath128 .
again , @xmath129 is cocommutative .
the situation that a group @xmath102 acts by @xmath0-automorphisms on a @xmath0-algebra as well as a lie algebra representation by @xmath0-derivations can be unified in terms of hopf @xmath0-algebras as follows : a _ @xmath0-action _ @xmath130 of @xmath107 on @xmath12 is a bilinear map @xmath131 such that @xmath132 and @xmath133 , i.e. @xmath12 is a left @xmath107-module , and @xmath134 , @xmath135 , and @xmath136 for all @xmath137 and @xmath138 .
then it is well - known and easy to see that for our two examples @xmath116 $ ] and @xmath129 this indeed generalizes and unifies the action by @xmath0-automorphisms and @xmath0-derivations , respectively .
the interesting relations are all encoded in the different coproducts . in principle and
probably even more naturally , one should consider coactions instead of actions of @xmath107 , see e.g. ( * ? ? ?
nevertheless , we stick to the more intuitive point of view where @xmath107 ` acts ' . now suppose we have @xmath0-algebras @xmath12 , @xmath13 with a @xmath0-action of @xmath107 .
let furthermore @xmath16 be a strong or @xmath0-morita equivalence bimodule .
then we call the bimodule _ @xmath107-covariant _ if there is an @xmath107-module structure on @xmath15 denoted by @xmath130 , too , such that we have the following compatibilities @xmath139 @xmath140 @xmath141 @xmath142 for all @xmath143 , @xmath43 , @xmath144 and @xmath137
. of course , in the case of ring - theoretic morita theory one only requires and . taking isometric isomorphism classes also respecting the action of @xmath107 gives the @xmath107-covariant flavours of the picard groupoids , denoted by @xmath145 , @xmath146 , and @xmath147 , respectively . since we can successively forget the additional structures we get the following commuting diagram of canonical groupoid morphisms : @xmath148 \vtriangle(0,0)/@{>}|\hole`>`>/<500,200>[{\sideset{}{^{\mathrm{str}}}{\operatorname{\mathsf{pic}}}}`{\sideset{}{^*}{\operatorname{\mathsf{pic}}}},`{\operatorname{\mathsf{pic } } } ; `` ] \morphism(0,525)<0,-250 > [ ` ; ] \morphism(500,325)<0,-250 > [ ` ; ] \morphism(1000,525)<0,-250 > [ ` ; ] \efig\ ] ]
now let us interpret the diagram on the level of picard groups and in our geometric situation : let e.g. @xmath149 and @xmath45 be as before , equipped with an action of @xmath106 by @xmath0-derivations .
then the kernel and the image of the group morphism @xmath150 encodes on which line bundles we can _ lift _ the @xmath106-action and if so , in how many different ways up to isomorphism .
analogously , in the strong situation one requires in addition compatibility with the hermitian fiber metric .
the case of @xmath151 $ ] leads to the question of existence and uniqueness of liftings of the group action on @xmath2 to a group action on @xmath152 by vector bundle automorphisms . in the strong case one requires the lift in addition to be unitary with respect to the fiber metric .
all this can easily be seen from the compatibility requirements , , , and applied to our situation .
clearly , all these lifting problems are very natural questions in differential geometry and have been discussed by various authors , see in particular @xcite , whence one can rely on the techniques developed there .
even though our approach does not give essential new techniques to attack the ( in general quite difficult ) lifting problem , it shines some new light on it and unreveals some additional structure of the problem , namely the groupoid structures together with the canonical groupoid morphisms .
moreover , this point of view embeds the lifting problem in some larger and completely algebraic context since neither the @xmath0-algebras have to be commutative nor has the hopf @xmath0-algebra to be cocommutative . as remarked already , we can even replace @xmath96 and @xmath79 by @xmath97 and @xmath153 , respectively , and incorporate in particular the formal star product algebras from deformation quantization as well .
in this last section we consider the case of @xmath154 more closely and develop some general results from @xcite slightly further .
assume that @xmath16 is a @xmath0-morita equivalence bimodule which allows for a lift of the actions of @xmath107 on @xmath12 and @xmath13 .
then it was shown in ( * ? ? ?
* thm 4.14 ) in full generality that the possible lifts are parametrized by the following group @xmath155 : we consider linear maps @xmath156 with the usual convolution product given by @xmath157 .
this makes @xmath156 an associative algebra with unit @xmath158 , see e.g. ( * ? ? ?
then we consider the following conditions for @xmath159 @xmath160 @xmath161 @xmath162 @xmath163 then @xmath155 is defined to be the subset of those @xmath159 which satisfy and it turns out that @xmath155 is a group with respect to the convolution product ( * ? ? ?
moreover , for unitary central elements @xmath164 one defines @xmath165 by @xmath166 .
then one obtains the exact sequence @xmath167 and the image @xmath168 is a central and hence normal subgroup .
thus we can define the group @xmath169 .
the parametrization of all possible lifts is obtained by a free and transitive group action @xmath170 of @xmath171 on the set of lifts given by @xmath172 where @xmath173 , @xmath114 and @xmath174 .
in fact , for @xmath107-covariantly @xmath0-morita equivalent algebras @xmath12 and
@xmath13 we have @xmath175 .
moreover , @xmath176 and @xmath130 give isomorphic actions iff @xmath177 for some @xmath178 whence the isomorphism classes of lifts are parametrized by the group @xmath179 which acts freely and transitively via on the isomorphism classes of lifts . while the above characterization works in full generality
we want to specialize now to lie algebra actions where @xmath154 .
first , it follows from ( * ? ? ?
a.7 ) that @xmath180 and hence @xmath181 since @xmath129 is cocommutative .
thus the values of @xmath182 are automatically central , essentially by .
moreover , the groups @xmath183 and @xmath184 are abelian . since the center @xmath185 is invariant under the @xmath106-action ( by derivations ! ) we can restrict @xmath186 to @xmath187 and obtain a chevalley - eilenberg cochain @xmath188 . evaluating the conditions , and on elements @xmath189
we find by a simple computation the following lemma : conversely , given @xmath191 we can construct an element @xmath192 with @xmath193 : let @xmath194 denote the @xmath195-th complexified tensor power of @xmath106 and define @xmath196 inductively by @xmath197 where @xmath198 . then a lenghty but straightforward computation using @xmath199 shows that @xmath200 passes to the universal enveloping algebra @xmath129 , viewed as a quotient of @xmath201 in the usual way , and fulfills to .
thus we have : note that this is in some sense surprising as the condition for @xmath202 to be a cocycle is linear while the condition for @xmath203 is highly _ non - linear_. it only becomes linear when evaluated on @xmath204 thanks to the fact that these elements are primitive , i.e. satisfy @xmath125
. thus a simplification like in theorem [ theorem : iso ] can not be expected for more non - trivial hopf @xmath0-algebras .
moreover , under the identification the elements @xmath205 give just the cocycles @xmath206 as usual .
note that in general @xmath207 are _ not _ ce - coboundaries .
thus , if we want to relate @xmath184 to lie algebra cohomology we have to assume an additional structure for @xmath12 : the first trivial observation is that for @xmath218 we have @xmath219 and for @xmath220 we clearly have @xmath221 .
thus in this case @xmath222 contains all anti hermitian ce - coboundaries in @xmath223 .
note however , that in general , @xmath222 is strictly larger . to measure this we consider those elements in @xmath224 which are not in the image of @xmath208
: we define the abelian group @xmath225 where the left hand side is of course only a symbol .
however , for @xmath45 we obtain indeed the @xmath226-integral first de rham cohomology of @xmath2 by the right hand side of which motivates our notation . since @xmath224 is mapped via @xmath227 into the cocycles @xmath223 and @xmath228 gives the coboundaries via we obtain a well - defined induced map @xmath229 collecting all the results we obtain the following statement : in particular , this applies to @xmath45 whence we obtain the full classification of the inequivalent lifts of the lie algebra action to line bundles . note that in general , @xmath184 does not depend on the line bundle itself but is universal for all line bundles | in this article we discuss some general results on the covariant picard groupoid in the context of differential geometry and interpret the problem of lifting lie algebra actions to line bundles in the picard groupoid approach .
* keywords : * morita equivalence , @xmath0-algebras , picard groupoid , hopf algebra actions .
* msc ( 2000 ) : * 16d90 , 16w30 , 16w10 , 53d55 |
extreme value behavior of gaussian processes is an important topic in probability theory and a crucial ingredient to many statistical inference procedures .
see for instance chernozhukov et al .
( 2014 ) for a very recent , general contribution to this topic . here
we are considering the extreme value behavior of gaussian fields on manifolds , which plays an important role in statistical inference .
in fact there is recent growing interest in the statistical literature in inference for manifolds such as integral curves ( koltchinskii et al . , 2007 ) , ridges or filaments ( hall et al . , 1992 , genovese et al
, 2012a , 2014 , chen et al . , 2013 , 2014a , 2014b , qiao and polonik , 2015a ) , level sets ( or level curves )
( lindgren et al . , 1995 ,
cuevas et al . , 2006 ,
chen et al . , 2015 ,
qiao and polonik , 2015b ) , or the boundaries of the support of a probability density function ( cuevas et al .
, 2004 , biau et al . , 2008 ) . in applications
these objects correspond to various types of geometric objects in geoscience ( fault lines ) , astronomy ( cosmic web ) or neuroscience ( fibers ) .
+ the derivation of asymptotic distributional results for such types of geometric objects often poses interesting technical challenges . in the case of density ridges in @xmath4 , qiao and polonik ( 2015a )
were able to derive such a distributional result for a smoothing based plug - in estimator .
note that density ridges in @xmath4 can be considered as curves indexed by a parameter , and the distributional result in qiao and polonik ( 2015a ) is _ uniform _ over the parameter . because of the pointwise asymptotic normality of the estimator and the uniform nature of the result , the extreme value behavior of a gaussian process on growing ( rescaled ) manifold comes into play there .
the corresponding result needed in qiao and polonik ( 2015a ) is proved here .
in fact we derive a much more general result , allowing for arbitrary dimension and more general non - stationarity .
+ also in other statistical literature the construction of uniform confidence bands for a target quantity rely on the asymptotic excursion probability of gaussian processes or fields on growing sets .
see , for instance , bickel and rosenblatt ( 1973a ) , konakov and piterbarg ( 1984 ) , rio ( 1994 ) , gin et al .
( 2003 ) , sharpnack and arias - castro ( 2014 ) . in this literature , the goal becomes to find the excursion probability of gaussian random processes or fields over growing sets in the form of @xmath5 where @xmath6 for each @xmath3 is a gaussian process or field , and @xmath7 is original set on which the estimation is constrained , and @xmath8 and @xmath9 is a parameter .
the quantities @xmath10 and @xmath11 need to be determined so that this limit is non - degenerate .
+ for instance , bickel and rosenblatt ( 1973a ) derive the asymptotic distribution of the quantity @xmath12}|\widehat{f}_n(x ) - f(x)|,$ ] where @xmath13 is a kernel density estimator based on a sample of @xmath14 independent observations from @xmath15 . in this case , @xmath7 is a compact set @xmath16 $ ] , and the parameter @xmath3 is the bandwidth of the kernel density estimator , so that @xmath17 $ ] which grows to the positive real line as @xmath18 the case of a multivariate kernel density estimator was treated later in rosenblatt ( 1976 ) , where @xmath19^d$ ] and thus @xmath20^d$ ] , growing to the positive quadrant for @xmath21 .
the corresponding derivations rely heavily on an approximation of @xmath22,$ ] by gaussian processes .
a similar idea underlies the derivations in qiao and polonik ( 2015a ) .
there , however , the processes in question are more complex .
the set @xmath7 is a certain manifold ( ridge line ) and a uniform nonparametric confidence region of @xmath7 itself is of interest .
it turns out that in order to achieve this , the main task is to find the limit of ( [ generalform ] ) with @xmath7 a smooth manifold , and this type of problem is considered below in a general set - up . +
from the perspective of probability literature , excursion probability of gaussian processes and fields is a classical and important topic . for the case of @xmath7 being an interval or a ( hyper- ) cube , the asymptotic distribution in ( [ generalform ] )
was studied in pickands ( 1969a ) , berman ( 1982 ) , leadbetter et al .
( 1983 ) , seleznjev ( 1991 ) , berman ( 1992 ) , seleznjev ( 1996 ) , hsler ( 1999 ) , hsler et al .
( 2003 ) , seleznjev ( 2006 ) , tan et al .
( 2012 ) , and tan ( 2015 ) . in this literature
the hausdorff dimension of @xmath7 is the same as the one of the ambient space for @xmath23 .
+ excursion probability for gaussian random fields over some more general ( but fixed ) parameter set is a classical subject and widely studied , e.g. adler ( 2000 ) , adler and taylor ( 2007 ) , azas and wschebor ( 2009 ) .
in particular , in piterbarg ( 1996 ) , mikhaleva and piterbarg ( 1997 ) , and piterbarg and stamatovich ( 2001 ) excursion probabilities of gaussian random fields over fixed manifolds can be found . in a recent work
the excursion probability of gaussian fields with some ( local ) isotropic properties indexed on a manifold is revisited in cheng ( 2015 ) .
+ this paper is deriving a result of type ( [ generalform ] ) , where the underlying process @xmath0 is a non - homogeneous , continuous , centered gaussian random field , and @xmath1 denotes a rescaled smooth , compact manifold .
our main result can be considered as a generalization of the classical bickel and rosenblatt result discussed above , as well as of a result by mikhaleva and piterbarg ( 1997 ) who considered a _ fixed _ manifold .
our proof combines ideas from both bickel and rosenblatt ( 1973a ) and mikhaleva and piterbarg ( 1997 ) .
+ the detailed set - up is as follows .
let @xmath24 with @xmath25 and @xmath26 .
let @xmath27 be a compact set and @xmath28 be a @xmath29-dimensional riemannian manifold with `` bounded curvature '' , the explicit meaning of which will be made clear later . for @xmath30
let @xmath31 and @xmath32 .
further let @xmath33 denote a class of non - homogeneous , continuous , centered gaussian fields indexed by @xmath34 our goal is to derive conditions assuring that for each @xmath35 we can construct @xmath36 with @xmath37 main result[mainresult ] first we define the notion of local stationarity used here .
the first definition can be found in mikhaleva and piterbarg ( 1997 ) , for instance .
+ a non - homogeneous random field @xmath38 is locally @xmath39-stationary , if the covariance function @xmath40 of @xmath41 satisfies the following property . for any @xmath42 there
exists a non - degenerate matrix @xmath43 such that for any @xmath44 there exists a positive @xmath45 with @xmath46 for @xmath47 and @xmath48 .
observe that this definition in particular says that @xmath49 for all @xmath50 .
since here we are considering random fields indexed by @xmath3 and study their behavior as @xmath21 , we will need local @xmath39-stationarity to hold in a certain sense uniformly in @xmath3 .
the following definition makes this precise .
consider a class of non - homogeneous random fields @xmath51 indexed by @xmath52 where @xmath53 is an index set .
we say @xmath54 is locally equi-@xmath55-stationary , if the covariance function @xmath56 of @xmath54 satisfies the following property . for any @xmath57 there
exists a non - degenerate matrix @xmath58 such that for any @xmath44 there exists a positive @xmath45 independent of @xmath3 such that @xmath59 for @xmath47 and @xmath48 .
an example for such a class of gaussian random fields ( with @xmath60 ) is provided by the fields introduced in qiao and polonik ( 2015a ) - see ( [ gaussianexample ] ) below .
+ we also need the concept of a condition number of a manifold ( see also genovese et al . ,
for an @xmath29-dimensional smooth manifold @xmath61 embedded in @xmath62 let @xmath63 be the largest @xmath64 such that each point in @xmath65 has a unique projection onto @xmath61 , where @xmath65 denotes the @xmath64-enlarged set of @xmath61 , i.e. the union of all open balls of radius @xmath64 and midpoint in @xmath66 .
@xmath63 is called _ condition number _ of @xmath61 in some literature . a compact manifold embedded in a euclidean space
has a positive condition number , see de laat ( 2011 ) , and references therein .
a positive @xmath63 indicates a `` bounded curvature '' of @xmath61 . as indicated in lemma 3 of genovese et al .
( 2012b ) , on a manifold with a positive condition number , small euclidean distance implies small geodesic distance .
+ now we state the main theorem of this section .
it is an result about the asymptotic behavior of the extreme values of locally equi-@xmath39-stationary continuous gaussian random fields indexed by a parameter @xmath3 as @xmath21 . as indicated above , our result generalizes theorem [ piterbarg ] in piterbarg and stamatovich ( 2001 ) and theorem a1 in bickel and rosenblatt ( 1973a ) .
+ for an @xmath67 matrix @xmath68 we denote by @xmath69 the sum of squares of all minors of order @xmath29 , and @xmath70 denotes the generalized pickands constant ( see section [ misc ] for a definition ) . at each @xmath71
let @xmath72 denote the tangent space at @xmath73 to @xmath61 and let @xmath74 be the normal space , which is a @xmath75 dimensional hyperplane .
+ [ probmain ] let @xmath27 be a compact set and @xmath31 for @xmath76 .
let @xmath77 be a class of gaussian centered locally equi-@xmath55-stationary fields with @xmath78 continuous in @xmath79 $ ] and @xmath80 .
let @xmath28 be a @xmath29-dimensional compact riemannian manifold with @xmath81 and @xmath32 for @xmath76 .
suppose that @xmath82 uniformly in @xmath83 , where all the components of @xmath84 are continuous and uniformly bounded in @xmath83 .
further assume the existence of positive constants @xmath85 and @xmath86 such that @xmath87 for any @xmath88 , define @xmath89 where @xmath90 is the covariance function of @xmath54 .
suppose for any @xmath88 , there exists a positive number @xmath91 such that @xmath92 in addition , assume that there exist a function @xmath93 and a value @xmath94 such that for any @xmath95 @xmath96^{2r/\alpha}\big|\leq v(\delta).\end{aligned}\ ] ] where @xmath97 is a monotonically decreasing function with @xmath98 and @xmath99 as @xmath100 for any @xmath101 .
for any fixed @xmath102 , define @xmath103,\end{aligned}\ ] ] where @xmath104 is a @xmath105 matrix with orthonormal columns spanning @xmath106 then @xmath107 * remarks .
* 1 . note that with ( [ ccprime ] ) , local equi-@xmath55-stationarity is equivalent to @xmath108 uniformly for @xmath109 and uniformly in @xmath3 .
an example of a function @xmath110 satisfying the properties of @xmath111 required in the theorem is given by @xmath112 for @xmath113 .
qiao and polonik ( 2015a ) use a special case of the above theorem . in that paper
a 1-dimensional growing manifold @xmath114 embedded in @xmath115 was considered .
the gaussian random field of interest there is @xmath116 where @xmath117 is a 2-dimensional wiener process , @xmath118 and @xmath119 are smooth functions , @xmath120 is a smooth kernel density function with the unit ball in @xmath115 as its support , and @xmath121 is an operator such that @xmath122 for any twice differentiable function @xmath123 .
it is shown in qiao and polonik ( 2015a ) that the assumptions formulated in that paper insure that the processes @xmath124 satisfy the assumptions of our main theorem in the special case of @xmath125 , @xmath126 , and @xmath127 for @xmath128 .
this in particular means that the function @xmath129 for @xmath113 works in this case .
the fact that this special function @xmath130 can be used there follows from the assumption that the support of @xmath131 ( and its second order partial derivatives ) is bounded .
this implies that the covariances of @xmath132 and @xmath133 become zero once the distance @xmath134 exceeds a certain threshold .
the derivation of the corresponding matrices @xmath43 can also be found in qiao and polonik ( 2015a ) .
the following notation and definitions are used below . given a set @xmath135 and a metric @xmath136 on @xmath135 ,
a set @xmath137 is an @xmath138-net if for any @xmath139 , we have @xmath140 and for any @xmath141 , we have @xmath142 .
let further @xmath143 , @xmath144 and @xmath145 .
let @xmath146 be the euclidean norm .
we also let denote @xmath147 denote @xmath29-dimensional hausdorff measure .
+ the proof is constructing various approximations to @xmath148 that will facilitate the control of the probability @xmath149 .
essentially the process @xmath54 on the manifold is linearized by first approximating the manifold locally via tangent planes , and then defining an approximating process on these tangent planes .
this idea underlying the proof is typical for deriving extreme value results for such processes ( e.g. see hsler et al .
we begin with some preparations thereby outlining the main ideas of the proof . + * ( i ) * _ partitioning @xmath1 : _ we partition the manifold @xmath1 as follows .
suppose that @xmath150 so that @xmath151 . for a fixed @xmath152 , there exists an @xmath153-net on @xmath114 with respect to geodesic distance with cardinality of @xmath154 .
a delaunay triangulation using the @xmath153-net results in a partition of @xmath155 into @xmath156 disjoint pieces @xmath157 the construction is such that @xmath158 , the norm of this partition , is @xmath159 , uniformly in @xmath3 .
it is known that for any @xmath160 with @xmath161 ( and for @xmath153 small enough ) such an @xmath153-net and a delaunay triangulation exist for compact riemannian manifolds ( see e.g. de laat 2011 ) .
( in the case of @xmath162 , the construction just described simply amounts to choosing all the @xmath163 many sets @xmath164 as pieces on the curve @xmath165 which has length at most @xmath153 . )
one should point out that while @xmath153 has to be chosen sufficiently small , it is a constant not depending on @xmath3 .
in particular this means that it does not tend to zero in this work . + * ( ii ) * _ ` small blocks - large blocks ' approach : _ for sufficiently small @xmath88 , let @xmath166 be the @xmath167-enlarged neighborhood ( using geodesic distance ) of the union of the boundaries of all @xmath164 .
the minus sign in the superscript indicates that this ( small ) piece will be ` cut out ' in the below construction .
we obtain @xmath168 ( ` large blocks ' ) and @xmath169 ( ` small blocks ' ) for @xmath170 .
geometrically we envision @xmath171 as a small strip along the boundaries of @xmath164 ( lying inside @xmath164 ) , and @xmath172 is the set that remains when @xmath171 is cut out of @xmath164 .
we have @xmath173 , uniformly in @xmath174 and @xmath3 .
the construction of the partition is such that the boundaries of the projections of all the sets @xmath175 and @xmath171 onto the local tangent planes are null sets , and thus jordan measurable .
this will be used below .
+ let @xmath176 and as a shorthand notation we use @xmath177 + approximating @xmath114 by @xmath178 leads to a corresponding approximation of @xmath179 by @xmath180 . even though the volume of @xmath181 i.e. the difference of @xmath114 and @xmath178 , tends to infinity as @xmath182 if we consider @xmath167 fixed , the difference @xmath183 turns out to be of the order @xmath184 .
thus we have to choose @xmath167 small enough .
+ * ( iii ) * _ refinement of the partition : _ let @xmath185 denote one of the sets @xmath175 and @xmath171 , and let @xmath186 be a cover of this piece constructed using the same delaunay triangulation technique as above , but of course based on a smaller mesh . as above , by controlling the mesh size , we can control the norm of the partition uniformly over @xmath3 , because of the uniform boundedness of the curvatures of the manifolds @xmath1 .
+ the probabilities @xmath187 are approximated by the sum of the probabilities @xmath188 , with @xmath189 the cover of @xmath172 introduced above .
it will turn out that the approximation error can be bounded by a double sum , using bonferroni inequality . to show this double sum is negligible compared with the sum
, we have to make sure the volume of @xmath189 is not too small , as long as @xmath189 is sufficiently small
. the double sum will be small if @xmath3 is small .
it turns out that @xmath187 essentially behaves like the tail of a normal distribution . +
* ( iv ) * _ projection into tangent space of refined partition : _ we approximate the small pieces @xmath189 on the manifold by @xmath190 , the projection onto the tangent space and correspondingly approximate the probabilities @xmath188 by the corresponding probabilities of a transformed field over @xmath190 .
more precisely , we choose some point @xmath191 on @xmath192 and orthogonally project @xmath192 onto the tangent space of @xmath155 at the point @xmath191 .
we denote the mapping by @xmath193 or simply @xmath194 and we let @xmath195 , which , as indicated above are jordan measurable by construction . the error generated from the approximation
is controlled by choosing the norm of the partitions given by the @xmath189 to be sufficiently small .
+ in the following , if @xmath185 is explicitly indicated in the context , then we often drop @xmath185 in the notation and simply write @xmath189 instead of @xmath192 . for simplicity and generic discussion
, we sometimes also omit the index @xmath196 of @xmath191 , @xmath189 and @xmath197 .
+ * ( v ) * _ discretizing the projection into the tangent space : _ the probabilities @xmath199 of the sets @xmath190 introduced in the previous step are approximated by replacing the probability of the supremum in @xmath199 by a maximum over a collection of ` dense grid ' on @xmath190 .
the accuracy of the approximation is controlled by choosing both @xmath200 and @xmath3 sufficiently small .
the construction of the dense grid is as follows : + let @xmath201 be linearly independent orthonormal vectors spanning the tangent space of @xmath155 at the point @xmath202 , and let @xmath203 denote the @xmath105 matrix with @xmath204 as columns . for a given @xmath200 consider the ( discrete ) set @xmath205 and let @xmath206 , which is a subset of @xmath207 .
note that the geodesic distance between any two adjacent points in @xmath208 is still of the order @xmath209 again due to the assumed uniformly positive condition number of the manifolds @xmath1 .
+ the collection of all sets of dense points in @xmath208 results in a set @xmath210 of dense points in @xmath211 .
it will turn out that the probability @xmath212 can be approximated by assuming the events @xmath213 to be independent . to make sure this approximation is valid
, @xmath167 may be not too small and @xmath200 may be not too small compared with @xmath3 .
+ putting everything together will then complete the proof .
+ * details of the proof .
* we now present the details by using the notation introduced above .
we split the proof into different parts in order to provide more structure .
note that the parts do not really follow the logical steps outlined above .
+ * part 1*. recall the definition of the refined partition @xmath214 of @xmath164 given in ( iii ) .
here we show that @xmath215 and that a similar approximation holds for @xmath164 replaced by @xmath216 . to this end
we will utilize the projections @xmath190 of @xmath189 onto the tangent space ( see ( iv ) ) as well as the approximation of @xmath190 by a set of dense points introduced in ( v ) .
+ the various asymptotic approximations in this step are similar to those in the proof of theorem 1 in mikhaleva and piterbarg ( 1997 ) , but here we consider them in the uniform sense . as indicated above , the uniform boundedness of the curvature of @xmath155 can be guaranteed due to the boundedness of the curvature ( positive condition number ) of @xmath217 . for any @xmath218
, there exists a constant @xmath219 ( not depending on @xmath3 ) such that if the volumes of all @xmath220 are less than @xmath221 , then we have @xmath222 where @xmath223 is the @xmath29-dimensional hausdorff measure .
on @xmath224 we consider the gaussian field defined as @xmath225 due to local equi-@xmath55-stationarity of @xmath54 , for any @xmath226 , the covariance function @xmath227 of the field @xmath228 satisfies @xmath229 for all @xmath230 , if the volume of @xmath231 is less than a certain threshold @xmath232 , which only depends on @xmath233 . by possibly decreasing @xmath232
further we also have @xmath234 for all @xmath235 .
note that this inequality holds uniformly over all @xmath224 under consideration , due to the curvature being bounded on @xmath114 .
+ on @xmath224 we introduce two homogeneous gaussian fields @xmath236 such that their covariance functions satisfy @xmath237 as @xmath238 .
thus if the volumes of all @xmath207 under consideration are sufficiently small then @xmath239 holds for all @xmath235 .
this can be achieved by possibly adjusting @xmath232 from above .
slepian s inequality in lemma [ slepian ] implies that @xmath240 and that @xmath241 for @xmath242 such that @xmath243 , denote @xmath244 by @xmath245 as a function of @xmath246 .
the covariance function of @xmath245 is @xmath247 as @xmath248 .
an application of lemma [ piece ] gives that for any @xmath249 and @xmath250 large enough @xmath251 similarly , by defining @xmath252 , we get @xmath253 combining ( [ s1s ] ) , ( [ slepianapp ] ) , ( [ applyuniformpiece1 ] ) and ( [ applyuniformpiece2 ] ) , we obtain for @xmath254 small enough and @xmath250 large enough that for any @xmath255 @xmath256 and since lemma [ hgamma ] says that @xmath257 as @xmath258 , we further have for @xmath200 sufficiently small that @xmath259 this in fact holds for any @xmath260 .
we now want to add over @xmath261 to this end observe that @xmath262 is a riemann sum , namely , for any @xmath44 , there exists a @xmath88 such that for @xmath263 , we have for @xmath3 sufficiently small that @xmath264 the selection of @xmath167 only depends on @xmath138 , and the uniformity comes from the fact that as @xmath265 , @xmath266 for any @xmath267 and @xmath268 and that @xmath84 is continuous in @xmath269 .
+ it follows from ( [ piecebound ] ) and ( [ intesum ] ) that for any @xmath255 and @xmath270 sufficiently small and @xmath250 large enough @xmath271 since the distribution of @xmath6 is symmetric , we also have @xmath272 we emphasize that these inequalities hold when the norm of the partition is below a certain threshold that is independent of the choice of @xmath3 .
+ following a similar procedure as above we see that ( [ piecesum ] ) and ( [ piecesum2 ] ) continue to hold ( for @xmath3 and @xmath273 sufficiently small and @xmath250 large enough ) if @xmath274 in ( [ piecesum ] ) is replaced by @xmath275 , and similarly , @xmath276 in ( [ piecesum2 ] ) is replaced by @xmath277 .
moreover , if we consider @xmath278 and @xmath279 , instead of @xmath280 , these inequalities continue to hold . in particular for @xmath164 we obtain @xmath281 where the @xmath282-term is uniform in @xmath170 as @xmath283 .
+ * part 2 . * here we show that @xmath284 as @xmath18 again we will use the various approximations introduced at the beginning of the proof
. + let @xmath285 denote the partition of @xmath164 constructed in ( iii ) .
this partition consists of closed non - overlapping subsets , i.e. their interiors are disjoint .
let further @xmath286 then obviously , @xmath287 we now use @xmath288 and we want to show that the double sum on the left - hand side is negligible as compared to the sum , so that we essentially have upper and lower bounds for @xmath289 in terms of @xmath290 to see this , first observe that it follows from ( [ sumtoint ] ) that for @xmath291 small enough we have as @xmath292 that @xmath293 we thus want to show that @xmath294 as @xmath295 the proof for a fixed manifold ( i.e. @xmath3 fixed ) can be found in the last part of mikhaleva and piterbarg ( 1997 ) . our proof for the more general case ( uniformly in @xmath3 ) is following a similar procedure .
it will turn out that we obtain the desired result if the norm of the partition given by the @xmath189 can be chosen arbitrarily small , uniformly in @xmath3 .
it has been discussed at the beginning of the proof that this is in fact the case .
+ let @xmath296 and @xmath297 are adjacent@xmath298 and @xmath299 , where non - adjacent means that their boundaries do not touch . note that @xmath300 in what follows we discuss the two sums on the right hand side of ( [ doublesum ] ) .
first we consider the case that @xmath301 are adjacent , i.e. @xmath302 . the developments in part 1 are here applied to @xmath189 , @xmath303 and @xmath304 , respectively .
we choose the points where the tangent spaces are placed to be the same for @xmath189 , @xmath303 and @xmath304 , i.e. , we choose this point to lie on the boundary of both @xmath189 and @xmath303 .
simply denote this point as @xmath305 .
then , by using the results from part 1 , for any @xmath44 , when @xmath306 is small enough and @xmath250 is large enough , then the bounds obtained as in part 1 result in @xmath307.\end{aligned}\ ] ] the sum of the right hand side of the above inequalities over @xmath308 again is a riemann sum that approximates an integral over @xmath164 . since @xmath82 uniformly in @xmath269 , and since the components of @xmath84 are continuous and bounded in @xmath83 , there exists a finite real @xmath309 such that @xmath310 hence as @xmath311 and @xmath312 , and noting that @xmath255 is arbitrary , we have @xmath313 next we proceed to consider the case that @xmath314 , i.e. @xmath315 are not adjacent on @xmath164 . to find a upper bound for @xmath316 , first notice that @xmath317 in order to further estimate this probability we will use the following borel theorem from belyaev and piterbarg ( 1972 ) .
[ borel ] let @xmath318 be a real separable gaussian process indexed by an arbitrary parameter set @xmath319 , let @xmath320 and let the real number @xmath321 be such that @xmath322 then for all @xmath23 @xmath323 there exists a constant @xmath324 such that @xmath325 i.e. , the distance between any two nonadjacent elements of the partition exceeds @xmath326 uniformly in @xmath327 $ ] .
this is due to the fact that the curvatures of the manifolds @xmath1 is ( uniformly ) bounded , and that @xmath328 is bounded away from zero uniformly in @xmath329 and @xmath3 .
see lemma 3 of genovese et al .
( 2012 ) for more details underlying this argument .
the latter also implies that we can find a number @xmath330 such that @xmath331 , the number of sets @xmath332 , satisfies @xmath333 for all @xmath3 .
assumption ( [ supgauss1 ] ) implies that @xmath334 we want to apply the above borel theorem to @xmath335 with @xmath336 and @xmath337 and @xmath338 . to this end
observe that @xmath339 and @xmath340 next we show that there is a constant @xmath321 such that @xmath341 for @xmath3 sufficiently small .
note that @xmath342 all the arguments in * part 1 * hold uniformly in @xmath3 as long as @xmath250 is large enough .
in other words , the conclusions there can be restated by replacing @xmath250 with @xmath23 where @xmath343 .
for instance , for any @xmath255 we can choose @xmath344 small enough such that @xmath345 holds for all @xmath346 and @xmath347 .
hence , since @xmath348 as @xmath349 , we can find @xmath321 such that @xmath350 for all @xmath346 when @xmath344 is sufficiently small .
the above borel inequality now gives ( for large enough @xmath250 ) that @xmath351 since the total number of elements in the sum in ( [ doublesum ] ) is bounded by @xmath352 , it follows from ( [ intersectionbound ] ) and ( [ psupbound ] ) that uniformly in @xmath174 ( recall that the @xmath353 depend on @xmath174 ) @xmath354 as @xmath355 by using the well - known fact that @xmath356 ( see cram@xmath357r , 1951 , page 374 ) .
+ considering ( [ sumb ] ) , ( [ doublesum ] ) , ( [ subsumbound2 ] ) and ( [ subsumbound1 ] ) and their respective conditions , we have @xmath358 where the @xmath282-term is uniform in @xmath174 .
+ combining ( [ sumtoint ] ) and ( [ protosum ] ) , we have for @xmath359}\max_{1\leq i \leq n_h}v_r(s_i^h)$ ] sufficiently small , that @xmath360 * part 3 . * note that from the expression of @xmath250 in ( [ thetaexp ] ) we have for any fixed @xmath102 @xmath361 as @xmath362 . + observing that @xmath363 ( uniformly in @xmath3 ) , and using ( [ phix ] ) we obtain for @xmath3 small enough that @xmath364 uniformly in @xmath365 here @xmath366 is from ( [ dmbound ] ) .
similarly , ( and again uniformly in @xmath174 ) we have @xmath367 uniformly in @xmath368 for some @xmath369 . collecting
what we have we get that uniformly in @xmath368 @xmath370 and @xmath371 * part 4 . * here we show that replacing @xmath372 by the dense ` grid ' @xmath210 ( see ( v ) ) leads to a negligible error in the corresponding extreme value probabilities . +
we write @xmath373 as @xmath374 our assumptions assure that @xmath375 , because @xmath376 and the ` mesh size ' of the curvilinear mesh on @xmath1 is @xmath377 , due to the construction of the triangulation and the uniformly bounded curvature on the manifolds @xmath1 .
+ with ( [ piecesum ] ) , ( [ dmbound ] ) and ( [ phix ] ) , we have @xmath378 uniformly in @xmath174 as @xmath362 . note that here and below we for brevity omit to indicate that the maxima ( or minima , respectively ) run over @xmath379 ( i.e. over all @xmath380 ) .
it follows that as @xmath21 @xmath381 it follows from ( [ piecesum ] ) and its version with the @xmath382 over the discrete set replaced by the @xmath383 over @xmath384 ( see discussion given below ( [ piecesum2 ] ) ) , that for any @xmath255 there exists thresholds for @xmath3 , @xmath200 and the norm of partitions , such that @xmath385\\ & \leq\epsilon\;\theta^{2r/\alpha}\psi(\theta)h_\alpha^{(r ) } \int_{j_{k , m_h}^\delta}\|d_s^hm_s^h\|_rds,\end{aligned}\ ] ] provided @xmath3 , @xmath200 and the norm of partitions are smaller then their respective thresholds . similarly , ( [ piecesum2 ] ) and its corresponding ` continuous ' version imply that for @xmath3 and @xmath200 smaller than their respective thresholds indicated in * part 1 * , we have @xmath386 consequently ,
if @xmath3 and @xmath200 and @xmath387 are small enough , we have @xmath388\nonumber\\ & \leq\sum_{k=1}^{m_h}\big[\mathbb{p}\big(\sup_{t\in j_{k , m_h}^\delta}x_h(t ) > \theta\big)+\mathbb{p}\big(\inf_{t\in j_{k , m_h}^\delta}x_h(t)<- \theta\big)-\mathbb{p}\big(\max_{t_j\in j_{k , m_h}^\delta}x_h(t_j ) >
\theta\big)\nonumber\\ & \hspace{1cm}-\mathbb{p}\big(\min_{t_j\in j_{k , m_h}^\delta}x_h(t_j)<- \theta\big)\big]\nonumber\\ & \leq2\epsilon\,\theta^{2r/\alpha}\psi(\theta)h_\alpha^{(r ) } \int_{\bigcup_{k\leq m_h}j_{k , m_h}^\delta}\|d_s^hm_s^h\|_rds\nonumber\\ & \leq2\epsilon\,\theta^{2r/\alpha}\psi(\theta)h_\alpha^{(r ) } \int_{\mathcal{m}_h}\|d_s^hm_s^h\|_rds.\end{aligned}\ ] ] to see the order of the upper bound in ( [ supdiff ] ) , by the dominated convergence theorem ( and using our assumption on the behavior of @xmath389 ) we have @xmath390 as a result of ( [ phix ] ) and ( [ integration ] ) , we can write for @xmath387 small enough that @xmath391 and @xmath392 as @xmath393 + * part 5 . * here we find an upper bound for the difference @xmath394 this step uses similar ideas as in the proof of lemma 5.1 in berman ( 1971 ) .
+ define a probability measure @xmath395 such that for any @xmath396 with @xmath397 , @xmath398 i.e. , under @xmath395 the vectors @xmath399 and @xmath400 independent for @xmath401 . by lemma [ bickellemma ] , the difference in ( [ pdiff ] )
can be bounded by @xmath402 ( note that here the notation @xmath403 is a shortcut for @xmath404 , and similarly for @xmath405 . ) for @xmath406 and @xmath407 with @xmath401 , it follows from the uniform boundedness of the curvature of the growing manifold that there exists a positive real @xmath408 such that @xmath409 , uniformly for all @xmath410 .
( similar arguments have been used above already . )
thus we obtain from assumption ( [ supgauss1 ] ) that there exists @xmath411 dependent on @xmath408 such that latexmath:[\ ] ] let @xmath461 . again , for @xmath441 orthogonal and due to isotropy of @xmath455 , we just write @xmath462 . we let @xmath463 assuming this limit exists , and for @xmath464 we simply write @xmath465 and @xmath466 instead of @xmath467 and @xmath468 , respectively . we have the following lemma from bickel and rosenblatt ( 1973b ) .
[ slepian ] ( * slepian s lemma ; see slepian , 1962 * ) let @xmath470 and @xmath471 be gaussian processes satisfying the assumptions of theorem [ borel ] with the same mean functions . if the covariance functions @xmath472 and @xmath473 meet the relations @xmath474 then for any @xmath23 @xmath475 [ zbigniew](*lemma 6.1 of piterbarg , 1996 * ) let @xmath41 be a continuous homogeneous gaussian field where @xmath477 with expected value @xmath478 and covariance function @xmath479 satisfying @xmath480 then for any compact set @xmath439 @xmath481 the next lemma is an extension of lemma a3 in bickel and rosenblatt ( 1973a ) , lemma 3 and and lemma 5 of bickel and rosenblatt ( 1973b ) and lemma 2.5 in pickands ( 1969 ) . its proof is also adapted from the three sources .
[ piece ] let @xmath41 be a centered homogeneous gaussian field on @xmath62 with covariance function @xmath480 let @xmath490 be a jordan measurable set imbedded in a @xmath29-dimensional linear space with @xmath491 . for @xmath492 let @xmath493 be a collection of points defining a mesh contained in @xmath490 with mesh size @xmath494 assume @xmath495 then @xmath496 and @xmath497 uniformly in @xmath498 where @xmath499 is the collection of all @xmath29-dimensional jordan measurable sets with @xmath29-dimensional hausdorff measure bounded by @xmath500 .
similarly , @xmath501 uniformly in @xmath502 the results in lemma 3 and and lemma 5 of bickel and rosenblatt ( 1973b ) are similar but they are only given for two - dimensional squares .
it is straightforward to generalized them to hyperrectangles and further to jordan measurable sets .
[ piterbarg](*theorem 2 of piterbarg and stamatovich , 2001 * ) let @xmath503 be a gaussian centered locally @xmath504-stationary field with a continuous matrix function @xmath505 .
let @xmath506 be a smooth compact of dimension @xmath29 .
then @xmath507 as @xmath343 , where @xmath508 is an @xmath105 matrix with columns the orthonormal basis of the linear subspace tangent to @xmath509 at @xmath305 .
+ [ bickellemma ] ( * lemma a4 of bickel and rosenblatt , 1973a * ) let @xmath510 let @xmath511 be @xmath512 nonnegative semi - definite matrices with @xmath513 for all @xmath196 .
let @xmath514 be a mean 0 gaussian vector with covariance matrix @xmath515 under probability measure @xmath516 or @xmath517 under @xmath518 .
let @xmath519 be nonnegative numbers and @xmath520 .
then latexmath:[\[\begin{aligned } koltchinskii , v. , sakhanenko , l. and cai , s. ( 2007 ) .
integral curves of noisy vector fields and statistical problems in diffusion tensor imaging : nonparametric kernel estimation and hypotheses testing .
statis . _ * 35 * , 1576 - 1607 .
leibon , g. and letscher , d. ( 2000 ) .
delaunay triangulation and voronoi diagrams for riemannian manifolds .
_ proceedings of the sixteenth annual symposium on computational geometry , scg 00 _ , acm new york , ny , usa , pp . 341349 .
piterbarg , v.i . and stamatovich , s. ( 2001 ) .
on maximum of gaussian non - centered fields indexed on smooth manifolds . in _
asymptotic methods in probability and statistics with applications ; statistics for industry and technology , _ eds : n. balakrishnan , i. a. ibragimov , v. b. nevzorov , birkhuser boston , boston , ma , pp . 189203 . | we consider a class of non - homogeneous , continuous , centered gaussian random fields @xmath0 where @xmath1 denotes a rescaled smooth manifold , i.e. @xmath2 and study the limit behavior of the extreme values of these gaussian random fields when @xmath3 tends to zero , which means that the manifold is growing .
our main result can be thought of as a generalization of a classical result of bickel and rosenblatt ( 1973a ) , and also of results by mikhaleva and piterbarg ( 1997 ) .
+ this research was partially support by the nsf - grant dms 1107206 + _ ams 2000 subject classifications . _
primary 60g70 , 60g15 .
+ _ keywords and phrases . _ local stationarity , extreme values , triangulation of manifolds |
bulk properties of strongly interacting matter under extreme conditions are the focus of the on - going rhic program .
deconfinement and chiral symmetry restoration@xcite are expected to take place in collisions of ultra - relativistic nuclei . because these phase transitions are multiparticle phenomena , a promising , albeit challenging , approach is the study of dynamics of large groups of final state particles .
the dynamics shows itself in the correlations and fluctuations ( texture ) on a variety of distance scales in momentum space .
the multi - resolution dynamic texture approach ( applied for the first time@xcite at sps ) uses discrete wavelet transform @xcite(dwt ) to extract such information . at present stage ,
the information is extracted in a comprehensive way , without any built - in assumptions or filters .
mixed events are used as a reference for comparison in search for dynamic effects .
event generators are used to `` train intuition '' in recognizing manifestations of familiar physics ( such as elliptic flow or jets ) in the analysis output , as well as to quantify sensitivity to the effects yet unidentified , such as critical fluctuations or clustering of new phase at hadronization .
star time projection chamber@xcite(tpc ) , mounted inside a solenoidal magnet , tracks charged particles within a large acceptance ( @xmath4 , @xmath5 ) and is well suited for event - by - event physics and in - depth studies of event structure .
the data being reported are obtained during the first ( @xmath3 gev ) year of rhic operation .
the minimum bias trigger discriminates on a neutral spectator signal in the zero degree calorimeters@xcite . by adding a requirement of high charged multiplicity within @xmath6 from the scintillating central trigger barrel , one obtains the central trigger .
vertex reconstruction is based on the tpc tracking .
only high quality tracks found to pass within 3 cm of the event vertex are accepted for the texture analysis .
discrete wavelets are a set of functions , each having a proper width , or scale , and a proper location so that the function differs from 0 only within that width and around that location .
the set of possible scales and locations is discrete .
the dwt transforms the collision event in pseudo - rapidity @xmath0 and azimuthal angle @xmath1 into a set of two - dimensional functions .
the basis functions are defined in the ( @xmath0 , @xmath1 ) space and are orthogonal with respect to scale and location .
we accumulate texture information by averaging the power spectra of many events .
the simplest dwt basis is the haar wavelet , built upon the _ scaling function _ @xmath7 for @xmath8 and 0 otherwise .
the function @xmath9 is the wavelet function . the experimental acceptance in @xmath0,@xmath1 , and @xmath2 ( ( @xmath6 , @xmath5 ) ) is partitioned into bins .
the @xmath0-@xmath1 partitions are of equal size , whereas in @xmath2 , the binning is exponential when more than one @xmath2 bin is used . in each bin
, the number of reconstructed tracks satisfying the quality cuts is counted .
the scaling function of the haar basis in two dimensions ( 2d ) @xmath10 is just a bin s acceptance ( modulo units ) .
the wavelet functions @xmath11 ( where the mode of directional sensitivity @xmath12 can be azimuthal @xmath1 , pseudo - rapidity @xmath0 , or diagonal @xmath13 ) are @xmath14 we set up a two dimensional ( 2d ) wavelet basis : @xmath15 where @xmath16 is the integer scale fineness index , and @xmath17 and @xmath18 index the positions of bin centers in @xmath1 and @xmath0 . then , @xmath19 with integer @xmath16 , @xmath17 , and @xmath18 are known @xcite to form a complete orthonormal basis in the space of all _ measurable functions _ defined on the continuum of real numbers @xmath20 .
we construct @xmath21 analogously to eq.[wavelet_2d ] .
[ haar ] shows the wavelet basis functions @xmath22 in two dimensions . at first glance
it might seem surprising that , unlike the 1d case , both @xmath23 and @xmath24 enter the wavelet basis in 2d .
[ haar ] clarifies this : in order to fully encode an arbitrary shape of a measurable 2d function , one considers it as an addition of a change along @xmath1 ( @xmath25 , panel ( b ) ) , a change along @xmath0 ( @xmath26 , panel ( c ) ) , and a saddle - point pattern ( @xmath27 , panel ( a ) ) , added with appropriate weight ( positive , negative or zero ) , for a variety of scales . the finest scale available is limited by the two track resolution , and , due to the needs of event mixing , by the number of available events .
the coarser scales correspond to successively re - binning the track distribution .
the analysis is best visualized by considering the scaling function @xmath21 as binning the track distribution @xmath28 in bins @xmath17,@xmath18 of fineness @xmath16 , while the set of wavelet functions @xmath29 ( or , to be exact , the wavelet expansion coefficients @xmath30 ) gives the difference distribution between the data binned with given coarseness and that with binning one step finer .
we use waili@xcite software to obtain the wavelet expansions . in two dimensions ,
it is informative to present the three modes of a power spectrum with different directions of sensitivity @xmath31 , @xmath32 , @xmath33 separately .
we define the * power spectrum * as @xmath34 where the denominator gives the meaning of spectral density to the observable .
so defined , the @xmath35 of a random white noise field is independent of @xmath16 .
however , for physical events one finds @xmath35 to be dependent on @xmath16 due to the presence of * static texture * features such as acceptance asymmetries and imperfections ( albeit minor in star ) , and non - uniformity of the @xmath36 shape . in order to extract the * dynamic * signal
, we use @xmath37 where the latter denotes power spectrum obtained from the * mixed events*. the mixed events are composed of the ( @xmath38 ) pixels of true events , so that a pixel is an acceptance element of the finest scale used in the analysis , and in no mixed event is there more than one pixel from any given true event .
the minimum granularity used in the analysis is @xmath39 pixels .. scale 1 : @xmath40 ; scale 2 : @xmath41 ; scale 3 : @xmath42 and so on . ]
systematic errors can be induced on @xmath37 by the process of event mixing .
for example , in events with different vertex position along the beam axis , same values of @xmath0 may correspond to different parts of the tpc with different tracking efficiency . that will fake a dynamic texture effect in @xmath0 . in order to minimize such errors
, events are classified into * event classes * with similar multiplicity and vertex position .
event mixing is done and @xmath37 is constructed within such classes .
only events with @xmath43 vertex lying on the beam axis within 30 cm from the center of the chamber are accepted for analysis . to form event classes , this interval
is further subdivided into five bins .
we also avoid mixing of events with largely different multiplicity .
therefore , another dimension of the event class definition is that of the multiplicity of quality tracks in the tpc . for central trigger events , the multiplicity range of an event class is typically 25 . for small multiplicity events taken with the minimum bias trigger ,
the multiplicity range per event class is 10 for multiplicities below 90 , and 20 in the range between 90 and 150 .
for larger multiplicities in the minimum bias sample , this multiplicity range was taken to be 50 .
events with less than 15 good quality tracks in the fiducial acceptance are ignored .
elliptic flow is a prominent large scale dynamic texture effect already well measured at rhic@xcite .
the dwt approach localizes elliptic flow on scales 2 and , to some degree , 3 of the @xmath44 . in this report , we ignore flow and concentrate on the @xmath0 observables .
multiplicity scans reveal dependence of the signals on the initial conditions : number of participants and binary collisions and the energy density and size of the interacting system .
such scans as a function of normalized multiplicity of quality tracks is divided by @xmath45 , where @xmath45 is such that 99% of minimum bias events have multiplicity of quality tracks less than @xmath45 .
the normalized multiplicity depends weakly on the experimental definition of minimum bias and on the track quality cuts . ]
are shown in fig .
[ vs_mult_c ] , [ vs_mult_p ] , and [ vs_mult_n ] .
the figures show the scale 1 ( large scale ) observables for tracks of both charges , and for the positive only and the negative only , respectively .
the horizontal error bars on the points are formed by multiplicity boundaries of event classes used in the analysis or , when a coarser re - binning of the multiplicity is done on the analyzed data to enhance the presentation , reflect that re - binning . the event mixing is always confined to the specific multiplicity and @xmath43-vertex position class , described in section [ analysis ] .
a comparison,@xmath46,@xmath47,@xmath48,@xmath49 ) and their antiparticles from the generator output are considered , provided that they fit into the star tpc fiducial @xmath0 acceptance @xmath50 .
momentum resolution and @xmath2 acceptance are not simulated . ] with hijing@xcite is done in order to understand the effects of energy - momentum conservation , jets and resonances on our measurements . for the regular hijing and hijing without jets ,
the multiplicity dependence is weak or absent .
this is understandable given the nature of hijing as a super - position of binary nucleon - nucleon collisions with no collective effects .
the regular hijing and hijing without jets show different magnitude of the signal .
this underscores the role of jets in creating local density fluctuations ( positive dynamic texture ) in hijing .
notice that this result is obtained without imposing a high @xmath2 cut .
the experimental data are close to regular hijing for peripheral events , but their texture becomes suppressed as multiplicity grows . to understand the nature of this suppression , we look at same charge particles ( fig .
[ vs_mult_p ] and [ vs_mult_n ] ) .
we see that the star data are not really intermediate between the regular hijing and hijing without jets , but display a more complex behavior .
this is seen from the fact that as the multiplicity grows , the dynamic texture becomes negative and stays roughly constant past a transition region which occupies about a half of the multiplicity range , whereas the `` no jets '' prediction points are consistent with 0 for all multiplicities .
how to explain the negative dynamic texture ?
recall the definition of the observable ( section [ analysis ] ) . at scale 1
, we are looking at the fluctuation in same - charge track occupancy between two `` pixels '' , each one unit of @xmath0 long , located in forward / backward hemispheres around mid - rapidity .
the negative @xmath51 means that in the mixed events , this occupancy fluctuation is stronger , i.e. a * correlation * takes place in the real events . and @xmath52 for positive and negative tracks .
in the following , the @xmath53 notation denotes coefficients in the expansion of an individual event track density @xmath54 into wavelet basis @xmath22 . for simplicity , we omit the indices .
we sum over locations but not over the scale index . averaging over many events
is assumed everywhere .
@xmath55 we measure independently @xmath56 ( fig.[vs_mult_p ] , modulo normalization ) and @xmath57 ( fig.[vs_mult_n ] , modulo normalization ) .
in order to test that eq.[eq : zero ] holds , we need to obtain the correlation terms between the opposite charges @xmath58 and @xmath59 .
this can be done by comparing power spectra of images filled with equal weights for positive and negative particles , and of those where negative particles are entered with a negative weight : @xmath60 which is true for @xmath52 as well . with this input , validity of eq.[eq : zero ] has been established and the `` paradox '' presented above resolved : it is the large scale correlation between positive and negative charges that accounts for the `` extra '' texture when the positive and negative sub - events are combined . ]
the tracking conditions in the chamber are different at high and low multiplicities , and the effect of that on the measurements in question needs to be understood .
we do this by processing hijing events through full geant and response simulation , and applying the actual experimental reconstruction to those events .
such processing ( done for a smaller number of high multiplicity hijing events than is shown in the figures ) reveals no significant difference for the reported observable .
thus , we are looking at an interplay of fluctuation and correlation effects , neither of which is trivial , as a function of initial conditions .
the signals become weaker for finer scales . at this point
one can only speculate about the nature of the correlation effect in central events , but the hijing simulation indicates that jet quenching ( or even total disappearance of jets ) does not account for the * negative * @xmath61 , even though it may be a prerequisite for its observation , given that jets work to create a * positive * texture .
what about bose - einstein correlation ? in the simulations , its effect can not be introduced into the multiparticle distributions event - by - event from the first principles , and in hijing it is not considered at all .
low values of @xmath62 ( @xmath63 mev@xcite ) , typical for both bose - einstein and coulomb effects , make it unlikely for these effects to be responsible for correlations with characteristic @xmath64 .
longitudinal expansion maps rapidity differences onto time differences ; the features characterized by larger rapidity differences must be formed early . in the color glass condensate picture , multiplicity fluctuations in different rapidity windows
are predicted to be correlated for large ( @xmath65 ) rapidity intervals@xcite .
this is seen as a consequence of the classical coherence of the gluon field .
such or a qualitatively similar effect is indeed required to explain the data .
the star measurements of auau data reveal a non - trivial picture of non - statistical correlations and fluctuations which is qualitatively different for peripheral and central collisions .
a possible interpretation of the central data is suppression of jets , combined with particle emission correlated over the length of the order of a unit of rapidity .
h. meyer - ortmanns , rev .
phys . * 68 * , 473 ( 1996 ) i. bearden _ et al . _
[ na44 collaboration ] , phys .
rev . c. 65 ( 2002 ) 044903
i. daubechies , _ ten lectures on wavelets _
( siam , philadelphia , 1992 ) and references therein .
k. h. ackermann _ et al .
_ [ star collaboration ] , nucl .
a * 661 * , 681 ( 1999 ) ; nucl .
a * 698 * , 408 ( 2002 ) . c. adler , a. denisov , e. garcia , m. murray , h. strobele and s. white , nucl .
instrum .
a * 470 * , 488 ( 2001 ) g. uytterhoeven _
et al . _ ,
waili : wavelets with integer lifting .
tw report 262 , department of computer science , katholieke universiteit leuven , belgium , july 1997 .
k. h. ackermann _ et al .
_ [ star collaboration ] , phys .
lett . * 86 * , 402 ( 2001 ) ; c. adler _ et al . _ [ star collaboration ] , phys .
c * 66 * , 034904 ( 2002 ) x. n. wang and m. gyulassy , phys .
d * 44 * , 3501 ( 1991 ) .
m. gyulassy and x. n. wang , comput .
commun . * 83 * , 307 ( 1994 ) c. adler _ et al . _ [ star collaboration ] , phys .
lett . * 87 * , 082301 ( 2001 ) y. v. kovchegov , e. levin and l. d. mclerran , phys .
c * 63 * , 024903 ( 2001 ) .
i am grateful to larry mclerran for informing me about this prediction during the conference . | dynamic fluctuations in the local density of non - identified hadron tracks reconstructed in the star tpc are studied using the discrete wavelet transform power spectrum technique which involves mixed event reference sample comparison .
the two - dimensional event - by - event analysis is performed in pseudo - rapidity @xmath0 and azimuthal angle @xmath1 .
hijing simulations indicate that jets and mini - jets result in signals , visible without high @xmath2 selection , when the dynamic texture analysis is applied . scanning a broad range of event multiplicities , we study the dependence of the signals on the initial conditions .
event structures are studied separately with positive and negative tracks , as well as both charges .
a change of regime is observed in auau collisions at @xmath3 gev as event multiplicity is increased : a long range @xmath0 correlation ( or suppressed fluctuation vis - a - vis mixed events ) is seen in same charge data .
this effect is qualitatively similar to one of the predicted manifestations of the color glass condensate . |
the different evolutive stages of massive stars have strong influence on their surroundings and , consequently , on the galactic evolution . during their lives ,
the massive stars inject large amounts of matter , energy and momentum into the interstellar medium ( ism ) . in the pre - main sequence ,
massive stars undergo a phase of very energetic mass loss giving rise to molecular outflows .
uv radiation , expansion of hii regions and fast stellar winds are present in the main sequence stage . in the short lived red supergiant ( rsg ) and
luminous blue variable ( lbv ) stages , the winds become slower and denser and ejections of up to some solar masses may be produced .
the wolf rayet ( wr ) stage , prior to sn explosions , is characterized by a copious mass loss ( typically @xmath5 m@xmath6 yr@xmath3 ) driven by a fast ( about 2000 km s@xmath3 ) and chemically enriched stellar wind .
massive stars are therefore continuously heating , ionizing , shocking and blowing - up the ism surrounding them .
garca - segura & mac low ( 1995 ) have modeled the evolution of the gas which surrounds massive stars , from the main sequence to the wr stage .
they only considered simple models for the mass ejection in the form of stellar winds in the different stages of evolution and predict the presence of large scale structures around the massive stars .
both the blown - up gas and the ejecta form multiple shells as a consequence of the different shockfronts produced inside the gas . during the main sequence phase
, the gas around massive stars is ionized , heated and highly evacuated by the fast , long - lived stellar wind and the lyman continuum radiation .
a hot cavity is surrounded by an expanding shell .
the ionization front may get trapped within it , the shell may start to recombine and become observable in the line radiation emitted by neutral atoms or molecules .
these large bubbles have been successfully detected in the 21-cm line of atomic hydrogen ( @xcite ; @xcite ; @xcite ; @xcite ; @xcite ) , in the form of huge shells of 3070 pc , expanding at 620 km s@xmath3 , with masses of hundreds ( or even thousands ) of solar masses .
some of the large iras shells found by marston ( 1996 ) might also be related to this type of structures .
once the massive star finishes the hydrogen burning in its nucleus and becomes a rsg or lbv , its stellar wind suddenly becomes more dense , and a significant amount of the stellar mass is deposited into the ism at velocities around 20 km s@xmath3 .
very probably , the formation of the optical ring nebulae starts in this brief stage .
when the star comes to a wr stage , the wind is accelerated to hypersonic velocities and rapidly reaches the rsg / lbv wind .
many of the optical nebulae possess arcs and have high inhomogeneities in the abundances of heavy elements ( @xcite , 1999 ) . however , the effects of the wr phase in the neutral ( neither atomic nor molecular ) gas have not been easily detected .
the structures formed during this phase would not be as large as those developed in the o - phase .
just in few cases , the neutral atomic hydrogen has been detected in interaction with ring nebulae ( @xcite ; @xcite ) .
vibrationally excited molecular hydrogen has only been detected towards ngc2359 .
the co j = @xmath0 emission has been observed around wr16 ( @xcite ) and in ngc2359 ( @xcite ) .
ngc2359 is a very interesting object , and many aspects of which has been thoroughly studied in the recent years .
this optical nebula is excited by hd56925(wr7 in the catalogue of van der hutch et al .
( 1988 ) , a wn4 star located in the outer milky way .
its distance varies from 3.5 to 6.9 kpc depending on the authors ( see @xcite for a discussion ) . the mass loss rate ( @xmath7 m@xmath6 yr@xmath3 ) and the terminal wind velocity ( 1545 km s@xmath3 ) have been measured by schneps et al .
( 1981 ) and rochowicz & niedzielski ( 1995 ) , respectively .
ngc2359 , a prototypical wind - blown bubble ( @xcite ) , is nearly spherical with several small filaments inside .
the nebula is not highly enriched of heavy elements as a whole , with the exception of the filaments ( @xcite ) .
this may indicate that the nebula was largely produced in the rsg / lbv and the wr stages of the exciting star . from the optical emission lines ( @xcite ) and radio recombination lines ( @xcite ;
@xcite ) , we know that the ionized gas has several components at velocities between 30 and 70 km s@xmath3 ( all velocities in this paper are referred to lsr ) .
an hii region partially surrounds the wind - blown bubble and is apparently limited by obscuring material .
schneps et al .
( 1981 ) have taken several profiles in the j = @xmath0line of co and detected the presence of three clouds at 37 , 54 and 67 km s@xmath3 .
the presence of broad co profiles at 54 km s@xmath3 and the morphology of the emission , just following the south - eastern edge of the nebula , suggests that this component has been interacting with the nebula .
st - louis et al .
( 1998 ) have analyzed the h@xmath8 emission in this region .
they have found the 10 s(1 ) line towards the southern border of the hii region , but they could not establish the nature of the excitation ( fluorescence or shocks ) of the h@xmath8 .
cappa et al .
( 1999 ) made a complete map of the ionized component at the 1465 mhz continuum , and traced the hi emission presumably connected with the nebula .
these hi features are probably associated with the wr stage of hd56925 ; so far , the large atomic main - sequence bubble predicted by the theory has not been yet reported .
the goal of this paper is to look into the history of the interaction of the wind and the radiation of hd56925 and its surroundings .
we have analyzed a public hi survey and studied for the first time the large scale interaction of hd56925 in the o - phase with the ambient ism .
based on the mapping of the j = @xmath0 and j = @xmath1 lines of co over the whole optical nebula , we study in detail the morphology and kinematics of the co around ngc2359 .
we also present several @xmath4co profiles towards selected positions in order to determine the global physical parameters of the molecular gas , such as kinetic temperature , density and opacity of the lines .
this study has allowed us to shed some light on the dominant excitation mechanism of the neutral gas in the wr stage .
we have analyzed a region of @xmath9 around ngc2359 using the 21 cm hi survey of hartmann & burton ( 1997 ) .
this survey was made using the 25 m radio telescope at dwingeloo .
the angular and velocity resolution of this dataset are 36 and 1.03 km s@xmath3 , respectively .
the survey has been done with a sampling of 30 in galactic coordinates and have a sensitivity in brigthness temperature of 70 mk . the observations of the co j = @xmath0 ( 115.271 ghz ) and j = @xmath1 ( 230.538 ghz ) lines and the j = @xmath0 ( 110.201 ghz ) line of @xmath4co towards ngc2359 and its surroundings were made during 2000 february , using the nrao 12 m radio telescope at kitt peak .
the half power beam width of the telescope was 54 and 27 at the rest frequency of the j = @xmath0 and the j = @xmath1lines , respectively .
the data were acquired simultaneously with two independent sis receiver channels with orthogonal polarizations .
the receiver temperatures were 250300k and 500600k for the j = @xmath0 and the j = @xmath1 lines , respectively .
we used filterbanks ( with 256 channels ) and the hybrid correlator mac ( in the 2 if mode , in parallel , 3072 channels ) as backends . in the j
= @xmath0line , the filterbanks provided 100 and 250 khz of resolution , corresponding to velocity resolutions of 0.26 and 0.65 km s@xmath3 , respectively .
for the same line , the mac correlator was setup with a total usable bandwidth of 300 mhz and with channels of 195.3 khz of resolution ( 0.51 km s@xmath3 ) .
in the j = @xmath1line , the filterbank with channel widths of 250 khz ( 0.33 km s@xmath3 ) and 500 khz ( 0.65 km s@xmath3 ) were used .
we also used for this line the mac correlator tuned with an usable bandwidth of 600 mhz and frequency resolution of 390.6 khz ( 0.51 km s@xmath3 ) .
mapping of the j = @xmath0 and j = @xmath1 lines was made using the `` on - the - fly '' technique .
we have obtained three individual maps in the j = @xmath0 line and six in the @xmath1 line .
these individual maps were later reduced with aips and combined weighting them by their inverse square of _
rms_. for our investigation , we finally used the mac data for the j = @xmath0 line and the 500 khz - filterbank data for the j = @xmath1 line .
typical final _
rms _ per grid point was 120 mk and 80 mk in the j = @xmath0 and j = @xmath1 lines , respectively .
we have also observed the j = @xmath0 line of @xmath4co and c@xmath10o at seven individual positions , with 710 minutes of integration time achieving a _
rms _ noise of 30 mk .
the pointing was checked at the begining of each observing session with five point cross maps .
the final data are pointed within 4 of uncertainty .
the data were calibrated using the chopper wheel method .
final intensity scale is in units of @xmath11 , e.g. , corrected by atmospheric attenuation and instrumental losses ( @xcite ) .
the calibration was also checked off - line by comparing individual spectra of standard sources .
we have analyzed the hi data in the velocity range 2080 km s@xmath3 , e.g. , the velocity range of the optical and the co emission . according to most of the galactic rotation models ( for example @xcite ) ,
this broad range in velocity is expected for the range of distances associated to hd56925(@xcite ) . within this velocity range
, we looked for the structures connected to the evolution of hd56925 .
1 depicts hi column density ( n@xmath12 ) maps as a function of the radial velocity , indicated at the top left corner .
the white circle near the center of the map sketchs the position and size of ngc2359 .
every map shown in fig .
1 has been obtained by the integration of the brightness temperature over 2.1 km s@xmath3 ( 2 channels ) and assuming optically thin emission .
we have found an outstanding structure which remains coherent across tens of velocity channels .
this structure is a huge cavity surrounded by a set of relative maxima in the n@xmath12 maps .
the centroid of this cavity remains approximately at the same point , and nearly off - centered from the nebula and the wr star .
the excentric position of the wr star with respect to the cavity is also observed in almost all the wr bubbles ( @xcite ) and might be due to the star motions and/or density variations in the original ism ( @xcite ) .
the structure reaches its maximum size at velocities between 61 and 65 km s@xmath3 . for receding and approaching velocities with respect to this velocity interval , the cavity diminishes its angular size , ending at @xmath13 51 and 76 km s@xmath3 with small relative maxima centered at the cavity .
the morphology of this structure has the typical ring
disk appearence of an expanding bubble . following the standard analysis of this type of shells ,
the largest size of the cavity indicates the geometrical dimensions and its radial velocity corresponds to approximately the systemic velocity of the gas before the expansion .
the full velocity range where the feature is observed gives a lower limit of twice the expansion velocity .
table 1 summarizes the parameters derived , as described above , for the expanding cavity .
following arnal et al .
( 1999 ) , we have also derived lower and upper limits to the mass of the hi shell .
these two values for the hi mass are the missing mass into the higher close contour and the neccesary mass to fill the cavity making it indistinguishable of the nearly environs .
all distance - dependent parameters were calculated assuming a distance of 5 kpc .
this hi shell is in geometrical coincidence with the iras shell reported by marston ( 1996 ) .
furthermore , the mass of gas predicted from the iras data falls within the hi estimates presented in table 1 . the co emission in the j = @xmath0 and j = @xmath1
lines are observed in three velocity components .
the components with the lowest radial velocity ( 3538 km s@xmath3 ) and with the highest radial velocity ( 6469 km s@xmath3 ) show narrow profiles of 1.52.5 km s@xmath3 width .
hereafter we will refer to these two components as ambient 1 ( a1 ) and ambient 2 ( a2 ) , respectively .
the third component , with radial velocities of 5058 km s@xmath3 , shows broader profiles than the components a1 and a2 , having linewidths of 45.5 km s@xmath3 .
we will refer to this component as broad component ( b ) .
2 shows the global spatial distribution of the co j = @xmath0 ( left panels ) and j = @xmath1 ( right panels ) integrated emission for the three components . for comparison , the co maps have been overlaid with a j band image of ngc2359 , obtained from the _ digitized sky survey_. at the top right corner of each panel , the velocity interval of integration is indicated . the mapped regions in both spectral lines are also depicted . in the j
= @xmath0line , almost all the optical nebula and its surroundings has been covered by our observations . in the j = @xmath1
line , however , we limited the maps at the east and the north - west because of our interest in studying in detail the interaction of the nebula with the molecular gas . the co emission for components a1 and a2 appear mainly to the east and to the south - east of the mapped regions , respectively .
although the spatial distribution for both components are found close in projection to the optical nebula , there is not a clear morphological correlation .
furthermore , there are no significant changes in the line widths and in the peak velocities over the regions where these components are observed .
in contrast , the broad component b is observed in clear correlation with the eastern and the southern edges of ngc2359 , and with broader profiles , varying from 4 to 5.5 km s@xmath3 wide .
the most intense co emission of component b is located beyond the outer edge of the hii region in ngc2359 , mainly to the south - east part , but also significant emission in the j = @xmath1 line is detected towards locations projected on the hii region ( middle right panel in fig . 2 ) .
for a comparison between the spatial distribution of the j = @xmath0 and @xmath1 lines of co for the component b , the fig .
3 shows and overlay of the emission in the j = @xmath0 line ( solid lines ) and the j = @xmath1 line ( dashed lines ) convolved to the resolution of the j = @xmath0 line .
this comparison shows that both lines are well correlated at the most intense and the external part of the nebula .
however , we note again that the j = @xmath1 line have significant extended emission to the inner part of ngc2359 , especially into the southern optical filament .
this is clearly ilustrated in the left and right panels in fig . 3 , where we show a sample of four co j = @xmath0 and @xmath1 line profiles toward the positions sketched by squares in the central map .
the lower left spectra show the mean profiles around the most intense region , while the other three panel show the co profiles in the periphery of the co emission .
there is a clear trend in the profiles . while the j = @xmath0 line is stronger than the j = @xmath1 line toward the most intense zone of emission , the j = @xmath0 line is much weaker than the j = @xmath1 line or it is not detected in positions closer to the hii region .
the @xmath1/@xmath0 intensity ratio varies from 0.81 near the peak to 1.52 towards the positions near the hiiregion .
in contrast , the @xmath1/@xmath0 intensity ratio is nearly constant in components a1 and a2 , with values around 0.50.6 .
while the components a1 and a2 do not have any remarkable kinematical features , the component b have slightly different peak velocities as a function of the position . fig .
4 shows four position - velocity diagrams , taken in the directions sketched in the map of the j = @xmath1emission at the top left corner of this figure .
each position - velocity diagram is identified by the slice number .
slice 1 shows the most intense part of component b at 5354 km s@xmath3 and the western border of the component a2 .
slice 1 also shows hints of a small velocity gradient in the component b. this is more clearly depicted in the slice 2 , where we find along the central part a velocity shift of @xmath13 2 km s@xmath3 .
both slices 1 and 2 remarks the spatial coincidence between the eastern border of component b and the western border of component a2 .
in fact , slice 3 shows the presence of a weak ( 4 @xmath14 ) `` bridge '' in the co emission which connects the components b and a2 .
this suggests a physical connection between both components .
slice 3 , parallel to the slice 2 but 1 to the south - west , does not show signs of this `` bridge '' , indicating the small size and weakness of this bridge .
the slice 4 show that the diffuse emission to the north have a radial velocity slightly lower than that of the main part of the component b. in order to study the excitation of the co line in the component b , we have observed seven positions in the @xmath4co and c@xmath10o j = @xmath0 lines , towards the southern part of ngc2359 .
the location of the observed points are shown in the fig .
5a , superimpossed on a co
j = @xmath1 map of the component b. five of these positions are located close to the slice 1 .
5b , 5c and 5d plot three of these @xmath4co j = @xmath0 profiles ( labeled with letters b , c and d in the circles of fig .
5a ) , together with the @xmath15co j = @xmath0spectra of the same points .
we find a significant variation in the @xmath15co/@xmath4co ratio .
this ratio is @xmath13 6 for the co emission peak ( fig .
5b ) and increases to values @xmath16 elsewhere .
another interesting result is that the @xmath15co/@xmath4co ratio does not only vary with position but also it changes with radial velocity within the beam towards the co emission peak ( fig .
one can see that both lines peak at different velocities .
the @xmath4co line appears shifted by @xmath13 12 km s@xmath3 with respect to that of @xmath15co .
this shift produces important changes in the @xmath15co/@xmath4co ratio as a function of the radial velocity .
the @xmath15co/@xmath4co ratio is @xmath17 30 at 5053 km s@xmath3 , and decreases to @xmath13 45 for radial velocities between 54 and 57 km s@xmath3 . in view of the small velocity gradient found in the component b
, we do not think that this shift can be due to a pointing error between the @xmath15co and @xmath4co data .
even in the worst possible case , 10 of pointing difference , we find that this ratio would change with radial velocity by at least a factor of 4 for velocities below and above 53 km s@xmath3 .
the hi structure found in this paper resembles the expanding shells often associated with o stars ( @xcite ) , wr stars ( @xcite ; @xcite ) and ob associations ( @xcite ; @xcite ) . in the following
we shall explore if the hi shell has been created by the stellar wind from hd56925 or from its o progenitor .
let us estimate , with a few assumptions , some important parameters to distinguish between the two possibilities .
we shall consider a spherical expanding bubble , with a radius of 50 pc , an expansion velocity of 12 km s@xmath3 , and a mass of 1500 m@xmath6 . with these values
, we estimate a dynamical time for the shell of 2.3 myr and a kinetic energy of @xmath18 erg .
the dynamical time was computed as a mean value between the adiabatic and the radiative case ( @xcite ) .
this time is in fact an upper limit , but a good approximation , to the age of the hi structure ( @xcite ) .
this age and the kinetic energy associated to the hi shell clearly indicate that its origin is due to the main - sequence phase of hd56925 . indeed ,
the most massive stars , greater than 25 m@xmath6 , remains a few million years at the main sequence ( @xcite ) and creates interstellar bubbles of 3060 pc in this stage ( @xcite ; @xcite ) .
it is thought that wr s are descendant of these massive stars .
on the other hand , the rsg , lbv or wr phases are short - lived , probably @xmath19 yr ( @xcite ; @xcite ) .
we therefore conclude that only the o progenitor of hd56925 had sufficient time to develop the hi expanding shell reported in this paper .
a similar conclusion was reached by marston ( 1996 ) concerning to the large iras shell found around several wr stars , including the one linked to hd56925 . by adopting typical values for the mass - loss rate of @xmath20 m@xmath6 yr@xmath3 and a wind velocity of 1000 km s@xmath3 for the o
progenitor star , we estimate that this star deposited into the ism during 2.3 myr a total energy of @xmath21 erg in the form of stellar wind .
if this star has blown up the shell which we are observed in hi , it implies a kinematical efficiency of nearly 10% .
although the uncertainties in these computations are large and hard to be estimated , this value for the efficiency is in good agreement with the models that predict this type of structures ( @xcite ; @xcite ) and reinforces our hypothesis of a main sequence origin of the hi shell .
we have estimated the physical parameters of the molecular gas from the co line emission by modelling the excitation of the co lines using the _ large velocity gradient _ ( lvg ) approximation .
for comparison with the model , we have smoothed the j = @xmath1 line data to the j = @xmath0 resolution . for component
b , the @xmath4co data was also taken into account in order to estimate the opacity of the @xmath15co lines .
six different components have been defined .
the first three are the ambient components a1 and a2 and the most extended emission of the component b ( hereafter b@xmath22 ) .
the fourth component considered is the one projected onto the southern hii region , named as b@xmath23 .
finally , we also considered the region around the co maxima in component b , where the @xmath4co behaviour is striking .
we have subdivided this component in the two velocity ranges having the greatest difference in the @xmath15co/@xmath4co ratio ( see fig .
5b ) , named as b@xmath24 ( for the velocity interval 5054 km s@xmath3 ) and b@xmath25 ( for the velocity interval 5458 km s@xmath3 ) .
the results derived from the lvg analysis are shown in table 2 .
this table gives estimates for the excitation temperature ( t@xmath26 ) , h@xmath8 density ( n(h@xmath8 ) ) , co column density ( n(co ) ) , the opacities of the co j = @xmath0 , co j = @xmath1 and @xmath4co j = @xmath0 lines , and the total mass of the regions ( m(h@xmath8 ) ) , derived from the two line ratios ( second and third columns ) and the @xmath4co intensities . a kinetic temperature ( t@xmath27 ) of 10k was assumed for all the regions but b@xmath23 .
the phyisical conditions in b@xmath23 seem to be very different from those of the other regions .
the large ratio between the j = @xmath1 and the j = @xmath0 lines is explained by optically thin emission with relatively high t@xmath27 .
the most likely t@xmath27 is about 80k and we present in table 2 the results derived for this t@xmath27 . components a1 and a2 are subtermally excited ( t@xmath26 @xmath28 t@xmath27 ) , with typical t@xmath26 of 56k .
we obtain higher excitation temperatures in component b. for b@xmath24 and b@xmath25 , we derived t@xmath26 similar to t@xmath27 , indicating that the lines are thermalized , and densities significantly higher than for a1 and a2 .
for b@xmath25 we did not apply the lvg method since this region has a very low value of the @xmath15co/@xmath4co ratio .
such low value for this ratio usually indicates the presence of a region of very high opactiy in the @xmath15co lines .
since we have found this low ratio toward only one observed position , we conclude that the emiting region corresponding to b@xmath25 is unresolved by the j = @xmath0 beam . to derive the physical properties in this component ,
we have assumed an optically thin emission for the @xmath4co and optically thick emission of the @xmath15co . for a @xmath15co/@xmath4co isotopic ratio of 60 ( @xcite ) , we derive an optical depth of 12 for the j = @xmath0 line of @xmath15co . from the @xmath15co j = @xmath0
line intensity and assuming that the emission in this line is thermalized to the t@xmath27 of 10k , we derive a beam filling factor for b@xmath25 between 0.1 and 0.2 . with the optical depth
we have estimated n(co ) . adopting a beam filling factor of 0.15 and a fractional abundance for co of @xmath29
we have also estimated n(h@xmath8 ) .
although our estimates of the h@xmath8 density can be uncertain because of the unknown t@xmath27 , the large differences in the h@xmath8 density between b@xmath25 and the other components indicate that a rather small region has a density larger than other components . in sect .
3.2 we have discussed the morphological and kinematical arguments for the relationship between the different co components and ngc2359 . in the previous section , we have derived their physical conditions . in this section
, we shall consider our observational findings together with those already published , in order to address the nature and the origin of the interaction of the molecular gas with the nebula and its exciting wr star .
the component a1 is located mainly to the ne of the map , while the component a2 is present to the se of the nebula .
both components are narrow ( 2 km s@xmath3 ) and do not show any morphological or kinematical structures that indicate any disturbance produced by the nebula or the wr star .
in contrast , the component b has some kind of ubiquity in this field , in agreement with optical emission lines like the [ ] .
it appears as a broad component ( 4.55 km s@xmath3 ) following the east and the south border of the hii region which surrounds the nearly - spherical bubble .
we noted that the co emission is stronger _ outside _ the optical border of the hii region , but significant emission is also present in projection _ onto _ the hii region .
we think that the component b is the only component which undoubtely interacts with ngc2359 , because of its morphology , the large width of the lines , the higher temperature and the striking kinematics .
as already mentioned , the eastern part and the southern border of the hii region are clearly bounded by the most intense component b. furthermore , it is also striking the presence of faint co emission perfectly correlated with the hii region .
these results are also in agreement with the vibrationally excited h@xmath8 emission observed at 2.2@xmath30 m ( @xcite ) .
the co emission is dense and more opaque outside the hii region , while onto the hii region it is less dense and has a higher temperature ( up to 80k ) .
this temperature is compatible with the dust temperature inferred by marston ( 1991 ) in the same region . at this point
we wish to consider the @xmath4co behaviour in the southern part of the component b. we have only detected this molecule near the peak of the co emission .
this indicates that this is a _
unique _ region where the @xmath15co/@xmath4co changes appreciably , most likely due to an increase of the opacity in the @xmath15co lines . only 1 to the east of this position ( essentially one beam ) there is no significant emission in @xmath15co from the component b , just at the western border of the component a2 .
this spatial anti - correlation between components a2 and b is also present toward the southern interface .
furthermore , there is a bridge in the co emission which seems to connect both components within a small area ( see slice 3 of fig . 4 ) , suggesting that both components are related .
the co in the hii region might be heated by the wr radiation field ( @xcite ) , but the interface between the components a2 and b may not be excited by the same mechanism .
all these observational facts can be explained in a scenario in which a shock of 1014 km s@xmath3 , driven by the expanding hot bubble , impacts on the component a2 . as a consequence of the shock
, the gas was accelerated up to the velocities of the component b. the large width of the component b when compared with the component a2 can be understood by means of an increase of the turbulent motion due to the energy injected by the shock . furthermore , the systemic velocity of the hi shell found at larger scales ( presumably near the rest velocity of the gas ) is also similar to the velocity of the component a2 , supporting this scenario .
the shock heats the gas and forms a _ thin layer _ of compressed material with large h@xmath8 density and with high opacity in the @xmath15co .
this thin shocked layer would remain unresolved by the beam .
the kinematics shown in fig . 4 and the low @xmath15co/@xmath4co ratio at 5356 km s@xmath3 towards the co peak are very suggestive of this thin shocked layer .
indeed , the large optical depth of the co lines can only be explained if this emission arises from a very small region unresolved by the beam , probably less than 0.2 pc wide .
furthermore , the large co column density of b@xmath25 as compared with b@xmath24 would indicate large densities for these radial velocities ( up to several 10@xmath31 @xmath32 ) . obviously , in the scenario of the shocked gas responsible for the component b , b@xmath25 would represent the material which has been recently shocked .
high angular resolution observations are needed to disclose this particular region and to check our interpretation .
although our hypothesis seems reasonable , it should be pointed out that this shocked region is small and it can not be spatially resolved in our data . another crucial point is the weakness of the `` bridge '' which connects the components a2 and b. higher angular resolution and more sensitive observations are needed to confirm the presence of the shocked layer and to fully study its properties . if this is the case , ngc2359 would have the first direct evidence of shocked molecular gas in a wr environment .
the o - stage of hd56925 have created in the ism the hi shell reported in this paper and the iras shell reported by marston ( 1996 ) . in consequence , the optical nebula and its surroundings , where we find most of the co emission , have a more recent origin than the large scale hi shell .
the hii region has a size that indicate a dynamical age of less than 10@xmath33 yr .
this age is comparable with the wr or a previous phase , such as lbv or rsg ( @xcite ) . on the other hand
, many optical shells of enriched material are found close to the wr star ( @xcite ) .
this leads us to think that the material now interacting with the hii region was ejected in previous episodes of rsg or lbv phase .
chemical studies using other molecular line transitions , especially those capable of tracing the molecular gas enriched by these phases , will definitively confirm or discard these ideas .
we have analyzed the large - scale 21 cm - hi emission in a @xmath34 field around the nebula ngc2359 , and found a large expanding bubble roughly centered at the nebula and its exciting star , the wolf - rayet hd56925 .
the systemic velocity of the bubble is @xmath13 64 km s@xmath3 and it is expanding at 12 km s@xmath3 . at an assumed distance of 5 kpc ,
the bubble has a size of @xmath35 pc and a hi mass between 700 and 2400 m@xmath6 . by simple estimates of energetics , ages and stellar wind parameters
, we think that this feature is a wind - blown bubble produced by the o - progenitor of the wr star .
the presence of a nearly spatially - coincident iras shell reinforces our hypothesis .
we have also mapped the co j = @xmath0 and @xmath1 emission adjacent to ngc2359 and found three co components in this region .
two of these components , with radial velocities of 34 ( component a1 ) and 67 ( component a2 ) km s@xmath3 , have narrow profiles ( up to 2 km s@xmath3 ) and do not show any morphological or kinematical effects which indicate the disturbance by the nebula or the wr star .
however , the third component ( the component b , emitting at 54 km s@xmath3 ) , clearly bounds ngc2359 at its southern and eastern border .
the profiles are broad , with linewidths of 45.5 km s@xmath3 .
a velocity gradient of a few km s@xmath3 is noted towards the south and south - eastern interface with the nebula .
a weak `` bridge '' in velocity of small angular extension seems to connect this component with the component b2 at the southern part .
we have observed the @xmath4co j = @xmath0 line at a few points in the region where the component b is detected .
we have found a striking variation of the @xmath15co/@xmath4co ratio both in position and in velocity .
this ratio seems to decrease near the peak of the co emission , for velocities higher than 53 km s@xmath3 .
this particular emiting region is not resolved by the beam , and is probably 0.2 pc wide . from the observational data , we have estimated the physical conditions for each co component using the lvg approximation .
the highest excitation temperature and density are found in the component b. it is remarkable that high temperatures ( up to 80k ) are derived for the co emission projected onto the hii region . in view of the kinematics , the morphology and the physical properties of the co , we think that the component a2 was shocked and accelerated by the expanding bubble to reach the radial velocities of the component b. the shock front is still acting at the southern part of the region . the origin of such shock might be related to the wr phase of hd56925 or , more probably , to previous episodes of rsg of lbv . from the @xmath15co/@xmath4co ratio we infer the presence of a thin layer of dense gas which represent the more recently shocked material . our data , however , can not resolve spatially the region where a thin layer of shocked material is expected .
co observations with better resolution and sensitivity is needed to disclose the main features of the region .
a1 & 0.5 & & 5 & 0.6 & 1.0 & 0.6 & 0.6 & & 96 + a2 & 0.5 & & 6 & 1.0 & 0.4 & 0.8 & 0.8 & & 28 + b@xmath22 & 0.8 & & 8 & 1.8 & 0.5 & 0.3 & 0.4 & & 160 + b@xmath24 & 1.0 & 30 & 10 & 5.0 & 1.6 & 0.6 & 0.8 & @xmath45 & 10 + b@xmath25 & 1.0 & 5 & 10 & 58 & 21 .
& 12 & & @xmath13 0.1 & 120 + b@xmath23 & 2.0 & @xmath17 30 & @xmath17 80 & @xmath46 & 0.1 & @xmath280.1 & 0.2 & @xmath47 0.1 & @xmath48 8 + + the lvg approximation was used in all cases but b@xmath25 . for this particular region , we have + estimated the size of the emiting region and taken into account the beam filling factor . + all masses were computed by assuming a distance of 5 kpc . | ngc2359 is an optical nebula excited by the powerful wind and the radiation of the wolf
rayet star hd56925 .
we have investigated the interaction between this massive star and the surrounding neutral gas by analyzing the large - scale 21cm - hi emission and by mapping the nebula in the j = @xmath0 and the j = @xmath1 lines of co. we found a conspicuous ( @xmath2 pc ) hi shell , expanding at 12 km s@xmath3 , likely produced during the main - sequence phase of the star .
the molecular gas towards ngc2359 shows three velocity components .
two of these components , a1 and a2 , have narrow linewidths ( 12 km s@xmath3 ) and radial velocities of 3538 and 6468 km s@xmath3 , respectively .
the third component is detected at radial velocities between 50 and 58 km s@xmath3 and has a broader profile ( up to 5.5 km s@xmath3 ) .
furthermore , this component is morphologicaly related with the nebula and has a velocity gradient of a few km s@xmath3 . we have also estimated the physical parameters of the molecular gas by means of a lvg modelling of the co emission .
the gas projected onto the southern hiiregion of the nebula has low co column density and is rather hot , probably up to 80k .
several profiles of the @xmath4co j = @xmath0 line near the peak of the emission , together with a weak emission bridge between the broad and one of the narrow components ( component a2 ) , suggest the presence of a shock front acting in the southern part of the nebula .
this shock was likely produced in a previous rsg / lbv phase of hd56925 . |
blazars have been shown to exhibit a sequence of properties as a function of source power ( fossati et al .
1998 ) . as the source power increases , the emission line luminosity and the ratio of compton to synchrotron luminosity increase , while the synchrotron peak frequency @xmath0 and the ic peak frequency decrease .
recent multiwavelength studies support this scheme ( e.g. kubo et al .
1998 ) , although the result that the ratio of the compton to synchrotron luminosity increases with source luminosity suffers from limited statistics , and should be only considered tentative .
the initial division of blazars into flat spectrum radio quasars ( fsrqs ) and bl lacertae objects ( bls ) was based on the equivalent width ( ew ) of the broad emission lines .
sources with ew @xmath1 5 @xmath2 were classified as fsrqs and sources with ew @xmath3 5 @xmath2 as bls .
this difference in the ew of the emission lines has been interpreted as absence of a substantial blr in bls , and has been used to advance the idea that in bls the gev - tev emission is due to ssc , while in fsrqs the gev emission is due to ec .
however , the ew criterion does not correspond to a dichotomy ( scarpa & falomo 1997 ) , because a weak blr is present in bls as well .
additionally , it is not the blr luminosity @xmath4 that is relevant for the ec luminosity , but the blr photon energy density @xmath5 that is measured in the comoving frame of the non - thermal source .
one needs to know not only @xmath4 , but also the blr radius @xmath6 and the location of the non - thermal emitter relative to the blr in order to quantify the relative importance of ssc versus ec emission .
the size of the blr has been measured only for radio quiet agns through reverberation mapping ( kaspi et al .
2000 ) , and it has been found to scale with the ionizing luminosity @xmath7 of the acrretion disk as @xmath8 . it was additionally found that , assuming keplerian motions for the blr clouds , the mass of the central object scales with the ionizing luminosity as @xmath9 .
here we assume the scalings apply also to radio loud agn and examine the implications for blazars .
following the formalism of internal shocks propagating in a conical jet ( rees 1978 ) , we scale the location @xmath10 and the radius @xmath11 of the non - thermal spherical emitter with the mass @xmath12 of the central object .
we assume that @xmath4 is a fraction of @xmath7 and scale the kinetic luminosity and the poynting flux of the blob with @xmath7 .
under this scheme , the radius of the blr scales as @xmath13 , and the distance of the blob from the center of the system scales as @xmath14 .
therefore , we assume that for powerful sources the blob radiates from inside the blr , @xmath15 .
however , if we consider increasingly weaker sources , @xmath6 falls faster than @xmath10 and , below some critical luminosity , the blob radiates from outside the blr . for a blob inside the blr , @xmath16 , where @xmath17 is the bulk motion lorents factor of the blob . for a blob outside the blr , the solid angle subtended by the blr , as seen by the blob ,
is reduced .
this affects @xmath5 in two ways .
the first is the usual geometric @xmath18 factor , and the second is a relativistic de - boosting that , for @xmath19 , gives @xmath20 ( dermer & schlickeiser 1994 ) . the second effect dominates if the blob is located less than a few @xmath6 away from the blr .
for example , for @xmath21 , @xmath22 at a distance of @xmath23 is smaller than @xmath22 inside the blr by a factor of @xmath24 due to the usual @xmath18 and by a factor of @xmath25 due to relativistic de - boosting .
we apply this scaling using a model ( georganopoulos & kirk 2000 ) for the blob which includes both particle acceleration and radiative losses with ic scattering losses treated in the thomson regime .
particles are accelerated in an acceleration zone and escape into a radiation zone , before they eventually escape out of the system .
this simplified picture is intended to represent particles which are accelerated by a shock front , escape it , and radiate downstream before the compressed plasma re - expands .
whilst undergoing acceleration , particles simultaneously suffer synchrotron losses and losses by compton scattering of both photons of external origin ( ec ) and synchrotron photons ( ssc ) from the radiation zone . in the radiation zone ,
ec , ssc and synchrotron losses occur until the particle finally escapes after a time @xmath26 . the electron energy distribution ( eed ) is computed self - consistently , taking account of these processes . our assumption about the scaling of the poynting flux implies a constant magnetic field , which we take to be @xmath27 g. the blob
is assumed to move with a lorentz factor @xmath28 at an angle @xmath29 with respect to the line of sight .
low energy particles ( @xmath30 ) are injected into the blob at a rate of @xmath31 @xmath32 s@xmath33 .
also @xmath34 . for @xmath35 erg s@xmath33 ,
reverberation mapping gives @xmath36 cm .
we assume that at this luminosity @xmath37 and @xmath38 cm . in fig.1 , we let @xmath7 vary from @xmath39 erg s @xmath33 to @xmath40 erg s@xmath33 , and follow the behavior of the system . in the lower panel @xmath6 , @xmath10 , and @xmath11
are plotted as a function of @xmath7 . for weak sources ,
the blob is located outside the blr ( @xmath41 ) . as the source power increases , the blob gradually approaches and eventually enters the blr . in the middle panel of fig.1
we see how this affects @xmath22 ( the magnetic field energy density @xmath42 is constant under the adopted scaling ) . for weak sources ,
@xmath22 is much smaller than both @xmath42 and the self - consistently derived synchrotron photon density @xmath43 .
gradually @xmath22 increases , and eventually , for bright sources , dominates over both @xmath42 and @xmath43 . in the upper panel of fig.1
we plot the self - consistently derived break @xmath44 in the eed .
note that @xmath44 decreases as the power of the source decreases . in fig.2
we plot basic observable quantities as a function of @xmath7 . in the lower panel
we plot the synchrotron luminosity @xmath45 , the ssc luminosity @xmath46 , and the ec luminosity @xmath47 . at low powers
, @xmath48 is much weaker than @xmath46 and @xmath45 that are roughly equal .
as the source power increases , @xmath48 gradually dominates over @xmath46 and @xmath45 , and we end up with an ec dominated source . in the middle panel we plot the peak frequencies of the three emission components . note how the synchrotron peak frequency @xmath49 decreases as the source power increases .
similar behavior is also seen for the ssc peak frequency @xmath50 and the ec peak frequency @xmath51 .
finally in the upper panel of fig.1 we plot the ratio @xmath52 ( solid line ) of the synchrotron to accretion disk luminosity and the ratio @xmath53 of the ic luminosity to synchrotron luminosity .
@xmath45 dominates over @xmath7 in weak sources in agreement with the lack of a thermal component and weak emission lines in weak sources like bls .
@xmath7 becomes more significant for powerful sources , again in agreement with the strong emission lines of fsrqs and the accretion disk signature observed in some fsrqs ( e.g. in 3c 279 , pian et al 1999 ) .
the dominance of the compton over the synchrotron emission increases as the source power increases , in agreement with observations ( e.g. kubo et al .
note , however , that the observational case for an increasing compton dominance with source power is based on poor statistics due to the limited sensitivity of the _
cgro_. in fig.3 we plot the synchrotron luminosity @xmath54 versus the synchrotron peak frequency @xmath55 for the blazars studied by sambruna , maraschi & urry ( 1996 ) and kubo et al .
( 1998 ) . on top of the data points we plot the model tracks as a function of @xmath7 for a range of observing angles .
note that these tracks represent two model derived quantities , @xmath45 and @xmath49 as @xmath7 varies .
the model covers rather well the observed parameter space .
given the track of the model under an angle @xmath56 , we do not expect that any powerful ( @xmath57 erg s@xmath33 ) sources with high peak frequencies @xmath58 hz exist .
the discovery of such sources would pose a serious problem for this model .
in fig.4 we plot the peak luminosities of the three emission mechanisms for 5 model sources with @xmath7 ranging from @xmath39 up to @xmath40 erg s@xmath33 .
the behavior of the model is in good agreement with the general characteristics of the observed luminosity sequence ( compare with fig.12 of fossati et al .
the synchrotron peak frequency decreases from @xmath59 hz down to @xmath60 hz , as the synchrotron power increases .
the compton peak frequency in weak sources is in the tev regime ( @xmath61 hz ) and it is due to ssc , while in powerful sources it is in the gev regime ( @xmath62 hz ) and it is due to ec .
according to the model , although in these bright sources the energy losses are dominated by ec , the hard x - ray emission is due to ssc , in agreement with resent observations ( kubo et al .
1998 ) .
we note here that fig.4 corresponds to sources that are oriented at an angle @xmath29 . in reality one should expect a range of angles , which will give rise to significant scattering around the presented trend .
this scattering is visible in the work of fossati et al .
( 1998 ) and indicates that a proper unification scheme for blazars should include the effects of orientation . | we present preliminary results of our work on blazar unification .
we assume that all blazars have a broad line region ( blr ) and that the size of the blr scales with the power of the source in a manner similar to that derived through reverberation mapping in radio quiet active galactic nuclei ( agns ) . using a self - consistent emission model that includes particle acceleration
we show that according to this scaling , in weak sources like mkn 421 , the inverse compton ( ic ) scattering losses are dominated by synchrotron - self compton scattering ( ssc ) , while in powerful sources , like 3c 279 , they are dominated by external compton ( ec ) scattering of blr photons . in agreement with other workers , we show that even in the powerful sources that are dominated by ec scattering , the hard x - ray emission is due to ssc . finally , we show that this scaling reproduces well the observed sequence of blazar properties with luminosity . |
the remarkable progress in the field of ultracold atomic gases has provided access to a number of systems which may display quite new physical properties .
one of the most striking example is provided by ultracold fermionic gases @xcite and the bec - bcs crossover . while the physics of bose - einstein condensates ( bec ) is known from superfluid @xmath1he ( and from ultracold bosonic atomic gases ) , and the one of bcs condensates from superconductors , fermionic gases provide systems which display a smooth continuous evolution between these two extremes .
this is made possible by the existence of feshbach resonances which allow to control the value of the scattering length @xmath2 merely by changing the applied magnetic field .
for ultracold gases the kinetic energy of the atoms is so small that only s - wave scattering is relevant , and it is fully characterized by the scattering length . since pauli principle forbids s - wave scattering between identical atoms , the scattering length corresponds in most of the experiments performed so far to scattering between atoms in different hyperfine states of a same element .
these hyperfine states are often called for convenience `` spin up '' and `` spin down '' . in the case of a wide feshbach resonance ( as it occurs for example for @xmath3li and @xmath4k ) , the closed channel responsible for the feshbach resonance may be omitted and the system is described by a single open channel hamiltonian , where the interaction is characterized by the single parameter @xmath2 . with the densities produced in experiments
this interaction has quite a small range compared to the interatomic distance .
this single channel hamiltonian is particularly interesting because it is at the same time very simple and highly non trivial , and moreover it is realized with an excellent precision in these ultracold fermionic gases .
hence we may hope , in getting full control and understanding of this hamiltonian , to gain knowledge which may be applicable to more complex hamiltonians , such as those encountered in condensed matter physics or in quark matter @xcite . in this context
a simple general expression for the energy of a system described by this hamiltonian has been found by tan @xcite , which involves only the momentum distribution @xmath5 of the particles together with their large @xmath6 behaviour
. however the details raise unanswered mathematical difficulties .
nevertheless the expression can be checked in limiting cases ( see appendix [ a ] for details ) .
this expression is of high interest since it is quite non trivial .
it is directly related to the well - known problem that , for a contact interaction , the kinetic energy presents a formal divergence , because the momentum distribution @xmath5 behaves as @xmath0 for large momentum ( as found for example in the perturbative calculation of belyakov @xcite ) , a feature merely linked to two - body physics .
naturally there is no divergence in the energy itself , because the interaction energy comes in to compensate this feature from the kinetic energy .
this is easily seen from a simple finite range interaction model ( for example the square well potential ) , with range @xmath7 .
the momentum distribution decreases more rapidly than @xmath0 , for momentum beyond @xmath8 which acts as a cut - off in momentum space .
the apparently singular situation is found when one lets @xmath9 , which should be taken as the definition of the contact interaction . in the formula found by tan
, the total energy appears essentially as the kinetic energy with this divergence problem removed .
indeed the high momentum part , responsible for the divergence , is subtracted out and an additional explicit contribution which contains the scattering length appears .
this matter has been taken up recently by braaten and platter @xcite who made use of the operator product expansion developped by k. g. wilson .
they have in this way derived the energy relation in a compact and formal way .
in contrast with tan approach the interaction comes in explicitly . they have also obtained the expression of the known @xcite adiabatic relation for this contact potential hamiltonian . this convenient expression in terms of the scattering length and of the coefficient of the @xmath0 tail in the momentum distribution has been pointed out by tan @xcite .
quite recently it has been used @xcite to obtain the number of closed - channel molecules in the two - channel model . also very recently it has been derived and studied in details by zhang and leggett @xcite . in this paper
we present first a derivation of this energy formula , which is simple , explicit , fairly short and avoids unnecessary and uncontrolled mathematical complications .
it is in line with tan approach , in that it deals only with the kinetic energy .
this is done in a careful way in order to avoid divergences .
our treatment is similar in spirit to ref.@xcite .
this simple proof allows straightforward generalizations to more complicated situations , with unequal masses , several kinds of particles and two - dimensional space .
next we show how , in a field theoretical approach , the @xmath0 tail in the momentum distribution and the energy formula arise in a natural way . finally , as a specific example , we consider also quantitative calculations within the ladder diagrams approximation in the normal state .
there are different ways to obtain the energy and we can compare the various approximate results , which gives an indication on the quality of the approximation .
in this paragraph we consider the case where we have only two kinds of particles which are ultracold fermions .
the volume of the system is assumed to be unity .
namely we have @xmath10 particles with mass @xmath11 and @xmath12 particles with mass @xmath13 .
we consider directly the case where the masses @xmath11 and @xmath13 are different since it does not make any problem .
the positions of the @xmath14 particles are denoted @xmath15 , while those of the @xmath16 particles are @xmath17 . for these ultracold fermions only s - wave scattering
has to be taken into account .
hence only interactions between @xmath14 and @xmath16 particles have to be considered .
generalizations are considered in the next section .
the hamiltonian reads @xmath18 with ( we take @xmath19 ) : @xmath20 and @xmath21 we assume the interaction potential @xmath22 to be short - range . for clarity and simplicity
we assume that it has a definite range @xmath7 and satisfies @xmath23 for @xmath24 .
however , just as in standard scattering theory @xcite , the results will hold for physical short - range potentials where the interaction decreases rapidly enough with interparticle distance .
we consider an eigenfunction @xmath25 of this hamiltonian , having for example in mind the ground state wavefunction .
however this is not necessary and we can as well consider excited states , which leads to the extension of the results to non zero temperature , as pointed out by tan @xcite .
let @xmath26 be the corresponding wavefunction , with proper symmetrization and normalization to unity .
we want to calculate the energy corresponding to this state : @xmath27 where @xmath28 is for @xmath29 with @xmath30 and @xmath31 .
we follow the basic principle put forward by tan @xcite . in eq.([en1 ] ) the potential @xmath22 comes in only in the regions of the integration domain where @xmath32 , where @xmath33 and @xmath34 .
the overlap between these regions is negligible for dilute gases where @xmath7 is small compared with the mean interparticle distance @xmath35 .
there are @xmath36 such regions .
compared to the total integration domain , the domain of these regions is of order @xmath37 . for equal populations
@xmath38 , this would be of order @xmath39 .
since we have @xmath40 also inside these regions , and that @xmath41 inside these regions has no singular behaviour ( this can be checked explicitly since this behaviour is obtained from a one - body schrdinger equation , as we use below ) , we may neglect the contribution of these regions .
actually this has to be done for consistency since @xcite keeping only s - wave scattering for ultracold gases implies that terms of order @xmath42 , corresponding to higher angular momenta , are neglected .
hence the energy is simply obtained by calculating the kinetic energy outside these regions : @xmath43 we introduce now the fourier transform @xmath44 of @xmath45 with respect to all variables : @xmath46 where we use the notation @xmath47 .
the one - particle density distributions @xmath48 and @xmath49 are given by : @xmath50 this leads to : @xmath51\,\varphi(\{{\bf k}_i\},\{{\bf q}_j\ } ) \int_{|{\bf r}_m - \boldsymbol{\rho}_n|>r_0}\;d{\bf r}_i\,d\boldsymbol{\rho}_j\;e^{i \sum_{i}({\bf k}_i-{\bf k}'_i).{\bf r}_i + i \sum_{j}({\bf q}_j-{\bf q}'_j).\boldsymbol{\rho}_j}\end{aligned}\ ] ] if we did not have the restrictions @xmath52 in the last integral , it would give a factor proportional to @xmath53 and we would find the standard expression for the kinetic energy : @xmath54 which displays a divergence for large wave vector . hence the restriction is crucial in order to avoid the divergence .
however , since we want to have the standard expression for the kinetic energy appearing , we will write : @xmath55 where for simplicity we have not written explicitely the exponential integrand appearing in eq.([eqetf ] ) .
as long as all the @xmath56 and @xmath57 integrations are not performed in eq.([eqetf ] ) , no singularity appears so we can handle the various terms in eq.([eqsep ] ) separately .
as we have just mentionned the first one leads to the standard kinetic energy expression , so we have just to handle the other terms quite carefully since the limit @xmath9 is singular .
note that , in the regions @xmath58 , we will deal with the analytic continuation of the expression of the wavefunction for @xmath52 which we will discuss below .
this is naturally quite different from its physical value in these regions , since , for these last one , the interaction @xmath22 has to be taken into account .
let us consider the term @xmath59 in eq.([eqsep ] ) . in eq.([eqetf ] ) only the terms @xmath60 in the bracket will be relevant , and for now on we consider only these ones .
basically we will deal carefully with this two - body problem .
accordingly it is convenient to think , as long as we are not back to the many - body problem , that only these two particles scatter .
equivalently we could handle all the @xmath61 terms simultaneously , but this is much more cumbersome presentation . since they are more convenient to express the condition @xmath62 , we introduce the relative position @xmath63 and the center of mass position @xmath64 , together with their conjugate variables the total momentum @xmath65 and the relative momentum @xmath66 with @xmath67 .
we have : @xmath68 where @xmath69 is the reduced mass .
the difficulties arise from the @xmath70 term in the large @xmath6 limit .
it is convenient for the discussion to introduce the fourier transform @xmath71 of the wavefunction with respect to the variables @xmath72 and @xmath73 , instead of @xmath74 and @xmath75 .
it is merely given by : @xmath76 making use , in eq.([eqetf ] ) and eq.([eqsep ] ) , of @xmath77 , together with @xmath78 and @xmath79 , we find that , for the @xmath80 term in eq.([eqch ] ) , we have to deal with : @xmath81 where @xmath82 is a short - hand for @xmath83 .
we have used the fact that integration over all unrestricted position variables gives corresponding @xmath84 functions for the corresponding wavevector variables , as we have mentionned above . the problem for large @xmath6 in this expression
is directly linked to the behaviour of the wavefunction for small relative distance @xmath85 , which we consider now .
when the relative distance @xmath85 is small compared to the mean interparticle distance @xmath35 , the dependence of the wavefunction @xmath26 on @xmath72 is given by the solution of the relative motion of the two - body problem .
for ultracold gases the energy corresponding to this motion is nearly zero and the corresponding wavefunction is , for @xmath24 , proportional to @xmath86 with @xmath87 where @xmath2 is the scattering length @xcite .
this form is actually valid provided @xmath85 is small compared to a typical particle wavelength , that is @xmath88 . since we have @xmath89 , there is a range of validity for this form for @xmath24 .
we are also interested in the case of large positive @xmath2 , where a two - body bound state exists with wavefunction proportional to @xmath90 . in this case the above form requires @xmath91 .
there is again a range of validity for this condition for @xmath24 since we are in practice interested in physical situations where the scattering length @xmath2 is large compared to the potential range @xmath7 .
hence for small @xmath85 we have : @xmath92 since in the evaluation of the @xmath93 term of eq.([eqsep ] ) , we will let @xmath9 , only large values of @xmath94 are relevant .
indeed if we considered only bounded values @xmath95 , this term would go to zero for @xmath9 ( and the @xmath6 integral would converge because of the cut - off @xmath96 ) .
hence we have to consider @xmath97 . in this case the dependence of @xmath71 on @xmath6 is entirely linked to the short distance behaviour on @xmath85 of @xmath26 given by eq.([eqptr ] ) .
we have in this limit : @xmath98 where @xmath99 and @xmath100 are the fourier transform of @xmath101 and @xmath102 respectively .
as it is well known , the fourier transform of @xmath103 is @xmath104 . on the other hand , for the calculation of @xmath105 in expression ( [ eqkterm ] )
, we do not have to take into account the constant @xmath106 in the wavefunction since it gives zero when we apply on it the kinetic energy operator @xmath107 ( corresponding to the factor @xmath108 ) .
all this fourier transform calculation can be done quite carefully by multiplying the wavefunction @xmath101 by a convergence factor @xmath109 , then letting @xmath110 .
this confirms the above results .
finally , when handling expression ( [ eqkterm ] ) , we have merely @xmath111 . physically
: @xmath112 is the isotropic probability distribution of wavevector @xmath94 in the relative motion of the two @xmath14 and @xmath16 particles we are considering .
we have found that its leading behaviour for @xmath97 is @xmath113 with : @xmath114 on the other hand , with respect to @xmath115 integration in expression ( [ eqkterm ] ) , it is more convenient to go back to @xmath72 space through @xmath116 , which gives a factor @xmath117 . since @xmath101 is isotropic , we can perform explicitely the angular @xmath72 integration @xmath118 .
finally , for expression ( [ eqkterm ] ) , we are left with the calculation of @xmath119 . this @xmath85 integration
is easily performed , leading to : @xmath120 we write now , with this result , the partial contribution @xmath121 to the energy eq.([eqetf ] ) , coming from the term @xmath59 in eq.([eqsep ] ) and where we retain only the @xmath80 term in eq.([eqch ] ) .
we obtain : @xmath122 in the first term in the right - hand side , we have put a cut - off @xmath96 since this integral diverges when @xmath123 .
this divergence will just compensate in the final result the above mentionned divergence in eq.([eqkinen ] ) .
hence the global result will be convergent as expected .
if in the last integral of eq.([eq11 ] ) we were setting @xmath124 , we would get @xmath125 and a divergent result .
instead this integral is perfectly convergent when we calculate it explicitly for @xmath126 and then take properly the @xmath9 limit .
indeed we see that it is not divergent because @xmath127 involves oscillatory functions .
the @xmath128 term merely gives a result proportional to @xmath129 , where @xmath130 is the dirac distribution . since @xmath126 the contribution of this @xmath128 term is zero . on the other hand , owing to : @xmath131 the @xmath132 term gives a contribution @xmath133 to the integral , leading to a contribution @xmath134 to @xmath121 .
as we have mentionned above , we could improve the presentation of our handling of the @xmath6 integration , to deal with perfectly defined integrals , by introducing a convergence factor @xmath135 and then let @xmath136 .
physically this would correspond to regularize the wavefunction in the @xmath137 limit .
this would confirm our above results .
we have not yet taken into account the @xmath138 term in eq.([eqch ] ) , because there is no singular behaviour associated with it . the corresponding contribution @xmath139 of the term @xmath59 in eq.([eqsep ] ) goes to zero as @xmath9 , as it is obvious directly and can be checked by following the same procedure as above .
finally one sees easily that , for all the other kinetic energy terms @xmath140 in the bracket of eq.([eqetf ] ) , the contribution from the term @xmath59 in eq.([eqsep ] ) also goes to zero when @xmath9 .
hence the total contribution @xmath141 of the @xmath59 term is merely @xmath142 .
we rewrite now the sum @xmath141 in terms of the variables corresponding to particle @xmath14 and @xmath16 , instead of the relative and center of mass variables .
we have : @xmath143 with again a cut - off @xmath96 understood for the summation over @xmath94 , and where we have introduced : @xmath144 related to @xmath145 by @xmath146 . with @xmath147 , we have : @xmath148 we make use of eq.([eqch ] ) to go back to the @xmath149 and @xmath150 variables .
in the first term we change the summation variables from @xmath151 to @xmath152 , making use of @xmath153 .
in particular we have @xmath154 .
however we can use the identity : @xmath155=0\end{aligned}\ ] ] where @xmath156 is any fixed vector ( this result is obvious by changing @xmath72 into @xmath157 ) , to write , @xmath158=0 $ ] at fixed @xmath159 . proceeding in the same way with the second term of eq.([eqe1 ] ) , we obtain : @xmath160 with a cut - off @xmath96 understood for the summation over @xmath149 and @xmath150 .
in conclusion we have calculated all the contributions coming from the presence of the term @xmath59 in eq.([eqsep ] ) .
these are just the three terms , proportional to @xmath145 , in eq.([eqfin11 ] ) . to summarize , for the expression of the energy , we are back to eq.([eqetf ] ) with the restriction @xmath161 removed , and the above three @xmath145 terms added .
we have just to repeat the same argument for all the other restrictions @xmath162 .
in this way we obtain @xmath163 analogous @xmath145 terms , which are naturally all equivalent after a change of variables . taking eq.([eqkinen ] ) into account , this leads to the final expression for the energy : @xmath164 + \frac{1}{2 m { _ { \downarrow}}}\sum_{{\bf q}}\left[q^2\,n { _ { \downarrow}}(q)-\frac{n_4}{q^2}\right ] + \frac{n_4}{8\pi \mu a}\end{aligned}\ ] ] where @xmath165 . indeed , from eq.([eqnup ] ) , @xmath48 and @xmath49 behave in this way for large @xmath6 since for example @xmath149 scatters with all the @xmath166 , and each scattering brings a contribution @xmath113 .
that is any of the @xmath10 particles scatters with any of the @xmath12 particles .
hence the summations in eq.([eqfin ] ) are perfectly convergent .
this formula is the simple generalization of the formula found by tan @xcite to the case where the two species of involved particles have different masses . in this derivation
we introduced a cut - off to manipulate separately each contribution .
this makes an easier presentation for the derivation .
however this is just a convenience .
we could avoid it by handling all the terms simultaneously .
the presentation would be much awkward , but we would only deal with well defined convergent integrals , without any need for a cut - off .
it is first worthwhile to note that the above derivation did not make use of the statistics of the particles .
hence the result is valid for bosons as well as for fermions .
naturally it is artificial for bosons to consider only scattering between different species , although it might just happen that this scattering is the dominant one .
nevertheless for bosons there is no reason to exclude scattering between particles belonging to the same species .
if we consider first the case of a single bosonic species , with @xmath167 particles of mass @xmath168 in the unit volume and with scattering length @xmath2 , we can follow the same procedure as in the preceding section .
we can write the equivalent of eq.([eqetf ] ) for a single species , with a restriction @xmath169 working now between any of the @xmath170 couples of particles .
taking care of the restrictions in the same way as in the preceding section , we end up with : @xmath171 + \frac{n_4}{8\pi m a}\end{aligned}\ ] ] with @xmath172 .
we can then generalize this result to any mixture of @xmath173 boson species with @xmath174 particles of mass @xmath175 in the unit volume ( @xmath176 ) and interaction between species @xmath177 and @xmath178 characterized by scattering lengths @xmath179 .
we find in the same way : @xmath180 + \frac{1}{16\pi}\sum_{i , j}\frac{n_{4ij}}{\mu_{ij } a_{ij}}\end{aligned}\ ] ] where @xmath181 is the reduced mass for the @xmath182 scattering , @xmath183 and @xmath184 .
naturally the formula is more complex for these mixtures , since it requires the knowledge of the @xmath185 constants @xmath186 associated with the @xmath187 scattering .
the same result works for ultracold fermionic mixtures , except that we have to set @xmath188 because of pauli principle .
while in 1d situations the kinetic energy converges , and indeed in exact solutions of many - body problems the energy is precisely calculated by evaluating the kinetic energy , there is in 2d a divergence analogous to the 3d case , except that the divergence is logarithmic in the 2d case .
we show here how the procedure followed in 3d can be extended to this 2d case .
actually there is not so much difference since the space dimensionality does not appear in the principle of the procedure .
the changes appear only when one comes to pratical matters .
we keep the same notations for the variables , but naturally we have to deal now with two dimensional variables and integrations .
first the expression of the wavefunction @xmath86 for the relative motion at small distance @xmath24 is modified into @xmath189 , which is solution of the 2d equation @xmath190 .
the length @xmath2 for which @xmath191 plays the role of the scattering length and is naturally obtained from the interaction potential .
the fourier transform of @xmath86 is @xmath192 , omitting the irrelevant fourier transform of @xmath193 .
similarly the large @xmath6 behaviour of @xmath194 is @xmath113 , with @xmath195 appearing in eq.([eqp4 ] ) instead of @xmath196 .
then the 2d angular integration gives @xmath197 where @xmath198 is the first kind bessel function .
hence , instead of eq.([eq11 ] ) , we obtain : @xmath199 where in the last expression we have changed to the variable @xmath200 , and we have naturally to take the cut - off @xmath123 at the end of the calculation , when the standard kinetic energy term is included , as we have done in section [ detder ] . the logarithmic divergence arising in this standard term
is expected to be compensated by the contributions of all the terms similar to eq.([eq112d ] ) . indeed , integrating by parts
, we have : @xmath201^{k_c}_{0}- r^2_0 \int_{0}^{k_c}dk\ , k\,\ln k \ln(kr_0)\ , j_0(kr_0)\end{aligned}\ ] ] to evaluate the first term we use @xcite ( when the integral is not absolutely convergent , we define it as above with an exponential convergence factor , with an extremely weak decreasing behaviour ) : @xmath202 ( where @xmath203 is the standard euler gamma function ) together with @xmath204 . note in particular that @xmath205 .
this gives @xmath206 , which is also consistent with the fact that the fourier transform of @xmath207 is @xmath208 .
hence we find indeed that eq.([eqterma ] ) provides the required compensating term to avoid a divergent result .
the other terms lead to constants , so we may take immediately in their expression the limit @xmath123 . first ,
except for the prefactor @xmath209 , the factor of @xmath193 in the second term of the integral in eq.([eq112d ] ) is : @xmath210 by the change of variable @xmath211 and by integrating by parts .
this means that the divergent contribution of the second term in eq.([eq112d ] ) is proportional to @xmath212 , as could be expected from dimensional analysis . finally in eq.([eq112d ] ) we have still a contribution proportional to @xmath213 : @xmath214 again by integrating by parts .
gathering this term and the second term in eq.([eqterma ] ) , we have : @xmath215 where , in the last step for the term proportional to @xmath216 , we have used again @xmath205 . the remaining integral
is calculated again by making use of eq.([eqbessel ] ) and is found to be @xmath217 , where @xmath218 is the euler constant . gathering the above results we obtain : @xmath219 finally , just as for the 3d case , we want to go back to the variables corresponding to particle @xmath14 and @xmath16 .
it may be seen that this does not lead to additional contributions .
this leads us finally to : @xmath220 + \frac{1}{2 m { _ { \downarrow}}}\lim _ { q_c \rightarrow \infty}\left[\sum^{q < q_c}_{{\bf q}}q^2\,n { _ { \downarrow}}(q)-\frac{n_4}{2\pi } \ln(q_ca)\right ] + n_4\frac{\ln 2 -c}{4\pi \mu } \end{aligned}\ ] ] where we recall that @xmath165
. this result can be checked explicitly in the molecular case ( see appendix a ) .
let us see now how the above expression for the energy arises in the field theoretic formalism . to be specific we will restrict ourselves for simplicity to the case of major interest , namely the one of two fermionic species with equal populations , so that @xmath221 . since all the @xmath14 and @xmath16 quantities are equal ,
we do not write explicitely this index . we consider also the non zero temperature case since it does not make any problem .
we deal first with the case of a normal system , and then extend the results to the superfluid case .
we consider first the large @xmath6 dependence of @xmath222 . for free fermions at temperature @xmath223 , the density distribution has an exponential tail proportional to @xmath224 .
however , as pointed out in ref .
@xcite , interactions modify this behaviour and give rise on general grounds to a @xmath0 dependence which dominates the exponential tail .
this is explicit in the weak coupling domain where the interaction can be treated perturbatively and the scattering length @xmath2 is small , as it has been done by belyakov @xcite at zero temperature : @xmath225 where @xmath226 is the fermi momentum @xmath227 . in the general case the distribution @xmath222 is obtained from the temperature green s function @xmath228 , where @xmath229 ( with @xmath167 an integer and @xmath230 ) is the matsubara frequency , by : @xmath231 where @xmath232 .
we separate out in this equation the free particle contribution by writing dysons s equation : @xmath233 where @xmath234 is the free particle green s function @xmath235^{-1}$ ] , with @xmath236 the free particle kinetic energy , and @xmath237 is the self - energy . the first term in eq.([eqdys ] )
gives in eq.([eqn ] ) the free particle contribution , namely the fermi distribution .
for large @xmath6 its exponential tail mentionned above is completely dominated by the algebraic decay @xmath0 that we will obtain .
hence we are left only with the second term . for large @xmath6 , implying a large kinetic energy for the particle , we expect the effect of interaction to be small in the same spirit as the born approximation in this regime .
since @xmath238 describes this effect we expect it to be small .
hence we may to lowest order replace in this second term @xmath239 by @xmath234 . on the other hand we replace , in a standard way @xcite , the summation over matsubara frequencies by a frequency integration over a contour @xmath240 encircling the imaginary axis in the anticlockwise direction .
this leads to the following expression for the dominant contribution to @xmath222 at large @xmath6 : @xmath241 where @xmath242 is the fermi distribution function .
the contour can be deformed into the sum of a contour enclosing the positive frequency @xmath243 half - plane and another contour enclosing the negative frequency @xmath244 half - plane , both being in the clockwise direction .
closing the contour at infinity in the @xmath243 half - plane is allowed by the presence of the fermi distribution @xmath245 , and closing it at infinity in the @xmath244 half - plane by the presence of the @xmath246 factor , which is then omitted since it does not play any other role .
now contributions from the double pole at @xmath247 will contain from the fermi distribution a factor @xmath224 which makes them negligible .
similarly @xmath248 has also singularities with frequencies which are large and positive , when @xmath6 is large .
the fermi distribution @xmath245 will again make their contribution exponentially small . on the other hand , as it is shown in details in appendix [ b ]
, @xmath248 has a pole at @xmath249 for large @xmath6 , that is deep in the @xmath244 half - plane . roughly speaking this pole appears since a @xmath14 particle with large @xmath94 and @xmath250 will scatter with a @xmath16 particle with essentially opposite parameters @xmath251 and @xmath252 .
this happens because , just as in section [ detder ] , the large values occur for the relative motion , but not for the center of mass motion
. then the @xmath16 particle propagator has a pole at @xmath253 , which produces a pole for @xmath248 at the same frequency . for this pole ,
we have merely @xmath254 and @xmath255 .
if we call @xmath256 the corresponding residue of @xmath248 , we obtain in this way : @xmath257 hence we see that the @xmath0 dependence of @xmath222 at large @xmath6 emerges quite naturally in this approach .
we note that our analysis may seem inconsistent , since we have first argued that @xmath248 should be small for large @xmath6 , but then considered a pole of @xmath248 , in the vicinity of which it is large .
the justification is that , as long as the contour is far away from the pole , @xmath248 is indeed small everywhere on the contour .
making use of the pole is then a convenient way to calculate the contour integral .
we consider now the expression of the energy .
it can be written in terms of the green s function by a standard formula @xcite which bears from the start a strong analogy with the expression found by tan @xcite : @xmath258 where again @xmath232 .
this equation includes an overall factor 2 , coming from summation over spin .
the essence of this formula is that , calculating @xmath259 by heisenberg equations of motion , one finds @xmath260 , that is the kinetic energy @xmath261 plus twice the interaction energy @xmath262 . adding the kinetic energy @xmath261 , which gives the @xmath263 term in eq.([eqengen ] ) , leads to @xmath264 . in practice eq.([eqengen ] ) is not so useful since one has to deal carefully with the divergent behaviour occuring for large @xmath265 and large @xmath6 , which is quite painful numerically . in this respect the number equation eq.([eqn ] ) is much more convenient .
now , to single out clearly the kinetic energy contribution , we can rewrite eq.([eqengen ] ) as : @xmath266 the first term is just the kinetic energy @xmath267 . in the second term , which is just @xmath262
, we can again replace the matsubara frequency summation by the same contour integrals as above .
moreover the quantity appearing in this summation is just @xmath268 .
however the term @xmath269 gives a zero contribution to these closed contour integrals .
hence we are left with : @xmath270 we can now analyze the behaviour of the interaction energy term for large @xmath6 , as we have done above for the particle number . in this range
@xmath248 is small so we may replace @xmath271 by @xmath272 .
hence we have to deal with the same expression as in eq.([eqnk ] ) , except that @xmath272 is not squared .
following the same arguments we obtain for large @xmath6 : @xmath273 hence we have shown explicitely that the large @xmath6 behaviour of the interaction energy comes in to cancel the divergent behaviour of the kinetic energy .
naturally this is the expected result .
accordingly we may write : @xmath274 + \sum_{\bf k } \left[\frac{m r_{\sigma}}{k^2}-\frac{1}{2i\pi } \int_{\mathcal c } d\omega \,f(\frac{\omega } { t})\,\sigma({\bf k},\omega)g({\bf k},\omega)\right]\end{aligned}\ ] ] where , from eq.([eqnkgdk ] ) and eq.([eqintgdk ] ) , both brackets give convergent integrals for large @xmath6 .
the first term is just the kinetic energy from which , according to eq.([eqnkgdk ] ) , the large @xmath6 behaviour has been subtracted . by comparison with eq.([eqfin ] ) , the second term in eq.([eqen2 ] ) is just @xmath275 , a result by no means obvious .
we are now in a situation where we can calculate the energy from three different formulas , as we explain just below . for the exact theory they would naturally give the same result .
however if we use an approximate scheme , the results are expected to be different .
the differences can , in some rough way , be seen as a measure of the errors resulting from the approximation .
hence it is of interest to make such a comparison .
specifically we have taken as an example the ladder approximation @xcite at unitarity and we have performed numerical calculations of the energy as a function of temperature in this approximation .
we use three different ways to obtain the energy .
our first expression for the energy is the general one eqs.([eqenc],[eqen2 ] ) .
the second one is the energy formula found by tan , that is eq.([eqfin ] ) with equal masses .
finally we use the non - interacting gas scaling relation @xmath276 between the grand potential @xmath277 and the energy @xmath278 , which is also valid at unitarity @xcite due to the lack of any microscopic energy scale .
we calculate the grand potential itself _ via _ the density @xmath279 ( obtained by integration of @xmath222 ) using the thermodynamic relation @xmath280 ( we have used the fact that @xmath277 goes to @xmath281 in the high temperature regime @xmath282 ) . as noted above this last method
is much more convenient numerically than the two others , since there are no convergence problems for large @xmath6 in the @xmath94 summation .
the results of these numerical calculations are shown in fig.[fig1 ] .
we have checked that we recover the proper high temperature behavior ( virial expansion ) with the three methods , as it should be .
this common evolution toward the virial result is clearer in the inset , where the results are shown with an extended scale , but this convergence is fairly slow .
the results of the three methods are fairly close , which is coherent with the fact that we expect the ladder approximation to be reasonably good , mostly toward higher temperatures .
indeed at lower temperature the differences between the three results are growing , mostly for the result of the tan formula .
this is clearly due to the proximity of the superfluid critical temperature , which occurs at @xmath283 in this approximation .
the major numerical problem with the use of eq.([eqfin ] ) is to obtain the proper coefficient of @xmath0 in the large momentum behaviour of @xmath222 .
this requires in particular to be sure that the asymptotic regime is reached numerically .
the same problem arises in eq.([eqen2 ] ) since one must find numerically that , for large @xmath6 , the kinetic energy is exactly balanced by the interaction energy . in the present case of the ladder approximation , we can check that the asymptotic regime as been reached since we have analytical expressions . in order to check numerically this large momentum behaviour of the contributions coming into the two brackets of eq.([eqen2 ] ) , we have performed , as we indicate specifically below , an expansion of the occupation number @xmath222 and of the interaction energy term up to order @xmath284 and @xmath285 respectively .
as indicated in the above subsection the self - energy is small in the large momentum and frequency limit and we can expand the green s function in powers of the self - energy .
hence we have for the occupation number and the interaction energy term : @xmath286 we start from equation ( b1 ) and following the arguments given above and detailed in appendix b , we neglect the contribution coming from the pole of the green s function , that gives exponentially small terms , and we only consider the contribution from the cut of the vertex @xmath287 on the real frequency axis .
following @xcite , we call this self - energy contribution @xmath288 : @xmath289 with @xmath290 , and @xmath291 .
the bose distribution effectively limits the frequency @xmath277 to be at most a few @xmath223 and the @xmath292-integral is therefore effectively bounded .
similarly to what is done in appendix b , we consider the singularities of the self - energy located at @xmath293 . in the large momentum @xmath294 and large frequency @xmath295 limit
, we expand the free green s function in eq.([eq : sigma_gamma ] ) in power of @xmath296 , leading to : @xmath297 where @xmath298 in this limit , the contribution to the self - energy due to the cut of the vertex @xmath287 is a sum of poles of order @xmath299 located at @xmath300 . in eq.([eq : n_k ] ) , only the second term is of interest , since the first one gives the fermi distribution which does not contribute to the algebraic tail .
so we use the expansion eq.([eq : sigma_dev ] ) in eq.([eq : n_k ] ) and find the following expansion for large wave vector @xmath6 : @xmath301 the first three terms of this expansion give the dominant contribution .
indeed , it is easily checked that the term of order @xmath302 in eq.([eq : n_k ] ) give contributions at least of order @xmath303 .
@xmath304 does not depend on @xmath94 while the @xmath94 dependence of @xmath305 vanishes after angular integration .
for the coefficient @xmath306 we can write @xmath307 . finally we find for the particle distribution in the large momentum limit : @xmath308 comparing with eq.([eqnkgdk ] ) we see that @xmath304 is identical to the coefficient @xmath309 defined in the above subsection .
we have checked the numerical convergence in the large @xmath6 limit by comparing the second term in our expansion eq.([nkfin1 ] ) with our numerical results .
the coefficients are identical with a precision of less than one percent .
we can perform the same analysis for the interaction energy term in eq.([eq : e_int_k ] ) and we find the following expansion : @xmath310 here also we have checked the numerical convergence with a precision of less than one percent .
compared to the normal state , we have now to modify the dyson s equation eq.([eqdys ] ) to take into account @xcite the anomalous self - energy @xmath311 and the anomalous propagator @xmath312 .
we have : @xmath313 again for large @xmath6 , we have to lowest order @xmath314 and @xmath315 small .
for such a large @xmath6 we expect this normal self - energy @xmath315 to behave as in the normal state , discussed in the preceding subsection . on the other hand
the anomalous self - energy is expected to go to a constant , without any singular behaviour .
this implies from eq.([eqdysonb ] ) that @xmath312 is small . in this case the first term in the right - hand side of eq.([eqdysonb ] ) is negligible and from the second term we have : @xmath316 substitution into eq.([eqdysona ] ) gives : @xmath317\end{aligned}\ ] ] hence , in addition to the pole at @xmath249 coming as discussed above in @xmath315 , the last term gives also a pole at the same place from the explicit factor @xmath318 .
this leads to the large @xmath6 behaviour : @xmath319\,\frac{1}{k^4}\end{aligned}\ ] ] where @xmath320 .
actually the additional term is well - known in the case of bcs theory which gives indeed , with standard notations , @xmath321 for large @xmath6 .
going now to the expression for the energy , there is naturally no modification in eq.([eqenb ] ) . on the other hand we have now , in the second term of eq.([eqenb
] ) involving @xmath322 , to make use of the appropriate dyson s equation eq.([eqdysona ] ) .
this leads , instead of eq.([eqenc ] ) , to : @xmath323\end{aligned}\ ] ] now the large @xmath6 analysis goes essentially as in the normal state , and is similar to the one for the particle distribution since there is only a factor @xmath234 which is different in the two calculations .
this factor leads to a factor @xmath324 in the present case , compared to the particle distribution , that is we have for large @xmath6 : @xmath325 \simeq -m\left [ r_{\sigma}+|\delta _ { \infty}|^2\right]\,\frac{1}{k^2}\end{aligned}\ ] ] and the generalization of eq.([eqen2 ] ) is : @xmath326}{2\,k^2}\right ] \\
\nonumber & + & \sum_{\bf k } \left[\frac{m \left [ r_{\sigma}+|\delta _ { \infty}|^2\right]}{k^2}-\frac{1}{2i\pi } \int_{\mathcal c } d\omega \,f(\frac{\omega } { t})\,\left[\sigma({\bf k},\omega)g({\bf k},\omega ) + \delta ( { \bf k},i\omega ) f^{+}({\bf k},i\omega)\right]\right]\end{aligned}\ ] ]
in this paper we have investigated the large momentum algebraic tail in the particle distribution and the energy formula found by tan associated with this tail .
we have provided a simple derivation of this energy formula , which rests on the fact that , in evaluating this energy for the short range potential under consideration , the interaction energy contribution is zero in most of phase space , which makes it negligible .
hence only the kinetic energy has to be calculated .
the basis of the derivation is the careful subtraction of the kinetic energy divergent contribution for interparticle distance less than the potential range .
this derivation is easily generalized to particles with different masses , to arbitrary mixtures , and to two - dimensional space .
we have then shown how the algebraic tail arises naturally in the field theoretical many - body approach , from the analytic structure of the self - energy . making use of these ingredients
we have shown how , starting from a standard general expression of the energy in terms of the green s function , one obtains a formula with a kinetic energy part which has the same structure as in the formula found by tan .
this has been done both in the normal and in the superfluid state .
finally we have taken the various exact formulas allowing to obtain the energy , and we have compared in the normal state at unitarity the resulting numerical values obtained within the ladder approximation .
we are quite grateful for discussions to l. tarruell and c. salomon . `` laboratoire de physique statistique de lecole normale suprieure '' is `` associ au centre national de la recherche scientifique et aux universits pierre et marie curie - paris 6 et paris diderot - paris 7 '' .
we are also very grateful to d. dhumires and n. regnault for their recurring help in numerical matters .
in the bec limit @xmath327 , one is led to consider the problem of a single molecule .
the normalized wave function is @xmath328 and the density distribution is given by the square of its fourier transform @xmath329 .
the coefficient of the @xmath285 tail is @xmath330 .
one checks that indeed the energy @xmath331 gives the proper binding energy @xmath332 . in the weak coupling limit @xmath333 the energy ( per unit volume )
can be expanded in powers of @xmath334 ( the contributions which would come from bcs pairing are exponentially negligible ) . to second order @xcite : @xmath335\end{aligned}\ ] ] where @xmath336 and @xmath337 .
second order perturbation gives also : @xmath338 where @xmath339 is a lengthy analytical expression given in refs.@xcite .
for large @xmath6 , @xmath340 with @xmath341 .
since there is no first order correction to the density distribution , the first order correction to the energy , in the energy relation eq.([eqfin ] ) , is merely given by the last explicit term @xmath342 , which coincides indeed with the ( mean field ) first order term in eq.([enwc ] ) .
to check this relation to second order , we need the coefficient @xmath343 of the tail in the third order contribution @xmath344 to the density distribution , which is not available
. however we can extract it from the adiabatic relation @xcite @xmath345 , which gives @xmath346 .
hence we have to check that : @xmath347 + \frac{2b_2nk_f^4a^2}{m}=b_2n \frac{k_f^2}{m}(k_fa)^2\end{aligned}\ ] ] which implies : @xmath348=-\frac{2}{3}\,b_2\end{aligned}\ ] ] we have checked numerically this equation with a seven digits precision corresponding to all the digits given above for @xmath349 .
finally in the 2d case , in the bec limit , one has again a single molecule problem .
the normalized wave function of the bound state with energy @xmath350 is @xmath351 . for small @xmath352 the bessel function @xmath353 , so that @xmath354 .
the density distribution is @xmath355 , so that @xmath356 . with these ingredients
eq.([eqfin2d ] ) is easily checked .
we discuss here in details the analytic structure of @xmath248 in order to justify the assumptions we have made above in section [ normst ] .
it is useful to consider first the case of the ladder approximation @xcite , where the structure is explicit . in this case
the self - energy is given by : @xmath357 where in the last step we have replaced the summation over bosonic matsubara frequencies @xmath358 , with @xmath359 being an integer , by an integral over the contour @xmath240 introduced above , with @xmath360 being the bose distribution function . the vertex @xmath361 is given by : @xmath362 \label{eqgam}\end{aligned}\ ] ] deforming contour @xmath240 as we have done above in section [ normst ] , we will from eq.([sigmagen ] ) express @xmath363 in terms of the singularities of @xmath361 and of @xmath364 , which are on the real frequency axis
. on one hand @xmath364 has a simple pole at @xmath365 . on the other hand the singularities of @xmath361
correspond first to the continuum of scattering states with energy @xmath366 , arising from the product of the two @xmath234 in eq.([eqgam ] ) .
in addition there is the possibility of a bound state of the two particles , corresponding to a zero of the right - hand side of eq.([eqgam ] ) . as a result
the spectrum of these singularities has some lower bound @xmath367 . on the other hand for large @xmath292
their frequencies are bounded from below by @xmath368 , corresponding to the kinetic energy of the mass center .
now let us first consider the contribution from the pole of @xmath369 in eq.([sigmagen ] ) , located at @xmath370 . the bose factor @xmath371 will produce a factor @xmath372 , which for large @xmath6 implies an exponentially small factor @xmath224 .
this makes the corresponding contribution to the self - energy irrelevant for the algebraic tail of @xmath373 .
the only way to avoid this is to have also a large @xmath374 , so @xmath375 is not large .
however in this case we will obtain a factor @xmath376 .
the singularities of @xmath248 we are looking for are obtained from this factor by continuing the imaginary frequency @xmath377 to the real @xmath250 axis .
however we know that , when @xmath292 is large , the singularities of @xmath361 are located at large frequencies of order of @xmath378 .
this implies that the corresponding singularities of @xmath248 will be for @xmath379 , that is @xmath380 since @xmath375 is not large . as explained below eq.([eqn ] ) these large @xmath250 singularities in @xmath248 give only exponentially small contributions to @xmath373 and again do not come in the @xmath285 tail we are looking for .
hence we conclude that the pole of @xmath234 in eq.([sigmagen ] ) gives no contribution to this tail , which comes accordingly only from the contribution of @xmath381 . for this contribution
the dependence of @xmath382 on frequency comes explicitely from the @xmath234 term , and it has poles for @xmath383 . here
@xmath384 runs over the frequencies of the singularities of @xmath381 , but in practice the bose factor limits them to some finite range @xmath385 .
this implies in particular that @xmath292 is bounded ( just as above the terms produced by the tail of the bose factor will necessarily have an exponential factor and do not contribute to the algebraic tail of @xmath373 ) .
hence , for large @xmath386 , we obtain from the @xmath234 term a single pole located at @xmath387 .
the corresponding residue is , from eq.([sigmagen ] ) : @xmath388 where @xmath389 is a clockwise contour enclosing only the singularities of @xmath381 and the last expression ( with @xmath390 ) is the real axis integral obtained by calculating the contour integral from the jump of the imaginary part of @xmath391 across the real frequency axis .
let us consider now the general situation where no approximation is made .
actually , provided we replace the bare propagator @xmath234 by the full propagator @xmath239 , eq.([sigmagen ] ) remains valid .
we have : @xmath392 indeed , considering the self - energy of , say , the @xmath14 particle , there will be in any diagram a first interaction with a @xmath16 particle , hence correspondingly a @xmath393 factor . merely isolating this factor and gathering the rest into @xmath394 produces eq.([sigmagen1 ] ) .
we may express @xmath394 in terms of the standard @xcite full vertex @xmath395 by isolating the hartree term in the self - energy .
this gives , for a contact interaction : @xmath396 where we have used a four - vector notation @xmath397 . here
the coupling constant @xmath398 is related to the scattering length @xmath2 and to the cut - off @xmath96 , related to the contact interaction , by @xmath399 .
physically @xmath400 describes the scattering of two particles @xmath6 and @xmath401 , and accordingly it will have the same qualitative properties as we have discussed above for @xmath361 ( where we have been quite general for purpose ) .
we consider the contributions in eq.([sigmagen1 ] ) from the singularities of @xmath239 .
either @xmath159 is small compared to @xmath94 , so @xmath402 is large and we can replace @xmath239 by @xmath234 , and as above we will have an exponential factor coming from the bose factor . or @xmath159 is comparable to @xmath94 , in which case the corresponding singularities of @xmath394 will start around @xmath378 and the resulting singularities for @xmath248 will be at least @xmath380 , which makes them irrelevant for the algebraic tail of @xmath373 .
hence we have only to consider the contributions of the singularities of @xmath394 in the calculation of eq.([sigmagen1 ] ) . again
those with large @xmath159 ( i.e. with small @xmath403 ) will be located at high frequency @xmath384 , and the bose factor produces an exponential factor which makes them irrelevant .
we are left only with contributions arising for @xmath159 small compared to @xmath94 , in which case @xmath239 can again be replaced by @xmath234 .
this produces for @xmath248 an explicit pole at @xmath387 , with a residue : @xmath404 actually it seems physically reasonable that , in the large @xmath94 limit , @xmath405 depends only on @xmath406 .
indeed this limit corresponds to a short range scattering between two @xmath14 and @xmath16 particles . in this case , in a way analogous to the one discussed in section [ detder ] for the energy formula , this scattering should be essentially described by two - body physics .
the situation is then similar to the one found in the ladder approximation .
the @xmath6 dependence is just the one which is already explicit , and we are left with a constant for the residue , given by a formula like eq.([eqresid ] ) with a modified @xmath381 . for a very recent review ,
see s. giorgini , l. p. pitaevskii and s. stringari , rev.mod.phys .
* 80 * , 1215 ( 2008 ) .
r. casalbuoni and g. nardulli , rev.mod.phys .
* 76 * , 263 ( 2004 ) . s. tan , arxiv : cond - mat/0505200 ; annals of physics * 323 * , 2952 ( 2008 ) . v. a. belyakov , sov .
jetp * 13 * , 850 ( 1961 ) .
e. braaten and l. platter , phys .
lett . * 100 * , 205301 ( 2008 ) .
l d. landau and e. m. lifshitz , _ quantum mechanics _ ( pergamon press , oxford , 1985 ) .
s. tan , arxiv : cond - mat/0508320 ; annals of physics * 323 * , 2971 ( 2008 ) .
f. werner , l. tarruell and y. castin , arxiv:0807.0078 s. zhang and a. j. leggett , arxiv:0809.1892 .
we have been aware of this paper when our work was essentially completed . for an introduction to many - body techniques and superconductivity ,
see for example a. a. abrikosov , l. p. gorkov and i. e. dzyaloshinski , _ methods of quantum field theory in statistical physics _
( dover , new york , 1975 ) and a. l. fetter and j. d. walecka , _ quantum theory of many - particle systems _ ( mcgraw - hill , new york , 1971 ) .
i. s. gradshteyn and i. m. ryzhik , _ table of integrals , series and products _
( academic press , 1980 ) .
r. sartor and c. mahaux , phys .
c * 21 * , 1546 ( 1980 ) .
a. perali , p. pieri , g. c. strinati and c. castellani , phys .
b * 66 * , 024510 ( 2002 ) .
r. combescot , x. leyronas and m. yu .
kagan , phys .
a * 73 * , 023618 ( 2006 ) and references therein .
t. l. ho , phys .
lett . * 92 * , 090402 ( 2004 ) . | we present a simple derivation of the energy formula found by tan , relative to the single channel hamiltonian relevant for ultracold fermi gases .
this derivation is generalized to particles with different masses , to arbitrary mixtures , and to two - dimensional space .
we show how , in a field theoretical approach , the @xmath0 tail in the momentum distribution and the energy formula arise in a natural way . as a specific example
, we consider quantitative calculations of the energy , from different formulas within the ladder diagrams approximation in the normal state .
the comparison of the results provides an indication on the quality of the approximation . |
by matching the luminosity , atmospheric parameters , and oscillation frequencies of models with the observed ones , asteroseismology is used to determine fundamental parameters of stars .
asteroseismology is also used to probe physical processes in stars and diagnose internal structures of stars @xcite .
asteroseismology is a powerful tool for studying the structure and evolution of stars .
stars with a mass larger than @xmath4 are considered to have a convective core during their main sequence ( ms ) stage . due to the fact that the overshooting of the convective core can bring more hydrogen - rich material into the core
, the evolution of a star could be significantly affected by the overshooting . thus determining the size of the convective core including the overshooting region is important for understanding the structure and evolution of stars .
however , the size of the convective core has never been determined directly from observed data of stars .
generally , the understanding of the size of the convective core derives from the computation of evolutionary models of stars . when seeking to probe the internal structures of stars with low-@xmath5 p - modes , the small separations , @xmath6 , @xmath7 , @xmath8 , and @xmath9 , and the ratios of the small separations to the large separations , , , @xmath10 , and @xmath11 ( * ? ? ?
* ; * ? ? ?
* and references therein ) , are considered to be the very useful diagnostic tools .
the small separations @xmath6 and @xmath7 are defined as @xcite @xmath12 and @xmath13 but in calculation , the smoother five - point separations are adopted .
stars with masses between about 1.1 and 1.5 have a convective core during their ms .
the discontinuity in density at the edge of the convective core increases with the evolution of the stars .
the rapid variation of density with depth in a stellar core can distort acoustic wave propagation in stellar interiors , producing a reflected wave @xcite .
the reflectivity can come from the rapid density change at the edge of the convective core @xcite .
for the modes with frequencies larger than a critical frequency , they can penetrate into the convective core .
partial wave reflection at the core boundary could lead to acoustic resonances in the convective core @xcite . as a consequence , at high frequencies ,
we would see a periodic variation in the small separations with frequency @xcite .
if this periodic component is determined from observations , it can be used for constraining the size of the convective core @xcite .
@xcite developed the theory of semiclassical analysis that can more accurately describe the low - degree p - modes and the small separations .
however , @xcite pointed out that their expressions for the perturbations in the phase shifts are not transparent enough to serve as a basis for simple estimates .
the effects of the convective core on @xmath6 , @xmath7 , , and are also studied by other authors @xcite .
the conclusion is that the ratios and can be affected by the presence of the convective core . in order to isolate the frequency perturbation produced by the edge of the convective core
, @xcite and @xcite defined a tool @xmath14 .
they have shown that the tool can potentially be used to infer information about the amplitude of the discontinuity in the sound speed at the edge of the convective core but it is unable to fully isolate the frequency perturbation .
@xcite show that the ratios and of a star with a convective core can be described by equation @xmath15 where the quantities @xmath16 and @xmath17 are two parameters , the @xmath18 is the frequency of @xmath19 mode whose inner turning point is located on the boundary between the radiative region and the overshooting region of the convective core . in this work ,
we propose a method to estimate the radius of the convective core of ms stars with masses between about 1.1 and 1.5 from observed frequencies of low - degree p - modes .
the estimated radius is comparable with that obtained from evolutionary model .
individual frequencies of p - modes of kic 9812850 have been extracted by @xcite .
the mass of kic 9812850 estimated by @xcite is @xmath20 .
thus kic 9812850 could have a convective core .
we determined the radius of the convective core of kic 9812850 in two ways .
one is estimated from the observed frequencies ; the other is determined from the best model for kic 9812850 . in section 2 , a formula that can be used to determine the radius of the convective core from oscillation frequencies
is proposed and is applied to different stars . in section 3 ,
based on finding the maximum likelihood of models of a grid of evolutionary tracks , the best - fit model of kic 9812850 is found out ; and then , we compare the radius of the convective core of the best model with that determined from oscillation frequencies .
finally , we give a discussion about the domain of the validity of the method and summary in section 4 .
the inner turning point , @xmath21 , of the mode with a frequency @xmath22 is determined by @xmath23 where @xmath24 is the adiabatic sound speed at radius @xmath21 , the value of the parameter @xmath25 is @xmath26 @xcite . for the modes with @xmath19 , the frequency @xmath18 of the mode
whose inner turning point is just located on the boundary between the radiative region and the overshooting region of the convective core can be estimated by equation ( [ eqb ] ) from observed frequencies and the ratios computed from the frequencies .
thus the radius of the convective core including the overshooting region , @xmath27 , can be determined by @xmath28 in the middle stage of ms stars with masses between about 1.1 and 1.5 , the magnitude of @xmath29 is of the order of about @xmath30 cm s@xmath31 .
thus for the ms stars extracted individual frequencies of low - degree p - modes , combining the equation ( [ eqb ] ) , the radius @xmath27 can be estimated by @xmath32 from the observed data . * by using the function of nonlinear curve fitting of origin software where chi - square fitting is used and hessian matrix is calculated , * from the observed frequencies @xcite and ratio of hd 49933 , * we obtained * that the value of * the parameter @xmath18 of equation ( [ eqb ] ) * is @xmath33 * for hd 49933*. from the observed frequencies of kic 6225718 @xcite and ratio , the value of @xmath18 is estimated to be about @xmath34 . from the frequencies of kic 2837475 and kic 11081729 given by @xcite , the value of @xmath18 is estimated to be @xmath35 for kic 2837475 and @xmath36 for kic 11081729 . using formula ( [ rc2 ] ) , one can obtain that the radius of the convective core including overshooting region is about @xmath37 for kic 6225718 , @xmath38 for hd 49933 , @xmath39 for kic 2837475 , and @xmath40 for kic 11081729 .
the models of the four stars were determined @xcite .
the radius of the convective core including overshooting region is around @xmath37 for the model 14 of kic 6225718 @xcite , @xmath41 for the model m52 of hd 49933 @xcite , @xmath42 for the model ma14 of kic 2837475 @xcite , and @xmath43 for the model ma6 of kic 11081729 @xcite .
the values of radius @xmath27 of these stars , estimated from observed data , are in good agreement with those obtained from the models of the stars .
this indicates that the radius of the convective core @xmath27 of the ms stars could be determined directly from observed frequencies and ratios and by using equations ( [ eqb ] ) and ( [ rc2 ] ) .
@xcite extracted the frequencies of low-@xmath5 p - modes of kic 9812850 . using the ratio computed from the frequencies and the observed frequencies @xmath44
, the values of @xmath16 , @xmath18 , and @xmath17 in equation ( [ eqb ] ) are estimated to be @xmath45 @xmath46 , @xmath47 , and @xmath48 , respectively , for kic 9812850 .
thus the radius @xmath27 of kic 9812850 is predicted to be about @xmath49 by formula ( [ rc2 ] ) .
in order to compare the value of @xmath27 estimated by equations ( [ eqb ] ) and ( [ rc2 ] ) for kic 9812850 with that of evolutionary model of kic 9812850 , we sought for the best model of kic 9812850 that match both non - seismic constraints and seismic characteristics in a grid of evolutionary models .
we used the yale rotation evolution code ( yrec ) @xcite to construct the models .
for the microphysics , the opal equation - of - state table eos2005 @xcite and opal opacity table gn93 @xcite were adopted , supplemented by the @xcite opacity tables at low temperature .
the models with a mass less than 1.30 take into account the diffusion and settling of both helium and heavy elements by using the diffusion coefficients of @xcite .
the standard mixing - length theory is adopted to treat convection . the mixing - length parameter @xmath50 is a free parameter in this work . for the sun ,
the value of the @xmath50 for the yrec is @xmath51 .
the distance of the overshooting of the convective core is defined as @xmath52 , where is a free parameter and @xmath52 is the local pressure scale - height .
the full mixing of material is assumed in the overshooting region .
the initial helium mass fraction is fixed at the standard big bang nucleosynthesis value 0.248 @xcite and 0.295 .
all models are evolved from zero - age ms to the end of ms .
the values of the input parameters , mass , @xmath50 , , and @xmath53 for the calculations are summaried in table [ tab1 ] .
the adiabatic oscillation frequencies @xmath22 of models were computed by using the pulsation code jig7 of @xcite .
the effects of the near - surface effects of a model on the frequencies were calculated by using the method of @xcite .
kic 9812850 is an f8 star @xcite .
the value of [ fe / h ] given by @xcite is @xmath54 , but that given by @xcite is @xmath55 for kic 9812850 . combining the value of @xmath56 of @xmath57 of the sun ,
the value of @xmath58 of kic 9812850 is estimated to be between 0.016 and 0.033 for the [ fe / h ] of @xcite , or in the range of 0.014 and 0.018 for the [ fe / h ] of @xcite .
the effective temperature of kic 9812850 is @xmath59 k @xcite or @xmath60 k @xcite .
the estimated atmospheric parameters of stars hotter than 6,000 k could be affected by the method of spectral analysis @xcite .
therefore , the atmospheric parameters determined by both @xcite and @xcite were considered in this work .
the parallax of kic 9812850 is in the range between about 5.9 and 17.5 mas @xcite .
the bolometric correction of kic 9812850 is estimated from the tables of @xcite .
the extinction of kic 9812850 is given by @xcite .
the visual magnitude of this star is @xmath61 mag @xcite .
thus the luminosity of kic 9812850 is estimated to be about @xmath62 . in order to find the best model for kic 9812850
, we calculated the likelihood function of all models .
the likelihood function is defined as @xcite @xmath63 where @xmath64^{2},\ ] ] the quantity @xmath65 indicates the observed , @xmath66 , @xmath58 , and @xmath22 , while the @xmath67 corresponds to the , @xmath66 , @xmath58 , and @xmath22 of models .
the quantity @xmath68 represents the observational error of @xmath65 .
the value of @xmath69 is 45 .
moreover , the values of classical and of models were also computed as a reference .
the and are defined as @xmath70^{2 } \label{chic}\ ] ] and @xmath71^{2 } , \label{chinu}\ ] ] respectively , where @xmath72 and @xmath68 denotes the observational error , @xmath73 corresponds to frequencies .
the observational error of @xmath74 is indicated by @xmath75 .
when the model evolves to the vicinity of the error - box of luminosity and effective temperature in the hertzsprung - russell diagram , the time - step of the evolution for each track is set as small as @xmath76 myr , which ensures that the consecutive models have an approximately equal . for a given mass , the model that maximizes @xmath77
is chosen as a candidate for the best - fit model .
table [ tab2 ] lists four models which have a larger @xmath77 in the calculations and shows that model m3 has the maximum @xmath77 .
figure [ fig1 ] compares the distributions of the observed and with those calculated from the models listed in table [ tab2 ] .
the distributions of and of kic 9812850 are reproduced well by models m2 and m3 .
the * right * panels of figure [ fig1 ] show that there are periodic variations in the differences between the observed ratios and those of models .
this may come from the effects of the helium ionization region and the base of the convective envelope on the observed frequencies @xcite .
model m3 not only maximizes the likelihood function in the calculations , but reproduces the distributions of observed and of kic 9812850 .
therefore , m3 is chosen as the best - fit model of kic 9812850
. moreover , figure [ fig1 ] shows that the distributions of and of m3 are reproduced well by equation ( [ eqb ] ) with @xmath78 , @xmath79 , and @xmath80 .
this indicates that ratios and can be described by equation ( [ eqb ] ) .
the value of @xmath27 of model m3 is @xmath3 , which is in good agreement with @xmath1 estimated by equation ( [ rc2 ] ) .
the radius of the convective core including overshooting region of the best model of kic 9812850 is successfully estimated by equations ( [ eqb ] ) and ( [ rc2 ] ) from observed oscillation frequencies .
moreover , the value of @xmath81 of kic 9812850 is @xmath82 @xcite , which is less than @xmath83 of @xmath18 .
the value of of m3 is @xmath84 , which is consistent with the deduction that if the value of @xmath81 of a star is less than the value * of * @xmath18 , the star may have a small @xcite .
* when * angular frequencies of modes are larger than * a critical frequency @xmath85 , * the modes can penetrates into the convective core * of stars*. assuming that the effects of the convective core on oscillations is related to @xmath86 , where @xmath16 is a free parameter , @xcite obtained the * equation * ( [ eqb ] ) as the result of fourier transform of @xmath86 .
thus the * equation * ( [ eqb ] ) is invalid for stars whose core is * radiative .
* figure [ fig2 ] shows the distributions of h mass fraction , adiabatic sound speed , and of * core - radiative models in different evolutionary stages . * the ratio decrease with increase in frequency .
the distributions can not be reproduced by * equation ( [ eqb ] ) .
the core of model s2 in figure [ fig3 ] is also radiative .
the distribution of of the model can not be reproduced by equation ( [ eqb ] ) too .
* @xcite pointed out that * the discontinuity in density * at the boundary of a convective core can distort acoustic wave propagation in stellar interior , producing a reflected wave .
the effects of the convective core on oscillations are related to the fact that the discontinuity reflects acoustic waves . * therefore , * the * equation * ( [ eqb ] ) is invalid for stars with a convective core but without the discontinuity in density or sound - speed at the edge of the convective core .
* model s3 in figure [ fig3 ] has a small convective core but has no an obvious discontinuity in density or sound speed at the edge of the convective core ( see figure [ fig3 ] ) .
the distribution of of the model can not be reproduced by equation ( [ eqb ] ) . while the model s4 has an obvious sound - speed discontinuity at the edge of the convective core .
the distribution of of model s4 is almost reproduced by equation ( [ eqb ] ) .
the cores of models s1 in figures [ fig4 ] and [ fig5 ] are also convective , but there is no an obvious discontinuity in density or sound speed of the models .
the distributions of of the models can not be reproduced by equation ( [ eqb ] ) too . while the distributions of of models with a convective core and an obvious discontinuity in sound speed at the edge of the convective core are almost reproduced by equation ( [ eqb ] ) ( see figures [ fig4 ] and [ fig5 ] ) . *
the modes with @xmath87 are considered to be able to reach the center of a star . according to equation ( [ tp ] ) , it is more difficult to arrive at the convective core for the modes with @xmath88 than for the modes with @xmath19
. thus the frequency @xmath89 could be the frequency of the @xmath19 mode whose inner turning point is located on the boundary between the radiative region and the overshooting region of the convective core .
@xcite show that there are periodic variations in small separations with period determined approximately by the acoustic diameter of the convective core , i.e. , the period @xmath90 the larger the value of @xmath27 , the longer the @xmath91 , i.e. , the smaller the frequency @xmath92 .
according to equation ( [ tp ] ) , the larger the value of @xmath27 , the smaller the frequency @xmath18 .
thus the frequency @xmath93 of roxburgh could be related to the frequency @xmath18 . the value of sound speed decreases from @xmath94 cm s@xmath31 to @xmath95 cm s@xmath31 in the convective core of model s3 of star with @xmath96 ( see figure [ fig4 ] ) .
if the @xmath97 in equation ( [ tc ] ) is replaced by @xmath29 , the value of @xmath93 can be estimated to be about @xmath98 . from equation ( [ rc1 ] )
, one can obtain @xmath99 which is very close to @xmath93 .
in stellar interior , sound speed decreases with the increase in radius .
the value of @xmath29 varies with the mass and the age of stars and is affected by overshooting .
the value of @xmath29 of the most of ms stars with masses between about 1.1 and 1.5 is mainly in the range of @xmath100 cm s@xmath31 ( see figures [ fig4 ] and [ fig5 ] ) . * for example , for * a star with @xmath96 , @xmath101 , @xmath102 , and @xmath103 , when it evolves from central hydrogen abundance @xmath104 to @xmath105 , the value of @xmath29 decreases from about @xmath106 cm s@xmath31 to @xmath107 cm s@xmath31 ; * for * a star with @xmath108 , @xmath101 , @xmath102 , and @xmath109 , when it evolves from @xmath110 to @xmath111 , the value of @xmath29 decreases from about @xmath112 cm s@xmath31 to @xmath107 cm s@xmath31 .
therefore , in the most of ms stage of stars with masses between about 1.1 and 1.5 , * taken a model in the middle stage of ms as a reference , * one can assume that there is a change of @xmath113 in @xmath29 , i.e. , @xmath114 cm s@xmath31 .
* for our sample , the relative uncertainty of @xmath18 determined by chi - square fitting from observed data is between about @xmath115 and @xmath116 .
but our sample is small .
the relative uncertainty of @xmath18 of other stars might be larger than @xmath116 . in order to estimate the uncertainty of the estimated @xmath27 of other stars when this method is applied to the stars , we assume that the relative uncertainty of @xmath18 for other stars is of the order of @xmath113 and apply the uncertainty of @xmath113 to all cases .
* as a consequence , the * relative * uncertainty of the estimated @xmath27 is about @xmath117 .
table [ tab3 ] shows that the values of the radius of the convective core determined by equations ( [ eqb ] ) and ( [ rc2 ] ) from observed frequencies of different stars are in good agreement with those obtained from the best models of the stars .
combining equation ( [ eqb ] ) , we propose here for the first time using formula ( [ rc2 ] ) to estimate the radius of the convective core including overshooting region of ms stars with masses between about 1.1 and 1.5 from observed frequencies and ratios .
the estimated values of the radius of the convective core of four stars are consistent with those of the best models of the four stars .
using the observed frequencies and ratios of kic 9812850 , equations ( [ eqb ] ) and ( [ rc2 ] ) predict the radius of the convective core of kic 9812850 is @xmath1 . in order to confirm this prediction
, we constructed a grid of evolutionary tracks . basing on finding the maximum likelihood of models
, we obtained the best - fit model of kic 9812850 with @xmath118 , @xmath119 , @xmath120 k , @xmath121 , @xmath122 gyr , @xmath123 , and @xmath124 .
the best model can reproduce asteroseismic and non - asteroseismic characteristics of kic 9812850 .
the value of the radius of the convective core of the best - fit model is in good agreement with that predicted by formula ( [ rc2 ] ) .
equations ( [ eqb ] ) and ( [ rc2 ] ) aid in understanding the structure of stars directly from the observed frequencies .
the author thanks the anonymous referee for helpful comments that helped the author improve this work , as well as the support from the nsfc 11273012 , 11273007 , the fundamental research funds for the central universities , and the hscc of beijing normal university .
alexander , d. r. , & ferguson , j. w. 1994 , apj , 437 , 879 ammons , s. m. , robinson , s. e. , strader , j. , laughlin , g. , fischer , d. , & wolf , a. 2006 , apj , 638 , 1004 appourchaux , t. , chaplin , w. j. , garca , r. a. , et al .
2012 , a&a , 543 , a54 basu , s. , chaplin , w. j. , & elsworth , y. 2010 , apj , 710 , 1596 benomar , o. , baudin , f. , campante , t. l. et al .
2009 , a&a , 507 , l13 brand@xmath125o , i. m. , cunha , m. s. , & christensen - dalsgaard , j. 2014 , mnras , 438 , 1751 brand@xmath125o , i. m. , cunha , m. s. , creevey , o. l. , & christensen - dalsgaard , j. 2010 , an , 331 , 940 bruntt , h. , basu , s. , smalley , b. , et al .
2012 , mnras , 423 , 122 chaplin , w. j. , basu , s. , huber , d. , et al .
2014 , apjs , 210 , 1 christensen - dalsgaard , j. , & houdek , g. 2010 , ap&ss , 328 , 51 cunha , m. s. , & metcalfe , t. s. 2007 , apj , 666 , 413 cunha , m. s. , & brand@xmath125o , i. m. 2011 , , 529 , a10 deheuvels , s. , bruntt , h. , michel , e. , et al .
2010 , , 515 , a87 demarque , p. , sarajedini , a. , guo , x .- j .
1994 , apj , 426 , 165 droege , t. f. , richmond , m. w. , sallman , m. p. , creager , r. p. 2006 ,
pasp , 188 , 1666 flower , p. j. 1996 , apj , 469 , 355 ge , z. s. , bi , s. l. , li , t. d. , liu , k. , tian , z. j. , yang , w. m. , liu , z. e. , yu , j. 2014 , mnras , 447 , 680 grevesse , n. , & sauval , a. j. 1998 , in solar composition and its evolution , ed .
c. fr@xmath126hlich et al .
( dordrecht : kluwer ) , 161 guenther , d. b. 1994 , apj , 422 , 400 guenther , d. b. , demarque , p. , & gruberbauer , m. 2014 , apj , 787 , 164 iglesias , c. , rogers , f. j. 1996 , apj , 464 , 943 kharchenko , n. v. 2001 , kfnt , 17 , 409 kjeldsen , h. , bedding , t. r. , & christensen - dalsgaard , j. 2008 , apjl , 683 , l175 liu , z. , yang , w. , bi , s. , et al .
2014 , apj , 780 , 152 mazumdar , a. , monteiro , m. j. p. f. g. , ballot , j. ,
2014 , apj , 782 , 18 metcalfe , t. s. , et al .
2014 , apjs , 214 , 27 molenda - zakowicz , j. , sousa , s. g. , frasca , a. , et al .
2013 , mnras , 434 , 1422 pinsonneault , m. h. , kawaler , s. d. , sofia , s. , & demarqure , p. 1989 , apj , 338 , 424 rogers , f. j. , & nayfonov , a. 2002 , apj , 576 , 1064 roxburgh , i. w. , & vorontsov , s. v. 1994 , mnras , 267 , 297 roxburgh , i. w. , & vorontsov , s. v. 1999 , in stellar structure : theory and test of connective energy transport , ed .
a. gimnez , e. f. guinan , & b. montesinos , asp conf .
( san francisco ) , 173 , 257 roxburgh , i. w. , & vorontsov , s. v. 2000a , mnras , 317 , 141 roxburgh , i. w. , & vorontsov , s. v. 2000b , mnras , 317 , 151 roxburgh , i. w. , & vorontsov , s. v. 2001 , mnras , 322 , 85 roxburgh , i. w. , & vorontsov , s. v. 2003 , a&a , 411 , 215 roxburgh , i. w. , & vorontsov , s. v. 2004 , in esa special publication , vol .
538 , stellar structure and habitable planet finding , ed .
f. favata , s. aigrain , & a. wilson ( noordwijk : esa ) , 403 roxburgh , i. w. , & vorontsov , s. v. 2007 , mnras , 379 , 801 silva aguirre , v. , ballot , j. , serenelli , a. m. , & weiss , a. 2011 , a&a , 529 , a63 silva aguirre , v. , basu , s. , brand@xmath125o , i. m. , et al .
2013 , apj , 769 , 141 spergel , d. n. , bean , r. , dore , o. , et al . , 2007 ,
apjs , 170 , 377 thoul , a. a. , bahcall , j. n. , loeb , a. 1994 , apj , 421 , 828 tian , z. j. , bi , s. l. , yang , w. m. , et al . 2014 , mnras , 445 , 2999 wright , c. o. , egan , m. p. , kraemer , k. e. , price , s. d. 2003 , aj , 125 , 359 yang , w. 2015 , arxiv:1508.00955 yang , w. 2016 , apj , 821 , 108 yang , w. , & bi , s. 2007a , apjl , 658 , l67 yang , w. , & bi , s. 2007b , , 472 , 571 yang , w. , & meng , x. 2010 , newa , 15 , 367 yang , w. , meng , x. , bi , s. , tian , z. , liu , k. , li , t. , li , z. 2012 , mnras , 422 , 1552 yang , w. , tian , z. , bi , s. , ge , z. , wu , y. , & zhang , j. 2015 , mnras , 453 , 2094 p0.70cmcccccccccccccc model & @xmath127 & & @xmath131 & @xmath132 & age & @xmath53 & @xmath133 & @xmath50 & & @xmath134 & @xmath135 & & & @xmath77 + & ( ) & ( k ) & ( ) & ( ) & ( gyr ) & & & & & ( ) & ( @xmath136hz ) & & + m1 & 1.44 & 6267 & 4.75 & 1.851 & 2.861 & 0.024 & 0.728 & 1.75 & 0.2 & 0.144 & 2364 & 1.5 & 0.7 & 1.8@xmath137 + m2 & 1.46 & 6188 & 4.58 & 1.864 & 2.950 & 0.028 & 0.724 & 1.75 & 0.2 & 0.147 & 2248 & 1.0 & 2.0 & 9.1@xmath138 + m3 & 1.48 & 6408 & 5.28 & 1.867 & 2.606 & 0.024 & 0.728 & 1.95 & 0.2 & 0.149 & 2246 & 0.7 & 1.6 & 1.0@xmath139 + m4 & 1.50 & 6320 & 5.07 & 1.880 & 2.620 & 0.030 & 0.722 & 1.95 & 0.2 & 0.152 & 2184 & 1.0 & 1.8 & 1.6@xmath140 + cccc star & @xmath18/@xmath136hz & @xmath27/@xmath2 & @xmath141/@xmath2 + kic 6225718 & @xmath34 & @xmath142 & 0.056 + kic 9812850 & @xmath47 & @xmath143 & @xmath3 + hd 49933 & @xmath33 & @xmath144 & 0.170 + kic 2837475 & @xmath35 & @xmath145 & 0.358 + kic 11081729 & @xmath36 & @xmath146 & 0.370 + @xmath147the uncertainty is estimated by assuming that there is an uncertainty of @xmath113 in @xmath97 and @xmath18 .
the values of @xmath27 are estimated by using equation ( [ rc2 ] ) , while the values of @xmath141 are obtained from the best models of the stars . | the determination of the size of the convective core of main - sequence stars is usually dependent on the construction of models of stars .
here we introduce a method to estimate the radius of the convective core of main - sequence stars with masses between about 1.1 and 1.5 @xmath0 from observed frequencies of low - degree p - modes .
a formula is proposed to achieve the estimation .
the values of the radius of the convective core of four known stars are successfully estimated by the formula .
the radius of the convective core of kic 9812850 estimated by the formula is @xmath1 @xmath2 . in order to confirm this prediction ,
a grid of evolutionary models were computed .
the value of the convective - core radius of the best - fit model of kic 9812850 is @xmath3 @xmath2 , which is in good agreement with that estimated by the formula from observed frequencies .
the formula aids in understanding the interior structure of stars directly from observed frequencies .
the understanding is not dependent on the construction of models . |
dry granular systems are known to exhibit both liquid and solid - like properties . while there are many instances where a granular assembly exhibits liquid - like properties @xcite , examples of granular systems illustrating solid - like properties are rare with one exception : the typical conical sandpile shape @xcite . to craft more complex shapes , e.g. , sand art
, it is necessary to add additional constraints .
these constraints can be added in the bulk or at the boundary . adding a small quantity of water to the sand introduces constraints in the bulk in the form of capillary bridges @xcite .
conversely , encapsulating dry grains in a container is an example of constraints that are applied at the boundary @xcite .
in general , encapsulation requires isolating materials from their surroundings and it is achieved either by ( i ) introducing another material in the interfacial region or ( ii ) facilitating processes at the interface that create self - encapsulation .
examples of the former range from the simple instance of a bag or a silo containing cereal grains to more complex examples of thin polymer films wrapping liquid droplets @xcite , texturing of liquid droplets ( ` liquid marbles ' ) by attaching hydrophobic powder to its surface @xcite and stabilizing emulsions by particles or surfactants @xcite .
in contrast , a liquid that oxidizes on contact with atmosphere to develop a stress bearing skin is an example of self - encapsulation @xcite . in this paper
we study the self - encapsulation of a granular system consisting of hydrophobic sand @xcite under water .
this system self - generates a _ skin _ which encapsulates dry hydrophobic sand - grains and stabilizes the trapped air ( bubble ) against the force of buoyancy .
removal of trapped air breaks down this encapsulation .
the paper also explores the mechanical properties of this system at multiple scales : from the pinning of the three phase contact line at the roughness scale of the particle , plastic flow at the grain - scale , to sample - spanning mechanical responses .
it may be noted here that while hydrophilic granular systems both dry and wet are widely studied @xcite , hydrophobic sand grains submerged in a non - wetting liquid like water remain largely unexplored @xcite even though such systems are of practical importance , especially in pharmaceutical , food and petroleum industries where newer encapsulation strategies are in great demand @xcite
. mechanical response of a system is usually described in a small neighborhood of a reference state . for a dry granular assembly
a ` state ' of a system is described in terms of the center of mass of the grains and the forces among them . for values of strain smaller than @xmath0 ,
dry granular assembles show elastic response , i.e. , the reference state can be restored by setting the applied forces to zero ( see p 92 of @xcite ) .
this elastic response comes from the reversible deformation of the region of contact between the grains .
however , to describe the state of the hydrophobic sand immersed in water the information regarding the center of mass of the grains is not sufficient .
we need to augment it with additional information about the detailed layout of the three phase contact lines and the local contact angles of the _ skin_. the three phase contact lines formed at the grain - water - air interface are immobilized ( pinned ) by defects present in the system @xcite .
any additional deformation of the _ skin _ causes the contact angles to change from their reference values .
this generates restoring forces in the system , i.e. , if the contact angle changes from @xmath1 to @xmath2 the restoring force per unit length is @xmath3 , here @xmath4 is the surface energy at the water - air interface .
the pinning and depinning of these contact lines determine the mechanical properties of the material and its mode of failure .
hence , the present study is an example of the more general problem of studying statics and dynamics of systems with elastic interfaces in a random environment @xcite .
as hydrophobic granular material , we use polyhedral shaped ` magic sand ' grains sourced from education innovation inc .
they are made by coating polyhedral shaped sand grains with a hydrophobic material .
alternatively , these hydrophobic particles can be made in the laboratory by coating similar sand particles with fluoropel pfc from cytonix llc silica @xcite .
when grains of magic sand are freely poured into water , they spontaneously form a cohesive lump : a typical example of which is shown in fig.[fig : intro](a ) .
the immediately evident features are : ( i ) the system retains its shape and ( ii ) the outer surface of the lump has a luster that originates from the total internal reflections of light at the pinned water - air interface @xcite .
on degassing of the system achieved by creating a partial vacuum over the liquid ( fig.[fig : intro](b ) , the lump begins to lose its lustre ( fig.[fig : intro](c ) and when this luster - free system is mechanically perturbed , it slumps to form a flat sand - bed inside water ( fig.[fig : intro](d ) .
a cross - sectional top - view of the granular lump in fig.[fig : intro](a ) is schematically shown in fig.[fig : intro](e ) . the grains ( red coloured with dots ) on the boundary together with the pinned water - interface constitute an encapsulating granular _ skin _ which provides the submerged lump its structural integrity .
the force of buoyancy of the trapped air and the weight of dry sand grains in the interior ( colored with green ) exerts stress on the _ skin _ causing it to be under tension .
obtained for a water - drop ( @xmath5 ) placed on a single hydrophobic grain surface .
( b ) advancing and receding contact angles are measured to be @xmath6 and @xmath7 respectively , for the same water droplet in a tilted configuration .
( c ) existence of isolated air bubbles on the topographic grain surface viewed at a larger magnification . ]
the scanning electron micrographs of the grain surface are shown in fig.[fig : intro](f ) and ( g ) , for two different values of magnification .
the micron - sized rough patches ( see fig.[fig : intro](g ) which come from the hydrophobic coating on the particle surface are crucial for the observed wetting properties of these sand grains @xcite .
the equilibrium contact angle is measured to be about @xmath8 for a @xmath9 water drop placed on a single hydrophobic grain ( see fig.[fig : contact_angle](a ) . for the same droplet in a tilted configuration ,
the advancing and receding contact angles are found to be @xmath10 and @xmath11 respectively , i.e. , the contact angle hysteresis is @xmath12 ; see fig.[fig : contact_angle](b ) . the higher magnification image of water - grain interface in fig.[fig : contact_angle](c ) shows presence of small isolated air bubbles trapped between the grain and the water which have similar equilibrium contact angles @xmath13 , i.e. , the surrounding water is in partial contact with the hydrophobic sand grains which creates a wenzel type wetting scenario @xcite .
washing these grains in acetone removes the hydrophobic coating and exposes the underlying smooth hydrophilic surface ( fig.[fig : intro](h ) .
in order to obtain a more comprehensive understanding of the mechanical properties of the system , we choose a relatively simple geometry which is a free - standing submerged cylindrical column built with magic - sand , using the following experimental steps . at first , by selective sieving we obtain grains of different mean sizes @xmath14 , ranging from @xmath15 to @xmath16 .
second , an acrylic tube of inner diameter @xmath17 is made to rest vertically on the flat base of an empty rectangular glass container .
the grains are thereafter gently poured into the tube upto a desired height @xmath18 .
this forms a _ wall - supported _ granular column shown in the left panel of fig.[fig : fig_modulation](b ) .
the typical packing fraction @xmath19 of the grain assembly is about @xmath20 .
third , and in a key step , the acrylic tube is gradually pulled out while simultaneously filling the glass container with water . consequently , the hydrostatic pressure partially compensates for the over - pressure due to weight of the grains .
this unique protocol prevents grains from clogging particles which are in constant physical contact with the wall of the acrylic tube are immobilized by the frictional interactions .
as the tube is withdrawn , it drags along particles that are in contact with it this exposes the next layer of particles to water , on which the _ skin _ forms .
the resulting particle rearrangements reduce the diameter by @xmath21 and increase the height of the column which is now partially submerged in water by @xmath22 .
the radial contraction happens within the first few seconds and does not evolve appreciably over time .
however , the axial strain @xmath23 keeps increasing in time ; see fig.[fig : fig_modulation](c ) . as the tube is pulled out , the number of sand grains which are in contact with it decreases .
this reduces the extent of particle rearrangements . as a result ,
the rate of change of @xmath22 decreases with time .
the fully submerged column has a length @xmath24 which is about @xmath25 larger than the initial length @xmath18 of the wall - supported column .
hence , the volume fraction of the submerged column is about @xmath26 lower than its value for the wall - supported case .
this free - standing submerged column is shown in the middle panel of fig.[fig : fig_modulation](b ) .
the column is anchored via capillary forces with the hydrophobic substrate at the bottom and is slightly flared close to the anchoring region . under an identical protocol , hydrophilic beach sand instead slumps to form a canonical sandpile @xcite .
the submerged column can be thought of as a thin cylindrical shell ( made from the _ skin _ ) containing non cohesive dry grains in its interior .
the _ skin _ is about a particle diameter thick and it comprises of the sand grains on the outer surface of the column and the pinned three phase contact line . using a du - noy ring based technique @xcite we measure the interfacial energy of the _ skin _ to be about . at a microscopic scale
this interfacial energy arises from the pinning of the three phase contact line to the defects present on the surface of the grains ( see fig . [
fig : contact_angle](c ) @xcite .
the cohesive coupling stiffness @xmath27 due to this pinning depends on the size of the defect @xmath28 and the typical separation @xmath29 between them . in the limit of strong but sparsely distributed defects , @xmath30 where @xmath31 is the surface energy at the water - air interface @xcite .
experimentally we observe that polyhedral grains with smooth coverage of hydrophobic coating consistently fail to sustain submerged columns .
this suggests that of the multiple scales present in the system , the micron - sized hydrophobic patches present on the surface of the grains are most effective in pinning the contact line .
the process of making the column introduces localized geometric imperfections which evolve slowly in time .
this is captured by the contour plots of the variation @xmath32 of the column diameter from its mean value along its length over four orders of magnitude in waiting time @xmath33 ( fig.[fig : fig_modulation](d ) . in fig.[fig : fig_modulation](d ) the blue regions correspond to regions with positive , @xmath34 , modulation and the orange regions correspond to regions with negative , @xmath34 , modulation .
the maximum @xmath32 is found to be around @xmath35 , nearly half - a - particle size @xmath36 ; whereas the width of the modulation in @xmath17 along @xmath24 is much larger , approximately 40 particle - lengths .
the slow temporal evolution of the column s deformation is indicative of creep in the presence of strong inter - grain friction , brought about by the competition between the stresses exerted on the _ skin _ by the dry grains in the interior @xcite and the interfacial energy density of the _ skin_. henceforth , we refer to it as the wrinkling of the _ skin_. in this section we make an order of magnitude estimation of an effective elastic constant for the column .
to do so , we apply increasing body forces to the column by gradually tilting the vertical water - filled glass container at a fixed rate ; for experimental details see fig.[fig : fig_modulation](a ) .
fig.[fig : fig_curvature](a ) displays five representative images of the column for increasing values of the tilt angle @xmath37 made by the base of the container with respect to the horizontal .
images corresponding to @xmath38 and @xmath39 capture the process of the column breaking which resembles a mode - i type transverse rupture .
as the column breaks from a finite column height @xmath40 , its upper - part traces out a distinct arc , @xmath41 ( the inset of fig.[fig : fig_curvature](b ) . here
@xmath42 is the failure angle made by the axis of the falling part with a line whose slope is @xmath43 and @xmath44 is the length of the falling part of the column .
the column for @xmath45 is about @xmath46 longer than its length for @xmath47 .
the variation of failure angle @xmath42 with @xmath37 changes slope at @xmath48 in a semi - log plot ( see fig.[fig : fig_curvature](b ) .
the large slope beyond @xmath49 indicates a rapid toppling of the upper part of the column .
we use @xmath49 to mark the onset of mechanical failure . the broken part of the column sinks as a single object .
one can not build an arbitrarily long and/or thin column .
if the gravitational energy @xmath50 of the column exceeds its bending energy @xmath51 , the column will break , here @xmath52 is the effective young s modulus of the column , @xmath53 is its cross sectional area , @xmath54 is moment of inertia of the column and @xmath55 and @xmath56 are the densities of sand and water , respectively . to estimate the maximum height @xmath57 of a column
we equate the gravitational energy to the bending energy to obtain , @xmath58 at this maximum height the column breaks in its upright position , i.e. , @xmath59 ; this allows us to construct an alternative estimate of @xmath57 from the experimentally observed linear variation @xmath60 where @xmath61 and @xmath62 are fitting parameters ; see fig.[fig : fig_scaling_b](a ) . for @xmath63
, we obtain @xmath64 for @xmath65 , @xmath66 .
this estimate of @xmath67 is consistent with experimental observations that stable @xmath68 diameter columns could only be built to the height of @xmath69 .
columns built greater than this height are unstable and fail in their upright position . equating the above two expressions for @xmath70 we obtain @xmath71 for the submerged column .
this is about two orders of magnitude smaller than that observed for wet sand containing @xmath72 water by volume @xcite .
[ [ heuristic - arguments - for - calculating - h_c ] ] heuristic arguments for calculating @xmath40 : + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + the falling column is visually similar to the well studied problem of a `` falling chimney '' that breaks mid - air due to tilting @xcite . for the case of a falling chimney
the tensile stress at the leading - edge and the compressive stress at the trailing - edge at small tilt angles is maximum at @xmath73 ( for tilt angles greater than @xmath74 it is about @xmath75 ) , and is usually the point from where the chimney fails @xcite .
though , one expects a similar spatial variation of the stress for the submerged column , when @xmath76 , we find the following ( i ) the height @xmath77 at which the column breaks varies quadratically with the column length @xmath78 ; see fig.[fig : fig_scaling_b](b ) and ( ii ) @xmath40 decreases with increasing column diameter ; see the inset of fig.[fig : fig_scaling_b](b ) . from a dimensional argument , for values of @xmath24
smaller than @xmath67 , we propose that the height @xmath77 is related to the column length @xmath79 by the following relation @xmath80 where @xmath81 is the ratio of the gravitational and the bending energies @xcite . since @xmath82 , @xmath83 must vary as @xmath84 , i.e. , @xmath85 a good agreement of this expression with a large set of experimental data , obtained for different lengths and diameters of the submerged column and for various grain sizes , are shown in fig.[fig : fig_scaling_b](c ) . here , we have taken the effective young s modulus of the column @xmath86 to be a constant , independent of the grain scale , which implies an approximate validity of a continuum mechanical description of the system in this range of deformation .
+ the functional dependence of the critical buckling stress on the geometrical parameters for a cylindrical column is different from that of a cylindrical shell . for the column
the critical stress is proportional to @xmath87 @xcite whereas for a shell it is proportional to @xmath88 @xcite . here
@xmath89 is the thickness of the shell . in the present experiments @xmath90 and @xmath91 , where the shell thickness @xmath89 is of the order of particle diameter @xmath92 .
thus , there are two critical stress values ; one that corresponds to the buckling of the shell ( _ skin _ ) and the other that corresponds to the buckling of the column .
since @xmath93 , for increasing values of stress , the buckling of the shell which causes the _ skin _ to wrinkle , precedes the failure of the column ( assuming that the elastic constant of the entire column is greater than or equal to the elastic constant of the _ skin _ alone ) .
the _ skin _ is an integral part of the column and its wrinkling influences the buckling of the column in the following way .
the wrinkles on the _ skin _ generates geometric imperfections of the column .
these imperfections ( i ) lowers the critical stress at which the system fails @xcite and ( ii ) influence the location from where the system fails . to study the role played by the wrinkles ( geometric imperfections of the column ) in determining the height @xmath40 from which the column fails we track the curvature @xmath94 of the column along its length for the leading edge till the column breaks up into two pieces . to calculate @xmath94 ,
the boundaries of the leading and the trailing edges are passed through a low pass filter that suppresses features smaller than a single grain size @xcite .
these edges are detected using an edge detection algorithm based on @xcite .
fig.[fig : fig_curvature_b ] shows the curvature data for the advancing edge .
the contour plot of the curvature clearly shows the existence of bands of high curvature along the length of the column .
these bands correspond to the wrinkling of the _ skin _
( fig.[fig : fig_modulation](d ) .
as the column tilts the lowest wrinkle develops into the most prominent imperfection of the column and it acts as a seed from where the failure of the column is initiated . the critical height @xmath40 from where the column breaks varies quadratically with the length @xmath24 of the column . since the location of the lowest wrinkle coincides with @xmath77 , it is natural to expect the wavelength associated with the wrinkles to have the same @xmath24 dependence as @xmath40 .
however , within the framework of shell buckling , the wavelength of wrinkles is independent of the length of the column @xcite .
so , a model based solely on buckling produced by static loading is inadequate to describe the experimental scenario . we speculate that the observed length dependence of @xmath40 is related to the perturbations imparted to the column by the mechanical noise associated with the process of tilting
this gives the loading a certain dynamic character which could in principle facilitate coupling of the various global length - dependent buckling modes to the wrinkling induced localized imperfections of the column @xcite .
this conjecture needs to be examined in future in greater detail .
the inset image in fig.[fig : fig_scaling_c ] displays an instantaneous distance @xmath95 at the apex of the _ visually identified wedge _ between the falling top and the anchored bottom parts of the column .
this wedge opens up from the trailing ( left ) edge and progresses towards the advancing ( right ) edge of the column .
the variation of this distance @xmath95 with time @xmath96 is plotted in fig.[fig : fig_scaling_c ] , here the time @xmath97 is measured onwards from the first observable instance of the wedge opening at the left edge between the falling top and the anchored bottom .
as can be seen , the wedge advances with an average velocity of @xmath98 . neglecting the viscous drag , the time required for a non - anchored column to topple
can be obtained by equating the toppling torque acting through the center of mass of the falling part of the column and the rate of change of the angular momentum .
this ` shortest ' falling time is about a second which is the same for the wedge to move across the sample .
accounting for the viscous drag will increase the estimate of the falling time . unlike conventional solids which break by developing cavities in the bulk
, the interior of the submerged column is made of dry grains that can not support an open gap .
the gap can be sustained only if water drains in and generates a confining _ skin _ which prevents individual particles from falling into it .
hence , the speed at which the wedge opens sets a lower limit on the tearing speed of the _ skin_. the inset of fig.[fig : fig_scaling_c ] shows that the particle size influences the speed @xmath99 at which the apex of the visually identified wedge between the falling top and the anchored bottom moves suggesting that the wedge opening is influenced by tearing of the _ skin_. figure [ fig : fig_curvature ] ( a ) shows that tilting of the column alters the shape of the column . to study the elastic and plastic response of the column
, we utilize an incremental stress - cycling protocol ; which is the mechanical equivalent of minor hysteresis loop in magnets @xcite . here , the column is recursively stressed such that the minor loops have three branches : ( i ) an increase@xmath100 branch where the tilt angle increases from 0 to @xmath101 at a fixed rate ( ii ) a clamp@xmath100 branch where the tilt angle is held constant at @xmath101 for a fixed period of clamping time and ( iii ) a decrease@xmath100 branch which corresponds to decreasing the tilt angle from @xmath101 to 0 , with the same rate as in ( i ) .
they are accessed sequentially . for each successive cycle , the maximum tilt angle @xmath102 increases linearly with @xmath103 , where @xmath104 is the numerical index for the cycle .
an instantaneous configuration of the submerged column is referred to as @xmath105 , where @xmath106 is the time spent by the system in the clamped state .
fig.[fig : fig_hist_loops]a shows the cyclic variation in the relative displacement @xmath107 as a function of the tilt angle @xmath108 for selective values of @xmath109 1 , 7 , 11 and 13 respectively ; here @xmath110 and @xmath111 is the distance of the mid - point of the top of the column from the point @xmath112 on the @xmath113axis ( the inset of fig.[fig : fig_curvature](b ) . for a first few cycles ( @xmath114 ) , i.e. , for smaller values of @xmath115 , the _ skin _ deforms in a reversible manner .
we limit our statement of the reversibility : it is entirely possible that while the _ skin _ may behave in a reversible manner , the dry grains in the interior of the column may not . for larger values of @xmath115 ( @xmath116 ) , the column shows noticeable irreversibility and hence , the hysteresis increases .
the parameter @xmath117 is a measure of the strain developed in each cycle while the parameter @xmath118 determines the loop area of each stress cycle ; which is a measure of the accumulated hysteresis and dissipation in the system , here @xmath119 .
the top and bottom panels of fig.[fig : fig_hist_loops]b show variation of these two parameters as a function of the maximum tilt angle @xmath120 , respectively .
the following observations can be made from fig.[fig : fig_hist_loops ] ( a ) and ( b ) : ( i ) the elastic regime of the column extends for @xmath121 beyond which the area of the hysteresis loop @xmath118 begins to increase abruptly .
this elastic regime ( non - shaded region of fig .
[ fig : fig_hist_loops](b ) can accommodate strains of the order of @xmath122 which is much larger than that observed for dry granular systems .
( ii ) in the elastic regime , the recursive stress cycling shows smooth variation in @xmath123 with @xmath37 .
however , beyond the elastic regime ( @xmath116 or @xmath124 , the shaded region of fig . [
fig : fig_hist_loops](b ) this variation becomes increasing jagged and clearly shows discrete ` jumps ' interspersed with smoothly varying sections .
these jumps are signatures of mechanical instabilities associated with contact line slippages .
they are related to the stress ( tilt ) induced reduction of the energy barriers associated with the underlying pinning potential which allows the ambient noise ( fluctuations ) to induce creep like motion of the contact line .
anomalies of elastic constants associated with the break down of linear elastic response generates similar jumps in the flow curves of disordered materials @xcite .
( iii ) the maximum size of the jump height seen in the quantity @xmath123 is about @xmath125 , i.e. , the contact line on an average moves by a particle scale .
further motion of the contact line is possible only by jumping to the next grain which is restricted by the presence of sharp edges at the grain corners @xcite .
( iv ) the increase in @xmath118 is caused by the cumulative effects of these jumps .
these jumps progressively generate local overhangs along the direction of the tilt .
this results in the column breaking forward unlike a rigid `` falling chimney '' which breaks backward @xcite .
the effects of stress assisted creep are best seen in the deformation of the column for the various clamp @xmath126 branches ( here @xmath37 is held constant at @xmath120 , see fig.[fig : fig_hist_loops](c ) .
the creep increases with increasing tilt angles ( forces ) , e.g. , see the @xmath127 branch in fig.[fig : fig_hist_loops](c ) .
even during creep the column evolves in a similarly punctuated stick - slip manner with jumps in the quantity @xmath128 being limited by the value @xmath129 .
as the occurrences of creep increase , the survival@xmath130time @xmath131 of a stationary stick phase decreases ; see the inset of fig.[fig : fig_hist_loops](c ) .
the jumps themselves are abrupt in the experimental time scale : they occur over a period less than , an order of magnitude shorter than the measured shortest survival time ( ) .
the breaking of the column at the macroscopic scale ( see fig.[fig : fig_scaling_c ] ) is visible only for values of @xmath132 greater than .
this region shows a rapid growth of @xmath123 , its initiation is marked by an arrow in fig.[fig : fig_hist_loops](c ) .
these results imply that the _ skin _ , comprising of the air - water - grain interface on the surface layer of grains , is primarily responsible for the mechanical response of the entire system and any modification of the _ skin _ will modify the behaviour of the system .
one striking modification is shown in fig.[fig : fig_vcaccum_column ] where we create a partial vacuum over the water surface , so that the trapped air within the column , one key component of the _ skin _ , escapes out from the system as bubbles .
the time - lapsed images in fig.[fig : fig_vcaccum_column ] ( a ) @xmath133 ( e ) demonstrate that here too the failure is initiated by bending at a finite critical height .
but , in contrast to the single tear shown above , the column shows large - scale plastic deformations .
its structural integrity is extended for values of @xmath42 greater than @xmath134 by forming a deformed neck , i.e. , a narrowing of the column s diameter , instead of forming a tear or a crack as above . here
, the failure proceeds via this necking instability whose spatial location is marked by an arrow in fig.[fig : fig_vcaccum_column](e ) . such large - scale ( many - grains wide , in this case ) plastic flow and necking instabilities are typical of ductile mode of failure in solids @xcite , in complete contrast with the brittle failure described above .
this change in failure - mode is brought about by the loss of entrapped air causing weaker cohesion in the column that results in conformational changes of a softer _ skin _ and finally leads to the ductile failure of the column . for completeness
, we note that a third , and limiting , mode of failure occurs if degassing continues for a long time . in the absence of air ,
the _ skin _ begins to lose its strength ( see the lustre - less lump in fig.[fig : intro](c ) .
the weak _ skin _ eventually fails to sustain the inward hydrostatic pressure and slumps under the slightest mechanical disturbance .
the water quickly drains in as the structure disintegrates into a heap of particles , seen in both fig.[fig : intro](d ) and fig.[fig : fig_vcaccum_column](f ) ; the hydrophobic sand then behaves like its hydrophilic counterpart ( right panel of fig.[fig : fig_modulation](b ) .
the removal of trapped air from the column demonstrates the existence of an underlying dynamical transition via which the failure - mode transforms from being brittle to being ductile , both of which are collective ( multi - particles ) in nature , and finally to a total disintegration of the _ skin _ that represents failure at the single - particle level .
this is analogous to dynamical transitions from collective response to single - particle response in a wide class of systems @xcite .
this paper provides a detailed study of mechanical properties of underwater granular structures made of hydrophobic sand , where the self - generated cohesive _ skin _ on the boundary of the structure encapsulates the dry grains inside it from the surrounding medium water . in our experimental results ,
three distinct length scales of the system are found .
( i ) the scale of the hydrophobic patches @xmath135 influences the strength of the pinning of air - water interface on the grain surfaces .
depinning of this interface produces macroscopically observable plastic deformations of the contact lines around ( ii ) the grain - scale @xmath136 of the sand particles .
( iii ) at the system scale @xmath137 , the compressive traction forces due to the column s own weight drives the wrinkling of the _ skin_. the regions of large curvature of the structure are the seeds from where the system - sized failure modes nucleate . by partially removing the trapped air from this structure , we also see that the collective failure can be tuned from brittle to ductile .
a more complete removal of air causes the _ skin _ to crumble completely into individual non - cohesive grains inside water .
these experimental findings imply the existence of a tunable dynamical transition between a collective and an individual ( single - grain ) mode of failure in this system .
we expect that these new results will help in engineering the granular encapsulation with desired material - properties in a variety of applications and , at the same time , it provides a deeper insight of the multi - scale mechanics , generic to granular materials . 49 ifxundefined [ 1 ] ifx#1 ifnum [ 1 ] # 1firstoftwo secondoftwo ifx [ 1 ] # 1firstoftwo secondoftwo `` `` # 1 '' '' @noop [ 0]secondoftwo sanitize@url [ 0 ]
+ 12$12 & 12#1212_12%12 @startlink[1 ] @endlink[0 ] @bib@innerbibempty @noop * * , ( ) link:\doibase 10.1103/physrevlett.99.188001 [ * * , ( ) ] link:\doibase 10.1146/annurev.fluid.40.111406.102142 [ * * , ( ) ] link:\doibase 10.1038/nphys834 [ * * , ( ) ] @noop * * , ( ) link:\doibase 10.1039/c2sm25883h [ * * , ( ) ] link:\doibase 10.1080/00018730600626065 [ * * , ( ) ] link:\doibase 10.1002/nag.476 [ * * , ( ) ] link:\doibase 10.1080/00018730500167855 [ * * , ( ) ] http://link.aps.org/doi/10.1103/physrevlett.114.018001 [ * * , ( ) ] @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) link:\doibase 10.1021/la0103822 [ * * , ( ) ] link:\doibase 10.1126/science.1074868 [ * * , ( ) ] @noop * * , ( ) in @noop _ _ , vol .
( , ) pp . @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) in \doibase http://dx.doi.org/10.1016/s0167-3785(07)80042-x[__ ] , , vol . ,
( , ) pp
. @noop _ _ ( , ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop _ _ , vol .
( , ) http://link.aps.org/doi/10.1103/physreve.86.031402 [ * * , ( ) ] link:\doibase 10.1364/josa.45.000572 [ * * , ( ) ] link:\doibase 10.1364/josa.69.001205 [ * * , ( ) ] @noop * * , ( ) @noop * * , ( ) link:\doibase 10.1103/revmodphys.57.827 [ * * , ( ) ] @noop * * , ( ) @noop * * , ( )
@noop * * , ( ) @noop * * , ( ) @noop _ _ ( , ) @noop * * , ( ) link:\doibase 10.1063/1.4812809 [ * * , ( ) ] @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) in @noop _ _ , vol .
( , ) pp .
@noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) link:\doibase 10.1103/physrevlett.105.154301 [ * * , ( ) ] | magic sand , a hydrophobic toy granular material , is widely used in popular science instructions because of its non - intuitive mechanical properties .
a detailed study of the failure of an underwater column of magic sand shows that these properties can be traced to a single phenomenon : the system self - generates a cohesive _ skin _ that encapsulates the material inside . the _ skin _ , consists of pinned air - water - grain interfaces , shows multi - scale mechanical properties : they range from contact - line dynamics in the intra - grain roughness scale , plastic flow at the grain scale , all the way to the sample - scale mechanical responses . with decreasing rigidity of the _ skin _ , the failure mode transforms from brittle to ductile ( both of which are collective in nature ) to a complete disintegration at the single grain scale . |
-5 mm recently , a host of important results on graph colouring have been obtained via the probabilistic method .
the first author presented an invited lecture at the 2002 international congress of mathematicians surveying a number of these results .
the recent monograph @xcite provides a more in depth survey of the topic .
this paper presents one example of a result proven using the method .
an instance of list colouring consists of a graph @xmath0 and a list @xmath5 of colours for each vertex @xmath6 of @xmath0 .
we are asked to determine if there is an _ acceptable _ colouring of @xmath0 , that is a colouring in which each vertex receives a colour from its list , and no edge has both its endpoints coloured with the same colour .
the _ list - chromatic number _ of @xmath0 , denoted @xmath7 is the minimum integer @xmath8 such that for every assignment of a list @xmath5 of size at least @xmath8 to every vertex @xmath6 of @xmath0 , there exist an acceptable colouring of @xmath0 .
the list - chromatic number was introduced by vizing @xcite , and independently erds et al .
this parameter has received a considerable amount of attention in recent years ( see , e.g. @xcite , @xcite ) . clearly , by definition , @xmath9 because @xmath10 precisely if an acceptable colouring exists when each @xmath11 is @xmath12 . however , the converse inequality is not true , e.g. @xmath13 as can be easily verified by considering figure [ fst ] .
in fact , there are bipartite graphs with arbitrarily high chromatic number ( indeed even for bipartite @xmath0 , @xmath7 is bounded from below by a function of the minimum degree which goes to infinity , see @xcite ) .
this shows that the gap between @xmath14 and @xmath15 can be arbitrarily large .
moreover it shows that @xmath15 can not be bound by any function of the chromatic number of @xmath0 .
this gives rise to the following intriguing question in the theory of graph colourings : find conditions which guarantee the equality of the chromatic and list - chromatic numbers .
there are many conjectures hypothesizing conditions on @xmath0 which imply that @xmath16 . probably , the most famous of these is the list colouring conjecture ( see @xcite ) which states that this is true if @xmath0 is a line graph .
one interesting example of a graph with @xmath17 was obtained in the original paper of erds et al .
they proved that if @xmath0 is complete @xmath8-partite graph with each part of size two then @xmath18 .
it took nearly twenty years until ohba @xcite noticed that this example is actually part of much larger phenomenon .
he conjectured ( cf .
@xcite ) that @xmath19 provided @xmath20 .
this conjecture if it is correct is best possible .
indeed , let @xmath0 be a complete @xmath8-partite graph with @xmath21 parts of size @xmath22 and one part of size @xmath23 .
then the number of vertices of @xmath0 is @xmath24 , the chromatic number is @xmath8 and it was proved in @xcite that list - chromatic number of @xmath0 is at least @xmath25 . in his original paper ohba obtained that @xmath26 for all graphs @xmath0 with @xmath27 .
his conjecture was settled for some other special cases in @xcite . recently
the result of ohba was substantially improved by the authors of this paper . in @xcite
they proved that ohba s conjecture is true for all graphs @xmath0 with at most @xmath28 vertices . in this paper
we want to improve this result for large graphs and prove that the conjecture is asymptotically correct .
more precisely we obtain the following theorem .
[ maint ] for any @xmath2 , there exist an @xmath3 such that @xmath26 provided @xmath29 . the rest of this paper is organized as follows . in the next section
we describe the main steps in the proof of theorem [ maint ] .
more precisely , we present our key lemma and show how to deduce from it the assertion of the theorem
. we will prove this lemma using probabilistic arguments . in sections 3 and 4
we discuss the main ideas we are going to use in the proof .
we present the details of the proof in section 5 .
finally , the last section of the paper contains some concluding remarks .
-5 mm in this section we present the main steps in the proof of theorem [ maint ] .
first we need the following lemma from @xcite whose short proof we include here for the sake of completeness . [ lessn ] for any integer @xmath30 , if @xmath31 then there exist a set of lists @xmath32 for which there is no acceptable colouring such that each list has at least @xmath30 elements and the set @xmath33 has size less than @xmath34 .
* proof . *
assume @xmath35 and choose a set of lists @xmath32 for which there is no acceptable colouring , in which each list has size at least @xmath30 and which minimizes @xmath36 .
now , if @xmath37 then we are done .
so , we can assume the contrary .
we consider the bipartite graph @xmath38 with bipartition @xmath39 and an edge between @xmath40 and @xmath6 precisely if @xmath41 . if there is a matching of size @xmath42 in @xmath38 then this matching saturates @xmath43 and points out an acceptable colouring for the list colouring instance in which no colour is used more than once . since , there is no such acceptable colouring , no such matching exists .
thus there must be a smallest subset @xmath44 of @xmath45 which is not the set of endpoints of a matching in this graph and this set must have at most @xmath42 elements .
clearly , @xmath44 contains at least two vertices .
now , by the minimality of @xmath44 there is a matching @xmath46 in @xmath38 of size @xmath47 whose endpoints in @xmath45 are in @xmath44 .
further , classical results in matching theory ( see e.g. theorem 1.1.3 of @xcite ) tell us that if @xmath48 is the set of endpoints of @xmath46 in @xmath43 then for @xmath49 , we have @xmath50 .
let @xmath51 be any vertex in @xmath52 and replace @xmath5 by @xmath53 for every vertex @xmath54 .
this yields a new list colouring problem in which the total number of colours in all lists is smaller than @xmath36 ( since all the new list are disjoint from b ) .
therefore by the minimality of our original choice , there exist an acceptable colouring of @xmath0 for this new lists colouring instance .
in particular this implies that we can obtain an acceptable colouring of @xmath52 for the original lists @xmath5 . since no colour in @xmath44
is used in this colouring , using the colouring of @xmath48 pointed out by @xmath46 yields an extension of this colouring to a colouring of @xmath0 in which no colour of @xmath44 appears more than once .
this contradicts our assumption that there is no acceptable colouring for this instance and proves the lemma .
* proof of theorem [ maint ] .
* let @xmath2 be a fixed constant and let @xmath0 be a graph satisfying @xmath55 .
we assume that @xmath56 and obtain a contradiction .
since adding an edge between vertices in different colour classes in an optimal colouring of @xmath0 does not change @xmath57 and can only increase @xmath7 , we will assume that @xmath0 is complete @xmath57-partite graph .
thus @xmath0 has a unique partition into @xmath57 stable sets .
we refer to these stable sets as parts rather than colour classes so as to avoid confusion with the colours used in our acceptable colouring of @xmath0 . now by lemma [ lessn ] , if @xmath58 then there is an instance of list colouring on @xmath0 for which no acceptable colouring exists , in which each list has length at least @xmath57 and such that the size of the union of all lists @xmath5 is less than @xmath34 .
this means that in an acceptable colouring at least one colour must be used on more than one vertex .
fortunately , it also implies that for every non - singleton part @xmath59 there is at least one colour which appears on @xmath5 for more than half the vertices of @xmath59 ( since each @xmath5 contains more than half the colours ) .
our proof approach is simple . for each non - singleton part @xmath59
, we choose some colour @xmath60 and colour with @xmath60 all the vertices of @xmath59 whose list contains @xmath60 ( thus we must insist that all the @xmath60 are distinct ) .
we complete the colouring by finding a bijection between the vertices not yet coloured and the colours not yet used so that each such colour is in the list of the vertex with which it is matched .
this yields an acceptable colouring in which for each part @xmath59 there is at most one colour @xmath60 used on more than one vertex of @xmath59 . to begin
, we consider the case when there is some part @xmath59 such that some colour appears on all the vertices of @xmath59 .
we show that we can reduce to a smaller problem by using any such colour for @xmath60 .
iteratively repeating this process yields a graph where no such @xmath59 exists and hence , in particular , there are no parts of size two . our choices for the remaining @xmath60 are discussed in the proof of the key lemma [ bruce ] which consists of the analysis of a probabilistic procedure for choosing the remaining @xmath60 .
unfortunately , before discussing this procedure we need to deal with some technical details .
so , to begin we show that we can assume that @xmath61 is empty for all parts @xmath59 of size bigger than @xmath62 in the partition of @xmath0 . to see this , let @xmath59 be a part of size at least @xmath22 such that @xmath63 .
then the graph @xmath64 has chromatic number @xmath65 and at most @xmath66 vertices and therefore also satisfies @xmath67 note that it also satisfies @xmath68 since otherwise we can obtain an acceptable colouring of @xmath0 from the lists @xmath5 .
indeed , let @xmath40 be a colour in @xmath69 . since @xmath70 , we know there is an acceptable colouring of @xmath64 from the lists @xmath71 . colouring all vertices in @xmath59 with @xmath40 we obtain an extension of this colouring to an acceptable colouring of @xmath0 from the original lists , a contradiction .
therefore we will consider the graph @xmath64 instead of @xmath0 and continue this process until we obtain a graph @xmath72 and an instance of list colouring on @xmath72 with the following properties .
* @xmath72 is @xmath73-partite graph which satisfies @xmath74 .
* each list @xmath75 has length at least @xmath73 and there is no acceptable colouring of @xmath72 from @xmath75 . * the size of the union of all lists is less than @xmath76 .
* @xmath77 is empty for all parts of size bigger than @xmath62 in the partition of @xmath72 .
since the size of the lists is @xmath78 we obtain that @xmath79 is non empty for any two vertices @xmath80 in @xmath72 . in particular
this implies that in the partition of @xmath72 there are no parts of size two .
note that the original graph @xmath0 has at most @xmath81 parts of size @xmath82 and each time we removed such a part the chromatic number of the remaining graph decreased by one .
therefore we decrease chromatic number of @xmath0 by at most @xmath81 and hence the remaining graph @xmath72 should have at least @xmath83 vertices .
so by choosing an appropriate bound on the size of @xmath34 we can make @xmath76 arbitrarily large .
this completes our discussion of parts @xmath59 for which some colour is in @xmath5 for all vertices @xmath6 of @xmath59 .
we turn now to the technical details necessary before we present the rest of the ideas needed in the proof .
let @xmath84 be the set of all the vertices in the singleton classes in the partition of @xmath72 .
pick @xmath85 to be a sufficiently large integer constant @xmath86 and let @xmath30 be an integer which satisfies @xmath87 since in the partition of @xmath72 there are no parts of size two , we obtain that @xmath88 and that @xmath89 .
set @xmath90 .
let @xmath38 be a bipartite graph with bipartition @xmath91 and an edge between @xmath40 and @xmath6 precisely when @xmath92 .
note that the degree of every vertex from @xmath84 in @xmath38 is at least @xmath93 .
therefore by well known results on zarankiewicz s problem ( see , e.g. , @xcite , problem 10.37 ) , @xmath38 contains a complete bipartite graph with @xmath30 vertices in @xmath84 and @xmath85 vertices in @xmath94 .
denote the set of vertices from @xmath84 and @xmath94 by @xmath95 and @xmath96 respectively and remove them from @xmath38 .
note that the bound on zarankiewicz s problem guarantees that we will continue to find a copy of the complete bipartite graph @xmath97 in @xmath38 until the minimal degree of a vertex in @xmath84 is @xmath98 thus in the end we obtain at least @xmath99 disjoint sets of colours @xmath100 and also @xmath8 disjoint sets of singleton partition classes @xmath101 , such that @xmath102 for every vertex @xmath103 .
denote by @xmath104 , by @xmath105 and let @xmath106 and @xmath107 be the sizes of @xmath108 and @xmath109 respectively . now using ( [ a ] )
we can obtain the following inequalities @xmath110 and @xmath111 in the above discussion and in particular in the last two inequalities we used that @xmath85 and @xmath30 are constants but @xmath76 ( and thus also @xmath73 ) tends to infinity . let @xmath48 be the union of some set of @xmath112 singleton partition classes which do not belong to @xmath113 .
such a set @xmath48 exists , since the number of singleton partition classes outside @xmath113 is at least @xmath114 .
note that we can obtain an acceptable colouring of @xmath48 with the lists @xmath115 greedily , since the size of @xmath115 is equal to @xmath116 .
let @xmath117 be the set of @xmath116 colours used to colour @xmath48 in one such acceptable colouring .
denote by @xmath118 and let @xmath119 for every vertex @xmath120 .
then to finish the proof it is enough to show the existence of an acceptable colouring of the @xmath121 from the set of lists @xmath122 . by definition ,
we have that @xmath123 and @xmath121 is a complete @xmath106 partite graph .
the number of vertices of @xmath121 satisfies @xmath124 so by choosing @xmath125 we obtain that @xmath126 .
this together with above discussion implies that @xmath121 satisfies all the condition ( @xmath62@xmath127 ) of the next lemma .
this lemma guarantees the existence of an acceptable colouring of @xmath121 from the set of lists @xmath122 and completes the proof of theorem [ maint ] .
[ bruce ] let @xmath128 be a constant and let @xmath129 and @xmath130 be integers with @xmath131 , @xmath132 , @xmath133 and @xmath134 .
suppose , in addition , that @xmath85 is fixed and @xmath130 ( and hence @xmath106 ) is a sufficiently large function of @xmath85 .
let @xmath0 be a complete @xmath106-partite graph on @xmath130 vertices and let @xmath5 be the set of lists of colours of size @xmath106 one for each vertex @xmath6 of @xmath0 such that the following holds .
1 . @xmath135 for any part @xmath59 of size bigger than one in the partition of @xmath0 .
@xmath0 contains a set of vertices @xmath113 of size @xmath107 such that the vertices in @xmath113 form parts of size one in the partition of @xmath0 .
the set @xmath113 is partitioned into @xmath8 parts @xmath136 each of size @xmath30 .
3 . @xmath0 contains no parts of size two and at most @xmath137 singleton parts which do not belong to @xmath113 .
there exist a set of colours @xmath138 of size @xmath106 and its partition @xmath139 into @xmath8 sets of size @xmath85 .
such that @xmath140 for every vertex @xmath141 .
in particular , for any subset of @xmath142 of size @xmath30 there exist an acceptable colouring of the vertices of @xmath143 which uses the colours in this subset .
the total number of colours in the union of all the lists @xmath5 is less than @xmath130 .
then there exist an acceptable colouring of @xmath0 from the lists @xmath5 .
we finish this section with discussion of the proof of lemma [ bruce ] .
we postpone all the details to the subsequent sections of the paper . * proof overview . *
the proof proceeds as follows : 1 .
we choose a random partition of each @xmath142 into two subsets @xmath144 of size @xmath30 and @xmath145 of size @xmath146 where these choices are made uniformly and independently .
we use the colours in @xmath144 to colour the vertices of @xmath147 which is possible by condition 4 of the lemma .
3 . we choose a ( random ) bijection between @xmath148 and the parts of @xmath0 not in @xmath113 in such a way that , for each part @xmath59 not in @xmath113 , @xmath59 is equally likely to correspond to each colour @xmath149 .
we denote by @xmath60 be the colour corresponding to @xmath59 .
4 . for each part
@xmath59 not in @xmath150 we colour every vertex @xmath6 of @xmath59 for which @xmath151 with the colour @xmath60 .
we match the set @xmath152 of vertices not yet assigned a colour with the set of colours not yet used ( i.e those colours not in @xmath108 ) so that every vertex is matched with a colour on its list .
we colour each vertex of @xmath152 with the colour with which it is matched .
if we successfully complete this five step process , we have an acceptable colouring as every colour not in @xmath153 appears on at most one vertex , and every colour in @xmath153 appears only on a subset of some part , and hence on an independent set of @xmath0 . to prove that we can find the colouring in this fashion ,
we need to describe and analyze our method for choosing the random bijection between the parts and the colours in @xmath108 made in steps i iv , in order to show that ( with positive probability ) we can complete the colouring by finding the desired matching in step v. a key tool will be hall s theorem which states that in a bipartite graph with bipartition @xmath154 , we can find a matching @xmath46 such that every vertex of @xmath155 is the endpoint of an edge of @xmath46 provided there is no subset @xmath84 of @xmath155 such that setting @xmath156 we have @xmath157 .
we remark that although steps i iii are presented as though they are separate processes performed sequentially , in the more complicated case of our analysis we will need to interleave these processes by first choosing some of the @xmath145 , then choosing the parts with which these colours will be matched , and finally completing step i and then step iii . to determine if we can find the desired matching in step v , we will need to examine the sets @xmath158 for the vertices of @xmath152 .
let @xmath38 be a bipartite graph with bipartition @xmath159 and an edge between @xmath40 and @xmath6 precisely when @xmath160 .
for each vertex @xmath6 , we let the weight of @xmath6 , denoted @xmath161 , be @xmath162 .
for any set @xmath107 of vertices we use @xmath163 to denote the sum of the weights of the vertices in @xmath107 .
this definition of weight is motivated by the following immediate consequence of hall s theorem : [ wobs ] if we can not find the desired matching in step 5 then there exist a subset @xmath84 of @xmath152 such that @xmath164 . in this case @xmath165 as well
. * proof .
* by hall s theorem there exist a subset @xmath84 of @xmath152 such that @xmath157 .
then , we obtain that @xmath166 will be crucial to the proof of the lemma . in the next section , by computing the expected value of the parameter , we show that the lemma holds if @xmath167 . in later sections
, we complete the proof using a more complicated analysis along the same lines .
-5 mm for each part @xmath59 which is not in @xmath150 , our choices in steps i and iii guarantee that each colour of @xmath108 is equally likely to be @xmath60 .
thus , for each vertex @xmath6 in such a part , the probability that @xmath6 is in @xmath152 , i.e. , @xmath169 , is @xmath170 . since @xmath171 ,
this is @xmath172 .
so , we have : @xmath173 so , if @xmath174 , then this expected value is less than or equal to one .
since the probability that a random variable exceeds its expected value is less than one , this implies that we can make the choices in steps i
iv so that @xmath175 , and hence by observation [ wobs ] , the desired matching can be found in step v. analyzing the behaviour of the ( random ) weights of various subsets of @xmath152 will allow us extend our proof technique to handle larger values of @xmath130 .
in doing so , the following definitions and observations will prove useful .
we let @xmath155 be the number of non - singleton parts . then by condition 3 of the lemma
, @xmath155 is at least @xmath176 .
since each non singleton colour class has at least three vertices and the total number of classes is @xmath106 , we obtain @xmath177 , i.e. , @xmath178 .
on the other hand , the analysis above shows that we can assume that @xmath179 and hence that @xmath180 .
thus , @xmath181 .
both these bounds on @xmath155 will be useful in our analysis .
note also that @xmath182
-5 mm let @xmath38 be a bipartite graph with bipartition @xmath183 and an edge between @xmath40 and @xmath6 precisely when @xmath160
. our first step will be to check hall s criterion for a fixed subset of colours @xmath184 in @xmath185 and show that the expected number of vertices in @xmath186 is less than @xmath187 . to begin
, we note that for any such @xmath6 , @xmath188 .
therefore , defining the set @xmath189 to be @xmath190 , we have that @xmath191 which , since @xmath192 , is at most @xmath193 . on the other hand , this estimate is not good enough to guarantee hall s criterion , since it still can be greater than @xmath187 . to improve on this bound , we use the fact that no colour @xmath40 appears on the list of all the vertices of any non - singleton part of @xmath0 . note that @xmath194 , @xmath195 , the number of non - singleton parts is @xmath155 and the total number of vertices is at most @xmath192 .
this altogether implies that for every @xmath40 , @xmath196 applying this fact for the @xmath40 in @xmath184 allows us to improve our bound on @xmath197 .
specifically , we note that summing this bound over all the colours @xmath40 in @xmath184 @xmath198 now , each vertex @xmath6 of @xmath189 contributes @xmath199 to exactly @xmath200 terms in the first sum in this equation , so its total contribution to the sum is 1 .
i.e. , we have : @xmath201 so , we do nt expect any particular set @xmath184 of colours to provide an obstruction to finding the desired matching in the bipartite graph @xmath38 in step v. however , we need to handle all the @xmath184 at once . in order to do so , we would like to prove that for each @xmath184 , the size of @xmath189 is highly concentrated around its expected value and hence is greater than @xmath187 only with exponentially small probability . as above , rather than focusing on all the @xmath184 we actually consider , for each colour c , the weight of the subset @xmath202 of @xmath152 consisting of those @xmath6 with @xmath40 on @xmath203 .
there are two major difficulties which complicate our approach .
* some of the parts @xmath59 can be very large making it impossible for us to obtain the desired concentration results directly ( e.g. , there could be a part of size exceeding @xmath204 ) . *
if @xmath203 is very small then @xmath199 is large and putting @xmath6 into @xmath152 can have a significant effect on the weight of the various @xmath202 .
this makes proving a concentration result directly impossible . in order to deal with these problems ,
we proceed as follows : 1 . we colour the `` big '' parts first , ignoring concentration in our computation and focusing only on the expected weight of the subset of @xmath152 intersecting the big parts .
we note that by considering the expected overall weight and not focusing on a specific @xmath202 , we only lose a factor of @xmath205 per part
. we will define big parts so that there are @xmath206 of them , and hence the total loss will not be significant .
2 . we treat @xmath6 with @xmath200 small separately using an expected value argument to bound the weight of the vertices in this set .
-5 mm in this section we will complete the proof of lemma [ bruce ] using the ideas which have already been discussed above .
we choose an integer @xmath207 so that @xmath208 which is possible because @xmath209 ( this holds , since @xmath85 and @xmath210 are fixed but @xmath106 tends to infinity ) and @xmath211 ( in fact it exceeds @xmath212 as we remarked at the end of section [ section 3 ] ) .
we call the largest @xmath213 parts in our partition of @xmath0 _ big _ , and the others _
small_. let @xmath214 be the union of the vertex sets of the big parts .
we will need the following lemma .
[ slem ] every small non - singleton partition class contains at least two @xmath6 which satisfy : @xmath215 * proof . * let @xmath59 be a small non - singleton colour class .
we already mentioned that every colour of @xmath138 is missed by a vertex of @xmath59 so @xmath216 .
now , since there are less than @xmath217 colours in total , every @xmath5 must contain at least @xmath218 colours in @xmath108 and so the largest @xmath158 in @xmath59 has at most @xmath219 elements .
thus , the sum of @xmath200 over the remaining vertices of @xmath59 is at least @xmath218 . since there are at least @xmath220 big colour classes , the largest small colour class has at most @xmath221 vertices .
so , the second largest @xmath203 has size at least @xmath222 .
this is the desired result . with this auxiliary result in hand
, we can now complete the proof .
we proceed as follows : 0.2 cm * first process : * we randomly choose @xmath207 of the @xmath223 and a partition of each of these into subsets @xmath144 of size @xmath30 and @xmath145 of size @xmath146 where these choices are all made independently and uniformly .
we then choose a uniformly random bijection between the @xmath213 colours in the union of these @xmath145 and the big parts .
0.2 cm * second process : * we chose a partition of each remaining @xmath142 into @xmath144 and @xmath145 where again these choices are uniform , independent , and independent of all the earlier choices .
we then choose a uniformly random bijection between the colours in these @xmath145 and the small parts .
0.2 cm denote by @xmath60 the colour which is assigned by the above bijection to the partition class @xmath59 .
use the colours in @xmath224 to colour the vertices of @xmath147 and for each part @xmath59 not in @xmath150 colour every vertex @xmath6 of @xmath59 for which @xmath225 with the colour @xmath60 .
let @xmath152 be a set of vertices not yet assigned a colour .
we set @xmath226 and @xmath227 .
note that @xmath228 is determined by our choices in the first process .
so , using a computation similar to that in ( [ expectation ] ) we obtain @xmath229 furthermore , by the definition of expectation , there exist at least one set of choices for the first process such that @xmath230 .
we condition on any such set of choices which ensures that this inequality holds .
we use @xmath231 and @xmath232 for the conditional probability of an event and conditional expectation of a variable for the second process , given this set of choices .
let @xmath233 be the union of the set of colours in the @xmath142 which were chosen in the first process .
at the end of section 3 we proved that @xmath146 is at least @xmath212 . therefore @xmath234 .
hence , we have that for every @xmath6 in a small part which is not in @xmath150 , @xmath235 clearly , this implies that for every subset @xmath84 of the set of vertices @xmath236 we have @xmath237 in particular , for every colour @xmath40 @xmath238 and also @xmath239 @xmath240 before we proceed with the proof , we need the following lemma . [ large - deviation ] for every color @xmath40 the probability that @xmath241 @xmath242 is @xmath243 .
* proof . * to prove the lemma we need the following variant of a standard large deviation inequality for martingales . since the proof of this inequality is essentially the same as other proofs which already appeared in the literature ( see , e.g. , section 3 of the survey @xcite )
, we will omit it here . given a finite set @xmath244 ,
let @xmath245 denotes the set of all @xmath246 permutations or linear orders on this set .
let @xmath247 be a family of independent random variables , where the random variable @xmath248 takes values in a finite set @xmath249 .
thus @xmath250 takes values in the set @xmath251 .
let @xmath252 be a random permutation independent from @xmath250 .
suppose that the non - negative real - valued function @xmath253 satisfies the following two conditions for every @xmath254 .
* for every @xmath255 , changing the value of a coordinate @xmath256 can change the value of @xmath257 by at most @xmath258 .
* swapping any two elements in permutation @xmath259 can change the value of @xmath257 by at most @xmath258 .
denote by @xmath260 the expected value of @xmath261 .
then for every @xmath262 we have that @xmath263 now fix a color @xmath40 and define the function @xmath261 to be @xmath264 where @xmath254 corresponds to the set of random choices for the second process .
more precisely , @xmath265 is a random partition of the set @xmath142 into subsets @xmath144 and @xmath145 and @xmath259 is a random bijection between the colors in these @xmath145 and the small parts of @xmath0 .
since we can fix one canonical ordering of these small parts we can assume that @xmath259 is just a random permutation of the set of colors which is , by definition , independent from the variables @xmath265 .
next , note that changing the outcome of the variable @xmath265 , i.e. , changing one particular @xmath145 can only affect vertices in at most @xmath146 small parts of @xmath0 . as we already mentioned in the proof of lemma [ slem
] , each small part contains at most @xmath266 vertices . since we considering only vertices @xmath6 satisfying @xmath267 , the weight of such a vertex is at most @xmath268 .
therefore , changing outcome of one @xmath265 can change the value of @xmath261 by at most @xmath269 . similarly swapping any two colors in @xmath259
can affect only vertices in two small parts of @xmath0 .
so again this can only change @xmath261 by at most @xmath270 . since the total number of random variables @xmath265 and also
the length of permutation @xmath259 are bounded by @xmath130 we have that in our case @xmath271 .
therefore it follows form the above large deviation inequality that @xmath272 this completes the proof of the lemma .
@xmath273 now , using the fact that the total number of colors is at most @xmath130 , we deduce from this lemma that with probability @xmath274 the following holds for every color @xmath40 @xmath275 @xmath276 in addition , we also want to satisfy the following inequality : @xmath277 @xmath278 since the probability that this last inequality fails is at most @xmath279 , there does indeed exist a set of random choices for the second process which satisfies simultaneously ( [ concentration ] ) and ( [ small - lists2 ] ) . fix any such set of choices .
then , combining the inequalities ( [ small - lists1 ] ) and ( [ small - lists2 ] ) together with the facts that @xmath280 and @xmath281 we obtain that @xmath282 @xmath283 note that , by lemma [ slem ] , every small non - singleton partition class contains at least two vertices @xmath6 such that @xmath284 .
since the number of small non - singleton partition classes is at least @xmath285 we obtain that @xmath286 @xmath287 here , in the last inequality we used that @xmath288 and @xmath289 .
note that a similar computation as in ( [ expectation ] ) shows that for any subset @xmath290 the expectation @xmath291 .
in particular , for @xmath292 we obtain @xmath293 combining this inequality with ( [ small - sets ] ) we have @xmath294 this completes our analysis of the weight of vertices with short lists .
we now consider the remaining vertices .
as we already mentioned , for every color @xmath40 and every non - singleton part of @xmath0 there is at least one vertex @xmath6 in this part such that @xmath295 .
since there are at least @xmath285 small non - singleton parts , a similar computations as in ( [ weight - color ] ) shows for every color @xmath40 that @xmath296 @xmath297 combining this inequality with ( [ v-color ] ) and ( [ concentration ] ) and using the fact that @xmath298 we will have that for every color @xmath40 @xmath299 @xmath300 recall that @xmath38 is a bipartite graph with bipartition @xmath159 and an edge between @xmath40 and @xmath6 precisely when @xmath160 .
let @xmath184 be any subset of colours in @xmath185 and denote by @xmath301 .
we complete the proof of the lemma by showing that the graph @xmath38 satisfies hall s condition , i.e. , @xmath302 for every set @xmath189 . then in step v we can match all uncoloured vertices in @xmath152 with the set of colours yet not used and and produce an acceptable coloring of @xmath0 .
first , note that any set @xmath184 of fewer than @xmath303 colours can not be an obstruction to the existence of the desired matching . indeed , if @xmath304 , then by observation [ wobs ] we have that @xmath305 . on the other hand , for every vertex
@xmath306 the size of @xmath203 is at most @xmath307 .
therefore we obtain a contradiction , since by ( [ small - sets1 ] ) @xmath308 turning to larger @xmath184 , we note next that the inequality ( [ small - sets1 ] ) yields : @xmath309 @xmath310 next , observe that the set of inequalities ( [ all - colors ] ) imply that for any set of colours @xmath184 @xmath311 this , together with the inequality ( [ small - sets2 ] ) yields that any set of colours @xmath184 of size at least @xmath312 satisfies @xmath313 thus we obtain that these larger @xmath184 also do not violate hall s condition and hence the desired matching of step v does indeed exist .
this completes the proof .
-5 mm in this paper we proved that for every @xmath314 and for every sufficiently large graph @xmath0 of order @xmath130 , the list chromatic number of @xmath0 equals its chromatic number , provided @xmath315 .
a more careful analysys of our methods yields that the value of @xmath316 in this result can be made as small as @xmath317 for any constant @xmath318 .
nevertheless the conjecture of ohba remains open for graphs with @xmath319 vertices and it seems one needs new ideas to tackle this problem . even to show that there is a constant @xmath320 such that @xmath321 for every graph @xmath0 with at most @xmath322 vertices , would be very interesting . in conclusion we would like to propose a related problem , which was motivated by ohba s conjecture .
let @xmath30 be an integer and let @xmath0 be a graph with at most @xmath323 vertices .
find the smallest constant @xmath324 such that for any such a graph @xmath0 its list chromatic is bounded by @xmath325 .
note that ohba s conjecture if true , implies that @xmath326 .
an additional intriguing question is to determine graphs with @xmath327 and for which the ratio @xmath328 is maximal .
here the case @xmath329 gives some indication that a complete multi - partite graph with all parts of size @xmath30 may have this property . | ohba has conjectured @xcite that if the graph @xmath0 has @xmath1 or fewer vertices then the list chromatic number and chromatic number of @xmath0 are equal . in this paper
we prove that this conjecture is asymptotically correct .
more precisely we obtain that for any @xmath2 , there exist an @xmath3 such that the list chromatic number of @xmath0 equals its chromatic number , provided @xmath4 4.5 mm * 2000 mathematics subject classification : * 05c15 , 05d40 . * keywords and phrases : * probabilistic method , graph coloring , list - chromatic number . |
one of the main features of many market and social processes is their bilateral structure and the need to match agents from one side of the market with the other side , e.g. students and schools in college admissions , employees and companies in the job market ( see * ? ? ?
* ) , men and women in online dating sites , advertisers and advertising slots in sponsored search ( see * ? ? ?
* ) . a class of `` two - sided matching model '' for studying such problems was first introduced by gale and shapley in 1962 in their seminal paper @xcite , in the context of college admissions and the marriage problem .
yet , the original model is quite general and can be easily adapted for other two - sided markets , such as the well - known national resident matching program ( nrmp ) for assigning medical students to residency positions in us . in gale and shapley s seminal paper ,
the college admissions problem is formulated as follows .
let there be a set of colleges @xmath2 and a set of students @xmath3 .
each college @xmath4 ( @xmath5 ) has a quota @xmath6 , that denotes the maximum number of students it can admit .
each student @xmath7 ( @xmath8 ) can apply to any number of colleges , represented by a _
strict _ preference ( ordering ) over the @xmath9 colleges .
each college also has a strict ranking for all students ( e.g. according to their test scores , interview performances or other criteria ) .
a _ matching _ is simply an assignment of students to colleges such that each college accepts no more students than its quota , and each student is admitted to at most one college . for convenience , we use @xmath10 to denote a dummy college that takes all unmatched students .
the college admissions problem can then be stated as the problem of designing an algorithm to arrive at a matching satisfying certain properties .
the _ mechanism _ that implements the solution can be considered as a blackbox that takes the students and colleges preferences ( denoted by @xmath11 ) as input , and outputs the matching , as depicted in fig .
[ mechanism_prev ] . in @xcite ,
two desirable properties are defined : _ stability _ and _ optimality_. stability of a matching means that it is not possible to find a student and a college that are not matched to each other , but both prefer each other more than their current match . in college admissions , optimality means student - optimality .
a stable matching is _ student - optimal _ if every student is at least as well - off as he / she is in any other stable matching .
college - optimality can be defined in a similar way , but it is never used .
gale and shapley proved that there always exists stable matchings .
their proof is by construction ( using the deferred acceptance ( da ) algorithm in fig . [ mechanism_prev ] ) . in the special case when the colleges preferences are all based on student examination results ,
the da algorithm is equivalent to a procedure that sorts all students according to merit and letting students choose the most preferred college ( if quota still allows ) in order , starting from the best student .
it is also shown in @xcite that the optimal stable matching is unique .
+ in practice , it is reportedly quite common that the boston mechanism ( bm ) ( see * ? ? ?
* ) is used instead .
the basic procedures of bm are as follows : in the first round each college @xmath12 only considers students who listed it as their first choice and accepts these students one by one according to @xmath12 s own preference until the quota of the college is filled up . in the second round , only the remaining student without any offers and colleges still having unfilled quota are considered .
each remaining college then considers students who listed it as the second choice and assigns offers to them one by one until the quota is exhausted .
the process goes on round by round until all quota are filled up or all students are already accommodated .
one major criticism concerning bm is its lack of stability
. judged by classic matching theory , matching results under bm are far from stable and thus many students are `` incentivized '' to make complicated strategies when submitting their preference lists .
in contrast , results under da are both stable and optimal for students .
students can feel free to reveal their true preference lists .
the transition from bm to da is therefore suggested in literature such as @xcite , which arguably would lead to significant efficiency gains for the whole community .
however , our investigation of some practical college admissions systems shows that _ certain hybrid mechanism may be more acceptable in society from the perspective of colleges enrollment concerns as well as students personal interest_. in this paper , we propose a generalized model for college admissions , which considers the tradeoff between students eligibility and interest by adjusting an additional parameter called _ reciprocating factor_. the larger the reciprocating factor is , the more would the interest factor counts when inspecting the applicants .
bm and da are merely two special cases of the generalized model when setting different reciprocating factors .
our model also extends the classic notion of stability and optimality by re - examining the formation of agents preferences .
the remainder of the paper is structured as follows . in section [ sec : ca_hk ] we elaborate the actual college admissions system used in hong kong which inspires our formal model . in section [ sec : gene_model ] we proposes the generalized model for stable matching . some of the important properties of generalized model are presented and discussed extensively in section [ sec : property ] .
section [ sec : simu ] compares the efficiency of generalized mechanism with the classic gale - shapley mechanism through simulation .
we give some further discussion and related work in section [ sec : discussion ] and conclude in section [ sec : conclusion ] .
the joint university programmes admissions system ( jupas ) is the central system for students to apply to the nine participating tertiary institutions in hong kong @xcite . in jupas
each student can apply to at most 25 programmes in order of preference .
these preferences are sub - divided and made known to the institutions in the form of five bands as shown in table 1 .
.correspondence between applicants band order and actual choice order [ cols="^,^",options="header " , ] the band order is made known to the programmes , however , the inner - band preferences are unrevealed . for example , in band a , the programme has no idea whether a student lists it as his / her first , second or third choice .
after aggregating the preference lists from all the applicants , each programme will make a `` _ _ merit order list _ _ '' for its applicants in accordance with its criterion for selection .
the rating criterion is determined independently by each programme : although many programmes would adopt a boston - like criterion which assigns band a students with highest priority , some programmes may also rate students only by their eligibility .
some unpopular programmes tend to employ the latter strategy if they find most excellent students have listed it as band b or band c choices rather than band a. finally , after all the _ _ merit order list__s and the applicants preference lists are sent to the jupas office , a central computer will automatically match the applicants with appropriate programmes .
the matching process applies the classic _ student - proposing da algorithm _ to give the students the best offer he / she can possibly obtain .
although the da algorithm is used in the last phase , the jupas mechanism as a whole is not equivalent to the classic _ gale - shapley student optimal mechanism _ in @xcite , where students would truthfully reveal their preference lists .
applicants in jupas face a similar problem like students in bm : their band a choices would receive higher priority than choices in other bands , although there are no discrimination over the multiple inner - band choices . actually in jupas the applicants are always advised to choose appropriate programmes according to their interests as well as their qualifications @xcite .
it is never a dominant strategy for students to always reveal their true preferences
. therefore jupas may be interpreted as a hybrid of bm regarding inter - band discrimination and da regarding the algorithm applied in the final phase .
the statement of `` gale - shapley student optimal stable mechanism is used in hong kong '' by @xcite is rather mis - informed and misleading : readers may falsely assume that applicants in hong kong could feel free to write down their true preferences , whereas in fact there is still room for students to manipulate their preference lists .
a natural question is why not replace jupas with da as suggested in most literature . to answer it , we formulate a model in the next section in order to justify such selection of the policy makers for sticking with the hybrid mechanism
we have mentioned in the previous section that programmes under jupas in hong kong have full right to determine how to rate students .
two factors are the most important : student eligibility and band order . in practice ,
most programmes put heavy weight on academic performance in determining students eligibility which makes the examination scores a very decisive factor in admission . to simplify the analysis , we assume each applicant would attend a standard examination and gain a total score which ranges from zero to the maximum mark . for band order , higher preference ( like band _ a _ )
indicates that the applicant is more interested in the programme while ranking certain programme as band _ d _ or _ e _ infers the lack of interest .
let @xmath13 denote the set of applicants , @xmath14 denote the set of programmes .
each student @xmath15 achieves a total score @xmath16 $ ] in the standard examination where @xmath17 is the full mark of the examination .
each programme @xmath18 has a quota @xmath19 .
when student @xmath20 applies to programme @xmath12 as his / her @xmath21-th choice , @xmath20 will obtain a bonus score @xmath22 which would promote his / her position in programme @xmath12 s _ merit order list_. generally , the bonus score should be a strictly decreasing function over _ preference order / ranking _ @xmath21 .
that is to say , the smaller @xmath21 is , the more bonus score it would bring . in practice , the programme director could make a corresponding table mapping each _ preference order _ to a certain bonus score for ease of reference .
all applicants are then sorted by their _ merit scores _ in each programme , where student @xmath20 s _ merit score _ in programme @xmath12 is computed according to the following equation : @xmath23\ ] ] the first term denotes the original score achieved by @xmath20 and the second term is the bonus score for students interest in @xmath12 . in case of tie when students share the same merit score , @xmath24 serves as the tie - breaker and student with higher @xmath24 obtains higher priority . finally ,
if all terms are equal , we break the tie by a random lottery .
we refer to @xmath25 as the _ reciprocating factor _ ( rf ) , a constant determined independently by each programme @xmath12 , reflecting its sensitivity towards applicant s _ preference order_. programmes with larger rf place more weight on applicants personal interests : other things being equal , students whose interests match with the programme are more favored . in the extreme case when @xmath26 of different programmes all equal to zero , the matching reduces to exactly the gale - shapley student optimal mechanism : preference order would not affect students positions in programmes ; in the contrary , when all @xmath26 are set to one , , there are actually an infinitely large amount of pairs @xmath25 and @xmath27 which can implement the boston mechanism , as long as it holds that for any @xmath12 and @xmath21 , @xmath28 and @xmath29 .
we thank an anonymous referee for pointing this out . ]
the matching works in the same way as the boston mechanism : the first choice gets the highest priority . for a general @xmath30 ,
say @xmath25 equals to @xmath31 , it means that programme @xmath12 would count @xmath32 of original score and @xmath33 of interest factor when evaluating the applicants .
after calculating the merit score for each applicant , each programme could generate a _ reciprocating preference _ list by comparing merits scores of the applicants .
these reciprocating preferences are then sent to the central college admissions system for further processing .
the complete procedures can be illustrated in fig .
[ mechanism_ext ] where students _ reciprocating preferences _ remain the same as their original preferences @xmath34 . + the reciprocating factor we propose in this paper gives programmes more flexibility in choosing a `` reasonable '' enrollment mechanism : * for programmes which hope to stick to the traditional boston - like scheme , there will be no need for any change since by default @xmath26 is set to one ; * for programmes whose sole objective is to raise the average score of newly admitted students , setting @xmath26 to zero would be their favorite strategy ; * for other elastic programmes concerning the students interest as well , a suitable @xmath26 between zero and one needs to be determined according to each programme s own admission policy in each admission year .
in last section we proposed the notion of _ reciprocating preference _ which can better reflect the selection criteria of individual programmes .
it it the student with higher merit score , rather than higher original exam score , that is more favored by each programme . with the change of the interpretation of agents preferences , important concepts such as stability and optimality
should also need to be re - defined _ from the perspective of reciprocating preferences_. we now give the formal definition as follows .
a matching is _ _ r - stable _ _ if it is not possible to find a student and a college that are not matched to each other , but both _ prefer _ each other more than their current match when judged by their _
reciprocating preferences_. a fundamental property of the generalized mechanism can be then presented as follows .
[ theo : r_stable ] matching outcome under the generalized mechanism is _ r - stable_. moreover , it is _ optimal _ among all possible r - stable outcomes .
the above theorem is easy to see by ignoring the preprocess phase in fig .
[ mechanism_ext ] and focusing on the matching phase where da always generates stable and optimal matching given any preference lists .
therefore from the perspective of reciprocating preferences , the matching outcome under jupas - like hybrid mechanism is still stable and optimal among all stable outcomes .
since bm is merely an extreme case of the generalized mechanism , we may easily get the following corollary through property [ theo : r_stable ] . the boston mechanism , known to be unstable by classic matching theory ( see * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , is actually r - stable with regard to colleges reciprocating preferences . in previous analysis , we assume all students would submit their true preferences . however , this assumption is unrealistic , especially in practical college admissions system like jupas where strategic behaviors is actually quite common .
we now show some key results for strategy analysis as follows .
[ theo : no_truthful_mech ] college admissions mechanism like jupas is not strategy - proof in general .
the exceptions are when all the programmes reciprocating factors are zero , it is the dominant strategy for students to reveal their true preferences .
we defer the detailed proof to appendix [ proof : no_truthful_mech ] .
when all programmes reciprocating factors are zero , the generalized model reduces to the classic gale - shapley student - optimal mechanism which is strategy - proof for students ( see * ? ? ?
* ; * ? ? ?
generally , when most programmes have positive reciprocating factors , students would act strategically : students with relatively lower examination scores may try to avoid some popular programmes where there would be lots of competitors with higher scores , and list those unpopular ones as their top choices instead in order to obtain more bonus scores there and increase the chances for admission .
it is worth pointing out that even though the generalized model is not strategy - proof , it does not imply that students would always have strong incentive to strategize . whether applicants strategies can work or not depends on how much information they know about other students behaviors .
for example , suppose there are five students @xmath35 and @xmath36 gets the lowest exam score ; there are two programmes @xmath37 and @xmath38 , both with reciprocating factors @xmath39 and quota @xmath40 .
if the preferences of students are strongly _ correlated _ , say all students consent that @xmath37 is better than @xmath38 , @xmath36 may try to avoid the popular programme @xmath37 and list @xmath38 as the first choice to get higher priority .
otherwise , @xmath36 may end up with no offers at all .
however , if students preferences are totally _ uncorrelated _ , which means different students hold independent views to the programmes , @xmath36 may know little about others preferences . in this scenario telling truth
might be the best choice of students .
the following property state this observation formally .
[ theo : bayes_ne ] suppose all programmes have equal quota and their reciprocating factors are independently drawn from a uniform distribution over any range @xmath41 $ ] ( @xmath42 ) .
besides , the preferences of all students are independently drawn from a uniform distribution over the set of all possible rank orderings ( i.e. , the uncorrelated environment ) .
then , truth - telling is a bayesian nash equilibrium for each student under the generalized matching mechanism . for readability
, we defer the detailed proof to appendix [ proof : bayes_ne ] . property [ theo : bayes_ne ]
shows that students would be forced to reveal their true preferences if they have no exact knowledge of other students realized preferences except the distribution .
this is one extreme case with totally uncorrelated preferences . at the other extreme
is the totally correlated case where students have complete information of other students true preferences .
property [ theo : dominated ] characterizes the possibilities of students strategies in the latter scenario . [ theo : dominated ] suppose all students have the same preference over colleges . besides , this is a common knowledge among all students . let the quota of their favorite college be @xmath43 .
then truth - telling is a _ dominated strategy _ for all students except the top @xmath43 students ranked according to their scores .
we present the proof in appendix [ proof : dominated ] . in practice
, students can only have partial knowledge to other students submitted preferences , probably through the history of admission data in each individual programme .
besides , students need to estimate the value of @xmath25 in each programme and carefully choose programmes they make strategy on since promoting preference in purpose for programmes which run da - like mechanism ( i.e. , @xmath26 is slightly above zero ) may have little effect in raising their merit scores in those programmes . on the other hand , in each admission year programmes would take lessons from admission outcomes of the previous year and adjust their policy on how to determine parameters like @xmath26 and @xmath44 .
these interactions form a complicated extensive - form game which may be repeated infinitely at the time granularity of each admission year .
this leaves a bunch of open questions to be settled such as : whether there are any equilibrium strategies and if they do exist , would the matching outcomes converge to any of these equilibrium states after finite rounds of games year by year . until now we have focused on students strategies in submitting preferences and assumed that college programmes would reveal their reciprocating factors nonstrategically .
while it seems intractable to characterize the dynamic game in the long run , we do obtain some analytical result for the single shot game in one particular admission year .
the following property helps relieve our concerns of programmes strategies so that we may refocus on the manipulations from applicants side .
[ theo : college_truth ] suppose after students submit their preferences , each college @xmath12 determines its true values of @xmath25 and @xmath22 independently .
then revealing @xmath25 and @xmath22 truthfully would be the dominant strategy for any programme @xmath45 in the _ ex post _ perspective .
we present the proof in appendix [ proof : college_truth ] .
here we take the ex post perspective may not hold since colleges would take students reaction into consideration before making any decisions .
it would be an interesting future work for analyzing the complicated interactions from the ex ante perspective . ] since when colleges make their strategies , students preferences are already submitted and fixed . in practice , when the admission authority evaluate the matching outcome each year , they can only conduct it based on the submitted preference lists of students
. it would be expensive ( or even impossible ) to obtain the true preferences of all applicants via survey or other methods .
thus for the single shot game , colleges would passively reveal their true parameters once they receive applicants preference lists .
in this section we have shown some fundamental properties of the generalized mechanism . to further investigate the degree of satisfaction participants perceive under different mechanisms , we implement simulations and present the social welfare results different mechanisms induce in the next section .
to define and compare social welfare under different matching outcomes , we need to quantify the utility of each participant / agent in the mechanism first . denoted by @xmath46 the set of students and @xmath14 the set of colleges .
@xmath47 is the set of all participants and @xmath48 is the set of all possible matching outcomes . for any agent @xmath49 ,
let @xmath50 be the reciprocating preference list is students original preference .
] of @xmath51 and @xmath52 be the set of participants matched to @xmath51 under certain outcome @xmath53 . for each agent @xmath54 ,
denote integer @xmath55 as the ranking agent @xmath56 appears in @xmath51 s preference list @xmath50 .
for example , @xmath57 means that @xmath56 is the first choice in @xmath51 s preference list .
we use @xmath58 to denote the unmatched case .
we assume that agent @xmath51 s utility is additive and only determined by the orders of the matching set in the preference list , which can be written as , @xmath59 where @xmath60 is non - increasing as integer @xmath61 increases .
intuitively , it means that higher order ( smaller @xmath21 ) would generate higher degree of satisfaction for agent @xmath51 .
the aggregate utility of students ( or colleges ) under outcome @xmath62 is then : @xmath63 the _ social welfare _
is defined as the aggregate utility of all participants in the mechanism , which can be written as follows : @xmath64 we say matching outcome @xmath65 is more _ efficient _ than @xmath66 if : @xmath67 we further say mechanism @xmath68 is more _ efficient _ than @xmath69 if @xmath70 can always induce a more efficient matching outcome than @xmath69 under any possible preference lists of agents . generally speaking
, an outcome would be more efficient if it induces higher ranked matching . in the context of college admissions ,
a mechanism which generates more first - choice matching for students is likely to be more efficient .
suppose there are @xmath71 students and @xmath72 colleges with just one quota in each college .
we assume the utility of each agent is as follows , @xmath73 thus a first - ranked matching would bring in utility of @xmath71 for either students or colleges .
notice that the efficiency upper bound is @xmath74 since there are at most five pairs of students and colleges matched with each other .
the preference lists of students are generated as follows : student @xmath20 evaluates each college @xmath12 by this formulae , @xmath75\ ] ] where @xmath76 denotes the social reputation of each college and @xmath77 denotes the individual preference of student @xmath20 , which is independently drawn from uniform distribution over @xmath78 $ ] .
the factor @xmath79 denotes the _ degree of correlation _ for students preferences . can be also regarded as a signal implying how much information students may know about other students preferences . if @xmath80 , each student has complete information to others preferences since they share exactly the same opinion over colleges . if @xmath81 , each student has independent opinion over colleges and thus has no posterior knowledge on others preferences . in general , students individual opinion would be more or less be affected by the common social opinion and each of them would have partial knowledge to other students preferences . ]
the preference list @xmath34 can therefore be deduced by comparing the value of @xmath82 for different @xmath12 , i.e. , @xmath20 prefers @xmath37 to @xmath38 if @xmath83 .
the preference lists of colleges are generated as follows : college @xmath12 evaluates each student @xmath20 by this formulae , @xmath84 where @xmath85 if @xmath12 is the @xmath21-th choice in @xmath34 .
@xmath24 is the exam score of student @xmath20 , which is independently drawn from uniform distribution over @xmath78 $ ] .
we generate the reciprocating factor @xmath25 for each college @xmath12 by the following distribution : @xmath86 the preference list @xmath87 can then be inferred by comparing the value of @xmath88 for different @xmath20 . in case of ties , namely , @xmath89
, @xmath90 is favored over @xmath91 if @xmath92 . after generating the reciprocating preference lists for both sides ,
the matching outcome can be obtained by applying the student - proposing da algorithm .
we then calculate the social welfare under the matching outcome .
in the simulation , we use @xmath93 $ ] , with a step size of @xmath94 . for each particular @xmath79
, we repeat the process of preference generation for @xmath95 times and compute the average values of aggregate utility and social welfare . for comparison
, we also calculate the average social welfare under pure gs mechanism , which can be easily implemented by just setting @xmath96 for each college @xmath12 in the distribution of reciprocating factors . as certain manipulation to the true preferences ,
then our comparison of social welfare of da and hybrid mechanism should have been based on the same ( true ) preferences of colleges . since the latter kind of comparison is already fully covered by property [ theo : college_truth ] theoretically
, we feel that there would be no need to conduct further simulations on it ( which would not produce any `` surprising '' results beyond theory ) . ]
[ uti_student ] presents the simulation results for aggregate utility of students over different degrees of preference correlation .
as we can see , the expected aggregate utilities under both mechanisms decrease as @xmath79 increases from zero to one .
the upper bound of @xmath97 is 50 since there are at most five students who can receive their first - choice offers from colleges .
when @xmath79 is small ( less than around @xmath98 ) , we can achieve about @xmath99 and @xmath100 of the upper bound under the jupas - like hybrid mechanism and the pure gale - shapley ( gs ) student optimal mechanism respectively .
as @xmath79 rises , the preference list of each student becomes more and more similar and there are more collision between students interest in colleges .
when @xmath79 is large enough ( greater than @xmath101 ) , the pre - determined social reputation of each college becomes the dominant factor in forming the preference lists of students .
that is to say , @xmath102 would be @xmath103 for all students .
thus college @xmath37 would always bring utility of @xmath71 to the student community , @xmath38 brings @xmath104 and so on , which forms this lower bound of @xmath105 .
we also notice in fig .
[ uti_student ] that in general @xmath97 is slightly larger under the hybrid mechanism than under the gs mechanism .
this result helps ease the concern that the jupas - like mechanism would hurt the interest of student community as a whole .
the intuition is that while some students with higher exam scores may get worse in the hybrid mechanism , other students with slightly lower scores would have more chances to enter the programmes / colleges in which they are really interested .
[ uti_college ] shows the result for aggregate utility of colleges . in the pure gs mechanism ,
since preferences of colleges are only determined by the exam scores of students , all colleges would share exactly the same preference list over students .
therefore the student with the highest score would always bring utility of @xmath71 to the college side , the student with the second highest score brings @xmath104 and so on .
that is why @xmath106 would be always equal to @xmath107 under the gs mechanism .
the upper bound of @xmath106 is also 50 , which occurs only if all five colleges realize their first choices . as shown in the figure
, we can achieve about @xmath108 of the upper bound under the jupas - like hybrid mechanism when @xmath109 $ ] .
as students preferences become more similar , colleges tend to have similar reciprocating preference , which means more conflict would occur among different colleges .
thus as @xmath79 continues increasing from about @xmath110 , the aggregate utility of colleges would decrease rapidly .
when @xmath79 is large enough ( greater than @xmath101 ) , all colleges would share the same preference over students .
thus student with the highest score would always bring utility of @xmath71 to colleges , student with the second highest score brings @xmath104 and so on , which forms the lower bound of @xmath111 .
+ the expected social welfare under different values of @xmath79 is shown in fig .
[ social_welfare ] . in the same way
, we obtain the upper bound of social welfare as @xmath112 . when @xmath79 is small ( less than @xmath113 ) and students
have various preferences over colleges , we achieve about @xmath108 and @xmath114 of the upper bound under the hybrid mechanism and the gs mechanism respectively .
when @xmath79 approaches to one and students share common opinion on colleges , the ratio would both decrease to @xmath32 .
this comparative result of social welfare helps justify the implementation of jupas - like mechanism in college admissions .
the transfer from the hybrid mechanism to the gs mechanism can only achieve the well - known incentive compatible property at the cost of potentially significant loss of efficiency , especially when students have independent opinions on different colleges .
as we have mentioned in the last section , when students preferences become more correlated , it would be easier for strategic students to manipulate the matching results successfully .
thus in the following simulation we use @xmath80 as an example to illustrate the effect of possible strategies of students .
when considering @xmath80 , the true preferences of students are all the same as @xmath115 .
we assume there are two types of students : the truthful ones and the strategic ones .
the truthful students would always submit their true preferences while the strategic students are likely to manipulate their submitted preference lists based on the information they already collect . in our simulation setting , the key information strategic students may infer is the reciprocating factor of each college . notice that in practical college admissions mechanism like jupas , students would know their own examination scores and overall performances of other students before submitting their applications . here in the simulation
we assume the strategic students would know exactly their score ranking out of all students . then one possible strategy would be as follows : @xmath116 for student with the highest score , since he / she can be assured to get into @xmath37 , there is no need to strategize . for other students , since @xmath37 is already occupied , they can shield their first choice and list @xmath37 as their last choice . thus the student would manipulate his / her first and second choice as @xmath117 and @xmath118 in order to obtain better opportunity .
[ fig : stu_gain_compare_random_alpha ] shows the expected utility of students of either type .
since the aggregate utility of all ten students are always @xmath119 under @xmath80 , the average utility per student would be @xmath120 when there are no strategic students , which serves as the benchmark for evaluating the effectiveness of possible strategies .
as we see , when there are only one strategic student , strategy @xmath121 would bring expected utility of @xmath122 , much more than the benchmark utility of @xmath120 when all students act truthfully .
as the number of strategic students increases , the expected utility per strategic student would decrease gradually , but still remain larger than @xmath120 .
this result validates the effectiveness of the proposed strategy @xmath121 under our simulation setting .
the existence of strategic students would make the students who act truthfully worse - off , however , from the figure we see this side - effect is bounded as the number of strategic students increase .
even when all other students act strategically , the remaining truthful student could still achieve about @xmath100 of the benchmark utility .
[ fig : uti_if_strategy_random_alpha ] presents the aggregate utility of both sides when strategic students exist .
the aggregate utility of students remains @xmath119 since all students share the same preference list as @xmath123 .
the aggregate utility of colleges ( the square - marked line ) is calculated based on the _ submitted _ preference lists of students , since in practice the true preferences of students are unrevealed to the public . when there are no strategic students , all students submit the same preference list .
thus all colleges would share exactly the same reciprocating preference over students , ordered solely by the scores of students .
this explains @xmath124 when the number of strategic students is zero since student with the highest score always brings utility of @xmath71 to colleges , student with the second highest score brings @xmath104 and so on . as there are more strategic students , colleges other than @xmath37 would probably enroll students with higher merit scores since strategic students list colleges as more favored choices .
thus the aggregate utility of colleges based on the submitted preferences would increase . for comparison
, we also show the aggregate utility of colleges based on the _ true _ preferences of students in the asterisk - marked line . in our setting of @xmath80
, the true reciprocating preferences of colleges would be all the same , ranking the students according to their exam scores .
when all students tell truth or all apply strategy @xmath121 , @xmath37 would enroll the student with the highest score , @xmath38 gets the student with the second highest score and so on .
thus @xmath125 .
in other cases , some strategic students with lower scores may receive better offer than they could achieve when telling truth .
for instance , @xmath118 may not receive the student with the fifth - highest score as anticipated .
therefore the aggregate utility of colleges would decrease from the initial value of @xmath119 .
however , this adverse effect is bounded as seen from fig .
[ fig : uti_if_strategy_random_alpha ] , because when there are too many strategic students in the system , it would be difficult for strategic students with very low scores to achieve better matching outcome .
[ whole_fig : diff_beta ] - show the expected utility each particular student gains under _ different degrees of preference correlation_. the horizontal axis denotes the score ranking of student , and each point drawn in the figure is the average value for one thousand iterations .
the possible strategy students could make is strategy @xmath121 .
we assume there is only one student behaving strategically while the rest of students would reveal their true preferences . from the figures we see that for the top one student , he / she would always achieve utility of ten since according to strategy @xmath121 he / she would just act truthfully and
can always get the first choice .
[ whole_fig : diff_beta ] verifies that truth - telling is an equilibrium strategy when students preference are totally uncorrelated : the students would not be better off when deviating unilaterally from the equilibrium , whereas under @xmath80 , truth - telling would be dominated by the strategy @xmath121 , as shown in fig .
[ whole_fig : diff_beta ] . for general value of @xmath126 ,
the students with higher ranking would tend to act truthfully while the students with lower ranking tend to behave strategically by choosing the less famous colleges . as the preferences of students becomes more correlated , more students would have incentive to make certain strategy rather than acting truthfully
. + +
* pros and cons regarding da and the generalized mechanism * we have shown that under complete information , students do have strong incentive to lie ; on the contrary , in the absence of any posterior knowledge , truth - telling is still the ne of the generalized mechanism . in the simulation , @xmath79 serves as an indicator of how much information each student may know about others preferences .
we would like to highlight through this paper the debate whether college side would all agree on applying da so as to achieve strategy - proofness property , or some colleges would be willing to enroll students with higher interests in risk of potential manipulation .
our generalized model just establishes a framework to admit such tradeoff between gain in efficiency and risk in truthfulness , and gives each college the freedom to adjust their policy through rf .
it is worth noting that the proposed model does not intend to repel the current trend of practice of da .
da is essentially encompassed in our framework and we just give it a second thought from the perspective of individual college .
to further understand the potential downside of da , let us consider the following scenario : if da is enforced in all participating programs of jupas , for those less popular programmes , lots of students with lower scores but more interest would be wiped out , and instead most quota of these programmes would be occupied by students with higher scores but less interest ( their scores are not high enough to get into popular programmes which they listed as top choices , so finally end up with an offer from bottom choices ) . one conventional policy in hong kong s universities
is that students are allowed to switch programme inside the university in the end of the first year in case students feel the current programme is not suitable for themselves after one year s study . in this case
most unsatisfied students enrolled by these unpopular programmes would apply to change .
if approving most of these applications , those programmes may suffer from high vacancy rate .
however , if rejecting most of them , the majority of students in the programme would feel unsatisfactory . this becomes a dilemma for the unpopular programmes .
the main objective of da is to achieve stability ; however , it may end up with `` unstable '' outcomes in the long run .
this explanation helps rationalize the current selection of hybrid manner in jupas rather than hasty replacement with da .
unfortunately , these concerns and debates from the college side are largely ignored in existing literature . *
extension to marriage problem * the generalization of college admissions mechanism also applies to the classic stable marriage problem discussed in the original paper of @xcite .
consider the situation when a woman faces two men proposals and has no clear idea which one she strictly prefers .
technically , we call there exists a _ tie _ in the woman s preference list . roughly speaking , the existing literature mainly provides two solutions to deal with marriage problem with tie .
a quick solution is just requiring the woman to flip a coin to produce a strict preference list so that the previous mechanism could be applied immediately .
the other solution concerns how to find the optimal matching outcome among all these artificial tie - breaking possibilities , for instance , the polynomial - time stable improvement cycles algorithm raised in @xcite .
however , suppose the woman has the wish : `` i ll choose the man who loves me most ! ''
yet the fact is that the first man has listed her as the first choice while the second man listed this woman as his last choice and was rejected by every other women in the previous rounds , assuming men - proposing da algorithm @xcite applied here .
obviously a `` reasonable '' matching mechanism should respect each participating agent wish and therefore always match the first man with the woman .
however , existing matching mechanisms provide no channel for agents to express such kind of reciprocating preferences , although mutual appreciation is a very natural and common factor in determining marriage mates .
our generalized model can be easily carried over to the above scenario . by analogy
, we can define the _ merit score _ of man @xmath9 in woman @xmath127 as follows : and @xmath127 in the equation , we may generate the merit score of woman @xmath127 in man @xmath9 . ]
@xmath128\ ] ] where @xmath129 is the initial rating score of @xmath9 in @xmath127 ; @xmath21 is an integer denoting @xmath127 s position in @xmath9 s initial preference list and @xmath130 is the bonus score which is decreasing over @xmath21 , i.e. , the smaller @xmath21 is , the more bonus score the man can achieve ; @xmath131 is the reciprocating factor of the woman , denoting @xmath127 s sensitivity to other men s evaluation to herself . if @xmath131 is set to zero , which is equivalent to the classic model where a woman only believes her own feeling and judgement
conversely , if @xmath131 equals to one , woman @xmath127 is extremely sensitive to men opinions on her and hopes to match with the one who loves her most . in general
a woman may set @xmath131 between zero and one to strike a balance between her initial feeling and men s appraisal to her . choosing a mate with mutual appreciation seems more `` reasonable '' and natural in practical marriage .
the parameter of _ reciprocating factor _ provides an opportunity for agents to more fully express their wishes or perceived payoff than in the classic model . finally by comparing the _ merit scores _ for different men
, we can reproduce the preference list of @xmath127 . in case of tie in merit scores ,
we resort to original score for tie breaking . * related work * some recent developments in matching theory share our concerns that the transition from the boston mechanism to the gs mechanism is not problem - free and object to the hasty rejection of the boston mechanism .
one main research direction is the analysis of efficiency in school choice ( sc ) setting where schools do not have _ strict _ preferences over students and have to largely rely on _ random lotteries _ to determine their preferences .
abdulkadirolu , che and yasuda first brought up the uncertainty factor of lotteries into efficiency consideration ( see * ? ? ?
they showed an elegant example when students share identical ordinal preference but differ in preference intensities , the boston mechanism can dominate the gs mechanism in terms of _ expected _ cardinal efficiency .
a new choice - augmented deferred acceptance ( cada ) mechanism was proposed accordingly which supports a greater scope of efficiency than the pure gs mechanism ( see * ? ? ?
the same authors further generalized the single example into a `` baseline model '' where students have common ordinal preferences and schools have no priorities in @xcite . besides , @xcite showed that the above analytical results could extend to more realistic cases such as weak priorities by simulation .
featherstone and niederle then classified the efficiency issue in sc into three categories ( see * ? ? ?
* ) : ex post : : : each student knows preferences of other students and lottery results in each school . the matching outcome as well as the efficiency are both deterministic .
interim : : : students know preferences of other students but remain unknown to the lottery results , i.e. , we would investigate the efficiency before the lotteries are drawn .
the distribution of lottery results would induce an expected , other than deterministic , value of efficiency .
ex ante : : : students only know the distribution of other students preferences and still remain ignorant of lottery results .
the authors concluded in the same paper that , when student preferences are uniformly distributed and schools are completely symmetric , the boston mechanism can first - order stochastically dominate the gs mechanism in terms of _ ex ante _ efficiency , both in theory and in the laboratory .
following the efficiency classification above , results in @xcite would all fall into the _ interim _ viewpoint with highly _ correlated _ student preferences , which complements the conclusion in @xcite under _ independent _ student preferences .
although sharing the same caution against a hasty replacement of the boston mechanism , our paper stands distinct from these above works in several aspects : * one _ key _
assumption for the above works is the _ weak _ or even no priorities in schools such that lotteries are largely relied upon in schools in order to break the tie .
it is this `` randomness '' that causes the potential _ ex ante _ efficiency loss of the gs mechanism .
however , in practical ca context where students scores rather than the random lotteries play the decisive role in admissions , the above assumption would no longer hold , so would the corresponding conclusions . *
our paper follows a distinctive and unique research direction and shows that even when the priorities in schools are _ strict _ , boston still exhibits some prominent properties such as respecting the interests of applicants .
the sociological consideration of agents preferences has been largely ignored in previous research of college admissions system . in terms of interdependent preferences ,
we notice a recent work of @xcite proposing `` interdependent values '' in two - sided matching which can be regarded as a complementary notion to our reciprocating preferences . in @xcite
the authors argue that a college @xmath12 s evaluation of a student @xmath20 could be affected by ( or depend on ) other colleges evaluation to this student @xmath20 , while we consider the scenario where @xmath20 s value to @xmath12 is dependent on @xmath12 s value to @xmath20 .
in this paper we propose a generalized matching mechanism which can incorporate both bm and da . inspired by a practical college admissions system ,
i.e. , jupas in hong kong , we propose a common parameter , namely _ reciprocating factor _ ( or @xmath26 ) , for the generalized matching model .
this parameter serves as a bridge between bm and da : when all @xmath26 are set to zero , the matching mechanism would be equivalent to pure bm ; when all @xmath26 equal to one , the matching mechanism reduces to pure da . practical systems like jupas can be regarded as a hybrid of bm and da with reciprocating factor between zero and one . in the context of college admissions ,
reciprocating factor is of practical significance for programmes to achieve the tradeoff between eligibility and real interest of enrolled students .
we have discussed the advantage and disadvantage of da and the generalized mechanism extensively and highlighted the debate from the perspective of colleges side . these potential concerns and doubts from colleges would help justify the current selection of hybrid system in hong kong . with regard to future works ,
our paper could be further improved and extended in several directions : 1 .
one major open question regarding the generalized mechanism is how history data of admission may affect the strategies and choices of current applicants . are there any equilibria for this extensive - form game that all participants would conform to ?
how would their behaviors in the system evolve in the long run ?
we hope that these strategic issues can be tackled in the future .
2 . how the reciprocating factors of colleges are distributed is another interesting direction which we have not yet investigated in details .
@xcite showed that both popular and unpopular colleges have motives of giving preferential treatment to applicants who rank them as top choices .
does this imply that different colleges tend to have similar reciprocating factors in practice ?
this question may not be addressed immediately since these `` inside '' information of colleges is typically inaccessible to the public .
we have showed some positive results of the hybrid system through simulation , however , simulation alone is insufficient to cover all complicated strategies of students . as the future work , we plan to design some lab experiments which would involve real participants to play and learn during the game .
we believe that these empirical results would provide more evidence and insight for supporting and spreading the adoption of hybrid system like jupas in hong kong .
it is easy to see that when all rf are zero , the generalized mechanism reduces to da which is strategy - proof . to prove a mechanism is non - truthful in general
, we just need to find one counter - example for any given rf which are not all zero .
suppose all colleges have zero rf except college @xmath12 with positive @xmath25 . besides , @xmath12 has quota for only one student .
let @xmath25 equal to any small @xmath132 .
suppose that there are two students @xmath90 and @xmath91 .
@xmath90 achieves a higher score @xmath133 and @xmath12 is his / her second choice school ; @xmath91 gets a lower score @xmath134 but lists @xmath12 as his / her first choice .
the merit score of @xmath91 in @xmath12 is then : @xmath135 if telling truth , @xmath90 s merit score in @xmath12 would be : @xmath136 by letting @xmath137 , we have the condition when @xmath12 prefers @xmath91 to @xmath90 : @xmath138 since @xmath139 and @xmath27 is a ( strictly ) decreasing function of choice order , we have @xmath140 .
we further assume that @xmath90 knows he / she would not succeed in his / her first choice college , thus the best he / she can achieve is getting into college @xmath12 . by listing @xmath12 as the first choice , @xmath90 would obtain a higher merit score in @xmath12 : @xmath141 which is always larger than @xmath142 . for student @xmath20 ,
let @xmath11 be its preference list where college @xmath12 is ranked @xmath21-th .
consider a new preference list @xmath144 of @xmath20 which exchanges the @xmath21-th choice with the @xmath56-th ( @xmath145 ) choice college .
a mechanism is called _ rank monotonic _ if the probability of @xmath20 being matched to @xmath12 is weakly higher under @xmath144 than under @xmath11 no matter how preferences and reciprocating factors are distributed .
consider the notation in the previous definition . also consider any state of the matching system ( i.e. , any possible scores , submitted preferences of students and rf of colleges ) where student @xmath20 submits @xmath11 and is matched to college @xmath12 . in this state , student @xmath20 is rejected by every college prior to its @xmath21-th choice in @xmath11 , and after @xmath12 enrolls @xmath20 , no other students can drive @xmath20 out of @xmath12 s quota in later rounds . mathematically ,
suppose @xmath12 has quota @xmath43 and @xmath20 is in the @xmath51-th position in @xmath12 s merit order list , then there will be at most @xmath146 students with higher position than @xmath20 in @xmath12 s list .
hence , if student @xmath20 had instead submitted @xmath144 , in the same state of the system , he also would have been rejected by every college prior to its @xmath56-th choice in @xmath144 . considering @xmath12 s merit order list , student @xmath20 can only get promoted since it lists @xmath12 as higher choice in @xmath144 . after getting into @xmath12 in the @xmath56-th round ,
since preferences of all other students and colleges remains the same as under @xmath11 , there will also be at most @xmath146 students with higher position than @xmath20 in @xmath12 s new list .
so @xmath20 can guarantee its admission to @xmath12 .
this implies that the probability of student @xmath20 being matched to college @xmath12 is weakly larger when it submits @xmath144 instead of @xmath11 .
if the probability of student @xmath20 being matched to its @xmath51-th choice college is independent of its submitted preference list , we say that its preference revelation problem exhibits _
college anonymity_. [ lemma : college_anonym ] suppose there are @xmath9 colleges with quota @xmath43 and @xmath147 students .
if the submitted preferences of all students other than @xmath20 , as well as the reciprocating factors of colleges are uniformly distributed , then the preference revelation problem of student @xmath20 exhibits _
college anonymity_. let student @xmath20 submit preference list @xmath34 where @xmath148 is its @xmath51-th choice college in list @xmath34 .
denoted by @xmath149 other student s preferences and @xmath150 the vector of each college s reciprocating factor .
let @xmath151 be the set of all possible @xmath149 . and
@xmath152^m$ ] is the @xmath9-dimensional space containing all possible values of @xmath153 .
for each particular @xmath154 , denoted by @xmath155 the region of @xmath153 in which @xmath20 is matched to its @xmath51-th choice under @xmath156
. then the probability of @xmath20 being matched to its @xmath51-th choice can be denoted as : @xmath157 now let student @xmath20 submit a different preference list @xmath158 .
this induces a permutation mapping @xmath159 from the set of colleges index @xmath160 to itself which is defined as follows : for any college @xmath161 @xmath162 , let @xmath163 be its ranking in list @xmath34 , then the image value @xmath164 is the index of college which is ranked @xmath163 in list @xmath158 . by a slight abuse of notation , we express this permutation as @xmath165 for conciseness . for each @xmath166 , by symmetry , we know that under + @xmath167 , student @xmath20 would still be matched to its @xmath51-th choice college in @xmath158 .
since @xmath159 is a one - to - one mapping , by this way of transformation we would have listed all the possibility for the matching .
thus the probability of @xmath20 being matched to its @xmath51-th choice in @xmath158 can be denoted as : @xmath168 since preference list is uniformly drawn , we have @xmath169 .
the integral parts in equation ( [ eqn : prob_i ] ) and ( [ eqn : prob_i_prime ] ) are equivalent since it is just a simple substitution of the variables , i.e. , they both equal to the following expression : @xmath170 we still use @xmath172 to denote the probability of @xmath20 being matched to its @xmath51-th choice college .
combining the conclusion in lemma [ lemma : rank_mono ] and lemma [ lemma : college_anonym ] , we have that @xmath173 for @xmath174 . the best response in this case
is clearly to put the favorite school in the first place , the second most favorite one in the second place and so on .
thus truth - telling would be the best response for students .
suppose there are @xmath9 colleges and @xmath147 students .
without loss of generality , we assume that all students share the same preference list : @xmath175 . let the quota of their favorite college @xmath37 be @xmath43 and reciprocating factor of @xmath37 is @xmath176 . by listing @xmath37 as their first choice , the top @xmath43 students judged by score ( denoted by @xmath177 with @xmath178 ) would still occupy the top @xmath43 vacancies in @xmath37 s merit score list .
this can be easily seen by the composition of merit score : @xmath179 for any other student @xmath7 ( @xmath180 ) listing @xmath37 as @xmath51-th choice , its merit score would be smaller than @xmath181 since @xmath182 and @xmath183 . for the rest students , since all the quota of @xmath37 is filled up , they would have no incentive to still list @xmath37 as their first choice . to prove that truth - telling ( denoted by @xmath184 ) is a dominated strategy , we only need to find one particular strategy which dominates @xmath185 .
we now focus on one of such strategies which requires students to always submit @xmath186 ( denoted by @xmath121 ) .
let @xmath20 be any of these remaining students and the utility of @xmath20 be @xmath187 when it is matched with @xmath4 ( @xmath188 ) .
denoted by @xmath172 the probability of @xmath20 being matched with @xmath4 under @xmath184 and @xmath189 the same probability under @xmath121 ( notice that @xmath190 and @xmath191 would be zero since @xmath37 s quota has already been occupied ) .
then the expected utility under @xmath184 would be no more than that under @xmath121 since : @xmath192 according to lemma [ lemma : rank_mono ] , the probability of @xmath20 being matched with @xmath4 ( @xmath193 ) would be higher under @xmath121 since @xmath4 s position gets promoted compared with @xmath184 .
thus @xmath194 and the inequality above holds .
this gives the proof that truth - telling is a dominated strategy . for analyzing colleges strategies
, we borrow the idea of _ dropping strategies _ and _ rejection chains algorithm _ from @xcite and show that it is impossible for colleges to find an effective dropping strategy and manipulate the matching successfully . thus the best matching colleges may achieve is via revealing their true preferences ( i.e. , their merit order lists ) . to complete our proof ,
we first briefly restate some basic definitions and lemmas in @xcite .
let @xmath87 be the reciprocating preference of college @xmath12 obtained through its true @xmath25 and @xmath22 .
@xmath90 and @xmath91 are two students applying to college @xmath12 .
[ lemma : dropping ] _ ( dropping strategies are exhaustive )
_ : if under certain strategy @xmath195 the mechanism produces matching outcome @xmath201 , then there must exist a dropping strategy of college @xmath12 that produces a matching that @xmath12 weakly prefers to @xmath201 according to its true preference @xmath87 . the detailed proof can be found in appendix b of @xcite .
lemma [ lemma : dropping ] implies that if there exists any successful strategy @xmath195 for college @xmath12 ( by manipulating its @xmath25 and @xmath22 ) , we can always find a dropping strategy which achieves at least the same improvement .
that is to say , if we can prove that for every dropping strategy , we can not make college @xmath12 better off , then @xmath12 would be forced to act truthfully . for any dropping strategies ,
the `` dropped ''
students in @xmath12 would re - apply to other colleges in the student - proposing da algorithm , which would cause a chain of rejection and acceptance in the subsequent stages .
this can be analyzed in details through the so - called _ rejection chains _ algorithm as follows . 1 .
if @xmath207 , * return * ; 2 . otherwise , let @xmath20 be the least preferred student by @xmath12 among @xmath208 , and let @xmath209 .
3 . iterate the following steps
@xmath20 continues to apply : 1 .
if @xmath20 has already applied to every college in @xmath34 , * go to * begin ; 2 .
otherwise , let @xmath210 be the most preferred college of @xmath20 among those which @xmath20 has not yet applied .
if @xmath211 , set @xmath212 then * return * ; 2 .
acceptance and/or rejection : 1 .
if @xmath210 has no vacant position and prefers each of its current mates to @xmath20 , then @xmath210 rejects @xmath20 , * go to * step 3 .
otherwise , @xmath210 accepts @xmath20 .
if @xmath210 has a vacant position , * go to * begin ; otherwise , @xmath210 rejects the least preferred student among those who were matched to @xmath210 .
let this rejected student be @xmath20 then * go to * step 3 ; therefore the proof can be now boiled down to answer whether algorithm 1 can possibly return to @xmath12 after applying any dropping strategy .
this is summarized in the following lemma as the final step of proof .
assuming that algorithm 1 returns to @xmath12 with strictly better @xmath203 after finite steps of iterations . by retrieving the running process of algorithm 1 , we would be able to reconstruct a complete chain of rejection and acceptance as follows ( denote @xmath218 ) : @xmath219 the above chain reflects the latest iteration before algorithm 1 returns with @xmath216 .
it starts from @xmath37 ( i.e. , @xmath12 ) rejecting @xmath90 , then @xmath90 applied to @xmath38 following its own preference @xmath220 ; after comparing @xmath90 and @xmath91 , @xmath38 decided to accept @xmath90 and reject @xmath91 ; then @xmath91 started its new application and so on .
finally , @xmath221 was rejected by @xmath222 and applied to @xmath223 ( i.e. , @xmath12 ) , which satisfies the stopping condition of algorithm 1 with @xmath216 . by inspecting the behavior of @xmath38 , we know that @xmath91 applied to it earlier than @xmath90 ( which implies @xmath91 obtained higher bonus score than @xmath90 in @xmath38 ) but is still less favored by @xmath38 .
the only explanation for this phenomenon is that @xmath90 achieved strictly better exam score than @xmath91 , i.e. , @xmath92 .
this rule also applies to other colleges in this chain . in the end , since we suppose @xmath12 achieves better matching through the dropping strategy , @xmath221 is regarded better than @xmath90 for @xmath12 , which implies that @xmath224 . in summary
, we have : @xmath225 by adding the left side and right side of those inequations , we derive @xmath226 , an apparent contradiction ! dtting , p. , henzinger , m. , weber , i. , 2010 . sponsored search , market equilibria , and the hungarian method . in : proceedings of the 27th international symposium on theoretical aspects of computer science .
287 - 298 .
iwama , k. , miyazaki , s. , 2008 .
a survey of the stable marriage problem and its variants . in : proceedings of the international conference on informatics education and research for knowledge - circulating society 2008 . 131 - 136 . | in considering the college admissions problem , almost fifty years ago , gale and shapley came up with a simple abstraction based on preferences of students and colleges .
they introduced the concept of stability and optimality ; and proposed the _ deferred acceptance _ ( da ) algorithm that is proven to lead to a stable and optimal solution .
this algorithm is simple and computationally efficient .
furthermore , in subsequent studies it is shown that the da algorithm is also strategy - proof , which means , when the algorithm is played out as a mechanism for matching two sides ( e.g. colleges and students ) , the parties ( colleges or students ) have no incentives to act other than according to their true preferences . yet , in practical college admission systems , the da algorithm is often not adopted . instead , an algorithm known as the _ boston mechanism _ ( bm ) or its variants are widely adopted . in bm
, colleges accept students without deferral ( considering other colleges decisions ) , which is exactly the opposite of gale - shapley s da algorithm . to explain and rationalize this reality ,
we introduce the notion of _ reciprocating preference _ to capture the influence of a student s interest on a college s decision .
this model is inspired by the actual mechanism used to match students to universities in hong kong .
the notion of reciprocating preference defines a class of matching algorithms , allowing different degrees of reciprocating preferences by the students and colleges .
da and bm are but two extreme cases ( with zero and a hundred percent reciprocation ) of this set .
this model extends the notion of stability and optimality as well . as in gale - shapley s original paper
, we discuss how the analogy can be carried over to the stable marriage problem , thus demonstrating the model s general applicability . * * jel classification**@xmath0c78
@xmath1 i23 @xmath1 i31 |
low - dimensional strongly correlated electron systems have attracted great attention in the last two decades .
the reason dates back to anderson s proposal @xcite that the @xmath5 version of the hubbard model might carry the basic mechanisms underlying the high - tc superconductivity observed in cuo@xmath0 compounds . despite
that this remains an open issue , the above suggestion fertilized intensive investigations on many related fundamental topics , such us itinerant electron magnetism , mott metal - insulator transitions and quantum critical phenomena . amongst several features of interest , we mention the possibility of realization of spiral @xcite , nagaoka @xcite and resonating - valence - bond ( rvb ) states @xcite , spatially separated phases @xcite and luttinger liquid behavior @xcite , which may present strong deviations from the landau fermi liquid theory .
in this work , we report numerical results of the hubbard model on the doped ab@xmath0 chain away from half filling , which show that its special unit cell topology greatly enriches the phase diagram found in the doped standard linear chain .
in fact , all features mentioned above are shown to be associated with well defined ground state ( gs ) phases of this doped chain . doped ab@xmath0-hubbard chains were previously studied through hartree - fock , quantum monte carlo and exact diagonalization ( ed ) techniques both in the weak and strong coupling limits @xcite , including also the @xmath5 model @xcite using density matrix renormalization group ( dmrg ) and recurrent variational anstzes , and the infinite - u limit @xcite using ed . in particular , these chains represent an alternative route to reaching two - dimensional quantum physics from one - dimensional systems @xcite . at half filling the ab@xmath0-hubbard chain exhibits a quantum ferrimagnetic gs @xcite , whose magnetic excitations have been studied in detail both in the weak and strong coupling limits @xcite , and in the light of the quantum heisenberg model @xcite .
further studies have considered the anisotropic @xcite and isotropic @xcite critical behavior of the ab@xmath0-quantum - heisenberg model , including its spherical version @xcite , and the statistical mechanics of the ab@xmath0-classical - heisenberg model @xcite . on the experimental side ,
the ab@xmath0 chain topology is of relevance to the understanding of the physics of some low - dimensional strongly correlated electronic systems .
one class is the line of trimer clusters present in fosfates with formula a@xmath6cu@xmath6(po@xmath7)@xmath7 , where a = ca @xcite , sr @xcite and pb @xcite .
the trimers have three cu@xmath8 ( @xmath9 ) paramagnetic ions antiferromagnetically coupled .
although the superexchange intertrimer interaction is much weaker than the intratrimer coupling , it proves sufficient to turn them bulk ferrimagnets
. another quasi - one - dimensional inorganic material closely associated with the ferrimagnetic phase of the ab@xmath0 chain is the nicu bimetallic chain @xcite .
these compounds display alternating ni@xmath8 ( @xmath10 ) and cu@xmath8 ( @xmath9 ) ions connected through suitable ligands in a line ; and are modeled by the alternating spin-@xmath11/spin-1 antiferromagnetic heisenberg chain @xcite .
we also would like to mention a more recently synthesized organic ferrimagnetic compound consisting of three @xmath9 paramagnetic radicals @xcite in its magnetic unit cell , as well as possible connections with the physics of the oxocuprates @xcite .
this paper is organized as follows : in sec .
ii we introduce the model system and the numerical techniques used to calculate several quantities suitable to characterize the occurrence of distinct phases as function of doping and coulomb coupling . in sec .
iii , we discuss spiral and nagaoka states at low hole doping , whose magnetic properties are shown to exhibit very interesting features in the weak and infinite - u limit , respectively . in sec .
iv we show that for higher hole doping the system phase separates , before reaching a mott insulating phase of short - range rvb states at @xmath2 . in sec .
v we discuss several features of the crossover region , which takes place before the luttinger liquid behavior observed for @xmath12 . finally , in sec .
vi we present a summary and some conclusions concerning the reported results .
the ab@xmath0 chain is a bipartite lattice with three sites ( named a , b@xmath13 and b@xmath0 ) per unit cell , as illustrated in fig .
[ disp](a ) .
the hubbard hamiltonian for a lattice with @xmath14 unit cells and @xmath15 sites reads : @xmath16+u\sum_{i=1}^nn_{i\uparrow}n_{i\downarrow},\ ] ] where @xmath17 and @xmath18 are the creation operators of an electron with spin @xmath19 at site a and in a bonding state between sites @xmath20 and @xmath21 of the cell @xmath22 , respectively , @xmath23 is the hopping amplitude and @xmath24 is the coulomb coupling . for @xmath25
, double occupancy is completely excluded and the hamiltonian takes the form : @xmath26p_g,\ ] ] where @xmath27 is the gutzwiller projector operator .
the model is invariant under the interchange of the @xmath28 sites of the same cell , a symmetry that implies in a well defined local parity ( @xmath29 ) for the gs wave function . as a result ,
in computing some quantities we find it convenient to use the _ effective linear chain _
( elc ) generated by the map illustrated in figs .
[ disp](a ) and [ disp](b ) , i. e. , any quantity @xmath30 associated with a @xmath28 site at cell @xmath22 of the elc is given by @xmath31 .
this mapping does not change the physical content of the gs and excited states , being used only to expose in a more clear fashion some properties of these states . in the tight - binding description ( @xmath32 )
this model presents three bands @xcite : one flat with @xmath14 odd parity states [ antibonding orbitals , @xmath33 ) ] and energy @xmath34 ; and two dispersive branches , @xmath35 with @xmath36 , @xmath37 , built from a sites and bonding ( even parity ) orbitals , as shown in fig .
[ disp](c ) . at half
filling ( @xmath38 , where @xmath39 is the number of electrons ) the gs total spin @xmath40 is degenerate , with @xmath40 ranging from the minimum value ( @xmath41 or @xmath42 ) to @xmath43 , where @xmath44 is the number of sites in the a ( b ) sublattice .
as proved by lieb @xcite the coulomb repulsion lifts this huge degeneracy and selects the @xmath45 ground state for any finite @xmath24 , giving rise to a ferrimagnetic gs @xcite . on the other hand , for @xmath25 , one hole ( @xmath46 ) and periodic boundary conditions ( bc s ) ,
the system satisfies the requirements of nagaoka s theorem for saturated ferromagnetism @xcite . for nagaoka ferromagnetism and lieb ferrimagnetism
the gs is homogeneous in parity with @xmath47 for any cell @xmath22 . due to this symmetry ,
the spectrum of the ab@xmath0 chain in the heisenberg limit ( @xmath48 , @xmath38 ) at the sector @xmath49 is identical to that of the alternating heisenberg spin-@xmath11/spin-1 chain @xcite .
here we focus on the effect of hole doping , @xmath50 , both in the weak coupling and the infinite - u limit , using exact diagonalization ( ed ) through the lanczos algorithm for closed bc s and dmrg for open bc s @xcite . in the ed procedure ,
the bc s are such to minimize the energy , except for @xmath51 and @xmath52 [ fig .
[ pdez](c ) ] in which the bc s ( periodic or antiperiodic ) are such that the fermi wave vector @xmath53 in the thermodynamic limit is included in the set of wave vectors for the finite system @xcite .
we used finite size dmrg for open chains with a sites in its extrema , keeping 364 to 546 states per block in the last sweep .
the maximum discarded weight in the last sweep was typically @xmath54 , except for odd phases and @xmath51 , where the discarded weight was @xmath55 . in the dmrg calculations we treated @xmath20 and @xmath21 as a composite site with 9 states for @xmath25 and 16 states for @xmath51
however , by considering the parity symmetry , we can decompose this supersite into the two possible symmetry sectors @xmath56 and @xmath57 . within this scheme ,
we have considered all parity symmetry sectors of the form @xmath58 , with @xmath59 contiguous cells of odd parity in one side of the open chain and @xmath60 contiguous cells of even parity in the other .
in addition , we have verified the stability of this phase separation against the formation of a mixed phase composed of smaller domains .
the energy is studied as function of @xmath59 for increasing number of states kept per block in order to localize the value of @xmath59 for which the energy is minimum , as shown in figs .
[ pdez](a ) and [ pdez](b ) .
the phase - separated boundaries are thus determined by the limiting dopings for which an inhomogeneous phase ( non - uniform parities ) is observed .
we have also developed a simple variational approach for @xmath25 and @xmath52 , which is explained in detail in appendix a. the results calculated using this approach are shown in figs .
[ pdez](d ) and [ sqsefl](c ) . in fig .
[ pdez](c ) ( @xmath51 ) and fig . [ pdez](d ) ( @xmath25 ) we present the average parity , @xmath61 as function of doping , computed using the above - mentioned methods . in both regimes , we observe the occurrence of an homogeneous phase near half filling with @xmath49 . for higher
doping , i. e. , @xmath62 [ @xmath63 and @xmath64 $ ] the system phase separates in one region with odd parity cells and the other with even ones . for @xmath65
the gs is homogeneous with @xmath66 . in order to present an overview of the conducting properties of the @xmath67 chain phases in the infinite - u limit
, we display in fig .
[ pdez](e ) the quantity @xcite @xmath68 calculated in the elc using ed , where @xmath69 , @xmath70 , @xmath71 is the electron density at site @xmath72 and @xmath73 is such that @xmath74 , with @xmath75 and @xmath73 co - primes .
the phase of @xmath76 corresponds to the gs expectation value of the position operator , while its modulus defines the localization length ; in an insulator , @xmath77 , as @xmath78 , while in a conductor , @xmath79 , for closed boundary conditions @xcite .
the increase of @xmath80 with system size for @xmath81 , as well as in the phase separated region , are evidences of insulating phases at these dopings .
these conclusions will be better fundamented by studying the drude weight using ed and the charge gap for larger systems with dmrg .
in figs . [ sqsefl](a ) and [ sqsefl](b ) we display the magnetic structure factor @xmath82 calculated at @xmath83 and @xmath51 using dmrg for the elc .
first , notice the presence of peaks at @xmath84 and @xmath85 revealing the ferrimagnetic order at half filling .
these peaks sustain up to two holes ( @xmath86 ) ; however , it is not clear whether the ferrimagnetic phase is robust against doping in the thermodynamic limit .
indeed , by increasing the hole doping , spiral peaks at @xmath1-dependent positions appear near @xmath84 and @xmath85 .
the analysis of the charge gap , @xmath87 suggests that these states are metallic , in opposition to the mott insulating ferrimagnetic state at @xmath88 .
it is worth mentioning that the occurrence of spiral phases in oxocuprates has been a challenging and topical subject @xcite . in fig .
[ sqsefl](c ) we present the gs total spin as function of doping for @xmath25 . for @xmath89 itinerant saturated ferromagnetism due to hole kinematics ( nagaoka mechanism ) is observed .
it is interesting to notice that our estimate for the upper hole density ( @xmath90 ) beyond which nagaoka ferromagnetism is unstable is in very good agreement with similar predictions for ladders @xcite and the square lattice @xcite .
we have also considered the presence of an aharonov - bohn flux @xmath91 for a closed chain through the gauge transformation : @xmath92 with @xmath93 .
the flux variation is equivalent to a change in the boundary condition : @xmath94 represents periodic and @xmath95 antiperiodic boundary conditions . in fig .
[ d0scn](a ) we present the dependence of the energy gap @xmath96 between the lowest energy state for a flux @xmath91 and that for saturated ferromagnetism ( @xmath97 ) as function of @xmath98 at @xmath99 .
we have identified many level crossings in this curve .
in fact , as the flux increases from @xmath100 , the total spin decreases from the maximum value , @xmath101 , to the minimum value @xmath102 ( @xmath9 ) for @xmath39 even ( odd ) , a behavior also observed in the square lattice @xcite .
notice that @xmath103 tends to saturation with system size , indicating that the level spacings decrease with @xmath104 .
these results suggest that the thermodynamic gs displays spontaneously su(2 ) symmetry breaking as a result of an ergodic combination of infinitely many states ( @xmath105 ) , including the singlet spiral state @xcite . in figs .
[ d0scn](b ) and [ d0scn](c ) we present the spin correlation function between cell spins @xmath106 and the magnetic structure factor @xmath107 as function of distance @xmath22 and wave vector @xmath108 , @xmath109 , respectively .
as we can observe , the saturated ferromagnetic and the spiral singlet states are adiabatically connected , such that all states contributing to the thermodynamic gs exhibit long - range ordering . in particular , as the flux increases from @xmath100 the peak of @xmath110 at @xmath84 ( saturated ferromagnetism ) steadily decreases , while the spiral state peak at @xmath111 increases .
we noted also that the charge structure factor @xmath112 where @xmath113 and @xmath114 is the electron occupation number at cell @xmath22 , is not affected by the flux variation and displays a peak at @xmath115 [ fig .
[ d0scn](d ) ] , where @xmath53 is the tight - binding _ spinless _ fermi wave vector @xcite , with @xmath116 , @xmath117 .
in the phase - separated regime the charge compressibility diverges following the linear dependence of the energy with doping . in figs .
[ cszb1b2 ] and [ nhs ] we present some properties of the gs in this regime calculated through dmrg for the elc .
first we notice that all these properties clearly exhibit some modulation on the same sublattice in the _ metallic _ odd parity region due to charge itinerancy
. in particular , this modulation is stronger in the @xmath51 spiral phase as evidenced by the correlation function @xmath118 shown in fig .
[ cszb1b2](a ) , but also noticed in the itinerant nagaoka phase ( @xmath25 ) as manifested by the site magnetization @xmath119 shown in fig .
[ cszb1b2](c ) . on the other hand , in the _ insulating _
even parity phase a flat behavior is observed , except for boundary and interface effects . these paramagnetic phases
[ see figs .
[ cszb1b2](b ) and [ cszb1b2](d ) ] are characterized by strong singlet correlations between spins at sites @xmath20 and @xmath21 at the same cell , i. e. , @xmath120 @xmath121 for @xmath51 @xmath122 , as shown in figs .
[ cszb1b2](b ) and [ cszb1b2](d ) .
in contrast , in the metallic phase this correlation varies very little with @xmath24 and indicates robust triplet correlations , i. e. , @xmath123 @xmath124 for @xmath51 @xmath122 .
notice that in the absence of hole hopping , even when restricted to a cell as in the insulating phase , the value of @xmath125 in a singlet ( triplet ) state should be @xmath126 ( @xmath127 ) .
the hole density @xmath128 is shown in figs . [
nhs](a ) and [ nhs](b ) . in the odd parity metallic phase
, holes do not occupy antibonding orbitals , whereas in the even parity insulating phase these orbitals are accessible for them . therefore , in the first case the hole densities at sites @xmath129 and @xmath130 are very similar .
this may also occur in the second case if double occupancy is excluded ( @xmath25 ) .
an illustration of the phase - separated regime for @xmath25 is shown in fig . [ nhs](c ) . in this coupling limit , unsaturated ferromagnetism
was suggested to occur in ladders @xcite and the square lattice @xcite as an intermediate phase between saturated ferromagnetism and paramagnetism as function of doping .
however , in the context of the @xmath5 model the situation is more complex and predictions of phase separation , both for ladders @xcite and the square lattice @xcite , and stripe formation for the square lattice @xcite have been reported . at @xmath2 , i. e. , one hole per @xmath129 site for open bc s using dmrg @xcite ,
the gs has even parity and is fully dominated by the mott insulating phase ( even parity ) illustrated in fig . [
nhs](c ) for @xmath25 .
the charge gap @xmath131 , where @xmath132/\delta n_e$ ] , @xmath133 ( @xmath134 ) , and @xmath135 , must be calculated with care .
first , notice that adding electrons to @xmath2 places the system in the phase - separated ( inhomogeneous ) region where the chemical potential @xmath136 is flat .
indeed , by comparing results using dmrg and ed calculations for @xmath25 , for which @xmath137 presents little finite size corrections [ fig .
[ dsdcsz13](a ) ] , we concluded that boundary effects are minimized by taking @xmath138 and placing the symmetry inverted cells at the chain center .
we thus find [ fig .
[ dsdcsz13](a ) ] @xmath139 ( @xmath140 ) for @xmath51 ( @xmath141 ) .
this problem is absent in the case of hole doping since the phase is homogeneous .
the extrapolated spin gap , @xmath142 characterized by symmetry inversion of a cell at the chain center , is also shown in fig .
[ dsdcsz13](a ) for @xmath51 ( @xmath143 ) and @xmath25 ( @xmath144 ) , with the spin gap at @xmath25 presenting little finite size dependence .
it is a quite massive excitation with the magnon localized at the odd symmetry cell , mostly at the b sites , as shown in fig .
[ dsdcsz13](b ) . in this context ,
sierra _ et al .
_ @xcite found @xmath145 using the @xmath5 model ( @xmath146 ) for @xmath147 , i. e. , @xmath148 .
we have confirmed this result by studying the @xmath24 dependence of @xmath149 using ed . in fig .
[ cor13 ] we show that the spin correlation functions at @xmath2 , calculated using dmrg , present a fast decay and can be fitted with the exponential form @xmath150}$ ] , where @xmath151 is the correlation length , @xmath22 is the cell index in the elc and @xmath152 denotes the central cell of the system .
this behavior is expected from the presence of a finite spin gap .
the values of @xmath151 for the correlations @xmath153 , @xmath154 and @xmath155 are @xmath156 0.4 ( 2.2 ) , 0.25 ( 0.45 ) and 0.39 ( 0.75 ) , respectively , for @xmath25 ( @xmath51 ) , with @xmath152 denoting the central cell .
thus , except for the correlation @xmath153 at @xmath51 , the correlation length is extremely short with spins correlated only within a cell .
further , the calculated bulk values of @xmath157 at @xmath2 are in very good agreement with those in the even phase of the separated region shown in figs . [ cszb1b2](b ) and [ cszb1b2](d ) .
the above results support a short - range - rvb ( sr - rvb ) @xcite state for the gs at @xmath2 , as illustrated in fig .
[ cor13](d ) . in this context , sierra _ et al . _
@xcite reached similar conclusions using the @xmath5 model on the ab@xmath0 chain , while giesekus has proved @xcite that a sr - rvb state is the gs of a non - bipartite lattice with the same local symmetry but a different hopping pattern .
we now focus on the behavior of the system for @xmath158 by considering a chain with closed boundary conditions and @xmath159 for @xmath25 using ed .
the first noticeable feature is the behavior of the spin correlation functions after doping the @xmath2 gs with two holes .
the value of @xmath160 ( where @xmath161 denotes an arbitrary cell ) changes from -0.41 to -0.28 .
this variation can be understood by considering that the two holes added to the system break two singlet bonds and reside predominately at b sites . in this picture
the correlation function would amounts to @xmath162 , which is close to -0.28 .
furthermore , the spin correlation functions shown in fig . [ cor13dopcross](a ) evidence the formation of long ranged bonds between electrons on @xmath28 sites , while the other correlations remain short ranged , as in the @xmath2 ground state .
this fact indicates that the electrons picked from the sr - rvb by hole doping are antiferromagnetically coupled and delocalized through the system , as illustrated in fig . [
cor13dopcross](b ) . in order to describe the system behavior for finite dopings , we display in fig .
[ cor13dopcross](c ) the correlation function @xmath163 and electronic densities as function of @xmath1 .
notice that for @xmath164 the electronic density at @xmath129 sites is almost fixed , while that at @xmath28 sites are monotonically depopulated .
as a consequence , @xmath163 continuously vanishes as the doping increases .
moreover , in fig .
[ cor13dopcross](d ) we show the relevant nearest - neighbor spin correlation functions .
these correlations display quite different magnitudes at @xmath2 , but their values approach each other for @xmath12 .
we thus consider the doping interval @xmath164 as a _ crossover region _
, where doping starts to build the luttinger liquid which is fully established for @xmath12 .
we have also calculated the charge compressibility @xmath165 through @xmath166,\label{ll2}\ ] ] where @xmath167 is the volume and @xmath168 is the electronic density ; the charge excitation velocity @xmath169 with @xmath170 and @xmath171 the system length ; and the drude weight @xmath172_{\phi_{min}},\label{ll4}\ ] ] where @xmath173 is the flux value that minimizes the energy @xcite . in an insulating phase
these quantities satisfy the limits below @xmath174 while for a metal @xmath175 , @xmath176 and @xmath177 are finite . as shown in fig .
[ csiud ] , at @xmath2 , @xmath175 and @xmath176 increases , while @xmath177 decreases with system size for both @xmath51 and @xmath25 , although the insulating character is better evidenced for @xmath25 due to its sizable charge gap , as shown in fig .
[ dsdcsz13](a ) . at the other commensurate density , @xmath178
, we can see the signals of an insulating phase for @xmath25 , while for @xmath51 we does not observe any especial behavior . in order to clarify this point ,
we have used dmrg to study the size dependence of the charge gap for larger systems at this doping . for a finite open chain
, the occupation of two holes per cell tends to @xmath178 in the thermodynamic limit . in fig .
[ dsdc23](a ) , we can clearly observe that for @xmath25 the system is in a mott insulating phase with @xmath179 ; however , the gap for @xmath51 is extremely small . in order to better understand the u - dependence of this gap ,
we have also calculated @xmath137 for intermediate values of @xmath24 , as also shown in fig .
[ dsdc23](a ) . in the inset of fig .
[ dsdc23](a ) we have fit @xmath180 using an expression similar to the limiting behavior of the charge gap as @xmath181 of the lieb - wu solution for a linear chain at half filling @xcite : @xmath182 , in which @xmath183 and @xmath184 are fitting parameters .
notice , however , that contrary to the lieb - wu solution @xcite , @xmath137 saturates to a finite value ( @xmath185 ) for @xmath25 . on the other hand , similarly to the linear chain at half filling @xcite , the data shown in fig .
[ dsdc23](b ) indicates the absence of spin gap at @xmath178 in the thermodynamic limit for both @xmath51 and @xmath25 . in the luttinger model ,
it is well known @xcite that @xmath175 , @xmath176 and @xmath177 are related through @xmath186 with @xmath187 where @xmath188 is the exponent governing the decay of the correlation functions . in order to probe the doped region for which the lower energy spectrum of the @xmath67 chain can be mapped onto the luttinger model , we consider the ratio @xmath189 which must be equal to one if the system is in the ll universality class @xcite .
since the @xmath67 chain is not strictly one - dimensional , care must be taken with the length scales ( @xmath167 and @xmath171 ) in eqs .
( [ ll2 ] ) , ( [ ll3 ] ) and ( [ ll4 ] ) . for @xmath32 ,
the orbitals at sites @xmath129 and bonding orbitals at sites @xmath28 are translationally equivalent and both build the dispersive branches shown in fig .
[ disp](c ) . in this case
, the system can be mapped onto a tight - binding linear chain with @xmath190 sites and a rescaled hopping parameter , @xmath191 , with @xmath192 . in order that eq .
( [ ll5 ] ) matches this result for @xmath193 , we must choose @xmath194 with @xmath195 ; or , likewise , @xmath196 and the dispersions as written in eq .
( [ disp ] ) . in both cases
@xmath197 , with @xmath198 .
consider , for example , the former option .
for @xmath32 , the charge excitation velocity is equal to the fermi velocity @xmath199 , which can be easily calculated as @xmath200 on the other hand , substituting the gs energy , @xmath201 into the continuous version of eq .
( [ ll2 ] ) , we obtain , @xmath202 using now eqs .
( [ ll6 ] ) and ( [ ll7 ] ) in eq .
( [ ll5 ] ) we find , as expected , @xmath192 .
we now turn to the interacting case using ed . as shown in fig .
[ ratiokr](a ) the ll character is quite clear for @xmath12 , while for @xmath164 we identify the crossover region .
the ed results for @xmath203 are presented in fig .
[ ratiokr](b ) .
notice that @xmath188 is close to 1 ( non - interacting fermions ) for @xmath51 ; while , @xmath188 is close to @xmath42 ( non - interacting _ spinless _ fermions ) for @xmath25 @xcite . in order to check these results , we used dmrg to calculate the elc spin correlation function @xmath204 whose asymptotic behavior should match that for the luttinger model @xcite : @xmath205^{1/2}}{l^{1+k_\rho}}. \label{cll}\ ] ] in eq .
( [ dmrgcorr ] ) we have considered an average over all possible pairs of sites separated by the same distance @xmath22 , a procedure that reduces open boundary effects . in figs .
[ cor23dop](a ) and [ cor23dop](b ) we show @xmath206 calculated at @xmath207 for @xmath51 and @xmath25 , respectively . also shown
are the fittings to @xmath206 using @xmath208 with @xmath197 and @xmath188 taken from the results shown in fig .
[ ratiokr](b ) after linear interpolation : @xmath209 ( @xmath51 ) and @xmath210 ( @xmath25 ) . motivated by a compromise between large values of @xmath22 and minimum boundary effects , we have considered intermediate values of @xmath22 in the fitting , which is quite good for both values of @xmath24 .
we thus conclude that the luttinger model correctly describes the low energy physics of the @xmath67 chain for @xmath12 .
in summary , the numerical results presented here have clearly evidenced the rich phase diagram exhibited by the hubbard model on the doped ab@xmath0 chain both for @xmath51 and in the infinite - u limit .
we have shown that at the commensurate dopings @xmath2 and @xmath211 the system display insulating phases , although for @xmath51 the charge gap @xmath137 is very small at @xmath178 , with indications that @xmath137 present an essential singularity as @xmath181 . for @xmath51 and @xmath212
the gs exhibit a ferrimagnetic phase reminiscent of the undoped regime , while for @xmath213 incommensurate magnetic correlations are observed . for @xmath25 and @xmath88 the gs total
spin is degenerate , whereas for @xmath214 hole itinerancy ( nagaoka mechanism ) sets a fully polarized gs . in this case , we have also observed the presence of an extensive number of low - lying levels with total spin ranging from the minimum value to @xmath215 and level spacing decaying with system size as @xmath104 . for higher doping , the system phase separates into coexisting metallic and insulating phases for @xmath216 ( with @xmath217 and @xmath218 ) .
the insulating state presents a finite spin gap and fully fills the system at @xmath2 , which is well described by a short - ranged - rvb state .
finally , a crossover region is observed for @xmath219 , while a luttinger liquid behavior is explicitly characterized for @xmath12 . in closing
, we would like to stress that the above - reported results might also stimulate further experimental and theoretical investigations on quasi - one - dimensional compounds displaying complex unit cell structures @xcite .
we acknowledge useful discussions with a. l. malvezzi and m. h. oliveira .
this work was supported by cnpq , finep , facepe and capes ( brazilian agencies ) .
in the metallic saturated ferromagnetic region ( parity symmetry -1 ) the energy as function of doping is known to have a non - interacting spinless fermion behavior : @xmath220 where @xmath221 , @xmath222 , and @xmath223 is the linear size of the system . on the other hand , in the insulating paramagnetic phase ( sr - rvb states with even parity symmetry ) at @xmath2 ( one hole per cell ) @xmath224 and the energy per cell @xmath225 is almost independent of the system linear size and can be estimated either by using ed or dmrg : @xmath226 let us now consider a phase - separated regime in which a paramagnetic phase with size @xmath227 coexists with a ferromagnetic one with size @xmath223 , so the energy per cell reads @xmath228 it is convenient to write @xmath229 as @xmath230 where @xmath231 , @xmath232 , @xmath233 and @xmath234 . using the above notation we rewrite eq .
( [ var1 ] ) in the form below @xmath235\label{var4}\ ] ] here we should notice the presence of a singularity at @xmath236 for any finite value of @xmath237 ( see fig . [ var ] ) . however , the region of physical values of @xmath59 is defined by @xmath238 @xmath239 in fig .
[ var ] we present @xmath240 for @xmath241 , in which the physical region is @xmath242 and can be found by eq .
( [ var_2 ] ) , with a minimum in @xmath240 for @xmath243 . the value of @xmath59 which minimizes the energy for a given @xmath1 , @xmath244 , satisfies the equation @xmath245_\delta=0 $ ] , which can be written as @xmath246 where @xmath247 the roots of eq .
( [ var5 ] ) are numerically calculated and conduct to @xmath248 we thus conclude that @xmath217 , which is in very good agreement with ed and dmrg calculations .
the magnetization is null at the even phase and maximum at the odd one .
we can thus derive the following expression for the gs total spin per unit cell : @xmath249\end{aligned}\ ] ] the dependence of the average parity @xmath75 on @xmath1 can also be easily written as @xmath250 finally , using eq .
( [ var11 ] ) for @xmath251 , the above results for @xmath75 and @xmath40 are plotted in figs .
[ pdez](d ) and [ sqsefl](c ) , respectively , and shown to be in excellent agreement with the ed and dmrg calculations .
30 p. w. anderson , science * 235 * , 1196 ( 1987 ) . h. j. schulz , phys .
lett . * 64 * , 1445 ( 1990 ) ; b. m. andersen and p. hedegard , phys .
. lett . * 95 * , 037002 ( 2005 ) .
y. nagaoka , phys .
147 * , 392 ( 1966 ) ; h. tasaki , prog . of theor .
phys . * 99 * , 489 ( 1998 ) and references therein , including those describing the occurrence of flat band , or nearly flat band , saturated ferromagnetism at finite @xmath24 .
saturated ferromagnetism in the square lattice was recently showed to occur in the infinite - u limit and a unit flux quantum per electron : y. saiga and m. oshikawa , phys .
lett . * 96 * , 036406 ( 2006 ) .
p. w. anderson , _ the theory of superconductivity in the high - tc cuprates _ ( princeton universitypress , princeton , 1997 )
. v. j. emery , s. a. kivelson , and h. q. lin , phys .
64 * , 475 ( 1990 ) .
e. dagotto , science * 309 * , 257 ( 2005 ) .
f. d. m. haldane , j. phys .
c * 14 * , 2585 ( 1981 ) ; j. voit , rep .
. phys . * 58 * , 977 ( 1995 ) .
a. m. s. macdo , m. c. dos santos , m. d. coutinho - filho , and c. a. macdo , phys .
lett . * 74 * , 1851 ( 1995 ) .
g. sierra , m. a. martn - delgado , s. r. white , d. j. scalapino , and j. dukelsky , phys .
b * 59 * , 7973 ( 1999 ) . y. watanabe and s. miyashita , j. phys . soc . jpn . * 68 * , 3086 ( 1999 ) .
m. a. martn - delgado , j. rodriguez - laguna , and g. sierra , phys .
b * 72 * , 104435 ( 2005 ) .
tian and t .- h .
lin , phys .
b * 53 * , 8196 ( 1996 ) . c. p. de
melo and s. a. f. azevedo , phys .
b * 53 * , 16258 ( 1996 ) . y. f. duan and k. l. yao , phys . rev .
b * 63 * , 134434 ( 2001 ) ; w. z. wang , b. hu , and k. l. yao , phys .
b * 66 * , 085101 ( 2002 ) .
r. r. montenegro - filho and m. d. coutinho - filho , physica a * 357 * , 173 ( 2005 ) . t. nakanishi and s. yamamoto , phys .
b * 65 * , 214418 ( 2002 ) ; c. vitoriano , f. b. de brito , e. p. raposo , and m. d. coutinho - filho , mol .
. liq . cryst . *
374 * , 185 ( 2002 ) .
f. c. alcaraz and a. l. malvezzi , j. phys .
a : math . gen . * 30 * , 767 ( 1997 ) .
e. p. raposo and m. d. coutinho - filho , phys .
* 78 * , 4853 ( 1997 ) ; phys .
b * 59 * , 14384 ( 1999 ) .
m. h. oliveira , m. d. coutinho - filho and e. p. raposo , phys .
b * 72 * , 214420 ( 2005 ) . c. vitoriano , m. d. coutinho - filho and e. p. raposo , j. phys . a : math .
35 * , 9049 ( 2002 ) .
j. b. anderson , e. kostiner , and f. a. ruszala , j. solid state chem . * 39 * , 29 ( 1981 ) .
m. drillon , m. belaiche , p. legoll , j. aride , a. boukhari , and a. moqine , j. magn .
magn . mater .
* 128 * , 83 ( 1993 ) .
a. a. belik , a. matsuo , m. azuma , k. kindo , and m. takano , j. solid state chem .
* 178 * , 709 ( 2005 ) .
m. matsuda , k. kakurai , a. a. belik , m. azuma , m. takano , and m. fujita , phys .
b * 71 * , 144411 ( 2005 )
. a. boukhari , a. moqine , and s. flandrois , mat .
* 21 * , 395 ( 1986 ) .
h. effenberger , j. solid state chem . * 142 * , 6 ( 1999 ) .
m. verdaguer , m. julve , a. michalowicz , and o. kahn , inorg . chem . * 22 * , 2624 ( 1983 ) ; y. pei , m. verdaguer , o. kahn , j. sletten , and j .-
renard , inorg . chem . * 26 * , 138 ( 1987 ) ; p. j. van koningsbruggen , o. kahn , k. nakatani , y. pei , j. p. renard , m. drillon , and p. legoll , inorg . chem .
* 29 * , 3325 ( 1990 ) .
s. k. pati , s. ramasesha , and d. sen , phys .
b * 69 * , 8894 ( 1997 ) ; s. yamamoto , s. brehmer and h .- j .
mikeska , phys .
b * 57 * , 13610 ( 1998 ) ; n. b. ivanov , phys . rev .
b * 62 * , 3271 ( 2000 ) ; s. yamamoto , phys .
b * 69 * , 064426 ( 2004 ) and references therein . y. hosokoshi , k. katoh , y. nakazawa , h. nakano , and k. inoue , j. am .
soc . * 123 * , 7921 ( 2001 ) ; k. l. yao , q. m. liu , and z. l. liu , phys . rev .
b * 70 * , 224430 ( 2004 ) ; k. l. yao , h. h. fu , and z. l. liu , solid state commun .
* 135 * , 197 ( 2005 ) .
s. a. kivelson , i. p. bindloss , e. fradkin , v. oganesyan , j. m. tranquada , a. kapitulnik and c. howald , rev .
phys . * 75 * , 1201 ( 2003 ) .
e. h. lieb , phys .
* 62 * , 1201 ( 1989 ) .
s. r. white , phys .
b * 48 * , 10345 ( 1993 ) ; u. schollwck , rev . mod . phys . *
77 * , 259 ( 2005 ) .
r. jullien and r. m. martin , phys . rev .
b * 26 * , 6173 ( 1982 ) .
raffaele resta and sandro sorella , phys .
82 * , 2560 ( 1999 ) ; a. a. aligia and g. ortiz , phys .
lett . * 82 * , 370 ( 1999 ) .
s. liang and h. pang , europhys .
lett . * 32 * , 173 ( 1995 ) ; m. kohno , phys . rev .
b * 56 * , 15015 ( 1997 ) ; h. ueda and t. idogaki , phys . rev .
b * 69 * , 104424 ( 2004 ) .
m. troyer , h. tsunetsugu , and t. m. rice , phys .
b * 53 * , 251 ( 1996 ) .
f. becca and s. sorella , phys .
lett . * 86 * , 3396 ( 2001 ) .
k. kusakabe and h. aoki , phys .
b * 52 * , r8684 ( 1995 ) .
t. koma and h. tasaki , j. stat .
phys . * 76 * , 745 ( 1994 ) ; r. arita and h. aoki , phys .
b * 61 * , 12261 ( 2000 ) .
s. r. white and d. j. scalapino , phys .
b * 61 * , 6320 ( 2000 ) .
e. eisenberg , r. berkovits , david a. huse , and b. l. altshuler , phys .
b * 65 * , 134437 ( 2002 ) .
d. s. rokhsar and s. a. kivelson , phys .
lett . * 61 * , 2376 ( 1988 ) .
a. giesekus , phys .
b * 52 * , 2476 ( 1995 ) .
r. m. fye , m. j. martins , d. j. scalapino , j. wagner , and w. hanke , phys .
b * 44 * , 6909 ( 1991 ) .
e. h. lieb and f. y. wu , phys .
lett . * 20 * , 1445 ( 1968 ) .
see , e. g. , c. a. hayward and d. poilblanc , phys .
rev b * 53 * , 11721 ( 1996 ) .
a. parola and s. sorella , phys .
lett . * 64 * , 1831 ( 1990 ) ; h. j. schulz , phys .
* 64 * , 2831 ( 1990 ) . c. d. batista and b. s. shastry , phys .
* 91 * , 116401 ( 2003 ) . | we present an extensive numerical study of the hubbard model on the doped ab@xmath0 chain , both in the weak coupling and the infinite - u limit . due to the special unit cell topology
, this system displays a rich variety of phases as function of hole doping ( @xmath1 ) away from half - filling . near half - filling ,
spiral states develop in the weak coupling regime , while nagaoka itinerant ferromagnetism is observed in the infinite - u limit . for higher doping the system
phase - separates before reaching a mott insulating phase of short - range rvb states at @xmath2 .
moreover , for @xmath3 we observe a crossover , which anticipates the luttinger liquid behavior for @xmath4 . |
recently , increasing attention has been paid on the understanding of linguistic systems by computational and analytical methods .
physicists , mathematicians , computer scientists and biologists apply their tools on the investigation of language capability , language change and language competition@xcite .
especially the similarity to biological systems opens the field of linguistics to methods of the far better understood evolutionary systems . for an overview
see refs.@xcite .
it is believed that a concept of `` universal grammar '' is fixed by some way in our genetic code enabling humans to learn languages fast during our childhood@xcite .
much attention has been paid on the evolution of this trait and the competition between different grammars@xcite . in these models
the different grammars compete by giving higher fitness to better competitors . because of
that it is possible to apply directly models already successfully used in evolutionary biology .
the competition between languages has been investigated in the last years analyzing the stability of an system of two or more languages@xcite as well as their size distribution@xcite . in our approach
we use a computational model _ without _ giving different fitness to agents with different language traits and without neglecting bilinguals .
models of that kind can be found in refs.@xcite .
this makes our model different from biological systems .
the knowledge of a different language does not imply a higher death probability ( we are aware that , unfortunately , there are quite many exceptions of that rule ) . in order to present an age structured model we build our one on the penna model@xcite . by the treatment of the genes like bits the model yields a rather realistic age
structure of a sexually reproducing population , and is additionally suitable for large population sizes due to its low computational cost .
the usage of an age structured model enables us to merge an age
dependent learning procedure with a model of language competition in order get more insight into their relation .
we will focus on the question whether a phase transition exists , as found in the models of refs.@xcite , between the state of dominance of one language and an uniform distribution ( also called fragmentation ) .
the paper is organized as follows : the next section explains the model , the following section presents the results of the model with respect to its parameter space .
the last section discusses the results and proposes further extensions of the model .
the model in this paper combines population genetics with age dependent language learning .
the presentation of the genetical part , with the purpose to yield a stable population having its characteristic age structure , will be provided only shortly to the reader . for a more detailed description of the penna model and its applications
we refer to refs.@xcite .
[ [ the - sexual - penna - model ] ] the sexual penna model + + + + + + + + + + + + + + + + + + + + + + is a individual based model where every agent is represented by its age and two strings ( diploid ) of 32 bits .
the bit strings are read in parallel .
every iteration after birth a new position on each of the bit
strings becomes visible .
a bit set to one corresponds to an harmful allele . at the beginning of a simulation
five positions are randomly chosen to be dominant ones .
if the position is dominant one bit of the two bit strings suffices to make the position represent a deleterious mutation . on the other positions both bits need to be set to one for deleterious effects . at the age
at which an agent has its third deleterious mutation it dies . in order to avoid population explosion
an additional factor limits its growth : the verhulst factor @xmath0 gives the probability of an agent to survive an iteration . here
@xmath1 is the current population size and @xmath2 the so called carrying capacity defining the maximum size of a population . after reaching the reproduction age
@xmath3 a female agent every iteration chooses randomly a male with age equal or older than @xmath4 in order to produce one offspring .
for each parent its two bit strings are randomly crossed and recombined .
both parents contribute one string to the offspring .
the offspring suffers one deleterious mutation at a randomly chosen position on each of its both new strings .
this high mutation rate lets the population reach a stationary distribution fast .
the two bit strings are initialized randomly .
generate a new language spoken by the child on a randomly chosen position.,scaledwidth=40.0% ] [ [ the - language - trait ] ] the language trait + + + + + + + + + + + + + + + + + + of an agent , that means its current ability to speak certain languages , is modelled in a similar way as the genetic strings of biological ageing .
the whole structure , a language should consist of , is neglected , and language is treated as an unit .
each agent contributes one additional bit string of @xmath5 bits representing its capability to speak a maximum of @xmath5 languages .
so an agent who has for instance its third bit set , speaks currently language @xmath6 ( eventually among others ) .
an agent can learn or forget a language only in youth during the first @xmath7 iterations after birth .
we call @xmath7 the _ maximum learning age_. older agents are not able to change their knowledge on languages .
the interaction between a young agent and a randomly chosen teaching agent works as follows : at first a random position on the language bit string is chosen .
the agent learns the language of that position if it is spoken by the teacher . if the agent already speaks more than one language it forgets the language of the chosen position if the teacher does not know it . in the other cases
no process of learning / forgetting is carried out .
let us illustrate the outcome of the interaction in computational terms : the language trait of an agent speaking zero or one language has an or operation with the teacher s one at the chosen position , otherwise an and operation .
the young agents s language traits are actualized in this way by @xmath8 different teachers at different positions per iteration .
@xmath9 means that the language traits are actualized once per iteration with a probability of @xmath10 . at birth
an offspring obtains its language trait as illustrated in figure [ fig : illustrate ] .
if both parents speak the same language the offspring will speak it as well . in the case that only one of the parents speaks a language the child learns it with a probability of @xmath10 .
additionally , the child can know a new language at birth with probability @xmath11 ( i.e. a randomly chosen bit position is set to one ) .
we call this probability @xmath11 of erroneous learning the _ mutation rate _ in the following because of its analogy to biological systems .
the bit string is initialized by chance at the beginning of a simulation .
this section will concentrate at first on the results of simulations with a trait of two languages ( @xmath12 ) .
the simulations are carried out for at least @xmath13 iterations or more for large population sizes .
all results present the final stationary state of a simulation .
a phase transition can be observed for the ratio between the number of speakers of different languages .
the ratio is averaged over the last @xmath14 time steps .
figure [ fig : phasetr1 ] shows the ratio versus the mutation rate , determining the control parameter , for different carrying capacities @xmath2 .
the simulations are carried out with a maximum learning age of @xmath15 and @xmath9 .
each point represents the outcome of a separate simulation .
the language ratio decreases fast at the critical point which is situated between @xmath16 and @xmath17 .
the transition separates the phase / state of about equal sizes for the two languages for small values of @xmath11 from the phase / state where one language dominates clearly the other one .
we can observe in figure [ fig : phasetr1 ] as well that the population size does not alter the shape of the transition .
thus this transition can be found for arbitrary population numbers and therefore corresponds to a real physical phase transition .
the language ratio increases strongly for low mutation rates .
a power law with an exponent of minus one is exhibited in figure [ fig : powerlaw ] .
this exponent is the same for all simulations presented here .
the parameters in the simulations are @xmath15 , @xmath9 and @xmath18 , i.e. the population consists of about @xmath19 agents . as well as with increasing amount @xmath8 of interaction per iteration.,scaledwidth=80.0% ] in figure [ fig : phase_interact ] the shape of the phase transition is compared for different values of @xmath7 and @xmath8 , parameters defining the amount of interaction between the agents . the carrying capacity @xmath20 fixes the population number at around @xmath21 .
the shape of the transition remains the same but is shifted to higher values of the mutation rate for increasing @xmath7 and @xmath8 .
the shift is a linear one with respect to each of both parameters .
the shift decreases for values of @xmath22 as expected due to the smaller number of individuals surviving up to ages much older than the reproduction age @xmath3 ( not shown ) .
the phase transition disappears and the only final state of the model is the dominant one for very large values of @xmath8 . in the following the results of simulations with a language trait consisting of more than two languages are compared to the previous ones .
the language ratio @xmath4 , the measure for a possible domination of one over the others , is now defined by @xmath23 where @xmath24 is the most spoken language , @xmath5 the number of languages , and @xmath25 the number of agents speaking language @xmath26 .
the simulations are carried out with @xmath27 and @xmath20 .
we used @xmath28 for @xmath29 , respectively in order to keep the interaction per language constant .
figure [ fig : numwords]a compares the shape of the transition for different numbers of languages .
the position of the critical point is shifted to smaller values of @xmath11 for more languages .
this means that the more languages compete against each other , the lower is the possibility to have a scenario where one language dominates the other ones .
figure [ fig : numwords]b compares the same simulations carried out with the initial state where all agents speak language @xmath30 .
the critical point is shifted to larger values of the mutation rate .
this dependence of the final state on the initialization , leading to a hysteresis , shows that the phase transition is of first order . here the jump from one phase into the other is located at the intersection of the transition curve and the curve for the ratio of two languages
. languages .
the critical point moves to smaller values of the mutation rate for increasing @xmath5 .
the phase transition is more abrupt . *
a * : for numbers of languages larger than two the curves collapse into one for values of @xmath11 in the left part where one language dominates the other ones . *
b * : the initialization with one language leads to a different result .
the transition shows a hysteresis for the simulations with more than two languages increasing with @xmath5.,title="fig:",scaledwidth=70.0% ] languages .
the critical point moves to smaller values of the mutation rate for increasing @xmath5 .
the phase transition is more abrupt . *
a * : for numbers of languages larger than two the curves collapse into one for values of @xmath11 in the left part where one language dominates the other ones .
* b * : the initialization with one language leads to a different result .
the transition shows a hysteresis for the simulations with more than two languages increasing with @xmath5.,title="fig:",scaledwidth=70.0% ] finally , we present a histogram of the hamming distances@xcite between all agents for simulations with 16 languages for different values of the mutation rate ( figure [ fig : hamming ] ) .
the parameters are the same as before .
the hamming distance is defined by the number of bits by which the two bit strings differ from each other . or , in other words , the number of bits which need to be changed to turn the language trait of an agent into that of the other one .
figure [ fig : hamming ] shows that the peak of the histogram of hamming distances turns from zero ( dominance ) to one ( no dominance ) at the critical point ( which here is located between @xmath31 and @xmath32 ) .
in our model of language competition a phase transition similar to the one for the competition of different grammars in ref.@xcite and the one for the competition of languages of ref.@xcite is observed .
nevertheless , working without fitness improvement for agents better adapted by their languages trait but with a sexual population with random mating makes the simulations useful to characterize current language competition .
the phase of an uniform distribution of the languages ( fragmentation ) suggests a scenario where these languages compete equally against each other maintaining a stable state of coexistence . in the other state one language dominates clearly the other ones .
the dependence of the phase transition point on the amount of language interaction as well as on the number of competing languages lets us make the following conclusions : there is no optimal value of maximum learning age for a language .
its increase shifts the critical point to larger values of the mutation rate .
thus a larger maximum learning age prevents the evolution of many languages where contact among speakers of different languages is strong .
the model shows also that a larger number of different languages decreases the possibility of dominance of one of them and makes the transition jump higher .
hence , the extinction of a language in a multilingual system can lead to a transition to the state where one language will begin to dominate all the other ones .
the phase transition analyzed in the presented model is of first order for simulations with many languages as found in the model of ref.@xcite .
the differences are that in our case , the transition is completely independent of the population size .
we do not know if the simplification of a language to a single bit leads to this discrepancy or if it is the usage of a model with sexual reproduction .
further analysis has to be done .
the model brought insight into the mechanisms of language competition but still much has to be done .
the results are similar to the ones of ref.@xcite .
nevertheless , our model presents another approach to understand language competition .
the design of an analytical model , for instance by diffusion equations similar to the ones used in ref.@xcite , without fitness would reveal if it is possible to obtain similar results with a mean field theory . the model presented in this paper can be extended by putting it on a lattice in order to investigate how the transition behaves with respect to a geographic distribution of languages .
another extension would be the substitution of random mating by assortative mating@xcite , a concept used frequently in the theory of biological speciation , in order to give different priorities to monolingual parents and bilingual ones . in order to understand language invasion ,
the model can be extended to include social structures .
i am funded by the daad ( deutscher akademischer austauschdienst ) and thank d. stauffer and s. moss de oliveira for comments on my manuscript . | the understanding of language competition helps us to predict extinction and survival of languages spoken by minorities .
a simple agent based model of a sexual population , based on the penna model , is built in order to find out under which circumstances one language dominates other ones .
this model considers that only young people learn foreign languages .
the simulations show a first order phase transition where the ratio between the number of speakers of different languages is the order parameter and the mutation rate is the control one . |
semiclassical laser theory , which neglects the quantum fluctuations of the electromagnetic field , is widely used to describe and simulate lasers @xcite . in principle
, it correctly describes the laser thresholds and frequencies , the spatial pattern of the lasing modes , and the laser output power , including all classical non - linear effects , such as spatial hole - burning , gain saturation , and mode and phase locking .
essentially the theory describes maxwell s equations in an open cavity , coupled to the non - linear polarization of the gain medium .
the gain polarization can be described using either a classical non - linear oscillator model @xcite , or a quantum - mechanical model of @xmath0 atomic levels in which the polarization and level populations obey the equations of motion of the quantum density matrix .
the simplest version of the theory , used widely in textbooks , is the two - level maxwell - bloch ( mb ) model @xcite ; however , most design and characterization simulations of lasers use models with @xmath1 or more levels . in addition , most theoretical solutions for the semiclassical laser equations employ a large number of simplifying assumptions in order to make them analytically tractable , most notably neglecting the openness of the cavity and/or treating only simple one - dimensional ( 1d ) or ring cavities , as well as approximating the non - linear interactions to cubic order . the results are typically not useful for quantitative modeling . until recently ,
the only useful way to obtain quantitative results for non - trivial laser structures was to integrate the semiclassical laser equations in space and time . for novel and interesting new laser structures with non - trivial 2d and 3d cavity geometries ,
such simulations are at the limits of computational feasibility , making it difficult to study a large parameter space or ensemble of designs . in the past five years a new approach to finding the stationary solutions of the semiclassical laser equations has been developed , known as steady state ab initio laser theory ( salt)@xcite .
salt treats the openness of the cavity exactly , and the multimode non - linear interactions to infinite order ( with two approximations , to be discussed below ) .
it is applicable to cavities of arbitrary complexity in 2d and 3d , although we will discuss here only the scalar wave equation coupled to a gain medium .
importantly , this approach eliminates the need to perform a time integration to steady - state , dramatically reducing the computational effort and allowing one to study cavities with high spatial complexity , such as 2d random lasers @xcite , photonic crystal lasers @xcite and chaotic disk microlasers @xcite .
salt was originally formulated to solve the steady - state two - level maxwell - bloch equations @xcite in the standard slowly - varying envelope approximation ( svea ) , and an iterative algorithm was developed @xcite to solve the resulting salt equations .
subsequently it was realized that the svea afforded no advantage in numerical solutions of the salt equations , so this approximation was dropped , leading to slightly different salt equations @xcite , which were used in all subsequent work @xcite .
recently an important generalization of the salt solution algorithm was developed , improving its performance for laser cavities with complex spatial index variation and inhomogeneous pumping @xcite . in its present form ,
salt only employs two approximations , the stationary inversion approximation ( sia ) and the rotating wave approximation ( rwa ) , which are both well satisfied for most lasers of interest . in ref .
@xcite , the results of salt were compared to the full time - dependent solution of the mb equations for a simple 1d cavity in the multimode regime , well above threshold .
excellent agreement was found in the parameter regime for which the sia holds .
this was , to our knowledge , the first demonstration of a frequency - domain method which agrees with exact time - domain methods for above - threshold multimode lasing .
previous applications of salt have focused on two - level gain media . in order for salt to be a useful modeling tool , it is necessary to demonstrate that it can be applied to @xmath0-level lasers .
in the current work , we show analytically that the steady - state equations for an @xmath0-level laser can be reduced to those for an effective two - level system , and hence solved using the efficient salt algorithm with essentially the same degree of computational effort .
we also explore how this effective two - level system differs from the ordinary two - level laser .
next , we present a numerical comparison between the results of salt calculations and exact @xmath0-level finite - difference time - domain ( fdtd ) calculations , for the same simple 1d laser studied in @xcite , as well as for a 1d random laser .
we note that a similar comparison between salt and fdtd has been performed for a four - level high - q single - mode photonic crystal laser in ref .
@xcite , with good agreement found . here
, we test salt s accuracy in treating the more challenging case of multimode lasing in low - q and random cavities .
we illustrate the approach using the semi - classical laser equations @xcite for the four - level atomic gain medium shown in fig .
[ schem ] : @xmath2 the rwa has already been made , and we have assumed a lasing structure with one or two directions of translational symmetry , so that tm and te polarizations are conserved and maxwell s equations reduce to a scalar helmholtz equation @xcite . @xmath3 and @xmath4 are the positive frequency components of the scalar electric and polarization fields respectively .
@xmath5 is the population density of level @xmath6 , @xmath7 is the frequency of the gain center , @xmath8 is the gain width ( polarization dephasing rate ) , @xmath9 is the dipole matrix element , @xmath10 is the pump rate , and @xmath11 is the decay rate from level @xmath12 to level @xmath6 .
the four levels are labelled from @xmath13 in order of increasing energy ( fig . [ schem ] ) .
the polarization equation ( [ poleqn ] ) is obtained from the four - level density matrix equation of motion , assuming that only the level @xmath14 transition will be inverted and lase . often in fdtd calculations a real classical oscillating dipole equation
is used to describe the polarization @xcite ; in appendix a , we show that this yields essentially the same results , with an appropriate identification of parameters . in the rate equations ( [ rbeqn])-([reeqn ] )
, the pump is coherently acting between levels @xmath15 and @xmath16 .
, scaledwidth=70.0% ] the polarization equation ( [ poleqn ] ) incorporates the inversion @xmath17 , similar to the polarization equation for a two - level laser . by assuming that the non - lasing populations are stationary , @xmath18
, we can show that @xmath19 obeys @xmath20 which is precisely the form of the inversion equation for the two - level medium @xcite .
the parameters @xmath21 and @xmath22 serve as an effective equilibrium inversion and inversion relaxation rate respectively , and are given by @xcite : @xmath23 where @xmath24 and @xmath25 is the total density of gain atoms .
( [ eq9 ] ) for the inversion in the absence of laser emission ( which here acts as the effective pump parameter ) has been discussed by siegman @xcite , while eq .
( [ eq8 ] ) for the effective relaxation rate has been derived for a special case by khanin @xcite .
these expressions have not been used previously to solve the four - level lasing equations in terms of the two - level solutions , as we do here .
if we use an incoherent pump , the @xmath26 terms in ( [ rbeqn ] ) and ( [ reeqn ] ) would be replaced by @xmath27 ; the effect of this would be the removal of the factor of @xmath28 preceding the ratio @xmath29 in the denominators of ( [ eq8 ] ) and ( [ eq9 ] ) . for typical laser systems ,
each denominator is dominated by the @xmath30 term , so the difference between coherently and incoherently pumping the system is negligible @xcite . given the parameters @xmath31 describing the four - level medium of fig .
[ fig1 ] , we can calculate the effective pump and relaxation rates , and feed those ( together with the cavity dielectric function , lasing transition frequency @xmath7 , and gain linewidth @xmath32 ) into the salt algorithm . that will yield the steady - state lasing properties of this four - level laser . a general @xmath0-level laser
can be treated via an analogous procedure .
suppose we have a gain medium with an arbitrary number of levels , @xmath0 .
we assume that there is only a single lasing transition , between two levels which we denote by @xmath33 and @xmath34 .
the population of each of the @xmath35 non - lasing levels obeys @xmath36 where the sums are taken over all of the non - lasing states and @xmath11 is the rate at which atoms transition from state @xmath12 to @xmath6 which is either a decay or pump rate depending on the relative energies of the states . in this way
, we can incorporate decay and pump processes between any levels .
we again assume that @xmath37 for all non - lasing levels .
then @xmath38 \rho_{jj } = \gamma_{iu } \rho_{uu } + \gamma_{il}\rho_{ll } , \label{rhoii matrix}\ ] ] where @xmath39 and @xmath40 is the kronecker delta .
the term in brackets on the left hand side corresponds to an @xmath41 matrix , which we denote as @xmath42 . upon inverting eq .
( [ rhoii matrix ] ) and substituting it into the equations of motion for the lasing levels , we obtain @xmath43 where @xmath44_{ij } \
, \gamma_{j , u / l } \\ b_u & = & - s_u + \sum_{ij } ( \gamma_{ui } - \gamma_{li } ) [ r^{-1}]_{ij } \gamma_{ju } , \\ b_l & = & s_l + \sum_{ij } ( \gamma_{ui } - \gamma_{li } ) [ r^{-1}]_{ij } \gamma_{jl}.\end{aligned}\ ] ] the details of this calculation are given in appendix b.
returning to the typical four - level case , we take note of two important physical regimes .
the first , the linear regime , is @xmath45 , for which one recovers the expected behavior that the equilibrium inversion increases linearly with the pump and that @xmath22 is a constant : @xmath46 in this case , varying the equilibrium inversion and the pump strength are essentially equivalent .
the second regime of interest , the non - linear regime , is @xmath47 , i.e. when the slow decay rate between the lasing levels is on the same order as the pump rate . in this regime
, @xmath22 increases with increasing pump and @xmath21 saturates with increasing pump : @xmath48 this regime is also interesting from the viewpoint of salt . as @xmath22 increases with @xmath10 , a laser could satisfy the inequality @xmath49 near threshold , leading to stationary inversion and an accurate solution via salt , but fail to satisfy the inequality as the pump becomes stronger , leading to a decrease in the accuracy of salt . for a system with an arbitrary number of levels , the first regime always occurs at sufficiently small pump values ; the second regime is obtainable if electrons in the upper lasing level are relatively long - lived compared to electrons in other levels .
for completeness we briefly outline salt .
the @xmath50 and @xmath51 fields are assumed to obey a multi - mode ansatz @xmath52 where the indices @xmath53 label the different lasing modes , and the field and polarization are now explicitly scalar quantities .
the total number of modes , @xmath54 , is not given , but increases in unit steps from zero as we increase the pump strength @xmath55 .
the values of @xmath55 at which each step occurs are the ( interacting ) modal thresholds , to be determined self - consistently from the theory .
the real numbers @xmath56 are the lasing frequencies of the modes ( henceforth @xmath57 ) , which will also be determined self - consistently .
we insert the ansatz ( [ mode ansatz ] ) into the two - level laser equations , and employ the stationary inversion approximation ( sia ) @xmath58 .
the result is a set of coupled nonlinear differential equations , which are the fundamental equations of salt @xcite : @xmath59 \psi_\mu(\vec{r } ) = 0 , \label{tsg1 } \\
d(\vec{r } ) = d_0(\vec{r } ) \ , \left[1 + \sum_{\nu=1}^m \gamma_\nu |\psi_\nu(\vec{r})|^2\right]^{-1}. \label{tsg2}\end{aligned}\ ] ] @xmath60 and @xmath61 are now dimensionless , measured in their natural units @xmath62 and @xmath63 , and @xmath64 is the lorentzian gain curve evaluated at frequency @xmath65 . note that these equations are time - independent ; eq .
( [ tsg1 ] ) is a stationary wave equation for the electric field mode @xmath66 , with an effective dielectric function consisting of both the `` passive '' contribution @xmath67 and an `` active '' contribution from the gain medium .
the latter is frequency - dependent , and has both a real part and a negative ( amplifying ) imaginary part .
it also includes infinite - order nonlinear `` hole - burning '' modal interactions , seen in the @xmath68 dependence of ( [ tsg2 ] ) .
in addition , we make the key requirement that @xmath66 must be purely out - going outside the cavity ; it is this condition that makes the problem non - hermitian .
it is worth noting that the sia is not needed until at least two modes are above threshold , so ( [ tsg2 ] ) is exact for single - mode lasing up to and including the second threshold ( aside from the well - obeyed rwa ) .
these equations are solved efficiently by projecting them onto a complete biorthogonal set of purely outgoing states with external wavevectors , @xmath69 , equal to the lasing frequencies .
we refer to these states as the _ threshold constant flux _ ( tcf ) states , because one member of the basis set is always equal to the ( non - interacting ) threshold lasing mode , leading to very rapid convergence of the basis expansion above threshold @xcite .
the major computational effort in solving the salt equations by this approach is in calculating the ( linear ) tcf states .
the salt solutions are obtained for successive values of the pump increasing from the first threshold ; at each step , the coefficients of the lasing modes in the tcf basis are obtained using a standard non - linear solver , with the coefficients from the previous step as an initial guess ( which is never far from the correct solution ) .
unlike fdtd , in the current version of salt one can not simply directly solve at a fixed pump value , well above threshold .
however , even with this limitation , salt is much more efficient , and provides substantial physical insight , as we will discuss below .
to perform a well - controlled comparison between salt and the four - level laser equations ( [ waveqn])-([reeqn ] ) , as well as @xmath0-level generalizations , we studied 1d microcavity lasers for which the fdtd calculations are tractable and fast enough to generate extensive steady - state data .
we first consider the same simple edge emitting uniform - index laser treated in refs .
@xcite , with a perfect mirror at the origin , active region of length @xmath70 terminating abruptly in air ( see schematic , fig . [ fig1 ] ) .
the simulations were carried out using standard fdtd for the electromagnetic field , and crank - nicholson discretization for the polarization and rate equations based on the method of bidgaray @xcite ( in which the polarization and inversion are spatially aligned with the electric field but updated at the same time steps as the magnetic field ) .
the reported modal intensities are calculated by fourier transforming the electric field at the cavity boundary after the simulation has reached steady state ( see [ efficiency section ] for a discussion of the steady - state criterion ) .
the lasing transition frequency @xmath7 is chosen so that @xmath71 , corresponding to roughly ten wavelengths of radiation within the cavity .
physical quantities are reported in terms of their natural scales , @xmath72 and @xmath73 .
in addition , we take @xmath74 and measure rates in dimensionless units , i.e. @xmath75 .
we note that the parameters chosen accurately reflect those of real microcavities at optical frequencies @xcite ; the complete set of simulation parameters is given in appendix d. ( effective pump ) in a 1d microcavity edge emitting laser ( schematic inset ) .
the cavity is bounded on one side by a perfect mirror and on the other side by air , and has uniform refractive index @xmath76 .
solid lines are results obtained by the time - independent salt method ; open circles are results of fdtd simulations with a coherently pumped four - level medium ( fig . [ schem ] ) ; solid triangles are results of fdtd simulations with a coherently pumped six - level medium with a lasing transition between @xmath77 and @xmath78 .
simulation parameters are given in appendix d. both the four - level and six - level media are chosen to satisfy sia .
the dephasing rate is @xmath79 .
the four - level system is in the linear regime described by eq .
( [ l1])-([l2 ] ) .
the six - level system is calculated using the formula in the appendix b , but is in the non - linear regime described by eq .
( [ gp1])-([gp2 ] ) .
the spectra at @xmath80 , and the gain curve , are shown in the upper left inset .
[ fig1],scaledwidth=70.0% ] as shown in fig .
[ fig1 ] , we find close agreement between salt calculations and fdtd simulations . at a representative pump strength @xmath81
, the mode intensities produced by salt differ from those of the four- and six - level fdtd simulations by @xmath82 , while the frequencies differ by @xmath83 .
the difference in mode frequencies between salt and fdtd also exists at the first lasing threshold , for which an analytical value can be calculated .
there , we find that the fdtd simulation has a @xmath84 error in the first mode frequency , while salt has a @xmath85 error ; this error arises from the spatial discretization of the cavity employed in both approaches @xcite .
it is worth emphasizing that salt treats the non - linearity to infinite order ; in the earlier work on the maxwell - bloch model @xcite it was shown that for this same cavity the common cubic approximation for the non - linearity fails both quantitatively and qualitatively .
these results demonstrate that so long as the system satisfies sia , the mapping between systems with an arbitrary number of levels to an effective two - level system is nearly exact , and salt is able to very accurately determine the steady state properties of the cavity .
if two cavities , each with an arbitrary number of levels , have the same effective parameters @xmath21 and @xmath22 , and otherwise have the same polarization relaxation rate and atomic transition frequency , the cavities are equivalent from the electromagnetic point of view , and will have identical lasing properties .
the six - level simulations shown in fig . [ fig1 ] occupy the non - linear parameter regime of eq .
( [ gp1])-([gp2 ] ) , i.e. @xmath22 is a linear function and @xmath21 a non - linear function of @xmath10 .
however , the unscaled modal intensity leaving the cavity is still , to leading order , linear in @xmath10 .
this can be seen by rearranging ( [ tsg2 ] ) , inserting the expressions for @xmath22 and @xmath21 , and noting that at the end of the cavity the inversion is roughly independent of the pump strength .
this result is discussed further in appendix c. ( effective pump ) are shown for a 1d microcavity edge emitting laser with @xmath79 and @xmath76 .
solid lines again represent results obtained from salt , while open circles represent fdtd simulations of a simple four - level system with @xmath86 .
triangles represent fdtd simulations of a six - level system in the non - linear parameter regime in which @xmath87 for @xmath88 , and thus satisfying sia , but @xmath89 for @xmath90 , and consequently no longer satisfying sia .
[ fig2],scaledwidth=70.0% ] the mapping between the @xmath0-level laser and two - level salt breaks down at large pump strengths , when the condition @xmath49 is violated due to the increase of @xmath22 with @xmath10 . in ref .
@xcite , following an argument by haken @xcite , it was demonstrated that violating this condition causes the sia for the two - level model to break down .
this effect can be seen in the four - level laser data in fig .
[ fig2 ] , where @xmath91 and accuracy is already lost for the third lasing mode . for the six - level data of fig .
[ fig2 ] , which is in the non - linear parameter regime , sia is satisfied and salt agrees with the fdtd simulations for small values of the normalized equilibrium inversion ; for larger values of @xmath21 , the salt and fdtd results begin to diverge . finally , to demonstrate that the mapping to an effective two - level model works equally well for a complex laser cavity , fig .
[ fig3 ] shows a comparison between salt and fdtd simulations for a four - level gain medium in a 1d random dielectric structure .
a number of studies have been published on random lasers using such simulations @xcite ; salt provides a much more efficient method for such studies , which often require generating a statistical ensemble of lasers . here
, the passive cavity dielectric function contains @xmath92 layers , alternating randomly between regions with refractive indices @xmath93 and @xmath94 .
each random layer was generated according to the formula @xmath95 where @xmath96 and @xmath97 are the average thicknesses of the layers , @xmath98 represents the degree of randomness of the cavity , and @xmath99 $ ] is a randomly generated number .
the gain medium was added uniformly to the entire cavity , and the coherent pump was likewise uniform .
the transition frequency was chosen such that @xmath100 , corresponding to roughly @xmath101 wavelengths inside of the cavity .
we find only small discrepancies between the salt and fdtd results , with @xmath102 difference in the modal intensities .
these differences did not vary significantly between different realizations of the random laser .
( effective pump ) .
solid lines represent salt results , and circles represent fdtd simulations for a four - level system with @xmath103 .
the refractive index distribution of the edge emitting random laser is described in the text .
the gain medium has @xmath79 and is in the regime described by eq .
( [ l1])-([l2 ] ) .
left inset : log - log plot of the indicated region where three modes turn on in close proximity .
right inset : schematic of the cavity structure .
[ fig3],scaledwidth=70.0% ]
in this section we present a set of benchmarks comparing the computational efficiency of salt to fdtd .
salt calculations enjoy three main advantages over fdtd simulations of the semiclassical laser equations .
first and foremost , salt directly finds the steady - state solutions , so no time integration is involved , which substantially decreases computational effort .
second , salt unambiguously determines how many modes are lasing at a given pump , whereas it can be difficult to determine , especially for multimode lasing , when an fdtd simulation has reached the steady - state with all modes that will lase on " .
third , within salt , with minimal additional computational effort , it is possible to monitor modes which are _ below _ threshold via a modified threshold matrix @xcite , and hence to ascertain if more modes are likely to turn on in some interval of pump .
it is important to note that the implementation of salt used in this study has also not yet been fully optimized .
for instance , the implementation of salt used in this study requires calculating the entire lasing intensity and frequency spectrum starting from the first lasing threshold .
however , this is not necessary and it is possible to implement salt to take an initial guess of the number of modes and their relative intensities and then allow the algorithm to flow to the correct solution as the salt algorithm has been demonstrated to be rather robust @xcite .
thus , while one might assume from fig .
[ rtfig ] that there would be a crossover pump value , high above threshold , where it would be more efficient to calculate the steady - state solutions using fdtd , this is not necessarily the case
. salt does have one disadvantage that fdtd does not , assuming the cavity is not in the chaotic regime .
the convergence time in fdtd is determined by the longest time scale in the problem , which is the greater of beating period between consecutive modes and the relaxation oscillation time .
these time scales are relatively independent of the number of modes lasing in the cavity , so the efficiency is largely independent of the pump in the multimode regime .
this is not the case for salt , as the computational time increases as @xmath104 where @xmath0 is the number of lasing modes .
however , as we see in fig .
[ rtfig ] , even a non - optimal implementation of salt is substantially more efficient than fdtd even when calculating the steady state of a single pump value . .
fdtd simulations that have not begun to lase are marked as crosses .
plot ( a ) shows data for @xmath105 , just above the first lasing threshold .
salt determined the steady - state single modal intensity in under three minutes , while the fdtd required @xmath106 minutes to reach steady state .
plot ( b ) shows data for @xmath107 , well above the third lasing threshold .
salt calculated all data up to and including this pump value in under 90 minutes , whereas fdtd required @xmath108 minutes for the first two modes to reach steady - state , with the third mode intensity ( green circles ) still fluctuating after @xmath109 minutes ( not shown ) .
[ rtfig],scaledwidth=100.0% ] calculating full modal intensity / frequency curves as a function of the pump strength , such as in fig .
[ fig1 ] , is generally much more efficient using salt .
for example , in order to generate the curves seen in fig .
[ fig1 ] , salt ran for a little under 2 processor hours . to generate
all of the fdtd data for the four level simulations took 267 processor days .
if one is attempting to explore a large parameter space of designs or system parameters , salt may make studies feasible which are simply impractical using fdtd , particularly in more realistic 2d and 3d structures . as mentioned before
, the bulk of the computational effort required for the salt algorithm , especially in higher dimensions , is in solving for the tcf states . while the difficulty of solving for the tcf states does scale with the dimensionality of the system , one only need solve the associated generalized eigenvalue problem @xmath110 ( even in higher dimensions ) to have a sufficiently complete basis for all pump values ,
whereas in fdtd one needs to solve an @xmath111 problem at each of many thousands , if not millions , of time steps .
furthermore , it is likely that switching to a finite element method in higher dimensions is not only possible , but likely to result in increased computational efficiency .
once one has a tcf basis library , using the salt algorithm to iterate above threshold does not directly scale with the dimensionality of the system .
finally , while current implementations of salt assume the electric field is perpendicular to the direction of wave propagation , the vectorial generalization of salt is under investigation .
we have found that using stationarity conditions on the non - lasing level populations , the rate equations for an @xmath0-level laser can be mapped to an effective two - level model , for which the steady - state multimode lasing properties are efficiently solvable using steady - state ab initio laser theory ( salt ) . using this mapping
, we found excellent agreement between salt and fdtd simulations for the modal frequencies , thresholds and above - threshold intensities , in the expected domain of validity .
salt is typically several orders of magnitude more efficient computationally than time domain solution of the laser - rate equations , assuming only steady - state properties are needed .
we thank robert j. tandy , hui cao , and peter bermel for helpful discussions .
this work was partially supported by nsf grant no . dmr-0908437 .
throughout this paper , we have used the density matrix equations of motion for the polarization field . however , much of the literature uses the classical oscillating dipole equation , in which the gain atoms are assumed to be dipoles undergoing harmonic oscillations and the quantum density matrix is neglected .
this appendix briefly demonstrates that the two models are equivalent , and derives the relevant parameter redefinitions .
a more thorough discussion has been given by boyd @xcite .
the polarization equation used here , @xmath112 is derived from the density matrix equation @xmath113 where @xmath114 and @xmath115 are the upper and lower lasing states , @xmath116 is the density matrix element @xmath117 and @xmath118 is the coupling constant of the lasing states to the electric field , which allows for the definition @xmath119 .
alternatively , following the notation of boyd @xcite , we could consider the equation of motion for @xmath120 , which is the expectation value of the dipole moment induced by the applied field , i.e. the classical oscillating dipole field .
thus @xmath121 using ( [ dme ] ) , and @xmath122 this can be rewritten as @xmath123 where @xmath124 is the inversion density of the lasing states . for @xmath125 , we can discard the term @xmath126 , resulting in the traditional form of the classical oscillating dipole polarization field equation .
this gives the classical coupling constant @xmath127 @xcite .
a similar analysis can used to show that the inversion equation can be rewritten as @xmath128 which demonstrates how the change in the inversion is dependent upon the classical polarization field .
therefore , so long as @xmath125 , a condition that is usually satisfied , these two formulations of the polarization dynamics are equivalent .
this appendix derives the effective equilibrium inversion and relaxation rate for a gain medium with an arbitrary number of levels .
we allow decays between any two levels , even if they are not adjacent .
we assume there is a single lasing transition , between levels @xmath33 and @xmath34 , which need not be adjacent .
the rate equation for an arbitrary _ non - lasing _ level in the system is @xmath129 where the summations are taken over all non - lasing levels .
here we do not distinguish between decay rates and pumping rates ; @xmath11 is simply interpreted as the rate at which level @xmath130 transitions into level @xmath6 , regardless of the energies of those states . if the populations of all the non - lasing transitions are stationary , i.e. @xmath131 , then we can rewrite ( [ non lasing rhodot ] ) as @xmath132 here , @xmath133 and @xmath40 is the kronecker delta .
inverting ( [ genrate ] ) gives @xmath134_{ij } ( \gamma_{ju } \rho_{uu } + \gamma_{jl } \rho_{ll } ) .
\label{rhoii}\ ] ] hence , we can express the total number density of gain atoms as @xmath135 where @xmath136_{ij } \gamma_{ju } , \\
t_l & = & 1 + \sum_{ij } [ r^{-1}]_{ij } \gamma_{jl}.\end{aligned}\ ] ] noting that @xmath124 , we can write the populations of the lasing states as @xmath137 from the equations of motion for the lasing levels , we have the inversion equation @xmath138 where @xmath139 and @xmath140 is defined similarly . inserting ( [ rhoii ] ) into this equation
gives @xmath141 where @xmath142_{ij } \gamma_{ju } , \\
b_l & \equiv & s_l + \sum_{ij } ( \gamma_{ui } - \gamma_{li } ) [ r^{-1}]_{ij } \gamma_{jl}.\end{aligned}\ ] ] plugging in ( [ rhouu ] ) and ( [ rholl ] ) now yields the inversion equation in the desired form ( [ efinv ] ) , with @xmath143
in this appendix we discuss the surprising result that the unscaled modal intensities of the six - level simulations discussed in fig .
[ fig1 ] as a function of the pump are approximately linear , as shown in fig .
[ intvpump ] , even though these six - level simulations are in the non - linear parameter regime .
in simple treatments of lasers @xcite the inversion is assumed to be clamped after the first lasing threshold
. however , a more detailed analysis lead to the inclusion of spatial hole - burning effects which forces the inversion in the cavity to change beyond the first lasing threshold in a non - linear manner .
this result then , that the unscaled modal intensity is linear in the pump rate , can be understood from the observation that at the end of the cavity , @xmath144 , all of the lasing modes have a maximum in their fields , and thus the inversion is effectively clamped at this point beyond the first lasing threshold , which can be seen in fig .
[ intvpump ] ( b ) . to understand this formally ,
we begin by rewriting ( [ tsg2 ] ) and removing the scaling factors @xmath145 and @xmath146 gives @xmath147 substituting in ( [ gp1 ] ) and ( [ gp2 ] ) , which are valid for this simulation , gives @xmath148 the inversion @xmath61 is a function of both position and the pump .
however , for @xmath149 corresponding to the cavity edge , @xmath61 should be mostly independent of the pump , as at this location every mode is at its maximum intensity and the effect of spatial hole - burning is most pronounced .
the fdtd simulation results , shown in fig .
[ intvpump ] , demonstrate that @xmath61 indeed varies very weakly with @xmath10 at the cavity edge . as a function of the pump . a cross sectional area of @xmath150
is assumed to calculate the power .
( b ) inversion as a function of the pump at the cavity boundary .
dashed lines in plots a and b correspond to the pump values shown in plot c. ( c ) inversion as a function of position within the cavity for three different pump values , cyan corresponds with @xmath151 , magenta with @xmath152 , and orange with @xmath153 to show the evolution of the inversion within the cavity as a function of the pump strength .
[ intvpump],scaledwidth=100.0% ]
in this appendix we list the parameters used in each of the fdtd simulations described above .
these constants are matrices , with @xmath11 denoting the decay rate from @xmath130 to @xmath154 .
these values are given in their dimensionless form , i.e. @xmath75 .
unlisted entries are zero .
we also note that throughout this paper @xmath155 denotes the ground state , so these matrices are @xmath156 indexed . for the four - level simulations in fig .
[ fig1 ] , @xmath157 the dipole matrix element is @xmath158 , and the number of gain atoms is @xmath159 .
the pump @xmath10 was varied between @xmath160 and @xmath161 .
thus , for an optical wavelength of @xmath162 , the requirement in fig .
[ fig1 ] that @xmath163 means that @xmath164 . using this length ,
the decay rates can be converted to their unit - full values as @xmath165 , @xmath166 , @xmath167 , and the pump at threshold is @xmath168 .
similarly , the dipole matrix element also acquires units of inverse time , and can be expressed as @xmath169 , which corresponds to a coupling constant in the classical oscillating dipole picture of @xmath170 . these constants can be seen to be similar to those used in other studies of optical microcavities @xcite . for the six - level simulations in fig .
[ fig1 ] , @xmath171 furthermore , @xmath172 , and the lasing transition is between levels @xmath77 and @xmath78 ( where the ground state is again @xmath173 and the states are numbered in order of increasing energy ) . | we show that steady - state ab initio laser theory ( salt ) can be applied to find the stationary multimode lasing properties of an @xmath0-level laser .
this is achieved by mapping the @xmath0-level rate equations to an effective two - level model of the type solved by the salt algorithm .
this mapping yields excellent agreement with more computationally demanding @xmath0-level time domain solutions for the steady state .
99 h. haken , _ light : laser dynamics _ vol . 2 ( north - holland phys .
publishing , new york , 1985 ) .
a. e. siegman , _ lasers _ ( university science books , mill valley - california , 1986 ) .
a. s. nagra and r. a. york , `` fdtd analysis of wave propagation in nonlinear absorbing and gain media , '' ieee trans .
antennas propag .
* 46 * , 334 - 340 ( 1998 ) . k. s. yee , `` numerical solution of the initial boundary value problems involving maxwell s equations in isotropic media , '' ieee trans antennas propag .
* 14 * , 302 - 307 ( 1966 ) .
h. e. treci , a. d. stone , and b. collier , `` self - consistent multimode lasing theory for complex or random lasing media , '' phys .
rev .
a * 74 * , 043822 ( 2006 ) .
h. e. treci , a. d. stone , and l. ge , `` theory of the spatial structure of nonlinear lasing modes , '' phys . rev .
a * 76 * , 013813 ( 2007 ) .
h. e. treci , l. ge , s. rotter , and a. d. stone , `` strong interactions in multimode random lasers , '' science * 320 * , 643 - 646 ( 2008 ) .
l. ge , y. d. chong , and a. d. stone , `` steady - state ab initio laser theory : generalizations and analytic results , '' phys . rev .
a * 82 * , 063824 ( 2010 ) .
h. cao , `` review on the latest developments in random lasers with coherent feedback , '' j. phys .
a * 38 * , 10497 - 10535 ( 2005 ) .
o. painter , r. k. lee , a. scherer , a. yariv , j. d. obrien , p. d. dapkus , and i. kim , `` two - dimensional photonic band - gap defect mode laser , '' science * 284 * , 1819 - 1821 ( 1999 ) .
s. chua , y. d. chong , a. d. stone , m soljaci , and j. bravo - abad , `` low - threshold lasing action in photonic crystal slabs enabled by fano resonances , '' opt .
express * 19 * , 1539 ( 2011 ) . c. gmachl , f. capasso , e. e. narimanov , j. u. nckel , a. d. stone , j. faist , d. l. sivco , and a. y. cho , `` high - power directional emission from microlasers with chaotic resonators , '' science * 280 * , 1556 - 1564 ( 1998 ) .
li ge , yale phd thesis , 2010 .
l. ge , r. j. tandy , a. d. stone , and h. e. treci , `` quantitative verification of ab initio self - consistent laser theory , '' opt .
express * 16 * , 16895 ( 2008 ) .
the equations are written for the tm case , the modifications for te are straightforward .
the observation that coherent and incoherent pumping are nearly equivalent for most systems , is invalid when the coherent pumping is supplied at a similar frequency to the atomic lasing transition and thus interactions between the lasing field and pumping field must be taken into account .
h. fu and h. haken , `` multifrequency operations in a short - cavity standing - wave laser , '' phys .
rev .
a * 43 * , 2446 - 2454 ( 1991 ) .
y. i. khanin , _ principles of laser dynamics _ ( elsevier , amsterdam , 1995 )
. b. bidgaray , `` time discretizations for maxwell - bloch equations , '' numer .
meth .
partial differential equations * 19 * , 284 - 300 ( 2003 ) . for any 1d cavity which is uniformly pumped the tcf states for solving salt can also be found using a transfer matrix method which does not require discretizing space .
we use a more general tcf solver in the calculations presented here which does discretize space .
x. jiang and c. m. soukoulis , `` time dependent theory for random lasers , '' phys .
rev .
lett .
* 85 * , 70 ( 2000 ) .
x. jiang , s. feng , c. m. soukoulis , j. zi , j. d. joannopoulos , and h. cao , `` coupling , competition , and stability of modes in random lasers , '' phys
. rev .
b * 69 * , 104202 ( 2004 ) .
r. w. boyd , _ nonlinear optics _ ( academic press , new york , 2008 ) . |
the cp symmetry is , without doubt , one of the fundamental symmetries in nature .
its breaking still carries a cloud of mystery in particle physics and cosmology .
indeed , cp symmetry is intimately related to theories of interactions between elementary particles and represents a cornerstone in constructing grand unified and supersymmetric models .
it is also necessary to explain the matter - antimatter asymmetry observed in universe .
the first experimental evidence of cp violation was discovered in the @xmath3 mixing and kaon decays @xcite . according to the cpt theorem , cp violation implies t violation .
the latter is tested through the measurement of the neutron electric dipole moment ( nedm)@xmath0 .
the upper experimental limit gives confidence that the nedm can be another manifestation of cp breaking . to investigate the cp violation phenomenon
many theoretical models were proposed . in the standard model of electroweak interactions
, cp violation is parametrized by a single phase in the cabbibo kobayashi maskawa ( ckm ) quark mixing matrix @xcite .
other models exhibiting a cp violation are given by extensions of the standard model ; among them , the minimal supersymmetric standard model @xmath4 includes in general soft complex parameters which provide new additional sources of cp violation @xcite .
cp violation can be also investigated in the strong interactions context through qcd framework .
in fact , the qcd effective lagrangian contains an additional cp - odd four dimensional operator embedded in the following topological term : @xmath5 where @xmath6 is the gluonic field strength , @xmath7 is its dual and @xmath8 is the strong coupling constant .
the @xmath9 quantity is a total derivative , consequently it can contribute to the physical observables only through non perturbative effects .
the nedm is related to the @xmath10-angle by the following relation : @xmath11 and consequently , according to the experimental measurements @xmath12 @xcite , the @xmath13 parameter must be less than @xmath14 @xcite .
the well known strong cp problem consists in explaining the smallness of @xmath1 . in this regard , several scenarios were suggested .
the most known one was proposed by peccei and quinn @xcite and consists in implementing an extra @xmath15 symmetry which permits a dynamical suppression of the undesired @xmath16-term .
this is possible due to the fact that the axial current @xmath17 is related to the gluonic field strength through the following relation @xmath18 .
the breakdown of the @xmath15 symmetry gives arise to a very light pseudogoldstone boson called axion .
this particle may well be important to the puzzle of dark matter and might constitute the missing mass of the universe @xcite . motivated by : ( a ) the direct relation between the @xmath19-angle and nedm @xmath20 , as it was demonstrated firstly in @xcite via the chiral perturbation theory and recently in @xcite within qcd sum rules formalism ; ( b ) the possibility to restore some broken symmetries by increasing the temperature ; we shall use the qcd sum rules at @xmath21 @xcite to derive thermal dependence of the ratio @xmath22 .
then we study its thermal behaviour at low temperatures and discuss the consequences of temperature effects on the restoration of the broken cp symmetry .
this paper is organized as follows : section 2 is devoted to the calculations of the nedm induced by the @xmath23 parameter from qcd sum rules . in section 3
, we show how one introduces temperature in qcd sum rules calculations .
we end this paper with a discussion and qualitative analysis of the thermal effects on the cp symmetry .
in the two later decades , qcd sum rules la svz @xcite were applied successfully to the investigation of hadronic properties at low energies . in order to derive the nedm through this approach
, many calculations were performed in the literature @xcite .
one of them , which turns out to be more practical for our study , has been obtained recently in @xcite .
it consists in considering a lagrangian containing the following p and cp violating operators : @xmath24 @xmath25 and @xmath26 are respectively two angles coming from the chiral and the topological terms and @xmath27 is the quark reduced mass given by @xmath27=@xmath28 .
the authors of @xcite start from the two points correlation function in qcd background with a nonvanishing @xmath26 and in the presence of a constant external electomagnetic field @xmath29 : @xmath30 @xmath31 is the interpolating current which in the case of the neutron reads as @xcite : @xmath32 where @xmath33 is a mixing parameter . using the operator product expansion ( ope ) ,
they have first performed the calculation of @xmath34 as a function of matrix elements and wilson coefficients and then have confronted the qcd expression of @xmath34 to its phenomenological parametrisation .
@xmath34 can be expanded in terms of the electromagnetic charge as@xcite : @xmath35 the first term @xmath36 is the nucleon propagator which include only the cp - even parameters @xcite , while the second term @xmath37 is the polarization tensor which may be expanded through wilson ope as : @xmath38 , where @xmath39 is an arbitrary lorentz structure and @xmath40 are the wilson coefficient functions calculable in perturbation theory @xcite . from this expansion ,
one keeps only the cp - odd contribution piece . by considering the anomalous axial current
, one obtains the following @xmath26 dependence of @xmath41 matrix elements @xcite : @xmath42 where @xmath43 and @xmath27 are respectively the quark and reduced masses .
the electromagnetic dependence of these matrix elements can be parametrized through the implementation of the @xmath44 , @xmath45 and @xmath46 susceptibilities defined as @xcite : + @xmath47 putting altogether the above ingredients and after a straightforward calculation @xcite , the following expression of @xmath37 for the neutron is derived : + @xmath48,\end{aligned}\ ] ] where @xmath49 is the physical phase and @xmath50 .
+ the qcd expression ( 2.7 ) will be confronted to the phenomenological parametrisation @xmath51@xmath52 written in terms of the neutron hadronic properties .
the latter is given by : + @xmath53 where @xmath54 is the neutron mass , @xmath55 is the quark charge .
a and @xmath56 , which originate from the phenomenological side of the sum rule , represent respectively a constant of dimension 2 and the neutron coupling constant to the interpolating current @xmath31 .
this coupling is defined via a spinor @xmath57 as @xmath58 .
the introduction of finite temperature effects may provide more precision to the phenomenological values of hadronic observables . within the framework of qcd sum rules ,
the t - evolution of the correlation functions appear as a thermal average of the local operators in the wilson expansion@xcite .
hence , at nonzero temperature and in the approximation of the non interacting gas of bosons ( pions ) , the vacuum condensates can be written as : @xmath59 where @xmath60 , @xmath61is the bose - einstein distribution and @xmath62 is the standard vacuum condensate ( i.e. at t=0 ) . in the low temperature region ,
the effects of heavier resonances @xmath63 can be neglected due to their distibution functions @xmath64@xcite . to compute the pion matrix elements ,
we apply the soft pion theorem given by : @xmath65|0>+ o({m^2_\pi \over \lambda^2}),\ ] ] where @xmath66 is a hadron scale and @xmath67 is the isovector axial charge : @xmath68 direct application of the above formula to the quark and gluon condensates shows that @xcite : + ( i ) only @xmath69 is sensitive to temperature .
its behaviour at finite t is given by : @xmath70 where @xmath71 with @xmath72 and @xmath73 is the pion decay constant ( @xmath74 ) .
the variation with temperature of the quark condensate @xmath75 results in two different asymptotic behaviours , namely : + @xmath76 for @xmath77 , and @xmath78 for @xmath79 .
+ ( ii ) the gluon condensate is nearly constant at low temperature and a t dependence occurs only at order @xmath80 . as usual , the determination of the ratio @xmath81 sum rules at non zero temperature is now easily performed through two steps . in the first step ,
we apply borel operator to both expressions of the neutron correlation function shown in eqs .
( 2.7 ) and ( 2.8 ) , where finite temperature effects were introduced as discussed above .
next step , by invoking the quark - hadron duality principle , we deduce the following relation of the @xmath82 induced nedm : @xmath83e^{m_n^2 \over m^2},\ ] ] where m represents the borel parameter . note that we have neglected the single pole contribution entering via the constant a , as suggested in @xcite .
+ the expression ( 3.5 ) is derived with @xmath84 which is more appropriate for us since it suppresses the infrared divergences . in fact , the ioffe choice @xmath85 which is rather more useful for the cp even case , removes the leading order contribution in the sum rules ( 2.7 ) . the coupling constant @xmath86 and the neutron mass @xmath87 which appear in ( 3.5 )
were determined from the thermal qcd sum rules . for the former
, we consider the @xmath88 sum rules in @xcite with @xmath84 and then we extract the following explicit expression of @xmath86 : @xmath89 within the pion gas approximation , eletsky has demonstrated in @xcite that inclusion of the contribution coming from the pion - nucleon scattering in the nucleon sum rules is mandatory .
the latter enters eq.(3.6 ) through the coupling constant @xmath90 , whose values lie within the range 13.5 - 14.3 @xcite . + numerical analysis
is performed with the following input parameters : the borel mass has been chosen within the values @xmath91 which correspond to the optimal range ( borel window ) in the @xmath92 sum rule at @xmath93 @xcite . for the @xmath94 and @xmath46 susceptibilities
we take @xmath95 @xcite and @xmath96 @xcite . as to the vacuum condensates appearing in ( 3.5 ) , we fix @xmath69 and @xmath97 to their standard values @xcite .
in the two above sections , we have established the relation between the nedm and @xmath98 angle at non zero temperature from qcd sum rules .
since the ratio @xmath99 is expressed in terms of the pion parameters @xmath73 , @xmath100 and of @xmath90 , we briefly recall the main features of their thermal behaviour .
various studies performed either within the framework of the chiral perturbation theory and/or qcd sum rules at low temperature have shown the following features : + ( i ) the existence of qcd phase transition temperature @xmath101 which signals both qcd deconfinement and chiral symmetry restoration @xcite .
+ ( ii ) @xmath73 and @xmath90 have very small variation with temperature up to @xmath101 .
so , we shall assume them as constants below @xmath101 .
however , they vanish if the temperature passes through the critical value @xmath101 @xcite .
+ ( iii ) the thermal mass shift of the neutron and the pion is absent at order @xmath102 @xcite .
@xmath103 shows up only at the next order @xmath104 , but its value is negligible @xcite . + by taking into account the above properties , we plot the ratio defined in eq . ( 3.5 ) as a function of t. from the figure
, we learn that the ratio @xmath105 survives at finite temperature and it decreases smoothly with t ( about 16@xmath106 variation for temperature values up to 200 mev ) .
this means that either the nedm value decreases or @xmath1 increases .
consequently , for a fixed value of @xmath1 the nedm decreases but it does not exhibit any critical behaviour . furthermore ,
if we start from a non vanishing @xmath98 value at @xmath93 , it is not possible to remove it at finite temperature .
we also note that @xmath107 grows as @xmath108 or @xmath94 susceptibility increases
. it also grows with quark condensate rising .
however this ratio is insensitive to both the @xmath46 susceptibility and the coupling constant @xmath90 .
we notice that for higher temperatures , the curve @xmath109 exhibits a brutal increase justified by the fact that for temperatures beyond the critical value @xmath101 , at which the chiral symmetry is restored , the constants @xmath73 and @xmath90 become zero and consequently from eq(3.5 ) the ratio @xmath110 behaves as a non vanishing constant .
the large difference between the values of the ratio for @xmath111 and @xmath112 maybe a consequence of the fact that we have neglected other contributions to the the spectral function , like the scattering process @xmath113 .
these contributions , which are of the order @xmath104 , are negligible in the low temperture region but become substantial for @xmath114 . moreover ,
this difference may also originate from the use of soft pion approximation which is valid essentially for low @xmath115 ( @xmath116 ) .
therefore it is clear from this qualitative analysis , which is based on the soft pion approximation , that the temperature does not play a fundamental role in the suppression of the undesired @xmath26-term and hence the broken cp symmetry is not restored , as expected .
this is not strange , in fact it was shown that more heat does not imply automatically more symmetry @xcite .
moreover , some exact symmetries can be broken by increasing temperature @xcite .
the symmetry non restoration phenomenon , which means that a broken symmetry at t=0 remains broken even at high temperature , is essential for discrete symmetries , cp symmetry in particular . indeed , the symmetry non restoration allows us to avoid wall domains inherited after the phase transition @xcite and to explain the baryogenesis phenomenon in cosmology @xcite .
furthermore , it can be very useful for solving the monopole problem in grand unified theories @xcite .
* acknowledgments * we are deeply grateful to t. lhallabi and e. h. saidi for their encouragements and stimulating remarks
. n. b. would like to thank the abdus salam ictp for hospitality and prof .
goran senjanovic for very useful discussions .
+ this work is supported by the program pars - phys 27.372/98 cnr and the convention cnprst / iccti 340.00
. 99 j. h. christensen , j. w. cronin , v. l. fitch and r. turlay , phys .
. lett . * 13*(1964)138 .
n. cabibbo , phys .
lett . * 10*(1963 ) 531 ; + m. kobayashi and t. maskawa , prog .
* 49*(1973 ) 652 .
d. a. demir , a. masiero , and o. vives , phys . rev . *
d61*(2000 ) 075009 .
i. bigi and n. g. uraltsev , sov .
jetp * 73*(2)(1991 ) 198 ; + i. bigi , surveys in high energy physics , * 12*(1998 ) 269 .
v. baluni , phys .
rev * d19*(1979)2227 .
r. crewther , p. di vecchia , g. veneziano , e. witten , phys .
lett * b88*(1979)123 .
r. m. barnett and al , phys .
d54*(1996 ) 1 .
r. d. peccei , hep - ph/9807516 .
r. golub and s. k. lamoreaux , hep - ph/9907282 .
r. d. peccei and h.r .
quinn , phys .
d16*(1977 ) 1791 .
g. lazarides and q. shafi , `` monopoles , axions and intermediate mass dark matter '' , hep - ph/0006202 .
m. pospelov and a. ritz , nucl . phys . * b558*(1999 ) 243 .
m. pospelov and a. ritz , phys .
lett . * 83*(1999 ) 2526 . m. a. shifman , a. i. vainshtein and v. i. zakharov , nucl . phys . *
b147*(1979 ) 385 . v.m .
khatsimovsky , i.b .
khriplovich and a.s .
yelkhovsky , ann . phys . * 186*(1988)1 ; + c. t. chan , e. m. henly and t. meissner , `` nucleon electric dipole moments from qcd sum rules '' , hep - ph/9905317 .
b. l. ioffe , nucl . phys .
* b188 * ( 1981 ) 317 ; + y. chung , h. g. dosch , m. kremer and d. schall , phys .
lett.*b102*(1981)175 ; nucl . phys .
* b197*(1982)55 m. a. shifman , a. i. vainshtein and v. i. zakharov , nucl . phys .
* b166*(1980 ) 493 .
ioffe and a. v. smilga , nucl . phys .
* b232 * ( 1984 ) 109 .
m. a. shifman , a. i. vainshtein and v. i. zakharov , nucl .
phys . * b166*(1980 ) 493 .
a. i. bochkarev and m. e. shaposhnikov , nucl
. phys . * b268*(1986)220 .
r. barducci , r. casalbuoni , s. de curtis , r. gatto and g. pettini , phys . lett . *
b244*(1990 ) 311 . j. gasser and h. leutwyler , phys . lett . * b184*(1987 ) 83 ; + h. leutwyler , in the proceedings of qcd 20 years later , achen,1992 , ed .
p.m. zerwas and h.a .
kastrup ( world scientific , singapore , 1993 ) .
y. koike , phys .
rev * d48 * ( 1993 ) 2313 ; + c. adami and i. zahed , phys.rev * d45*(1992 ) 4312 ; + t. hatsuda , y. koike and s.h .
lee , nucl .
* b394 * ( 1993)221 .
s. mallik and k. mukherjee , phys .
rev * d58 * ( 1998 ) 096011 ; + s. mallik , phys . lett . * b416*(1998 ) . h.g .
doch , m. jamin and s. narison , phys . lett . * b220*(1989 ) 251 ; + m. jamin , z. phys . * c37*(1988)625
. v. l. eletsky , phys . lett . * b245*(1990 ) 229 ; phys . lett . *
b352*(1995 ) 440 .
proceeding of the workshop `` a critical issue in the determination of the pion nucleon decay constant '' , ed .
jan blomgten , phys .
scripta * t87 * ( 2000)53 .
v. m. belyaev and y. i. kogan , sov . j. nucl
. phys . * 40*(1984 ) 659 . i. i. kogan and d. wyler ,
* b274*(1992 ) 100 . c. a. dominguez , c. van gend and m. loewe , phys
. lett . * b429*(1998 ) 64 ; + v.l .
eletky and i.i .
kogan , phys .
rev * d49 * ( 1994)3083 .
s. gupta , hep - lat/0001011 ; a. ali khan et al .
, hep - lat/0008011 . s. weinberg , phys .
rev * d9*(1974 ) 3357 .
r. n. mohapatra and g. senjanovic , phys .
* d20 * ( 1979 ) 3390 ; + g. dvali , a. melfo , g. senjanovic , phys.rev . * d54 * ( 1996)7857 and references therein . ya .
b. zeldovich , i. yu . kobzarev and l. b. okun , jetp . *
40*(1974 ) 1 ; + t. w. kibble , j. phys . * a9*(1976 ) 1987 , phys .
rep . * 67*(1980 ) 183 .
a. sakharov , jetp lett .
* 5 * ( 1967 ) 24 .
g. dvali , a. melfo and g. senjanovic , phys .
lett . * 75*(1995 ) 4559 .
figure : temperature dependence of the ratio @xmath117 | we use the method of thermal qcd sum rules to investigate the effects of temperature on the neutron electric dipole moment @xmath0 induced by the vacuum @xmath1-angle
. then , we analyze and discuss the thermal behaviour of the ratio @xmath2 in connection with the restoration of the cp - invariance at finite temperature . 22.5cm16.8cm-.4cm-.9 cm = 6pt plus 2pt minus 1pt addtoresetequationsection = 18.6pt plus 0.2pt minus 0.1pt addtoresetequationsection |
berry introduced the term quantum chaology in his 1987 bakerian lecture @xcite as the study of semiclassical but non - classical behaviour of systems whose classical motion exhibits chaos .
he illustrated it with the statistics of energy levels , following his earlier work with tabor @xcite and related developments from the study of a range of systems .
in the regular spectrum of a bound system with @xmath0 degrees of freedom and @xmath1 constants of motion , the energy levels are labelled by @xmath1 quantum numbers , but the quantum numbers of nearby energy levels may be very different . in the case of an irregular spectrum , such as for an ergodic system where only energy is conserved , we can not use quantum number labelling .
this prompted the use of energy level spacing distributions to allow comparisons among different spectra @xcite .
it was known , eg from the work of porter @xcite , that the spacings between energy levels of complex nuclei and atoms with @xmath1 large are modelled by the spacings of eigenvalues of random matrices and that the wigner distribution @xcite gives a very good fit .
it turns out that the spacing distributions for generic regular systems are negative exponential , that is random ; but for irregular systems the distributions are skew and unimodal , at the scale of the mean spacing .
mehta @xcite provides a detailed discussion of the numerical experiments and functional approximations to the energy level spacing statistics , alt et al @xcite compare eigenvalues from numerical analysis and from microwave resonator experiments , also eg .
bohigas et al @xcite and soshnikov @xcite confirm certain universality properties .
also , miller @xcite provides much detail on a range of related number theoretic properties , including random matrix theory links with l - functions .
forrester s online book @xcite gives a wealth of analytic detail on the mathematics and physics of eigenvalues of infinite random matrices for the three ensembles of particular interest : gaussian orthogonal ( goe ) , unitary ( gue ) and symplectic ( gse ) , being the real , complex and quaternionic cases , respectively .
the review by deift @xcite illustrates how random matrix theory has significant links to a wide range of mathematical problems in the theory of functions as well as to mathematical physics . from @xcite
, we have definitions for the three cases of interest : goe : : : a random real symmetric @xmath2 matrix belongs to the gaussian orthogonal ensemble ( goe ) if the diagonal and upper triangular elements are independent random variables with gaussian distributions of zero mean and standard deviation @xmath3 for the diagonal and @xmath4 for the upper triangular elements .
gue : : : a random hermitian @xmath2 matrix belongs to the gaussian unitary ensemble ( gue ) if the diagonal elements @xmath5 ( which must be real ) and the upper triangular elements @xmath6 are independent random variables with gaussian distributions of zero mean and standard deviation @xmath4 for the @xmath5 and @xmath7 for each of the @xmath8 and @xmath9 gse : : : a random hermitian @xmath2 matrix with real quaternionic elements belongs to the gaussian symplectic ensemble ( gse ) if the diagonal elements @xmath10 ( which must be real ) are independent with gaussian distribution of zero mean and standard deviation @xmath7 and the upper triangular elements @xmath11 and @xmath12 are independent random variables with gaussian distributions of zero mean and standard deviation @xmath13 for each of the @xmath14 and @xmath9 then the matrices in these ensembles are respectively invariant under the appropriate orthogonal , unitary and symmetric transformation groups , and moreover in each case the joint density function of all independent elements is controlled by the trace of the matrices and is of form @xcite @xmath15 where @xmath16 is a normalizing factor .
barndorff - nielsen et al @xcite give some background mathematical statistics on the more general problem of quantum information and quantum statistical inference , including reference to random matrices .
here we show that gamma distributions provide approximations to eigenvalue spacing distributions for the goe distribution comparable to the wigner distribution at the scale of the mean and for the gue and gse distributions , except near the origin
. that may be useful because the gamma distribution has a well - understood and tractable information geometry @xcite as well as the following important uniqueness property : [ hwangthm ] for independent positive random variables with a common probability density function @xmath17 having independence of the sample mean and the sample coefficient of variation is equivalent to @xmath18 being the gamma distribution .
it is noteworthy also that the non - chaotic case has an exponential distribution of spacings between energy levels and that the sum of @xmath1 independent identical exponential random variables follows a gamma distribution and moreover the sum of @xmath1 independent identical gamma random variables follows a gamma distribution ; furthermore , the product of gamma distributions is well - approximated by a gamma distribution . from a different standpoint ,
berry and robnik @xcite gave a statistical model using a mixture of energy level spacing sequences from exponential and wigner distributions .
monte carlo methods were used by car et al .
@xcite to investigate such a mixture .
car et al .
established also the best fit of goe , gue and gse unit mean distributions , for spacing @xmath19 using the generalized gamma density which we can put in the form @xmath20^ { \beta+1}}{\left[\gamma((1+\beta)/\omega)\right]^{\beta+2 } } \ \ { \rm and } \ \
b(\beta , \omega)= \left[\frac{\gamma((2+\beta)/\omega ) } { \gamma((1+\beta)/\omega)}\right]^\omega \ .
\nonumber \nonumber \end{aligned}\ ] ] then the best fits of ( [ ggamma ] ) had the parameter values @xcite [ cols="^,^,^,^",options="header " , ] [ gengamapprox ] and were accurate to within @xmath21 of the true distributions from forrester @xcite .
observe that the exponential distribution is recovered by the choice @xmath22 these distributions are shown in figure [ ggamgamfits ] along with corresponding fits of the gamma distribution .
the two classes of spectra are illustrated in two dimensions by bouncing geodesics in plane billiard tables : eg in the de - symmetrized ` stadium of bunimovich ' with ergodic chaotic behaviour and irregular spectrum on the one hand , and on the other hand in the symmetric annular region between concentric circles with non - chaotic behaviour , regular spectrum and random energy spacings @xcite .
( 300,200)(0,0 ) ( 0,0 ) it turns out that the mean spacing between eigenvalues of infinite symmetric real random matrices the so called gaussian orthogonal ensemble ( goe)is bounded and therefore it is convenient to normalize the distribution to have unit mean ; also , the same is true for the gue and gse cases .
barnett @xcite provides a numerical tabulation of the first 1,276,900 eigenvalues .
in fact , wigner @xcite had already surmised that the cumulative probability distribution function at the scale of the mean spacing should be of the form : @xmath23 which has unit mean and variance @xmath24 with probability density function @xmath25 remarkably , wigner s surmise gave an extremely good fit with numerical computation of the true goe distribution , cf .
mehta @xcite appendix a.15 , and with a variety of observed data from atomic and nuclear systems @xcite . from mehta
@xcite p 171 , we have bounds on the cumulative probability distribution function @xmath26 for the spacings between eigenvalues of infinite symmetric real random matrices : @xmath27 here the lower bound @xmath28 has mean @xmath29 and variance @xmath30 and the upper bound @xmath31 has mean @xmath32 and variance @xmath33 ( 300,200)(0,0 ) ( 0,0 ) the family of probability density functions for gamma distributions with dispersion parameter @xmath34 and mean @xmath35 for positive random variable @xmath36 is given by @xmath37 with variance @xmath38 then the subset having unit mean is given by @xmath39 with variance @xmath40 these parameters @xmath41 are called natural parameters because they admit presentation of the family ( [ gammapdfs ] ) as an exponential family @xcite and thereby provide an associated natural affine immersion in @xmath42 @xcite @xmath43 this affine immersion was used @xcite to present tubular neighbourhoods of the 1-dimensional subspace consisting of exponential distributions @xmath44 so giving neighbourhoods of random processes .
the maximum entropy case has @xmath45 and corresponds to an underlying poisson random event process and so models spacings in the spectra for non - chaotic systems ; for @xmath46 the distributions are skew unimodular .
the unit mean gamma distribution fit to the true goe distribution from mehta @xcite has variance @xmath47 and hence @xmath48 in fact , @xmath49 is a geodesic coordinate in the riemannian 2-manifold of gamma distributions with fisher information metric @xcite ; arc length along this coordinate from @xmath50 to @xmath51 is given by @xmath52 plotted in figure [ cdfs ] are the cumulative distributions for the bounds ( [ bounds ] ) ( dashed ) , the gamma distribution ( thick solid ) fit to the true goe distribution with unit mean , and the wigner surmised distribution ( [ w ] ) ( thin solid ) .
the corresponding probability density functions are in figure [ pdfs ] : unit mean gamma distribution fit ( thick solid ) to the true goe distribution from mehta @xcite appendix a.15 ( points ) , the wigner surmised density ( thin solid ) and the probability densities for the bounds ( dashed ) , respectively , @xmath53 ( 300,200)(0,0 ) ( 0,0 ) ( 86,170)gse ( 70,135)gue ( 65,94)goe ( 25,144)@xmath54
berry has pointed out @xcite that the behaviour near the origin is an important feature of the ensemble statistics of these matrices and in particular the goe distribution is linear near the origin , as is the wigner distribution . moreover , for the unitary ensemble ( gue ) of complex hermitian matrices , near the origin , the behaviour is @xmath55 and for the symplectic ensemble ( gsm , representing half - integer spin particles with time - reversal symmetric interactions ) it is @xmath56 from ( [ unimeangamma ] ) we see that at unit mean the gamma density behaves like @xmath58 near the origin , so linear behaviour would require @xmath59 which gives a variance of @xmath60 whereas the goe fitted gamma distribution has @xmath61 and hence variance @xmath62 this may be compared with variances for the lower bound @xmath63 @xmath30 the upper bound @xmath64 @xmath65 and the wigner distribution @xmath66 @xmath67 the gamma distributions fitted to the lower and upper bounding distributions have , respectively , @xmath68 and @xmath69 figure [ ggamgamfits ] shows the probability density functions for the unit mean gamma distributions ( dashed ) and generalized gamma distribution ( solid ) fits to the true variances for left to right the goe , gue and gse cases ; the two types coincide in the exponential case , @xmath70 shown dotted .
the major differences are in the behaviour near the origin .
figure [ gamma1249 g ] shows unit mean gamma distributions with @xmath71 for the best fits to the true variances of the spacing distributions for the goe , gue and gse cases , as points on the affine immersion in @xmath42 of the @xmath72-manifold of gamma distributions , cf .
the information metric provides information distances on the gamma manifold and so could be used for comparison of real data on eigenvalue spacings if fitted to gamma distributions ; that may allow identification of qualitative properties and represent trajectories during structural changes of systems .
the author is indebted to rudnick @xcite for pointing out that the gue eigenvalue spacing distribution is rather closely followed by the distribution of zeros for the riemann zeta function ; actually , hilbert had conjectured this , as mentioned along with a variety of other probabilistic aspects of number theory by schroeder @xcite .
this can be seen in figure [ zz2mguegam ] which shows with unit mean the probability distribution for spacings among the first 2,001,052 zeros from the tabulation of odlyzko @xcite ( large points ) , that for the true gue distribution from the tabulation of mehta @xcite appendix a.15 ( medium points ) and the gamma fit to the true gue ( small points ) , which has @xmath73 the grand mean spacing between zeros from the data was @xmath74 the coefficient of variation @xmath75 and variance @xmath76 table 1 shows the effect of location on the statistical data for spacings in the first ten consecutive blocks of 200,000 zeros of the riemann zeta function normalized with unit grand mean ; table 2 shows the effect of sample size . for gamma distributions
we expect the coefficient of variation to be independent of sample size and location , by theorem [ hwangthm ] . * remark * the gamma distribution provides approximations to the true distributions for the spacings between eigenvalues of infinite random matrices for the goe , gue and the gse cases .
however , it is clear that gamma distributions do not precisely model the analytic systems discussed here , and do not give correct asymptotic behaviour at the origin , as is evident from the results of car et al .
@xcite who obtained excellent approximations for goe , gue and gse distributions using the generalized gamma distribution ( [ ggamma ] ) .
the differences may be seen in figure [ ggamgamfits ] which shows the unit mean distributions for gamma ( dashed ) and generalized gamma @xcite ( solid ) fits to the true variances for the poisson , goe , gue and gse ensembles .
the generalized gamma distributions do not have a tractable information geometry and so some features of the gamma distribution approximations may be useful in studies of qualitative generic properties in applications to data from real systems .
it would be interesting to investigate the extent to which data from real atomic and nuclear systems has generally the qualitative property that the sample coefficient of variation is independent of the mean .
that , by theorem [ hwangthm ] , is an information - theoretic distinguishing property of the gamma distribution .
it would be interesting also to know if there is a number - theoretic property that corresponds to the apparently similar qualitative behaviour of the spacings of zeros of the riemann zeta function , tables 12.1 , 12.2 .
since the non - chaotic case has an exponential distribution of spacings between energy levels and the sum of @xmath1 independent identical exponential random variables follows a gamma distribution and moreover the sum of @xmath1 independent identical gamma random variables follows a gamma distribution , a further analytic development would be to calculate the eigenvalue distributions for gamma or loggamma - distributed matrix ensembles .
information geometrically , the riemannian manifolds of gamma and loggamma families are isometric , but the loggamma random variables have bounded domain and their distributions contain the uniform distribution , which may be important in modelling some real physical processes .
h. alt , c. dembrowski , h.d .
graf , r. hofferbert , h. rehfield , a. richter and c. schmit .
experimental versus numerical eigenvalues of a bunimovich stadium billiard : a comparison . _
e _ 60 , 3 ( 1999 ) 2851 - 2857 .
g. le car , c. male and r. delannay .
nearest - neighbour spacing distributions of the @xmath79-hermite ensemble of random matrices .
_ physica a _ ( 2007 ) 190 - 208 .
also their erratum : _ physica a _ 387 ( 2008 ) 1713 .
hwang and c - y .
hu . on a characterization of the gamma distribution :
the independence of the sample mean and the sample coefficient of variation .
_ annals inst .
_ 51 , 4 ( 1999 ) 749 - 753 . * steven j. miller : random matrix theory , random graphs , and l - functions : how the manhatten project helped us understand primes .
ohio state university colloquium 2003 .
+ ` http://www.math.brown.edu/~sjmiller/math/talks/colloquium7.pdf ` . *
steven j. miller .
random matrix theory models for zeros of l - functions near the central point ( and applications to elliptic curves ) .
brown university algebra seminar 2004 .
+ ` http://www.math.brown.edu/~sjmiller/math/talks/rmtandntportrait.pdf ` .
* number theory in science and communication . with applications in cryptography , physics , digital information , computing , and self - similarity*. springer series in information science , 3@xmath81 edition , springer , berlin 1999 | quantum counterparts of certain simple classical systems can exhibit chaotic behaviour through the statistics of their energy levels and the irregular spectra of chaotic systems are modelled by eigenvalues of infinite random matrices .
we use known bounds on the distribution function for eigenvalue spacings for the gaussian orthogonal ensemble ( goe ) of infinite random real symmetric matrices and show that gamma distributions , which have an important uniqueness property , can yield an approximation to the goe distribution . that has the advantage that then both chaotic and non chaotic cases fit in the information geometric framework of the manifold of gamma distributions , which has been the subject of recent work on neighbourhoods of randomness for general stochastic systems .
additionally , gamma distributions give approximations , to eigenvalue spacings for the gaussian unitary ensemble ( gue ) of infinite random hermitian matrices and for the gaussian symplectic ensemble ( gse ) of infinite random hermitian matrices with real quaternionic elements , except near the origin .
gamma distributions do not precisely model the various analytic systems discussed here , but some features may be useful in studies of qualitative generic properties in applications to data from real systems which manifestly seem to exhibit behaviour reminiscent of near - random processes .
+ * keywords : random matrices , goe , gue , gse , quantum chaotic , eigenvalue spacing , statistics , gamma distribution , randomness , information geometry * |
parton - hadron duality or duality for short is one of the central concepts in contemporary particle physics .
it is invoked to connect quantities evaluated on the quark - gluon level to the ( observable ) world of hadrons .
it is used all the time , more often than not without explicit reference to it .
a striking example of the confidence the hep community has in the asymptotic validity of duality was provided by the discussion of the width @xmath2 .
there was about a 2% difference in the predicted and measured decay width , which lead to lively debates on its significance vis - a - vis the _ experimental _ error .
no concern was expressed about the fact that the @xmath3 width was calculated on the quark - gluon level , yet measured for hadrons .
likewise the strong coupling @xmath4 is routinely extracted from the perturbatively computed hadronic @xmath3 width with a stated theoretical uncertainty of 0.003 with translates into a theoretical error in @xmath5 of about 0.1% .
there are , however , several different versions and implementations of the concept of duality .
the problem with invoking duality implicitly is that it is very often unclear which version is used . in @xmath0 physics in particular when determining @xmath6 and @xmath7 the measurements have become so precise that theory can no longer hide behind experimental errors . to estimate theoretical uncertainties in a meaningful way one has to give clear meaning to the concept of duality ;
only then can one analyze its limitations . in response to the demands of @xmath0 physics
a considerable literature has been created on duality over the last few years , which we want to summarize .
we will emphasize and illustrate the underlying principles ; technical details can be found in the references we list .
duality for processes involving time - like momenta was first addressed theoretically in the late 70 s in references @xcite and @xcite .
we sketch here the argument of poggio , quinn and weinberg since it contains several of the relevant elements in a nutshell .
the cross section for @xmath8 can be expressed through an operator product expansion ( ope ) of two hadronic currents .
one might be tempted to think that by invoking qcd s asymptotic freedom one can compute @xmath9 for large c.m .
energies @xmath10 in terms of quarks ( and gluons ) since it is shaped by short distance dynamics
. however production thresholds like for charm induce singularities that vitiate such a straightforward computation .
this complication can be handled in the following way .
consider the correlator of two electromagnetic currents : @xmath11 where @xmath12 can be written in terms of a spectral function @xmath13 using an unsubtracted dispersion relation : @xmath14 it is well known that , to leading order in @xmath15 , @xmath13 is related to the total cross section for @xmath16 hadrons @xmath17 relying on qcd s asymptotic freedom one computes the correlator ( [ corr ] ) in terms of quarks and gluons for @xmath18 in the deep euclidean domain @xmath19 ; @xmath18 is chosen euclidean so that one avoids a proximity to singularities induced by hadronic thresholds like for charm production etc . from the spectral function @xmath20 calculated in the euclidean regime one can infer the cross section for physical , namely minkowskian values of @xmath18 .
however , one can not obtain it as a point - for - point function of @xmath18 , only averaged or ` smeared ' over an energy interval , which can be written symbolically as follows : ( s_euclid ) _
s_0^s_0+s ds(e^+e^- hadrons ) this feature is immediately obvious : for the smooth @xmath18 dependence @xmath20 has to be compared to the measured cross section @xmath21 hadrons as a function of @xmath18 , which has pronounced structures , in particular close to thresholds for @xmath22- and @xmath23-production .
this simple illustration already points to the salient elements and features of duality and its limitations : * an ope description in terms of quark and gluon degrees of freedom for the observable under study is required .
* this ope has to be constructed in the euclidean domain . *
its results are analytically continued to the minkowskian domain with the help of a dispersion relation . *
this extrapolation implies some loss of information ; i.e. in the notation given above @xmath24 where @xmath25 denotes the smearing which is an average using a smooth weigth function @xmath26 ; it generalizes the simplistic use of a fixed energy interval : @xmath27 * some contributions that are quite insignificant in the euclidean regime and therefore can not be captured through the ope can become relevant after the analytical continuation to the minkowskian domain , as explained later on .
for that reason we have used the approximate rather than the equality sign in eq.([ansatz ] ) .
* one can make few universal statements on the numerical validity of duality .
how much and what kind of smearing is required depends on the specifics of the reaction under study .
the last item needs expanding right away .
the degree to which @xmath28 can be trusted as a theoretical description of the observable @xmath29 depends on the weight function , in particular its width .
it can be broad compared to the structures that may appear in the hadronic spectral function , or it could be quite narrow , as an extreme case even @xmath30 . it has become popular to refer to the first and second scenarios as _ global _ and _ local _ duality , respectively .
other authors use different names , and one can argue that this nomenclature is actually misleading . below
we will describe these items in more detail without attempting to impose ex cathedra a uniform nomenclature ourselves .
irrespective of names , a fundamental distinction concerning duality is often drawn between semileptonic and nonleptonic widths .
since the former neccessarily involves smearing with a smooth weight function due to the integration over neutrino momenta , it is often argued that predictions for the former are fundamentally more trustworthy than for the latter .
however , as we shall see , such a categorical distinction is overstated and artificial ; also it is not needed for the discussion in the following chapters . of much more relevance
is the differentiation between distributions and fully integrated rates . no real progress beyond the more qualitative arguments of refs . @xcite and @xcite occurred for many years . for
as long as one has very limited control over nonperturbative effects , there is little meaningful that can be said about duality violations . yet
this has changed for heavy flavour physics with the development of heavy quark expansions , since within this ope framework we can assess nonperturbative effects as well as duality violations .
the remainder of this note will be organized as follows : in the next section we shall give a more precise definition of what is meant with `` parton hadron duality '' , which then allows us a discussion of possible violations of duality in section [ sec : violation ] . based on this
we give hints on how to check the concept of duality in section [ sec : validty ] , before presenting conclusions .
in order to discuss possible violations of duality we have to give first a more precise definition of this notion , which requires the introduction of some theoretical tools .
we follow closely the arguments given in the extensive reviews of ref .
@xcite and @xcite .
the central ingredient into this definition is the method of the wilsonian _ operator product expansion _ ( ope ) frequently used in field theory to perform a separation of scales . in practical terms
this means that we can write @xmath31 for @xmath32 .
the following notation has been used : @xmath33 denotes a state that could be the vacuum as for @xmath34 considered above or a @xmath0 meson when describing semileptonic beauty decays .
denote electromagnetic and weak current operators for the former and the latter processes , respectively ; for other decays like nonleptonic or radiative ones one employs different @xmath36 operators ; the @xmath37 are local operators of increasing dimension
. the operator of lowest dimension yields the leading contribution . in @xmath38 annihilation
it is the unit operator @xmath39 , for @xmath0 decays @xmath40 .
they produce ( among other things ) the naive partonic results .
yet the ope allows us to systematically improve the naive partonic result .
the coefficients @xmath41 contain the contributions from short distance dynamics calculated perturbatively based on qcd s asymptotic freedom .
following wilson s prescription a mass scale @xmath42 has been introduced to separate long and short distance dynamics ; both the coefficients and the matrix elements depend on it , their product of course not .
the perturbative expansion takes the form @xmath43 and is performed in terms of quarks and gluons .
the expectation values for the local operators provide the gateways through which nonperturbative dynamics enters .
the crucial point is that the ope result is obtained in the euclidean domain and has to be continued analytically into the minkowskian regime relating the ope result to observable hadronic quantities .
as long as qcd is the theory of the strong interactions , it does not exhibit unphysical singularities in the complex @xmath44 plane , and the analytical continuation will not induce additional contributions . to conclude :
_ duality between @xmath45 and @xmath28 arises due to the existence of an ope that is continued analytically_. it is just a restatement of qcd s basic tenet as the theory of the strong interactions that hadronic observables can be expressed in terms of quark - gluon degrees of freedom _ provided _ all possible sources of corrections to the simple parton picture are properly accounted for .
it is thus misleading to refer to duality as an additional assumption .
up to this point our discussion was quite generic . to specify it for semileptonic @xmath0 decays
one chooses the current @xmath46 to be the weak charged current driven by @xmath47 or @xmath48 .
the expansion parameter for inclusive semileptonic decays is given by the energy release @xmath49 $ ] for @xmath50 $ ] .
for the exclusive mode @xmath51 it is @xmath52 and @xmath53 with the latter yielding the numerically leading contributions .
one of the main applications of the heavy quark expansion is the reliable extraction of @xmath6 and @xmath7 .
one wants to be able to arrive at a meaningful estimate of the theoretical uncertainty in the values obtained .
there are three obvious sources of theoretical errors : 1 .
unknown terms of higher order in @xmath54 ; 2 . unknown terms of higher order in @xmath55 ;
3 . uncertainties in the input parameters @xmath54 , @xmath56 and the expectation values .
duality violations constitute uncertainties _ over and above _ these ; i.e. they represent contributions not accounted for due to * truncating these expansions at finite order and * limitations in the algorithm employed .
these two effects are not unrelated .
the first one means that the ope in practice is insensitive to contributions of the type @xmath57 with @xmath58 denoting some hadronic scale ; the second one reflects the fact that under an analytic continuation the term @xmath57 , which is quite irrelevant for @xmath59 though not neccessarily for @xmath60 !
turns into an oscillating rather than suppressed term sin@xmath61 . of course we do not have ( yet ) a full theory for duality and its violations .
yet we know that without an ope the question of duality is ill - posed .
furthermore in the last few years we have moved beyond the stage , where we could merely point to folklore .
this progress has come about for the following reasons : * we have refined our understanding of the physical origins of duality violations as due to * * hadronic thresholds ; * * so - called ` distant cuts ' ; * * the suspect validity of @xmath53 expansions . *
we understand the mathematical portals through which duality violations can enter , namely that the innocuous euclidean quantity @xmath57 transmogrifies itself into the much more virulent minkowskian quantity @xmath62 under analytical continuation . *
the quantity @xmath63 is actually innocuous for beauty , yet not necessarily for charm quarks .
the ` euclidean ' quantity @xmath64 the formfactor for @xmath65 at zero recoil , which is given by an expansion in @xmath53 , could be vulnerable to such a duality violation
. however , the heavy mass expansion for exclusive quantities such as @xmath64 is not directly given by an ope , thus this argument may not apply in this case . *
we have come up with an increasing array of field - theoretical toy models , chief among them the t hooft model , which is qcd in 1 + 1 dimensions in the limit of @xmath66 .
it is solvable and thus allows an unequivocal comparison of the ope result with the exact solution .
* for the analysis of @xmath47 transitions we also have the small - velocity expansion as a powerful tool .
we will not go into any details here , since they can be found in the literature @xcite .
the models do exhibit duality violations , but only highly suppressed ones conforming to general expectations .
there had been claims of sizeable duality violations in the previous literature ; those have been analyzed carefully in ref.@xcite , where their flaws are pointed out explicitly .
based on general expectations as well as on analyzing the models one finds that indeed duality violations are described by highly power suppressed ` oscillating ' terms of the form [ dv ] t(m_q ) ~ ( ) ^k sin(m_q ) for some integer power @xmath67 .
more generally one can state : * the primary criterion for addressing duality violation is the existence of an ope for the particular observable . * duality will not be exact at finite masses .
it represents an approximation the accuracy of which will increase with the energy scales in a way that depends on the process in question .
* limitations to duality can enter only in the form of an oscillating function of energy or @xmath56 ( or have to be exponentially suppressed ) .
duality violations _ can not be blamed for a systematic excess or deficit in the decay rates .
_ for example , no duality violation can convert @xmath56 into @xmath68 in the full width parametrically , only for discrete values of @xmath56 .
* the ope equally applies to semileptonic as well as nonleptonic decay rates .
likewise both widths are subject to duality violations .
the difference here is quantitative rather than qualitative ; at finite heavy quark masses corrections are generally expected to be larger in the nonleptonic widths . in particular ,
duality violations there can be boosted by the accidental nearby presence of a narrow hadronic resonance .
similar effects could arise in semileptonic rates , but are expected to be highly suppressed there .
* it is not necessary to have a proliferation of decay channels to reach the onset of duality , either approximate or asymptotic .
instructive examples are provided by the so - called small - velocity kinematics in semileptonic decays and by nonleptonic rates in the t hooft model .
putting everything together it has been estimated with considerable confidence at least by the authors of ref.@xcite that _ duality violations in the integrated semileptonic width of @xmath0 mesons can not exceed the fraction of a percent level .
_ as such we do not envision it to ever become the limiting factor in extracting @xmath6 and @xmath7 since the uncertainties in the expression for the semileptonic width due to fixed higher order contributions will remain larger than this level . the oscillatory nature of duality violating contributions is a main ingredient in this conclusion .
it also shows that duality violations could become quite sizeable if an only partially integrated width
let alone a distribution is considered .
generally , for distributions the expansion parameter is not the heavy mass , rather it is a quantity such as @xmath69 $ ] where @xmath70 is e.g. the rescaled charged lepton energy of a semileptonic decay . from equation ( [ dv ] ) one would expect that contributions the form @xmath71)/[m_q ( 1-x)]^k$ ] appear in differential distributions .
if in the future we were to find a discrepancy between the measured and predicted values for , say , a cp asymmetry in @xmath0 decays , we had to check very diligently all ingredients upon which the prediction was based , in particular the values for @xmath72 and @xmath73 , before we could make a credible claim to have uncovered new physics .
this means one needs a measure for potential duality violations that is not based purely on theoretical arguments .
most theoretical uncertainties are systematical rather than statistical .
as it is the case for experimental systematics the most convincing way to establish control over them is to determine the same quantity in independent ways and analyze their consistency .
the heavy quark expansions lend themselves naturally to such an approach since it is their hallmark that they allow the description of numerous decay rates in terms of a handful of basic parameters , namely quark masses and hadronic expectation values .
again the situation is very similar as for the perturbative series : once the coupling constant is determined ( e.g. @xmath74 ) from a measurement , one may use this as an input to all other perturbative calculations , thereby predicting other measurements .
if a prediction obtained in this way fails , one would conclude that higher order effects have to be unusually large or that there is another deeper reason why a perturbative treatment does not apply .
such independent determinations of the same quantity of course probe the overall theoretical control that we have established . by themselves
they do not tell us whether a failure found is due to unusally large higher order contributions or to a breakdown in duality .
the fact that both the inclusive and exclusive methods for extracting @xmath6 yield consistent values and that the theoretical corrections one had to apply are both nontrivial and essential for the agreement
is such a test .
we want to point to two other such tests that have become available , namely concerning the @xmath1 quark mass and its kinetic energy expectation value .
the @xmath1 quark mass has been extracted from beauty production at threshold in @xmath38 annihilation by several authors @xcite .
their findings can be stated in terms of two definitions of quark masses : * the ` kinetic mass ' is defined by = - - + o ( _ s^2,_s ) , normalized at 1 gev ; it is well - defined in full qcd and does not suffer from a renormalon ambiguity ; equivalently one can use [ lambdadef ] | ( ) _ m_q . where @xmath75 is the mass of the @xmath76 ground state .
* the pole or hqet mass which is a very popular choice , although it is not well - defined in full qcd since it suffers from the renormalon ambiguity .
if appropriate care and caution are applied one can still use it in calculation ; as a rule of thumb one has for its relatioship to the kinetic mass : |_hqet = |(1 ) - 0.255 where the parameter @xmath77 is defined in the same way as in ( [ lambdadef ] ) : |^hqet _ m_q [ m(h_q ) - m_q^pole ] the results are completely consistent within the stated uncertainties of 1 - 2 % and can be summarized as follows : m_b^kin(1)|_e^+e^-|bb = 4.57 0.06 |(1 ) = 0.71 0.06 [ mkinav ] or m_b^pole = 4.82 0.06 |^hqet = 0.45 0.06 the techniques employed in the analysis differ somewhat from author to author ; the full agreement in their findings is thus quite re - assuring .
one should keep in mind , though , that these determinations share their experimental input to a large degree . the value stated in eq.([mkinav ] ) could thus be subject to some systematic shift from the true value .
arguments based on the small - velocity sum rules indeed suggest that @xmath78 could lie a bit above 4.6 gev .
the @xmath1 quark mass also affects the shape of lepton energy and hadronic mass spectra in semileptonic ( and photon spectra in radiative ) @xmath0 decays
. its value can therefore be obtained from the measured lepton energy and hadronic mass moments , which encode the shape of these spectra .
the delphi and cleo collaborations have presented data as shown in the figures [ fig1 ] and [ fig2 ] .
again it is pleasing to see that the different moments indeed yield completely consistent values although this is not truly surprising since they are highly correlated .
delphi finds m_b^kin(1)|_mom & = & 4.59 0.08 0.01 + & & |(1 ) = 0.69 0.08 0.01 + m_b^pole|_mom & = & 4.88 0.10 0.02 + & & |^hqet = 0.40 0.10 0.02 it is again reassuring that their fit results are consistent with eq.([mkinav ] ) cleo measures truncated lepton energy moments and states their findings in terms of hqet parameters m_b^pole|_mom & = & 4.88 0.03 0.06 0.12 + & & |^hqet = 0.39 0.03 0.06 0.12 the main news here is how well the values for @xmath79 extracted from two sources agree that are quite different in their experimental as well as theoretical systematics , namely weak @xmath0 decays and the electromagnetic production of beauty at threshold .
a similarly pleasing picture emerges from determining the kinetic expectation value .
based on qcd sum rules and sv sum rules one had inferred for the infrared stable quantity @xmath80 : _
^2(1 ) 0.45 0.1 ( ) ^2 . using the rule of thumb for the relation to the hqet parameter @xmath81 @xcite - _ 1 _ ^2(1 ) - 0.18
( ) ^2 this estimate translates into - _ 1 0.27 0.1 ( ) ^2 the aforementioned delphi and cleo analyses also yield values for this quantity , namely _
^2(1 ) & = & 0.31 0.07 0.02 ( ) ^2 + - _ 1 & = & 0.15 0.07 0.03 ( ) ^2 + - _ 1 & = & 0.25 0.02 0.05 0.14 ( ) ^2 the fact that the parameters extracted in different way and form different observables yield consistent values for the quark mass and the kinetic energy parameter indicates that no anomalously large higher order corrections or unexpectedly sizeable duality violating contributions are present .
from all what we know currently from purely theretical considerations duality violations should be safely below one percent in the semileptonic branching ratio .
this is likely to remain in the noise level of theoretical uncertainties due to terms of order @xmath82 and higher and of higher order perturbative contributions .
hence we do not see any need to assign some additional uncertainty to the extraction of @xmath83 from a possible duality violation in inclusive decays .
this should not be seen as an ex cathedra statement . when more and more types of moments will be measured with more and more accuracy even separately in the decays of @xmath84 , @xmath85 and @xmath86 mesons , additional constraints will be placed on the same set of heavy quark parameters .
this will provide highly nontrivial tests of our theoretical control .
we gratefully acknowledge illuminating discussions with n. uraltsev . this work has been supported by the national science foundation under grant number phy00 - 87419 , bt the dfg research group `` quantenfeldtheorie , computeralgebra und montre carlo simulationen '' and by the german minister for education and research bmbf und grant number 05ht1vkb1 and the dfg merkator program . 99 e. c. poggio , h. r. quinn and s. weinberg , phys .
* d 13 * , 1958 ( 1976 ) .
m. greco , g. penso and v. srivastava , phys .
* d 12 * ( 1980 ) 2520 .
m. a. shifman , `` quark - hadron duality , '' published in the boris ioffe festschrift at the frontier of particle physics / handbook of qcd , ed .
m. shifman ( world scientific , singapore , 2001 ) , arxiv : hep - ph/0009131 . i. i. bigi and n. uraltsev , int .
j. mod .
phys . * a 16 * , 5201 ( 2001 ) [ arxiv : hep - ph/0106346 ] .
m. shifman , in _ continuous advances in qcd _ ,
a. smilga ( world scienific , singapore 1994 ) , p.249 [ hep - ph/9405246 ] , and in _ particles , strings and cosmology _ eds . j. bagger et al . , ( world scientific , singapore 1996 ) p.69 [ hep - ph/9505289 ] . n. isgur , phys . lett .
* b 448 * , 111 ( 1999 ) [ arxiv : hep - ph/9811377 ] .
a. le yaouanc , d. melikhov , v. morenas , l. oliver , o. pene and j. c. raynal , phys .
* b 517 * , 135 ( 2001 ) [ arxiv : hep - ph/0103339 ] . b. grinstein and r. f. lebed , phys . rev . *
d 57 * , 1366 ( 1998 ) [ arxiv : hep - ph/9708396 ] .
k. melnikov and a. yelkhovsky , phys .
d 59 * ( 1999 ) 114009 ; + a. hoang , phys .
* d 61 * ( 2000 ) 034005 ; + m. beneke and a. signer , phys .
* b 471 * ( 1999 ) 233 ; + .
h. kuhn and m. steinhauser , nucl .
* b 619 * , 588 ( 2001 ) [ erratum - ibid .
* b 640 * , 415 ( 2002 ) ] [ arxiv : hep - ph/0109084 ] .
gambino , these processdings m. battaglia _ et al .
_ , cern - th/2002 - 290 , arxiv : hep - ph/0210319 .
r. a. briere _ et al . _
[ cleo collaboration ] , contributed to 31st international conference on high energy physics ( ichep 2002 ) , amsterdam , the netherlands , 24 - 31 jul 2002 , determination of the heavy quark expansion parameters , arxiv : hep - ex/0209024 . n. uraltsev , talk given at 31st international conference on high energy physics ( ichep 2002 ) , amsterdam , the netherlands , 24 - 31 jul 2002 , arxiv : hep - ph/0210044 . | we summarize the current view on parton - hadron duality as it applies to @xmath0 meson decays .
it is emphasized that an ope treatment is essential for properly formulating duality and its limitations .
duality violations are unlikely to become the limiting factor in describing semileptonic @xmath0 width vis - a - vie higher order corrections .
the consistent extraction of the @xmath1 quark mass from @xmath0 production and decays provides a striking example of the theoretical control achieved . * parton hadron duality in @xmath0 meson decays * and thomas mannel + _ institut fr theoretische teilchenphysik , + universitt karlsruhe , d76128 karlsruhe , germany _ + _ contribution to the ckm workshop , + held at cern , geneva , feb.15.-18.2002 _ |
the structure of light nuclei has always provided a certain fascination , given that a full range of structural properties are displayed from spherical shell model like structure via deformation to clustering . in particular
even - even nuclei , composed of equal numbers of protons and neutrons offer the possibility that the nucleus may be decomposed into a collection of @xmath5-particles .
this was the picture developed by ikeda @xcite , who speculated that at the point that the @xmath7 decay threshold was encountered the associated cluster degree of freedom should be liberated .
correspondingly , this picture would suggest that the @xmath8be ground state should possess an @xmath5-@xmath5 cluster structure , and the 3@xmath5 cluster structure should appear at an excitation energy close to 7.3 mev in @xmath9c .
indeed , it has been recently speculated that certain states above the 3@xmath5 decay threshold in @xmath9c ( e.g. the 7.65 mev state ) , or the 4@xmath5 decay threshold in @xmath10o , should possess properties reminiscent of a bose gas @xcite .
these highly developed cluster systems have been shown to have an important impact on the structure of neutron - rich isotopes of beryllium and carbon .
for example , in the case of @xmath11be the properties of the valence neutron may be explained in terms of the sharing of the neutron between the two @xmath5-cores in a manner which is reminiscent of the covalent binding of atomic molecules @xcite . for valence neutrons based on @xmath5-particle cores , linear combinations of the @xmath12-type orbits
lead to analogues of the atomic @xmath13 and @xmath14-bonds .
indeed , such a description provides a useful basis for the understanding of the low lying spectroscopy of the @xmath1be nucleus , with the ground state structure forming a rotational band with @xmath14-like character .
the presence of 3@xmath5 cluster structures in @xmath9c provides an extension to the ideas developed for the two - centre systems .
calculations which explicitly contain the molecular - orbital structure @xcite show that unlike the linear 3@xmath5 system , which is unstable against bending modes and thus collapse into more compact structures , the 3@xmath15 systems are stable against the bending mode @xcite .
thus , there would appear to be an opportunity for observing linear chain configurations , partially stabilized by the valence neutrons in covalent type orbits .
a recent analysis of available data by milin and von oertzen has provided tentative evidence for the existence of such a structure in @xmath6c @xcite . via a careful examination of previous experimental spin and parity assignments ,
two rotational bands were proposed ( @xmath16 ) whose deformations indicated a chain - like structure and which were split in energy due to the intrinsic asymmetry of the 3@xmath17 structure . this analysis provided information on such rotational structures up to @xmath1817 mev .
although , such a picture is appealing the relationship between the states considered by milin and von oertzen has not been demonstrated since the spins and parities remain to be confirmed . in the present paper
we present a measurement of the @xmath19li(@xmath11be,@xmath6c@xmath20be+@xmath5)@xmath3h reaction .
the @xmath5-transfer process provides a possible mechanism by which the chain - like structures in @xmath6c may be populated .
the measurement covers the excitation energy range of 12 to 28 mev , the interval over which the higher energy and spin members of the two rotational bands are believed to exist .
the measurements were performed at the australian national university s 14ud tandem accelerator facility . a 70 mev @xmath11be beam , of intensity 3 ena , was incident on a 100 @xmath21g@xmath22 li@xmath23o@xmath24 foil .
the integrated beam exposure was 0.45 mc .
reaction products formed in reactions of the @xmath11be beam with the target were detected in an array of four charged particle telescopes .
these telescopes contained three elements which allowed the detection of a wide range of particle types , from protons to @xmath25=4 to 5 nuclei .
the first elements were thin , 70@xmath21
m , 5@xmath265 @xmath27 silicon detectors segmented into four smaller squares ( quadrants ) .
the second elements were position - sensitive strip detectors with the same active area as the quadrant detectors , but divided into 16 position - sensitive strips .
these strips were arranged so that the position axis gave maximum resolution in the measurement of scattering angles . finally , 2.5 cm thick csi detectors were used to stop highly penetrating light particles . these detector telescopes provided charge and mass resolution up to be , allowing the final states of interest to be unambiguously identified . the position and energy resolution of the telescopes was @xmath281 mm and 300 kev , respectively .
calibration of the detectors was performed using elastic scattering measurements of @xmath1be from @xmath29au and @xmath9c targets .
the four telescopes were arranged in a cross - like arrangement , separated by azimuthal angles of 90@xmath30 .
two opposing detectors were located with their centres at 17.3@xmath30 and 17.8@xmath30 ( detectors 1 and 2 ) from the beam axis and with the strip detector 130 mm from the target , whilst the remaining pair were at the slightly larger angles of 28.6@xmath30 and 29.7@xmath30 ( detectors 3 and 4 ) , 136 mm from the target .
the detection system employed in the present measurements allowed a measurement of the charge , mass , energy and emission angle ( and hence a determination of the momentum ) of each particle .
this provides a complete determination of the reaction kinematics in the case of a three particle final state .
figure [ etot ] shows the total energy spectrum of the three final state particles assuming a @xmath31h recoil ( @xmath32 ) , after gating on @xmath11be+@xmath5 in the particle identification spectra .
the large peak in this spectrum corresponds to the observation of the @xmath19li(@xmath1be,@xmath11be@xmath5)@xmath3h reaction ( q=-2.468 mev ) , where all particles are produced in their ground states ( @xmath11be only possesses one state bound against particle decay ) .
the energy resolution in the total energy spectrum is @xmath281.9 mev , considerably worse than the intrinsic energy resolution of the detectors ( 300 kev ) .
this is due to in fact the uncertainty imposed by the measured momenta of the two detected particles limits the resolution of the reconstructed momentum of the undetected triton .
given the relatively small mass of the recoil particle the uncertainty in momentum is magnified in energy .
the first peak at lower total energy corresponds to reactions involving the oxygen and carbon components in the target : @xmath10o(@xmath11be,@xmath1be@xmath5)@xmath9c ( q=-7.162 mev ) and @xmath9c(@xmath11be,@xmath1be@xmath5)@xmath33be ( q=-7.367 mev ) , predominantly the former , with the lowest energy peak coinciding with the 4.4 mev excitation of the @xmath9c recoil .
unfortunately , the underlying background means that these channels can not be analyzed . total energy spectrum for @xmath34he+@xmath1be coincidences detected in telescopes 1 and 2 , assuming a @xmath31h recoil .
the highest energy peak corresponds to the observation of the @xmath0li(@xmath11be,@xmath11be@xmath5)@xmath31h reaction .
statistical uncertainties are presented only.,scaledwidth=75.0% ] by selecting only the events associated with the @xmath19li(@xmath1be,@xmath11be@xmath5)@xmath3h reaction the @xmath6c excitation energy can be calculated from the relative velocity of the two decay fragments .
figures [ exc]a ( top ) and [ exc]b ( bottom ) show such spectra for the small angle detector pair ( 1 and 2 ) and the larger angle pair ( 3 and 4 ) respectively .
as expected these spectra sample slightly different excitation energy ranges due to the minimum opening angle required for the fragments to reach the detectors .
given the presence of three particles in the final state the possibility remains that at least some of the yield in figures [ exc]a and [ exc]b arises from decays of either @xmath0li into t+@xmath5 or @xmath9b to t+@xmath1be .
both of these decay processes were reconstructed and only evidence for decays of @xmath0li was found , where the states at 4.63 and 6.68 mev were observed in the coincidence spectra for detectors 3 and 4 .
this contribution was removed from the spectrum in figure [ exc]b by eliminating all data corresponding to @xmath0li excitations less than 7 mev .
@xmath6c excitation energy spectra for decays detected in detectors 1 and 2 ( a ) top , and 3 and 4 ( b ) bottom .
observed peaks are labelled with excitation energies ( mev ) .
the details of the peaks observed in these spectra are presented in table [ 13cexc ] .
error bars represent statistical errors.,scaledwidth=75.0% ] .[13cexc ] excitation energies of @xmath6c states @xmath5-decaying into the ground state of @xmath1be .
the uncertainty in these energies is 100 kev .
the previous measurements are from the tabulations of @xcite .
note that the excitation energy resolution is 180 to 200 kev from an analysis of the 13.4 mev peak , and therefore widths are only quoted where the experimental resolution permits , and the experimental resolution has been unfolded .
the states shown in italics are states observed at the same energies as those in the present measurements , but have different experimental widths . [
cols="^,^,^,^,^ " , ] the angular distributions in figure [ expad ] appear to show two features , the first is that the distributions for the five states at 14.1 , 14.6 , 16.8 , 18.7 and 21.3 mev all have a similar character , and secondly it is distinct from that of the distributions for the 13.4 and 23.9 mev states .
this difference is in terms of the reduced oscillatory structure for the distributions of the latter two states , and that the distributions for these two states appear to possess a gradient which differs to that of the remaining states .
this feature would suggest that the two sets of states possess different characters , either in terms of their parity or the angular momentum transfer . in the present case
positive parity states would be populated by odd angular momentum @xmath35 values ( @xmath35 being the angular momentum of the orbit into which the @xmath5-particle is transferred ) , whilst negative parity states would be populated by even @xmath35 values .
the spins which could be populated in the transfer of an @xmath5-particle with this angular momentum would be in the range @xmath36 .
thus , for example , for the sequence of states 3/2@xmath37 , 5/2@xmath37 , 7/2@xmath37 , 9/2@xmath37 and 11/2@xmath37 the @xmath35 values would be ( 1,3 ) , ( 1,3 ) , ( 3,5 ) , ( 3,5 ) and ( 5,7 ) respectively .
here @xmath38 , the initial angular momentum of the @xmath5-particle in @xmath0li , is equal to 1 . the transferred angular momentum @xmath39 is given by @xmath40 , and is even for positive parity states and odd for those with negative parity .
the ( @xmath0li , t ) reaction has a long history in the population of @xmath5-cluster states in light nuclei , and the angular distributions for differing @xmath35-values have been collated ( in particular for @xmath35 = 0 and 2 ) by bethge @xcite , together with those for the ( @xmath41li , d ) @xmath5-transfer reaction .
typically , the ( @xmath0li , t ) angular distributions are less oscillatory than those for the ( @xmath41li , d ) reaction , and are reminiscent of those that appear in figure [ expad ] .
the reduced structure in the ( @xmath0li , t ) angular distributions is related to the presence of two transfer angular momentum values , @xmath39 , for each @xmath35 @xcite , for @xmath420 .
nevertheless , the compilation of the @xmath5-transfer reaction in @xcite indicates the sensitivity to the orbital angular momentum of the orbit into which the @xmath5-particle is transferred .
these data indicate that the second maximum for @xmath35=0 and @xmath35=2 transfers appear at centre - of - mass angles 20@xmath43 - 30@xmath43 and 40@xmath43 - 50@xmath43 , respectively . via the properties of bessel functions the corresponding locations of the maxima for
@xmath35=1 and @xmath35=3 transfers would appear at 30@xmath43 - 40@xmath43 and 47@xmath43 - 57@xmath43 , respectively .
for the five states with similar angular distributions the secondary maxima shift a little with increasing excitation energy ( as indicated by the dashed line in figure [ expad ] ) , but lie within the region 29@xmath30 to 41@xmath30 .
we note that the secondary maxima appear to be less strong in the case of the 14.1 and 14.6 mev states
. the observed small change of about 10 degrees of the maxima with excitation energy may reflect the changing amplitudes of the two contributing angular momentum transfer values .
the data thus appear to indicate that the five states possess the same value of @xmath35 which may be either 1 or 2 . from the perspective of the excitation energy of the observed
states the latter is most likely .
however , we note that the angular distributions in @xcite were measured at centre - of - mass energies approximately a factor of 2 smaller than in the present case , which may result in some variations of the distributions in @xcite and those measured here
. these data would suggest that the five states at 14.1 , 14.6 , 16.8 , 18.7 and 21.3 mev all correspond to a similar angular momentum transfer and thus may be structurally linked , whereas the two states at 13.4 and 23.9 mev may possess an alternative character .
this observation appears to be inconsistent with the assignments made by milin and von oertzen .
we have attempted to reproduce the experimental angular distributions using the dwba formalism .
however , the calculated distributions were found to be overly sensitive to the details of the interaction potentials , and moreover required an approximation to bound state wave - functions for the unbound resonant states . these features of the calculations
have not permitted spin information to be accurately and unambiguously extracted from the present data .
a measurement of the @xmath19li(@xmath11be,@xmath1be@xmath5)@xmath3h at @xmath4=70 mev has been performed using an array of charged particle detector telescopes .
the measurement of the @xmath6c excitation energy spectrum corresponding to decays into @xmath5+@xmath11be produced evidence for excited states between 12 and 28 mev .
evidence is found for three new broad resonances at 18.7 , 21.3 and 23.9 mev .
given the nature of the reaction process , @xmath5 transfer onto the @xmath5+@xmath44+@xmath5 cluster nucleus @xmath11be , it is possible that some of the states observed in the present measurement may be linked with the 3@xmath5+@xmath44 chain structure suggested by milin and von oertzen @xcite .
an analysis of the angular distributions suggests that the group of states at excitation energies 14.1 , 14.6 , 16.8 , 18.7 and 21.3 mev correspond to the @xmath5-transfer to a common orbital in @xmath6c .
the measurements suggest that there may be inconsistencies in the assignments made in reference @xcite , in identifying the chain - structure .
particularly the resonance at 14.6 mev observed in the present data is strongly populated , and thus would be expected to feature strongly in the rotational systematics .
further , the measurements indicate that the states may not be ordered into the bands suggested by table [ matkovon ]
. clearly , spin determinations are imperative in order to fully understand the nature of the states above the @xmath11be+@xmath5 decay threshold . due to the large number of reaction amplitudes , resulting from the presence of non - zero spin particles ,
the angular correlations measured here were featureless .
it is possible that reactions such as @xmath9c(@xmath1be,@xmath6c)@xmath33be and @xmath10o(@xmath1be,@xmath6c)@xmath9c , possessing spin - zero nuclei in the initial and final state will provide a more promising avenue for resolving this issue .
alternatively , an analysis of the @xmath41li(@xmath11be,@xmath6c)@xmath45h reaction should produce angular distributions which are more oscillatory than those observed in the present study .
the authors would like to acknowledge the assistance of anu personnel in running the accelerator .
this work was carried out under a formal agreement between the u.k .
engineering and physical sciences research council and the australian national university .
pjl , bjg and klj would like to acknowledge the epsrc for financial support .
k. ikeda , suppl .
physics ( japan ) extra numbers , 464 ( 1968 ) .
a. tohsaki , h. horiuchi , p. schuck , and g. ropke , phys .
lett . , * 87 * 192501 ( 2001 ) .
w. von oertzen , z. phys .
, * a354 * 37 ( 1996 ) .
w. von oertzen , z. phys . , * a357 * 355 ( 1997 )
. w. von oertzen , nuovo cimento a , * 110 * 895 ( 1997 ) .
n. itagaki , s. okabe , k. ikeda , and i. tanihata , phys .
c , * 64 * 014301 ( 2001 ) .
n. itagaki , _ private communication _ m. milin , and w. von oertzen , eur .
j. a , * 14 * 295 ( 2002 ) .
f. ajzenberg - selove , nucl .
* a523 * 1 ( 1991 ) . m. freer , nucl . instr . meth . *
a383 * , 463 ( 1996 ) . c. lee , d.d .
caussyn , n.r .
fletcher , d.l .
gay , m.b .
hopson , j.a .
liendo , s.h .
myers , m.a .
tiede , and j.w .
baker , phys .
c , * 58 * 1005 ( 1998 ) .
goss , s.l .
blatt , d.r .
parsignault , c.d .
porterfield , and f.l .
riffle , phys .
c , * 7 * 1837 ( 1973 ) .
w. tornow , _ et al .
_ , j. phys .
g , * 11 * 379 ( 1985 ) .
collins , _ et al . _ ,
phys . , * a481 * 494 ( 1988 ) . k. bethge , ann . rev .
, * 20 * 255 ( 1970 ) .
f. phlofer , h. ritter , r. bock , g. brommundt , h. schmidt and k. bethge , nucl .
, * a147 * 258 ( 1970 ) . | a study of the @xmath0li(@xmath1be,@xmath2he@xmath1be)@xmath3h reaction at @xmath4=70 mev has been performed using resonant particle spectroscopy techniques and provides a measurement of @xmath5-decaying states in @xmath6c .
excited states are observed at 12.0 , 13.4 , 14.1 , 14.6 , 15.2 , 16.8 , 17.9 , 18.7 , 21.3 and 23.9 mev .
this study provides the first measurement of the three highest energy states .
angular distribution measurements have been performed and have been employed to indicate the transferred angular momentum for the populated states .
these data are compared with recent speculations of the presence of chain - like structures in @xmath6c . , , , , , , , ,
, , nuclear reactions , @xmath0li+@xmath1be , @xmath6c levels deduced , @xmath2he+@xmath1be decay , angular distributions , molecular structure |
the electron - phonon coupling strongly affects vibrational modes of nano- and micro - electro - mechanical systems .
much interest have attracted the effects of this coupling related to the reduced dimensionality of the electron system , as they make it possible to reveal interesting consequences of the electron correlations at the nanoscale , the coulomb blockade being a simple example , cf .
@xcite and references therein .
much less attention has been paid to the consequences of the electron - phonon coupling , which are related to the discreteness of the vibrational spectrum of a nanosystem , but emerge in the absence of size quantization of the electron motion .
one of such consequences , which we study in this paper , is the coupling - induced change of the vibration nonlinearity .
strong nonlinearity is a generic feature of vibrations in small systems @xcite .
its easily accessible manifestation is the dependence of the mode frequencies on the vibration amplitudes .
this dependence corresponds to the self - action " of the mode , and its familiar analog in bulk crystals are acoustic solitons @xcite ; however , the nonlinearity required for observing such solitons usually is sufficiently strong only for high - frequency phonons .
also , the change of the eigenfrequency with the mode amplitude is of interest for modes with a discrete frequency spectrum , such as standing waves in mesoscopic systems , but not for propagating waves with a quasi - continuous spectrum . much attention
have been recently attracting si - based nano- and micromechanical systems , see @xcite and references therein . in such systems
there was observed an unexpectedly large change of the amplitude dependence of the vibration frequency with the varying electron density @xcite .
when the doping level was increased from @xmath0 @xmath1 to @xmath2 @xmath1 , the nonlinearity parameter increased by more than an order of magnitude . moreover ,
the nonlinearity change was different for the vibrational modes with different spatial structure . in this paper
we develop a theory of the nonlinearity of vibrational modes in semiconductor nano- and micro - mechanical systems with high electron density .
we show that the electron - phonon coupling can lead to a strong self - action of the vibrational modes , which in turn significantly modifies the amplitude dependence of the mode frequencies .
we find the dependence of the effect on the electron density and temperature . for bulk semiconductors , the effect of the electron - phonon coupling on the elastic properties , including the three - phonon coupling ,
was first analyzed by keyes @xcite .
the analysis referred to @xmath3-ge and was based on the deformation potential approximation .
the idea was that deformation lifts the degeneracy of the equivalent electron valleys , which leads to a redistribution of the electrons over the valleys . in turn ,
such redistribution changes the speed of sound depending on the direction and polarization of the sound waves and also affects the sound speed in the presence of uniaxial stress .
this theory was extended to silicon and the corresponding measurements were done by hall @xcite .
however , hall also observed the change of the speed of transverse sound waves and the effect of stress on sound propagation in the geometries , where these effects are due to shear deformation and do not arise in the deformation potential model .
a theory of the change of the linear shear elastic constant in silicon due to the intervalley redistribution of the electrons was developed by cerdeira and cardona @xcite .
as we show , in mesoscopic systems the strain - induced redistribution of the electrons over the valleys of the conduction band leads to the previously unexplored strong fourth - order nonlinearity of the vibrational modes .
this nonlinearity gives a major contribution to the amplitude dependence of the vibration frequency .
the redistribution also leads to a temperature dependence of the frequencies .
the magnitudes of the effects sensitively depends on the mode structure .
we describe them for several types of modes , including those studied in the experiment @xcite and qualitatively compare the results with the observations .
the theoretical results refer to both degenerate and nondegenerate electron systems .
specific calculations are done for silicon resonators . in sec .
[ sec : model ] we give , for completeness , the expressions for the mode normalization and the amplitude - dependent frequency shift of coupled nonlinear modes in a nano- or micro - system . in sec .
[ sec : e - ph_coupling ] and appendix [ appendixa ] we provide expressions for the electron - phonon coupling induced change of the elasticity parameters , including the parameters of quartic nonlinearity . in sec .
[ sec : explicit_form ] we discuss the asymptotic behavior of the parameters of quartic nonlinearity for low and high electron density and give their explicit form for silicon . in sec .
[ sec : simple_modes ] we calculate the nonlinear frequency shift for several frequently used vibrational modes in single - crystal silicon systems and show the dependence of this shift on the electron density and temperature .
the explicit analytical expressions are given in appendices [ sec : lame ] and [ sec : extension ] .
[ sec : conclusions ] contains concluding remarks .
of primary interest for nano- and micro - mechanical systems are comparatively low - frequency modes with wavelength on the order of the maximal size of the system .
examples are provided by long - wavelength flexural modes of nanotubes , nanobeams , and nano / micro - membranes , or acoustic - type modes in microplates or beams .
these modes are easy to excite and detect .
we will enumerate them by index @xmath4 .
their dynamics is described by the elasticity theory @xcite .
the spatial structure of the displacement field of a mode @xmath5 in the harmonic approximation is determined by the boundary conditions .
we will choose @xmath5 dimensionless , so that in our finite - size system @xmath6 here , @xmath7 is the volume of the system .
we assumed that the mode eigenfrequencies @xmath8 are nondegenerate ; including degenerate modes is straightforward . for simplicity
, we also assumed that the system is spatially uniform ; an extension to spatially nonuniform systems is straightforward as well .
we emphasize the distinction of the normalization ( [ eq : normalization ] ) from the conventional normalization for bulk crystals , where @xmath4 corresponds to the wave vector and the branch number , and the normalization integral is independent of the volume .
the normalization ( [ eq : normalization ] ) is convenient for the analysis of low - frequency modes with the discrete spectrum characteristic of mesoscopic systems .
such modes are standing waves , and therefore vectors @xmath9 can be chosen real .
the low - frequency part of the displacement can be written as @xmath10 functions @xmath11 give the mode amplitudes . in the harmonic approximation
the dynamics of the standing waves is described by the hamiltonian @xmath12 where @xmath13 is the momentum of mode @xmath4 and @xmath14 is the mass of the system .
the anharmonicity of the crystal leads to mode - mode coupling . within the elasticity theory this coupling
is described by the terms in the hamiltonian , which are cubic and quartic in the strain tensor .
we will not consider higher - order terms , which are small for the mode amplitudes of interest . from the expansion
, we obtain the nonlinear part of the hamiltonian in the form @xmath15 equation is essentially an expansion in the ratio of the mode amplitudes to their characteristic wavelength , which is of the order of the appropriate linear dimension of the system .
this is why mesoscopic systems are of particular interest , as here vibrations of low - frequency eigenmodes become nonlinear for already small vibration amplitudes . a familiar consequence of nonlinearity in nano- and micromechanical systems is the dependence of the vibration frequency of a mode on its own amplitude and on the amplitudes of other modes , see ref . for a review . in particular
, the change @xmath16 of the mode frequency due to the vibrations of the mode itself , @xmath17 , is @xcite @xmath18a_\nu^2,\end{aligned}\ ] ] where @xmath19 and we kept the terms of the first order in @xmath20 and the second order in @xmath21 . the nonlinear mode coupling ( [ eq : anharmonic_hamiltonian ] ) leads also to the frequency shift due to thermal vibrations of the modes .
the dominating contribution to this shift for low - frequency modes comes from their coupling to modes with frequencies @xmath22 , which have a much higher density of states .
this shift is described by an expression that is similar to eq .
( [ eq : duffing_shift ] ) with @xmath23 replaced by @xmath24 and placed under the sum over @xmath25 , in the classical limit .
we will consider the vibration nonlinearity due to the electron - phonon coupling in multi - valley semiconductors with cubic symmetry , silicon and germanium being the best known examples .
in such semiconductors , the energy valleys of the conduction band are located at high - symmetry axes of the brillouin zone .
strain lifts the symmetry and thus the degeneracy of the valleys .
the simplest mechanism of the electron - phonon coupling is the deformation potential . here
, the energy shift @xmath26 of valley @xmath27 is determined by the deformation potential parameters @xmath28 and @xmath29 of the coupling to a uniaxial strain along the symmetry axis of the valley and to dilatation , respectively . in terms of the strain tensor @xmath30
we have @xmath31 , where @xmath32 , with @xmath33 being the unit vector along the symmetry axis of the valley .
we use the hat symbol to indicate tensors and symbol @xmath34 " to indicate tensor products .
the analysis below is not limited to the deformation potential approximation .
an important extension will be discussed using silicon as an example .
we assume that the strain varies in time and space slowly compared to the reciprocal rate of intervalley electron scattering and the intervalley scattering length , respectively .
then the electron system follows the strain adiabatically .
the electron density @xmath35 in valley @xmath27 is decreased or increased depending on whether the bottom of the valley goes up or down . in the single - electron approximation and for the deformation potential coupling ,
the electron free energy density for a given strain is @xmath36 + n^{(\alpha ) } ( \rb ) \xi^{(\alpha)}_{ij } \ep_{ji}(\rb)\}$ ] where @xmath37 $ ] is the free energy density for electrons with density @xmath38 in a valley in the absence of coupling to phonons .
the electro - neutrality requires that the total electron density summed over the valleys be constant . the free energy density @xmath39 has to be minimized over @xmath35 to meet this constraint .
this gives the change of the electron chemical potential @xmath40 due to strain @xmath41 .
the resulting increment of the electron free energy density has the form of a series expansion in the strain tensor , @xmath42 here @xmath43 , and @xmath44 are tensors of ranks 2 , 4 , 6 , and 8 , respectively . they are contracted with the tensor products of the strain tensor @xmath41 .
respectively , @xmath45 are the electronic contributions to the linear ( for @xmath46 ) and nonlinear ( for @xmath47 ) elasticity parameters of the crystal .
these contributions are isothermal , but since the change of the mode frequencies from the electron - phonon coupling is small and the nonlinearity is also small , the difference with the adiabatic expressions can be disregarded . to the third order in @xmath41 the expression for @xmath48 in terms of the shift of the valleys
was found by keyes @xcite in the analysis of sound wave propagation .
however , to find the parameters of the quartic nonlinearity of resonant modes in small systems , which is of primary interest to us , we also need to keep quartic terms in eq .
( [ eq : free_energy_expansion ] ) . as seen from the explicit form of the parameters of the expansion ( [ eq : free_energy_expansion ] ) given in appendix [ appendixa ] , @xmath49^{k-1}$ ] ( @xmath50 ) , where @xmath51 is the electron chemical potential in the absence of strain
; it is determined by the total ( summed over the valleys ) electron density @xmath3 .
of central importance for the analysis is that parameter @xmath52 for electron densities @xmath53 and room temperatures , i.e. @xmath54 as a consequence , the coefficients at the nonlinear in @xmath41 terms in eq .
( [ eq : free_energy_expansion ] ) quickly increase with the increasing order of the nonlinearity [ the overall series ( [ eq : free_energy_expansion ] ) is converging fast because of the smallness of the strain tensor ] .
the increase of @xmath45 with @xmath55 allows us to keep in @xmath41 only the terms linear in the lattice displacement , i.e. , to set @xmath56 , where @xmath57 and @xmath58 are the components of the displacement and the coordinates , respectively . indeed , in this case a @xmath55th term of the series ( [ eq : free_energy_expansion ] )
is of order @xmath55 in the displacement . if we included the quadratic in @xmath59 term into one of the @xmath41 tensors in the @xmath55th term , this term would become of order @xmath60 in the displacement .
however , for linear @xmath41 the @xmath61th term in the series ( [ eq : free_energy_expansion ] ) is also of the @xmath61th order in the displacement , but is larger by factor @xmath62 . for linear @xmath41 ,
the total strain is a sum of partial contributions of strain from individual modes . for mode @xmath4 ,
such partial contribution is expressed in terms of the scaled displacement @xmath5 [ see eq .
( [ eq : displacement ] ) ] as @xmath63 , where @xmath64 $ ] .
we note that , in contrast to the dimensionless strain tensor @xmath41 , tensor @xmath65 has dimension [ length]@xmath66 . from eq .
( [ eq : free_energy_expansion ] ) we find the electronic contributions to the nonlinearity parameters @xmath67 in hamiltonian ( [ eq : anharmonic_hamiltonian ] ) , @xmath68 where @xmath69 ; tensors @xmath70 are independent of @xmath71 . similarly , the electronic contribution to the eigenfrequency is @xmath72 generally , the term @xmath73 leads to mode mixing ; however , if the mode frequencies are nondegenerate , this mixing is weak and can be disregarded , to the leading order in the electron - phonon coupling .
one can see that the effect of the static stress @xmath74 can be disregarded as well .
the frequency change ( [ eq : frequency_change ] ) depends on temperature because of the temperature dependence of @xmath75 .
the nonlinearity ( [ eq : nonlin_params ] ) also leads to a temperature dependence of the mode eigenfrequency .
together they modify the temperature dependence of the mode eigenfrequencies compared to that of undoped crystals .
this modification often weakens the temperature dependence of the eigenfrequencies , which proves very important for applications of micro - mechanical systems in devices that work in a broad temperature range @xcite .
equations ( [ eq : free_energy_expansion ] ) - ( [ eq : frequency_change ] ) are generic and apply beyond the deformation potential approximation .
this is of particular importance for silicon . here ,
the electron band valleys lie on the @xmath76-axes close to the @xmath77-points on the zone boundaries where two electron energy bands cross .
lattice strain can lead to a band splitting at @xmath77-points and a shift of the valleys @xcite .
importantly , this shift results from a shear strain , which does not lead to a linear in the strain shift in the deformation potential approximation .
the valley shift is quadratic in @xmath41 in this case , as explained in appendix [ appendixa ] , which corresponds to an effectively two - phonon coupling .
the coupling parameter @xmath78 is quadratic in the strain - induced band splitting , see eq .
( [ eq : general_shift ] ) .
it is large , much larger than the constant @xmath28 .
therefore the arguments given below eq .
( [ eq : strong_coupling ] ) apply in this case as well . for purely shear strain in silicon , terms of odd order in @xmath41 in @xmath48 , eq .
( [ eq : free_energy_expansion ] ) , vanish .
[ cols="^,^,^,^ , < " , ] [ table : tensors ]
tensors @xmath79 can be obtained by minimizing the free energy density of the electron system for a given strain and expanding the result in a series in @xmath41 . a general procedure that allows one to find the components @xmath79 for @xmath80
is described in appendix [ appendixa ] . using the symmetry arguments ,
the elasticity tensors are conveniently written in the contracted ( voigt ) notation where the symmetric strain tensor is associated with a six - component vector .
then the nonlinear elasticity tensors @xmath81 and @xmath82 become tensors of rank three and four in the corresponding vector space .
we use notation @xmath83 for tensors @xmath84 in these notations to emphasize that we are calculating corrections to the nonlinear elasticity tensors due to the electron - phonon coupling .
the explicit expressions for the nonlinear elasticity tensors @xmath85 are given in table [ table : tensors ] .
they refer to silicon and include the contributions that come from both the deformation potential coupling and from the splitting of the electron bands due to shear strain . in the deformation potential approximation ,
the components of the third - rank tensor @xmath85 , which determine the cubic in the strain terms in the free energy , were found earlier @xcite . therefore we give only the components that contain a contribution from shear strain .
the fourth - rank tensor @xmath85 determines the quartic in the strain terms in the free energy and has not been discussed before , to the best of our knowledge .
we give all independent components of this tensor .
it is expressed in terms of the derivative of the electron density @xmath3 over the chemical potential in the absence of strain @xmath51 , which is a familiar thermodynamic characteristic .
it is intuitively clear that the considered effect of the change of the electron density in different valleys in response to strain should be related to the derivative @xmath86 . interestingly , because we consider nonlinear response to strain , the expressions in table [ table : tensors ] contain also higher - order derivatives of @xmath3 over @xmath51 .
as we will see , this leads to a nontrivial behavior of the nonlinear frequency shift with varying temperature and density .
the considered mechanism of the strain - induced inter - valley electron redistribution does not contribute to the components @xmath87 and @xmath88 , therefore @xmath89 .
the expressions for @xmath85 simplify in the case of low doping ( or high temperature ) , where the electron gas is strongly nondegenerate , and in the opposite case of a strongly degenerate electron gas . for a nondegenerate gas , where the chemical potential in the absence of strain is @xmath90 ,
we have in table [ table : tensors ] @xmath91 with @xmath92 .
the @xmath51-dependent factors @xmath93 in @xmath94 and its derivatives cancel each other in the expressions for @xmath85 and drop out from these expressions .
the dependence of @xmath85 on density is then just linear , @xmath95 .
parameters @xmath96 in table [ table : tensors ] depend only on temperature , @xmath97 and @xmath98 .
the decrease of the nonlinear elasticity parameters with increasing temperature in a nondegenerate electron gas is easy to understand .
the effect we consider is determined by the competition between the energetically favorable unequal population of the electron energy valleys in a strained crystal and the entropically more favorable equal valley population . with increasing temperature
the entropic factor becomes stronger , leading to a smaller population difference and thus smaller effect of the electron system on the vibrations . for strong doping , where @xmath99 , we have @xmath100 , and then @xmath101 with @xmath92 .
therefore parameters @xmath102 in table [ table : tensors ] become temperature independent , with @xmath103 , and @xmath104 .
the results on the asymptotic behavior of the corrections to nonlinear elasticity are not limited to silicon . since parameters @xmath105
are given by the coefficients in the general expansion of the free energy in strain , ( [ eq : total_f ] ) , these results can be applied to the nonlinear elasticity induced by the electron - phonon coupling in other multi - valley semiconductors . to illustrate this point , in appendix [ sec : germanium ]
we give @xmath85 tensor in germanium .
the difference between the asymptotic behavior of the tensors @xmath85 in the limits of nondegenerate and strongly degenerate electron gas can lead to a peculiar density and temperature dependence of the nonlinear frequency shift of the vibrational modes .
it comes from the coefficients @xmath102 containing higher - order derivatives of @xmath3 with respect to @xmath51 . in the transition region
@xmath106 , thinking of the competition between the entropic and energetic factors does not provide a simple insight into the behavior of @xmath85 , as both the energy and the entropy are complicated functions of density and temperature .
the nonlinear elasticity tensors in table [ table : tensors ] give the doping - induced contributions to the nonlinearity parameters of the eigenmodes of micro- and nanomechanical systems .
these contributions are described by eq .
( [ eq : nonlin_params ] ) . as mentioned before , an important characteristic of the mode nonlinearity is the dependence of the mode frequency on the vibration amplitude . to the leading order
, it is given by eq .
( [ eq : duffing_shift ] ) .
this dependence has a contribution from the nonlinearity of an undoped crystal , which is quadratic in the parameters of the cubic nonlinearity ; for example , if the latter is described by the grneisen constant , the corresponding contribution is quadratic in this constant .
it is typically small .
there is also a contribution from the quartic nonlinearity ; the parameters of such nonlinearity are not known in undoped crystals and are not expected to be large . respectively , the amplitude dependence of the vibration frequency for low - frequency modes in weakly doped single - crystal micro - mechanical systems is relatively weak @xcite .
a feature of the doping - induced nonlinearity described by table [ table : tensors ] is that the quartic in the strain term in the free energy has a large coefficient compared to the cubic term , cf .
( [ eq : strong_coupling ] ) and the discussion below this equation .
therefore , in eq .
( [ eq : duffing_shift ] ) for the amplitude dependence of vibration frequency one can keep only the duffing nonlinearity constant @xmath107 .
the contribution from the cubic nonlinearity terms @xmath108 can be disregarded . for a mode @xmath4 ,
the doping - induced contribution to @xmath107 is equal to @xmath109 in eq .
( [ eq : nonlin_params ] ) . to find the dependence of the mode frequency on the vibration amplitude we go through the following steps .
first , we find the normal modes of interest for the given geometry of the system , with account taken of the boundary conditions , and normalize the displacements @xmath5 as indicated in eq .
( [ eq : normalization ] ) .
we use @xmath5 to find the strain tensor @xmath110 .
the result is substituted into eq .
( [ eq : nonlin_params ] ) and is convoluted with tensor @xmath82 , giving the value of @xmath107 , which is then used in eq .
( [ eq : duffing_shift ] ) to find the frequency dependence on the vibration amplitude @xmath16 .
of particular interest is the relative frequency shift @xmath111 . to find this shift to the leading order
, one can disregard nonlinearity when calculating the eigenfrequency @xmath8 . then , from eq .
( [ eq : duffing_shift ] ) , @xmath112 where @xmath113 is the full tensor of linear elasticity , which includes the major term of the linear elasticity of the undoped crystal and the doping - induced correction @xmath75
. an important feature of the relative shift @xmath111 is its scaling with the size of the system . the vibration amplitude @xmath114 in eq .
( [ eq : scaling ] ) can be scaled by the lateral dimension @xmath115 , for example the length of a nanobeam or a nanowire for an extension mode , or the size of the square for a lam mode , or the diameter of a disk for a breathing mode in a disk . respectively , we write @xmath116 .
then , if one takes into account the explicit form ( [ eq : nonlin_params ] ) of the parameter @xmath117 , one finds from eq .
( [ eq : scaling ] ) that the ratio @xmath118 is independent of the system size for the aforementioned modes . in this estimate
we used that the tensors @xmath119 are material parameters and are independent of the geometry .
we also used that the modes of interest have typical wavelength @xmath120 , and therefore @xmath65 scales as @xmath121 .
most of the experiments in nano- and micromechanics are done with nanobeams , nanowires , membranes , or thin plates . in such systems
the thickness is much smaller than the length or , in the case of membranes or plates , the lateral dimensions .
then , from the boundary condition of the absence of tangential stress on free surfaces @xcite , it follows that the strain tensor @xmath41 weakly depends on the coordinate normal to the surface .
this simplifies the denominator in eq .
( [ eq : scaling ] ) , making it proportional to the thickness .
similarly , from eq .
( [ eq : nonlin_params ] ) @xmath107 is also proportional to the thickness , and the thickness drops out of eq .
( [ eq : scaling ] ) .
the explicit expressions for @xmath122 and @xmath107 that determine the denominator and the numerator in eq .
( [ eq : scaling ] ) , respectively , are given in appendices [ sec : lame ] and [ sec : extension ] for lam and extension modes .
these expressions are cumbersome , and it is convenient to use symbolic programming to obtain them . the scaled ratio @xmath118 that characterizes the relative nonlinear frequency shift is shown in fig .
[ fig : n_dependence ] for several modes that are often used in single - crystal silicon mems .
this ratio depends on the type of the mode and the crystal orientation .
figure [ fig : n_dependence ] refers to high - symmetry crystal orientations , in which case the modes have a comparatively simple spatial structure and the surfaces can be made smooth .
we used the values @xmath123 ev @xcite , @xmath124 ev , the effective mass for density of states @xmath125 @xcite , and the temperature - dependent linear elasticity parameters given in ref .
@xcite .
figure [ fig : n_dependence ] shows that the electron - redistribution induced nonlinearity of vibrational modes is very strong . for the ratio of the vibration amplitude to the system size @xmath126 and the mode eigenfrequency @xmath127 mhz ,
the frequency change can be as a large as @xmath128 khz .
this explains , qualitatively , the observations @xcite .
a quantitative comparison with the experiment @xcite is complicated , as the observations refer to different samples .
our preliminary results show an excellent quantitative agreement with the data obtained for the same sample at different temperatures and for different types of modes @xcite .
the nonlinear frequency shift displays several characteristic features , as seen from fig .
[ fig : n_dependence ] .
one of them is the strong dependence of the shift on the type of the mode and the crystal orientation .
for both the lam and the extension mode , the shift is much stronger for crystals cut out in @xmath76 direction than in @xmath129 direction .
this is a consequence of the electron energy valleys lying along the @xmath76 axes , making the system more responsive " to the lattice displacement along these axes .
interestingly , in the both configurations the shifts for the lam modes are larger than for the extension modes .
a somewhat unexpected feature is the nonmonotonic dependence of the nonlinear frequency shift on the electron density and temperature .
the nonmonotoncity occurs in the range where the electron system is close to degeneracy , @xmath130 , and it strongly depends on the crystal orientation .
it is much stronger for crystals cut in @xmath76 than @xmath129 directions . for a crystal cut in @xmath129 direction ,
both the density and temperature dependence of the shift are monotonic in the case of the lam mode , whereas for the extension mode the nonmonotonicity is weak .
the nonmonotonicity of the frequency shift stems from the behavior of the parameters @xmath131 in the range @xmath106 . as seen from table [ table : tensors ] , parameter @xmath132 exponentially increases with the increasing @xmath133 for negative @xmath133 , but for large positive @xmath133 it falls off as @xmath134 .
it has a pronounced maximum for @xmath135 .
parameter @xmath136 also displays a maximum , which occurs for @xmath137 .
in contrast , parameters @xmath138 depend on @xmath133 monotonically . the results of appendices [ sec : lame ] and [ sec : extension ] show that , for the lam and extension modes in crystals cut in @xmath76 direction , the relative shift @xmath111 is determined by coefficient @xmath132 , which explains the nonmonotonicity of the shift . for crystals cut in @xmath129 , the shift of the lam mode is fully determined by coefficient @xmath139 and is monotonic , whereas for the extension mode the expression for the shift has contributions from @xmath132 , @xmath136 , and @xmath139 that partly compensate each other , leading to a comparatively small shift all together and its weak nonmonotonicity .
the results of this paper show that the electron - phonon coupling strongly affects the nonlinearity of vibrational modes in semiconductor - based nano- and micromechanical systems .
the mechanism of the effect is the strain - induced redistribution of the electrons between the valleys of the conduction band .
the redistribution results from lifting the degeneracy of the electron energy spectrum by the strain from a vibrational mode .
the analysis refers to the range of temperatures where the rate of intervalley scattering strongly exceeds the frequencies of the considered modes . in this case
the valley populations follow the strain adiabatically .
the change of the valley populations is a strongly nonlinear function of the strain tensor .
the respective expansion of the free energy in the strain is an expansion in the strain multiplied by the ratio of the electron - phonon coupling energy ( in particular , the deformation potential ) to the chemical potential of the electron system or the temperature .
this ratio is large , @xmath140 .
it is this parameter that makes the nonlinearity of the vibrational modes in doped semiconductor structures strong . of special interest in nano- and micromechanical systems
is the amplitude dependence of the vibration frequency . to the leading order
, it is determined by the quartic terms in the expansion of the free energy in strain .
these terms are comparatively large in doped crystals .
we have calculated the nonlinear elasticity tensor that describes the electron contribution to the terms in the free energy , which are quartic in the strain .
the explicit expressions for the tensor components refer to semiconductors with the valleys on @xmath76 axes , in particular , to silicon .
we have also found this tensor for germanium . in silicon , along with the deformation potential coupling , an important role is played by the coupling to shear strain .
such strain lifts the band degeneracy at the zone boundary and is effectively described by a two - phonon coupling .
we show that this coupling also leads to strong nonlinearity of vibrational modes .
the parameter of the electron coupling to shear strain in silicon is not easy to access in the experiment @xcite .
measurements of the nonlinear frequency shift provide a direct means for determining this parameter . in particular , the nonlinear frequency shift of the fundamental lam mode in a silicon plate cut along @xmath129 axes
is determined by this parameter only , except for small corrections from the nonlinearity of the undoped crystal .
we found that the nonlinear frequency shift strongly depends on the type of a vibrational mode and the crystal orientation .
we also found that the ratio of the frequency shift to the squared vibration amplitude can be profoundly nonmonotonic as a function of electron density and temperature .
the results provide an insight into the experimentally observed strong mode nonlinearity in doped crystals @xcite . in terms of applications , they enable choosing the appropriate range of doping and the temperature regime to optimize the operation of nano- and micromechanical resonators
. we are grateful to t. kenny for attracting our attention to the problem and for stimulating discussions .
we benefited from useful discussions with j. atalaya , d. heinz , p. polunin , s. w. shaw , and y. yang .
this research was supported in part by the us defense advanced research projects agency ( grant no .
fa8650 - 16 - 1 - 7600 ) .
the major effect of a strain on the electron free energy comes from the shift of the energy valleys .
we will assume that valley @xmath27 is shifted in energy by @xmath26 and the shift is small , @xmath141 , where @xmath51 is the chemical potential in the absence of strain .
we further assume that the vibrations are slow compared to the time it takes the electron system to come , locally , to thermal equilibrium for given values of @xmath26 , i.e. , the temperature and the chemical potential are the same in all valleys . since for high electron densities the thermal conductivity is high , the change of the temperature compared to the ambient temperature can be disregarded ; also , as mentioned in the main text , the electron density @xmath3 summed over all valleys is constant .
expanding the electron free energy density to the 4th order in the strain - induced shifts @xmath26 , we find that , in an @xmath142-valley semiconductor , the change @xmath48 of the free energy density is @xmath143 + \frac{1}{6}\frac{f''_{1/2}}{f_{1/2}}\left[\overline { \delta_\ep^3 } -3\overline { \delta_\ep^2}\;\overline\delta_\ep + 2(\overline\delta_\ep)^3 \right ] \nonumber\\ & + \frac{1}{8}\frac{{f''_{1/2}}^2}{f_{1/2}f'_{1/2}}\left[(\overline { \delta_\ep^2})^2 -2\overline { \delta_\ep^2}\,(\overline\delta_\ep)^2 + ( \overline\delta_\ep)^4 \right]+ \frac{1}{24}\frac{f'''_{1/2}}{f_{1/2}}\left [ 4\overline { \delta_\ep^3}\;\overline\delta_\ep - \overline { \delta_\ep^4 } - 6\overline { \delta_\ep^2}\,(\overline\delta_\ep)^2 + 3(\overline\delta_\ep)^4 \right].\end{aligned}\ ] ] here , @xmath144 .
we use the standard notation @xmath145 $ ] ; primes indicate differentiation over @xmath146 , for example , @xmath147 .
function @xmath94 and its derivatives are calculated for @xmath92 .
equation ( [ eq : total_f ] ) immediately gives the tensors @xmath148 of the expansion of the free energy increment ( [ eq : free_energy_expansion ] ) if one expresses the shift @xmath26 of the valleys in terms of the strain tensor . in the deformation potential approximation the relation between @xmath26 and @xmath41
is given in the main text , see also eq .
( [ eq : general_shift ] ) below . in the case of si crystals , which are often used in micromechanical resonators ,
an important contribution to @xmath26 comes from the shear - strain induced splitting of the electron energy bands at the zone boundary .
shear strain does not lead to the valley shift in the deformation potential approximation .
the overall shift of valley @xmath27 , to the lowest order in the coupling that causes it ( i.e. , to the first order in the deformation potential where its contribution is nonzero and to the second order in the band splitting for shear strain ) is @xcite : @xmath149 here we use that silicon has six valleys located at the @xmath76 axes , and we chose the coordinate axes @xmath150 along @xmath76 . respectively , the valley index @xmath27 takes on three values that correspond to the @xmath150 axes ( the valleys lying on the same axis , but in the opposite directions , are equivalent ) .
the strain @xmath151 , which enters the second term in the right - hand side of eq .
( [ eq : general_shift ] ) , is a component of the strain tensor @xmath30 with @xmath152 such that @xmath153 and @xmath154 .
the parameter @xmath155 is the interband matrix element of the electron - phonon coupling calculated for the electron conduction bands @xmath156 and @xmath157 at the @xmath77 point on the boundary of the brillouin zone , where the bands cross ; @xmath158 is the energy separation between the bands @xmath156 and @xmath157 at the value of the wave vector @xmath159 that corresponds to the conduction band minimum .
parameter @xmath78 is the effective deformation potential of two - phonon coupling to shear strain .
the numerical value of @xmath78 is not well known .
the experimental data give @xmath160 ev @xcite and the numerical data on the band splitting give @xmath161 ev @xcite so that @xmath78 is in the range of @xmath162 ev ; this is essentially an order of magnitude estimate . in calculating @xmath48 in eq .
( [ eq : total_f ] ) we kept terms that are quartic in @xmath41 .
the components of the tensors @xmath45 in eq .
( [ eq : free_energy_expansion ] ) are expressed in terms of @xmath48 as @xmath163 tensors @xmath84 are symmetric with respect to the interchange of indices @xmath164 and the pairs @xmath165 .
for the considered long - wavelength strain , tensors @xmath45 are independent of coordinates .
the corrections @xmath166 to the linear elasticity tensors were found previously @xcite and are not discussed in this paper .
in this section we provide the corrections to the nonlinear elastic constants of germanium , which are due to the redistribution of the electrons over the valleys .
germanium has four equivalent valleys in the conduction band , which are located on the boundary of the brillouin zone along @xmath167 axes .
we use the voigt notation and write the components of the corrections to the nonlinear elasticity tensor @xmath85 in the frame where the axes @xmath168 are along the @xmath76 directions of the crystal . using the results of appendix [ appendixa ] , we obtain @xmath169 the notations are the same as in appendix [ appendixa ] and in table [ table : tensors ]
. the electron - phonon coupling does not contribute to the other third- and fourth - order elastic constants .
corrections @xmath170 and @xmath171 for germanium were found by keyes @xcite ; however , his final expression for @xmath171 differs from eq .
( [ eq : ge ] ) by a factor of 4 ( our expressions for @xmath170 coincide with ref .
parameters @xmath172 and @xmath173 have not been found before , to the best of our knowledge . in the limiting cases , corrections @xmath172 and
@xmath173 have the same dependence on temperature and electron density as constant @xmath132 discussed in sec .
[ sec : limiting ] .
we consider a square plate with side @xmath115 and thickness @xmath174 made out of a single crystal with cubic symmetry .
if the crystal is cut out along @xmath76 or @xmath129 axes , one of the simplest modes is the first lam mode @xcite .
the normalized displacement field is @xmath175 here , @xmath146 and @xmath176 axes are in the lateral plane along the sides of the square , axis @xmath177 is perpendicular to the plate and @xmath178 . calculating the strain tensor for the displacement ( [ eq : lame_displacements ] ) and substituting the expressions into eqs .
( [ eq : nonlin_params ] ) and the relation @xmath179 for the plate cut out along @xmath180 axes we obtain , in voigt notation for the elasticity tensors , @xmath181 if we consider silicon and take into account only the contribution @xmath85 to the nonlinear elasticity tensor @xmath182 , with the account taken of table [ table : tensors ] , the expression for @xmath107 simplifies to @xmath183 for the lam mode cut along the @xmath129 axis , if the tensors are calculated in the axes @xmath76 , we have @xmath184 note that only coupling to shear strain contributes to the nonlinearity parameter @xmath107 in this case .
we consider the fundamental extension mode in a thin beam of length @xmath115 with a rectangular cross - section of area @xmath185 .
the beam is cut along a symmetry axis , and the sides are also along symmetry planes of a cubic crystal . from the free - surface boundary conditions ,
the normalized displacement field is @xcite : @xmath186 this expression takes into account transverse compression that accompanies beam extension and uses the smallness of the beam cross - section ; corrections @xmath187 are disregarded .
the transverse compression in a cubic crystal cut in a symmetric direction is described by poisson s ratios @xmath188 and @xmath189 .
generally , they do not coincide . in eq .
( [ eq : displace_extension ] ) the transverse coordinates @xmath176 and @xmath177 are counted off from the center of the beam for the longitudinal direction of the beam @xmath76 and the sides parallel to @xmath190 planes , the poisson parameters are equal , @xmath191 and @xmath192 . in this case
( [ eq : nonlin_params ] ) and ( [ eq : frequency ] ) give @xmath193.\end{aligned}\ ] ] the expression for @xmath107 is simplified if in the nonlinear elasticity tensors we take into account only the contribution from the electron - phonon coupling as given in table [ table : tensors ] and also allow for the interrelation between different components of the tensor @xmath85 .
then for a silicon beam @xmath194 for extension along @xmath129 axis , with one side parallel to @xmath190 plane and the other side parallel to @xmath195 plane , the poisson s ratios @xmath196 and @xmath197 are given in ref .
@xcite . then eqs . ( [ eq : nonlin_params ] ) and ( [ eq : frequency ] ) give @xmath198.\end{aligned}\ ] ]
if in the nonlinear elasticity tensor @xmath182 we take into account only the contribution @xmath85 from the electron - phonon coupling , in the case of a silicon beam the expression for @xmath107 simplifies to @xmath199 expressions ( [ eq : extension_100 ] ) and ( [ eq : extension_110 ] ) were generated using a computer code to calculate the sums and integrals in eq .
( [ eq : nonlin_params ] ) .
35ifxundefined [ 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.92.166801 [ * * ( ) , 10.1103/physrevlett.92.166801 ] link:\doibase 10.1103/physrevb.70.193305 [ * * , ( ) ] @noop * * , ( ) link:\doibase 10.1103/physrevb.76.165317 [ * * , ( ) ] @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) link:\doibase 10.1103/physrevlett.110.066804 [ * * , ( ) ] @noop _ _
( , ) in @noop _ _ , ( , ) pp . @noop * * , ( ) http://stacks.iop.org/1742-6596/92/i=1/a=012002 [ * * , ( ) ] link:\doibase 10.1038/srep03244 [ * * , ( ) ] @noop * * , ( ) in @noop _ _ ( ) pp .
@noop * * , ( ) @noop * * , ( ) link:\doibase 10.1103/physrev.161.756 [ * * , ( ) ] @noop * * , ( ) @noop _ _ , ed .
( , ) @noop _ _ , ed . ( , ) @noop * * , ( ) @noop * * , ( ) @noop _ _ ( , ) @noop * * , ( ) @noop _ _ ( , ) @noop * * , ( ) @noop * * , ( ) @noop ( ) @noop * * , ( ) @noop * * , ( ) @noop _ _ ( , ) http://dx.doi.org/10.1038/32842 [ * * , ( ) ] | we study the effect of the electron - phonon coupling on vibrational eigenmodes of nano- and micro - mechanical systems made of semiconductors with equivalent energy valleys .
we show that the coupling can lead to a strong mode nonlinearity .
the mechanism is the lifting of the valley degeneracy by the strain .
the redistribution of the electrons between the valleys is controlled by a large ratio of the electron - phonon coupling constant to the electron chemical potential or temperature .
we find the quartic in the strain terms in the electron free energy , which determine the amplitude dependence of the mode frequencies .
this dependence is calculated for silicon micro - systems .
it is significantly different for different modes and the crystal orientation , and can vary nonmonotonously with the electron density and temperature . |
within the framework of quantum chromodynamics ( qcd ) , inelastic scattering between a proton and antiproton is described as a hard collision between their constituents ( partons ) .
after the collision , the outgoing partons manifest themselves as localized streams of particles or `` jets '' . predictions for the inclusive jet cross section have improved in the early nineties with next - to - leading order ( nlo ) perturbative qcd calculations @xcite and new , accurately measured parton density functions ( pdf)@xcite .
d has recently measured and published @xcite the cross section for the production of jets as a function of the jet energy transverse to the incident beams , @xmath7 .
the measurement is based on an integrated luminosity of about 92 pb@xmath1 of @xmath2 hard collisions collected with the d detector @xcite at the fermilab tevatron collider .
this result allows a stringent test of qcd , with a total uncertainty substantially reduced relative to previous results @xcite .
we also measure the ratio of jet cross sections at two center - of - mass energies : 630 ( based on an integrated luminosity of about 0.537 pb@xmath1 ) and 1800 gev @xcite .
experimental and theoretical uncertainties are significantly reduced in the ratio .
this is due to the large correlation in the errors of the two cross section measurements , and the suppression of the sensitivity to parton distribution functions ( pdf ) in the prediction .
the ratio of cross sections thus provides a stronger test of the matrix element portion of the calculation than a single cross section measurement alone .
previous measurements of cross section ratios have been performed with smaller data sets by the ua2 and cdf @xcite experiments .
jets are reconstructed using an iterative jet cone algorithm with a cone radius of @xmath8=0.7 in @xmath9@xmath10 space , ( pseudorapidity is defined as @xmath11 $ ] ) @xcite . the offline data selection procedure , which eliminates background caused by electrons , photons , noise , or cosmic rays , follows the methods described in refs . @xcite .
the jet energy scale correction , described in @xcite , removes instrumentation effects associated with calorimeter response , showering , and noise , as well as the contribution from spectator partons ( underlying event ) .
the energy scale corrects jets from their reconstructed @xmath7 to their `` true '' @xmath7 on average .
an unsmearing correction is applied later to remove the effect of a finite @xmath7 resolution @xcite .
the resulting inclusive double differential jet cross sections , @xmath12 , for @xmath13 and @xmath14 ( the second region for comparison to ref .
@xcite ) , are compared with a nlo qcd theoretical prediction @xcite .
discussions on the different choices in the theoretical calculation : pdfs , renormalization and factorization scales ( @xmath15 ) , and clustering algorithm parameter ( @xmath16 ) can be found in refs .
@xcite .
figure [ fig_3 ] shows the ratios @xmath17 for the data ( @xmath18 ) and jetrad nlo theoretical ( @xmath19 ) predictions based on the cteq3 m , cteq4 m and mrst pdf s [ 4,5 ] for @xmath13 .
( the tabulated data for both @xmath13 and @xmath14 measurements can be found in ref .
@xcite . )
the predictions are in good quantitative agreement with the data , as verified with a @xmath20 test , which incorporates the uncertainty covariance matrix @xmath21 . here
@xmath22 and @xmath23 represent the @xmath24-th data and theory points , respectively .
the overall systematic uncertainty is largely correlated .
table [ tab : table2 ] lists @xmath25 values for several jetrad predictions using various parton distribution functions @xcite .
the predictions describe both the @xmath13 and @xmath14 cross section very well .
the measurement by d and cdf are also in good quantitative agreement within their systematic uncertainties @xcite .
-0.8 cm .
@xmath25 comparisons between jetrad and @xmath26 and @xmath27 data for @xmath28 , @xmath29=@xmath30 , and various pdfs .
there are 24 degrees of freedom . [ cols="^,>,>",options="header " , ] -1.2 cm [ tab : table2 ] -1 cm -0.8 cm [ fig_3 ] -2 cm
d has made a preliminary measurement of the pseudorapidity dependence of the inclusive jet cross section .
figure [ fig : forward ] shows the ratios @xmath17 for the data ( @xmath18 ) and jetrad nlo theoretical ( @xmath19 ) predictions using the cteq3 m pdf set for @xmath32 and @xmath33 .
the measurements and the predictions are in good qualitative agreement .
a detailed error analysis is currently being completed .
a simple parton model would predict a jet cross section that scales with center - of - mass energy . in this scenario , @xmath34 , plotted as a function of jet @xmath35 , would remain constant with respect to the center - of - mass energy .
figure [ fig : ratmu ] shows the d measurement of @xmath36 ( stars ) compared to jetrad predictions ( lines ) .
there is poor agreement between data and nlo qcd calculations using the same @xmath15 in the numerator and the denominator ( probability of agreement not greater than 10% ) .
the agreement improves for predictions with different @xmath15 at the two center - of - mass energies @xcite . in conclusion
, we have made precise measurements of jet production cross sections . at @xmath0=1800 gev
, there is good agreement between the measurements and the data .
the ratio of cross sections at @xmath0=1800 and 630 gev , however , differs from nlo qcd predictions , unless different renormalization scales are introduced for the two center - of - mass energies .
w. t. giele , e. w. n. glover , and d. a. kosower , phys .
letters * 73 * , 2019 ( 1994 ) ; s. d. ellis , z. kunszt , and d. e. soper , phys .
letters * 64 * , 2121 ( 1990 ) ; f. aversa _ et al .
_ , phys . rev .
letters * 65 * , ( 1990 ) | we have made a precise measurement of the inclusive jet cross section at @xmath0=1800 gev . the result is based on an integrated luminosity of 92 pb@xmath1 collected at the fermilab tevatron @xmath2 collider with the d detector .
the measurement is reported as a function of jet transverse energy ( 60 gev @xmath3 500 gev ) , and in the pseudorapidity intervals @xmath4 0.5 and [email protected] .
a preliminary measurement of the pseudorapidity dependence of inclusive jet production ( @xmath4 1.5 ) is also discussed .
the results are in good agreement with predictions from next to leading order ( nlo ) quantum chromodynamics ( qcd ) .
d has also determined the ratio of jet cross sections at @xmath0=630 gev and @xmath0=1800 gev ( @xmath6 ) .
this preliminary measurement differs from nlo qcd predictions . |
there are many physically interesting examples of quantum resonances in atomic physics and quantum optics . to mention one , the state of a cold gas of atoms localized in a trap may be metastable , since the trap may be not strictly confining . in typical bose - einstein condensation experiments ,
the shape of the trap usually varies slowly over time scales small compared to the lifetime of the metastable state , yet larger than a typical relaxation time ( see for example @xcite ) .
this is an example of an adiabatic evolution of shape resonances .
while there has been much progress in a time - independent theory of quantum resonances ( see @xcite ) , there has been relatively little work on a time - dependent theory of quantum resonances ( see @xcite ) .
surprisingly , and inspite of its relevance to the interpretation of many experiments and phenomena in atomic physics , the problem of adiabatic evolution of quantum resonances received very litte attention , so far ; ( but see @xcite ) . in this paper , we study the adiabatic evolution of three general types of quantum resonances .
this is a first step towards a rigorous understanding of resonance- and metastability phenomena , such as hysteresis in magnets and sisyphus cooling of atomic gases ; ( see for example @xcite ) .
we first consider the adiabatic evolution of so called shape resonances .
more specifically , we consider a quantum - mechanical particle in a potential well , say that of a quantum dot or a locally harmonic trap , with the property that the shape of the potential well changes over time scales which are small compared to the time needed for the particle to escape from the well .
the analysis of this problem is based on a precise form of the time - energy uncertainty relation , see @xcite , and the standard adiabatic theorem in quantum mechanics , @xcite . in our approach ,
we obtain an explicit estimate on the distance between the true state of the system and an _ instantaneous _ metastable state .
our approach can also be applied to study the time evolution of the state of an electron in a @xmath0 ion moving in a time - dependent magnetic field which changes over time scales that are small compared to the ionization time of the ion ; ( see @xcite for a discussion of this example in the time - independent situation ) . the second class of examples concerns quantum resonances that appear as isolated complex eigenvalues of spectrally deformed hamiltonians , such as the n - body stark effect ;
( see for example @xcite ) .
our analysis is based on balslev - combes theory for dilatation analytic hamiltonians , @xcite , and on an adiabatic theorem for generators of evolution that are not necessarily normal or bounded , @xcite .
this approach , too , yields explicit estimates on the distance between the true state of the system and an instantaneous metastable state . the third class of examples concerns resonances that emerge from eigenvalues of an unperturbed hamiltonian embedded in the continuous spectrum after a perturbation has been added to the hamiltonian .
typical examples of such resonances arise when a small quantum - mechanical system , say an impurity spin , is coupled to an infinite , dispersive medium , such as magnons ( see for example @xcite for relevant physical models ) .
our approach to such examples is based on an extension of adiabatic theorems _ without a spectral gap condition _ , @xcite .
our results also cover the case of resonance crossings .
further details of applications where our assumptions are explicitly verified for various physical models will appear in @xcite .
* acknowledgements * was is grateful to an anonymous referee for pointing out references @xcite .
in this section , we study the time evolution of the state of a quantum - mechanical particle moving in @xmath1 under the influence of a potential well which is not strictly confining .
the potential well is described by a time - dependent function on @xmath1 @xmath2 where @xmath3 is the adiabatic time scale , @xmath4 is the time , @xmath5 is rescaled time , @xmath6 is a parameter characterizing the width and height of the well , and @xmath7 is a function on @xmath8 that is twice differentiable in @xmath9 and smooth in @xmath10 ; see below for precise assumptions on the potential .
we assume that @xmath3 is small compared to the escape time of the particle from the well .
will be estimated later in this section . ] by introducing an auxiliary adiabatic evolution , we obtain precise estimates on the difference between the true state of the particle and an instantaneous metastable state .
our analysis is based on the generalized time - energy uncertainty relation , as derived in @xcite , and on the usual adiabatic theorem in quantum mechanics , @xcite .
the hilbert space of the system is @xmath11 its dynamics is generated by the time - dependent hamiltonian @xmath12 where @xmath13 is the @xmath14-dimensional laplacian . and
planck s constant @xmath15 we make the following assumptions on the potential @xmath16 for @xmath17 where @xmath18 is an arbitrary , but fixed compact interval of @xmath19 * the origin @xmath20 is a local minimum of @xmath21 for all @xmath17 and , without loss of generality , @xmath22 for @xmath23 * the hessian of @xmath24 at @xmath20 is positive - definite , with eigenvalues @xmath25 and @xmath26 is a constant independent of @xmath27 * consider a smooth function @xmath28 with the properties that @xmath29 for @xmath30 and @xmath31 for @xmath32 where @xmath33 for @xmath34 we define the rescaled function @xmath35 by @xmath36 we assume that , for all @xmath34 @xmath37 uniformly in @xmath38 , where @xmath39 is the hessian of @xmath40 at @xmath20 and @xmath41 is a finite constant independent of @xmath23 * @xmath7 is smooth , polynomially bounded in @xmath42 and bounded from below , uniformly in @xmath23 moreover , @xmath7 is twice differentiable in @xmath23 we also assume that @xmath43 where @xmath44 is a finite constant . note that under these assumptions , @xmath45 is a potential well of diameter of order @xmath46 and height @xmath47 let @xmath48 and @xmath49 where depends on the parameters @xmath50 and @xmath51 but we drop the explicit dependence to simplify notation . ]
@xmath52 .\ ] ] note that @xmath53 where @xmath54.\ ] ] it follows from assumptions ( a3 ) and ( a4 ) , and ( [ deltav ] ) that @xmath55 uniformly in @xmath17 for some polynomial @xmath56 of @xmath57 denote by @xmath58 the projection onto the eigenstates of @xmath59 corresponding to the @xmath60 eigenvalue of @xmath61 it follows from assumptions ( a3 ) and ( a4 ) that , for @xmath62 small enough , @xmath63 is twice differentiable in @xmath64 as a bounded operator for @xmath65.$ ] denote by @xmath66 the propagator generated by @xmath67 which solves the equation exists as a unique unitary operator with domain @xmath68 a common dense core of @xmath69 @xmath70 suppose that the initial state of the system is given by a density matrix @xmath71 which is a positive trace - class operator with unit trace .
then the state of the particle at time @xmath72 is given by the density matrix @xmath73 which satisfies the liouville equation @xmath74\ ] ] and @xmath75 the solution of ( [ liouville ] ) is given by @xmath76 let @xmath77 be an orthogonal projection onto a reference subspace @xmath78 and let @xmath79 denote the probability of finding the state of the particle in the reference subspace @xmath80 at time @xmath81 this probability is given by @xmath82 we are interested in studying the adiabatic evolution of a state of a particle which initially , at time @xmath83 is localized inside the well .
such a state may be approximated by a superposition of eigenstates of @xmath84 ( defined in ( [ h1 ] ) ) .
the initial state of the particle is chosen to be given by @xmath85 where @xmath86 are the eigenprojections onto the states corresponding to the eigenvalues @xmath87 of @xmath88 @xmath89 with @xmath90 for some finite integer @xmath91 we let @xmath92 be the propagator of the auxiliary evolution generated by @xmath61 it is given as the solution of the equation @xmath93 moreover , let @xmath94 then @xmath95 solves the equation @xmath96 where @xmath97 as follows from ( [ dynamicsequation ] ) , ( [ h1 ] ) , ( [ auxevol1 ] ) and ( [ wdef ] ) .
then @xmath98 where @xmath99 we define @xmath100 we have the following proposition .
* proposition 2.1
* _ suppose assumptions ( a1)-(a4 ) hold .
then @xmath101 for @xmath102 where @xmath103 is defined in ( [ pn ] ) , @xmath104 in ( [ tildeh ] ) , @xmath105 in ( [ tildep ] ) , @xmath106 and @xmath107 _ the proof of proposition 2.1 is given in the appendix , and it is based on the generalized time - energy uncertainty relation derived in @xcite . before stating an adiabatic theorem for shape resonances , we want to estimate the time needed for the quantum - mechanical particle to escape from the potential well if its initial state is given by ( [ initialstate ] ) .
note that , for each _ fixed _ value of @xmath17 the spectrum of @xmath108 @xmath109 is formed of the eigenvalues @xmath110 where @xmath111 with corresponding eigenfunctions @xmath112 where @xmath113 are hermite functions normalized such that @xmath114 recall that the hermite functions decay like a gaussian away from the origin , @xmath115 and a finite constant @xmath116 ( see for example @xcite ) .
it follows from analytic perturbation theory ( lemma a.1 in the appendix ) that the eigenstates of @xmath59 decay like a gaussian away from the origin .
moreover , it follows from assumption ( a3 ) that @xmath117 is supported outside a ball of radius @xmath118 let @xmath119 denote the kernel of @xmath120 whose modulus decays like a gaussian away from the origin for arbitrary finite @xmath3 ; see lemma a.1 in the appendix . for each fixed @xmath17
the following estimate follows from lemma a.1 , assumptions ( a3)-(a4 ) and ( [ tildeh ] ) . @xmath121
^ 2)|\nonumber\\ & = & \int dx dy |\pi^n_1(x , y;s)|^2 ( \delta v_{\epsilon,\theta}(x , s)- \delta v_{\epsilon,\theta}(y , s))^2 \nonumber \\ & \le & c_{\epsilon , n } e^{-\mu_\epsilon \theta^{2/3}},\label{lifetime}\end{aligned}\ ] ] where @xmath122 is proportional to @xmath123 @xmath124 is a finite constant independent of @xmath17 ( for finite @xmath125 appearing in ( [ initialstate ] ) and fixed @xmath62 ) .
let @xmath126 which , by ( [ lifetime ] ) and ( [ initialstate ] ) , is a lower bound for the time needed for the particle to escape from the well .
inside the well .
note that one may also directly use time - dependent perturbation theory to estimate the time needed for the particle to escape from the well , see @xcite . ]
we now introduce the generator of the adiabatic time evolution for each eigenprojection , @xmath127 , \ ] ] and the corresponding propagator @xmath128 which satisfies @xmath129 by assumptions ( a1)-(a4 ) , it follows that ( [ uadiabeq ] ) has a unique solution , @xmath130 which is a unitary operator . from the standard adiabatic theorem in quantum mechanics @xcite , we know that @xmath131}\| u_a^n(s,0)-u_1(s,0)\| & = & o(\tau^{-1}),\label{adiabtheorem1}\end{aligned}\ ] ] for @xmath132 for @xmath133 where @xmath134 is given in ( [ lifetime1 ] ) , it follows from ( [ auxiliaryuncertainty ] ) , ( [ tildep ] ) , ( [ intertwiningprop ] ) and ( [ adiabtheorem1 ] ) that @xmath135 let @xmath136 the _ instantaneous _ metastable state of the particle inside the well . by ( [ initialstate ] ) , ( [ auxiliaryuncertainty2 ] ) and ( [ instmeta ] )
, we have that , for @xmath137 @xmath138}|p_s - tr ( { { \mathsf p}}{\widetilde{\rho}}_s)|\le a /\tau
+ b\tau/\tau_l , \ ] ] where @xmath139 and @xmath140 are finite constants .
this proves the following theorem for the adiabatic evolution of shape resonances .
* theorem 2.2 ( adiabatic evolution of shape resonances ) * _ suppose assumptions ( a1)-(a4 ) hold for some @xmath3 satisfying ( [ tauchoice ] ) . then @xmath141 _ in other words , over time scales that are small compared to the escape time @xmath142 given in ( [ lifetime1 ] ) , of the particle from the potential well , the true state of the particle which is initially localized inside the well , as given by the choice ( [ initialstate ] ) , is approximately equal to the instantaneous metastable state given in ( [ instmeta ] ) .
we remark that a similar analysis can be applied to study the adiabatic evolution of the metastable state of the electron of an @xmath0 ion moving in a time - dependent magnetic field ; ( see @xcite for a discussion of this model in the time - independent case ) .
in this section , we discuss the adiabatic evolution of quantum resonances which appear as isolated eigenvalues of spectrally deformed hamiltonians . examples of such resonances include ones of the stark effect and the n - body stark effect ( see for example @xcite ) .
our analysis is based on balslev - combes theory for dilatation analytic hamiltonians and on an adiabatic theorem for nonnormal and unbounded generators of evolution .
the main result of this section is theorem 3.3 , which gives an estimate on the distance between the true state and an instantaneous metastable state when the adiabatic time scale is much smaller than the lifetime of the metastable state .
consider a quantum mechanical system with hilbert space @xmath143 and a family of selfadjoint hamiltonians @xmath144 which are given by @xmath145 with fixed dense domain of definition , where @xmath146 and @xmath147 is the ( generally time - dependent ) unperturbed hamiltonian , while @xmath148 is a perturbation bounded relative to @xmath108 unless specified otherwise ; see footnote after assumption ( b1 ) below . here , @xmath149 $ ] is the rescaled time .
let @xmath150 denote the one - parameter unitary group of dilatations . for fixed @xmath151
we assume that there exists a positive @xmath152 independent of @xmath65,$ ] such that @xmath153 extends from real values of @xmath50 to an analytic family in a strip @xmath154 for all @xmath65.$ ] the spectrum of @xmath155 is assumed to lie in the closed lower half - plane for @xmath156 the relation @xmath157 holds for real @xmath50 and extends by analyticity to the strip @xmath158 we make the following assumptions .
* @xmath159 is an isolated or embedded simple eigenvalue of @xmath147 with eigenprojection @xmath160 we assume that , for each fixed @xmath65 $ ] and @xmath161 @xmath159 is separated from the essential spectrum of @xmath162 we also assume that the corresponding eigenprojection @xmath163 is analytic in @xmath50 for @xmath164 and strongly continuous in @xmath50 for @xmath165 * for @xmath166 let @xmath167 denote the reduced hamiltonian acting on @xmath168 and let @xmath169 be its corresponding eigenvalue .
then @xmath170 we assume that @xmath169 is differentiable in @xmath65.$ ] * for each fixed @xmath65 $ ] and fixed @xmath50 with @xmath161 there is an annulus @xmath171 centered at @xmath159 such that the resolvent , @xmath172 exists for each @xmath173 and @xmath174 * let @xmath175 be an arbitrary contour in @xmath176 enclosing @xmath159 and @xmath177 for @xmath178 then , for @xmath179 the spectral projection @xmath180 satisfies @xmath181 we assume that @xmath182 is twice differentiable in @xmath65 $ ] as a bounded operator , for fixed @xmath183 * _ rs ( rayleigh - schrdinger ) expansion . _
the perturbation @xmath184 for @xmath154 is densely defined and closed , and @xmath185 we define @xmath186 on a core of @xmath187 for @xmath188 and @xmath189 small enough , the iterated resolvent equation is @xmath190 for @xmath191 ( depending on the model ) , where @xmath192 we assume that the individual terms in ( [ resolventeq ] ) are well - defined , and that @xmath193 defined in ( [ resolventterms ] ) are analytic in @xmath50 in the strip @xmath161 for @xmath194 and @xmath195 and strongly continuous in @xmath196 this assumption is satisfied for @xmath197 in dilatation - analytic systems where @xmath198 is bounded relative to @xmath199 ; see , e.g. , @xcite .
moreover , this assumption holds for arbitrary @xmath200 if @xmath159 is an isolated eigenvalue of the unperturbed hamiltonian @xmath108 as in the case of discrete eigenvalues of coulumb systems , with @xmath198 a perturbation describing the stark effect , @xcite .
the rs - expansion for @xmath182 implies that , for @xmath161 @xmath201 where @xmath202 is analytic in the strip @xmath161 and strongly continuous in @xmath196 in other words , the spectral projection onto the resonance state is only defined up to a certain order @xmath203 in the coupling constant @xmath204 this is to be expected since resonance states decay with time .
we now show that , for each _ fixed _
@xmath65,$ ] the projections @xmath205 can be regarded as projections onto approximate metastable states , up to an error of order @xmath206 .
denote by @xmath207 the eigenstate of @xmath147 with corresponding eigenvector @xmath208 and let @xmath209 we have the following proposition for approximate metastable states , for each _ fixed _
@xmath65;$ ] see @xcite .
* proposition 3.1 ( approximate metastable states ) * _ assume that ( b1)-(b5 ) hold , and fix @xmath65.$ ] let @xmath210 be supported close to @xmath159 with @xmath211 in some open interval containing @xmath212 then @xmath213 for small @xmath151 where @xmath214 and @xmath215 for @xmath216 where @xmath217 is a finite constant , independent of @xmath65.$ ] _
although the proof of proposition 3.1 is a straightforward extension of the results in @xcite , it is sketched in the appendix to make the presentation self - contained .
choosing @xmath218 in ( [ decayisolated ] ) gives @xmath219 in particular , for @xmath220 @xmath221 this motivates considering @xmath222 as approximate _ instantaneous _
metastable states , up to an error term of order @xmath223 in the next subsection , we recall a general adiabatic theorem proven in @xcite . consider a family of closed operators @xmath224 acting on a hilbert space @xmath225 with common dense domain of definition @xmath226
let @xmath227 be the propagator given by @xmath228 for @xmath229 we make the following assumptions , which will be verified in the application we consider later in this section . * @xmath227 is a bounded semigroup , for @xmath230 ie , @xmath231 where @xmath232 is a finite constant . * for @xmath233
the resolvent set of @xmath234 , let @xmath235 .
assume that @xmath236 is bounded and differentiable as a bounded operator on @xmath143 , and that @xmath237 is bounded , where the @xmath238 stands for differentiation with respect to @xmath4 .
assume that @xmath239 for @xmath240 , and that it is perturbed _ slowly _ over a time scale @xmath3 such that @xmath241 , where @xmath242 $ ] is the rescaled time .
the following two assumptions are needed to prove an adiabatic theorem . *
the eigenvalue @xmath243 is isolated and simple , with @xmath244 where @xmath245 is a constant independent of @xmath65 $ ] , and @xmath246 is continuously differentiable in @xmath65 $ ] .
* the projection onto @xmath246 , @xmath247 where @xmath248 is a contour enclosing @xmath246 only , is twice differentiable as a bounded operator .
note that , since @xmath246 is simple , the resolvent of @xmath249 in a neighborhood @xmath250 of @xmath251 contained in a ball @xmath252 centered at @xmath246 with radius @xmath253 is @xmath254 where @xmath255 is analytic in @xmath250 .
we now discuss our general adiabatic theorem .
let @xmath256 be the propagator given by @xmath257 for @xmath258 moreover , define the generator of the _ adiabatic time evolution _
, @xmath259 \ ; , \ ] ] with the corresponding propagator @xmath260 which is given by @xmath261 for @xmath258 it follows from assumption ( c4 ) that @xmath262}\|[\dot{p}_{\lambda(s ) } , p_{\lambda(s)}]\|\le c,\ ] ] for some finite constant @xmath44 , and hence by perturbation theory for semigroups , @xcite chapter ix , and assumption ( c1 ) , @xmath263 defined on the domain @xmath264 exists and is unique , and @xmath265 for @xmath266 , where @xmath267 we are in a position to state our adiabatic theorem .
* theorem 3.2 ( a general adiabatic theorem ) * _ assume ( c1)-(c4 ) .
then the following holds .
_ * @xmath268 + for @xmath269 ( the intertwining property ) .
* @xmath138 } \| u_\tau ( s,0)-u_a(s,0 ) \| \le \frac{c}{1+\tau},\ ] ] for @xmath270 and @xmath44 a finite constant .
in particular , @xmath138 } \| u_\tau ( s,0)-u_a(s,0 ) \|=o(\tau^{-1}),\ ] ] for @xmath132 we refer the reader to @xcite for a proof of theorem 3.2 . * remark . *
assumption ( c1 ) can be relaxed , but the result of theorem 3.2 will be weakened .
suppose @xmath234 generates a quasi - bounded semigroup , ie , there exist finite positive constants @xmath232 and @xmath271 such that @xmath272 then ( ii ) in theorem 3.2 becomes @xmath138}\| u_\tau ( s,0)- u_a(s,0)\| \le c \frac{e^{\tau\gamma}}{\tau},\ ] ] for @xmath273 we consider a quantum mechanical system satisfying assumptions ( b1)-(b5 ) , subsection 3.1 .
denote by @xmath274 the propagator corresponding to the deformed time evolution , which is given by @xmath275 for @xmath276 and @xmath165 we make the following assumption on the existence of the deformed time evolution , which can be shown to hold in specific physical models ; see @xcite and @xcite , chapter ix .
* for fixed @xmath50 with @xmath277 @xmath278 exists and is unique as a bounded semigroup with some dense domain of definition @xmath226 in particular , there exists a finite constant @xmath232 such that @xmath279 the generator of the deformed adiabatic time evolution is given by @xmath280,\ ] ] and it generates the propagator @xmath281 for @xmath276 and fixed @xmath50 with @xmath282 for fixed @xmath50 with @xmath277 assumptions ( b4 ) and ( b6 ) and perturbation theory for semigroups , @xcite , imply that @xmath283 exists and @xmath284 where @xmath285 is a finite constant independent of @xmath286.$ ] assumptions ( b1)-(b6 ) in subsection 3.1 imply assumptions ( c1)-(c4 ) in subsection 3.2 , with the identification @xmath287 for fixed @xmath50 with @xmath282 we consider a reference subspace corresponding to a projection @xmath77 which is dilatation analytic , ie , @xmath288 extends from real values of @xmath50 to a family in a strip @xmath289 moreover , we assume that the initial state of the quantum mechanical system is @xmath290 where @xmath222 has been defined in ( [ psin ] ) .
we are interested in estimating the difference between the true state of the system and the _ instantaneous _ metastable state defined in ( [ psin ] ) when @xmath291 varies over a time scale smaller than the lifetime of the metastable state , @xmath292}({im}\lambda_g(s))^{-1}\sim g^{-2}.\ ] ] more precisely , we are interested in comparing @xmath293 to @xmath294 this is given in the following theorem .
* theorem 3.3 ( adiabatic evolution of isolated resonances ) * _ suppose assumptions ( b1)-(b6 ) hold .
then , for @xmath189 small enough and for @xmath295 @xmath296 _ _ proof .
_ this result is a consequence of theorem 3.2 .
since assumptions ( b1)-(b6 ) hold , we know that , for fixed @xmath50 with @xmath161 @xmath297}\| u_a ( s,0,\theta)-u_\tau ( s,0,\theta)\| & \le & \frac{c}{\tau},\label{propagatordiff}\end{aligned}\ ] ] for @xmath298 where @xmath44 is a finite constant . for @xmath299 and @xmath161we have @xmath300 @xmath301 _ remarks . _
* to estimate the survival probability of the true state of the system , choose @xmath302 where @xmath222 is defined in ( [ psin ] ) .
* one may also estimate the difference between the true expectation value of a bounded operator @xmath139 and its expectation value in the instantaneous metastable state , provided the operator @xmath139 is dilatation analytic .
similar to the proof of theorem 3.3 , one can show that @xmath303 for @xmath304 * the results of this section can be extended to study the quasi - static evolution of equilibrium and nonequilibrium steady states of quantum mechanical systems at positive temperatures , e.g. , when one or more thermal reservoirs are coupled to a small system with a finite dimensional hilbert space ; see @xcite for further details .
in these applications , the generator of time evolution is deformed using complex translations instead of complex dilatations .
* assumption ( b6 ) can be relaxed .
fix @xmath50 with @xmath178 suppose that @xmath305 generates a quasi - bounded semigroup , @xmath306 where @xmath232 and @xmath307 are positive constants and @xmath189 is the coupling constant .
it follows from assumption ( b4 ) that @xmath308}\| [ \dot{p}_g(s,\theta)p_g(s,\theta)]\| \le \frac{c}{\tau}\ ] ] for finite @xmath309 together with ( [ quasibdsemigr ] ) , this implies that @xmath310 where @xmath285 is a finite constant .
then , under assumptions ( b1)-(b6 ) , the result of theorem 3.3 becomes @xmath311 for @xmath312 * the results of this section can be extended to study _
`` superadiabatic '' _ evolution of quantum resonances . in the last decade
, there has been a lot of progress in studying superadiabatic processes ( see for example @xcite and references therein ) .
depending on the smoothness of the generator of the time evolution , superadiabatic theorems give improved estimates of the difference between the true time evolution and the adiabatic one . very recently , and after the submission of this paper , superadiabatic theorems with a gap condition have been extended to evolutions generated by nonselfadjoint operators @xcite .
using superadiabatic theorems and methods developed in @xcite , the results of this section can be extended to longer time scales under additional regularity assumptions on the hamiltonian .
further details will appear in @xcite .
in this section , we study the case of resonances which emerge from eigenvalues of an unperturbed hamiltonian embedded in the continuous spectrum after a perturbation has been added to the hamiltonian .
such resonances arise , for example , when a small system , say a toy atom or impurity spin , is coupled to a quantized field , e.g. to magnons or the electromagnetic field .
the main result of this section is theorem 4.1 , which is based on an extension of the adiabatic theorem without a spectral gap ; see for example @xcite .
the results of this section are more general than section 3 , since the perturbation is not restricted to be dilatation analytic . consider a quantum mechanical system with a hilbert space @xmath143 and a family of time - dependent selfadjoint hamiltonians @xmath313 such that @xmath314 where @xmath315 is the unperturbed hamiltonian with fixed common dense domain of definition @xmath316 and @xmath317 is a perturbation which is bounded relative to @xmath315 in the sense of kato.@xcite we assume that the variation of the true hamiltonian , @xmath318 in time is given by @xmath319 where @xmath65 $ ] is the rescaled time . we make the following assumptions on the model .
* @xmath320 is a generator of a contraction semigroup for @xmath65 $ ] with fixed dense core .
let @xmath321 for @xmath322 the resolvent set of @xmath323 we assume that @xmath324 is differentiable in @xmath64 as a bounded operator , and @xmath325 is bounded uniformly in @xmath65.$ ] this assumption is sufficient to show that the unitary propagator generated by @xmath320 exists and is unique .
* @xmath159 is a simple eigenvalue of @xmath147 which is embedded in the continuous spectrum of @xmath108 with corresponding eigenvector @xmath326 @xmath327 furthermore , the eigenprojection @xmath328 corresponding to @xmath159 is twice differentiable in @xmath64 as a bounded operator _ for almost all _ @xmath65,$ ] and is continuous in @xmath329,$ ] as a bounded operator .
* let @xmath330 and , for a given operator @xmath139 on @xmath225 denote by @xmath331 its restriction to the range of @xmath332 @xmath333 let @xmath334 for each @xmath65,$ ] we have @xmath335 we note that @xmath336 * _ instantaneous metastable states .
_ let @xmath337 be supported in a neighborhood of @xmath212 for each _ fixed _
@xmath65,$ ] we have @xmath338 where @xmath339 and @xmath340 @xmath44 is a finite constant independent of @xmath65,$ ] for some @xmath341 note that @xmath342 equation ( [ quasiexpdecay ] ) uniquely defines the _ instantaneous _ resonance state , up to an error @xmath343[multiblock footnote omitted ] a physical example where assumptions ( d1)-(d4 ) may be satisfied is a small system interacting with a field of noninteracting bosons or fermions , for example , a spin system coupled to a time - dependent magnetic field ; see @xcite for further details on the relevant model of a toy atom interacting with the electromagnetic radiation field .
we are interested in the adiabatic evolution of the quantum resonance over time scales which are much smaller than the lifetime of the resonance .
we will prove an adiabatic theorem _ without a spectral gap condition _ for quantum resonances for weak coupling @xmath189 ( see @xcite ) .
let @xmath256 be the propagator given by @xmath344 with some dense domain of definition @xmath226 eixstence of @xmath345 as a unique unitary operator follows from assumption ( d1 ) and theorem x.70 in @xcite .
moreover , we introduce the generator of the adiabatic time evolution @xmath346.\ ] ] the propagator corresponding to the approximate adiabatic evolution is given by @xmath347 with domain of definition @xmath226 note that @xmath263 exists as a unique unitary operator due to assumptions ( d1 ) and ( d2 ) .
we have the following theorem , which is an extension of the results in @xcite .
* theorem 4.1 ( adiabatic theorem for embedded resonances ) * _ suppose assumptions ( d1)-(d4 ) hold . then , for small enough coupling @xmath189 and large enough @xmath348 @xmath349 and @xmath138 } \| u_\tau ( s,0)- u_a^0(s,0)\| \le \frac{a}{\tau^{1/2 } } + bg\tau^{1/4 } + c(\tau^{-1/4}),\label{c2}\ ] ] where @xmath139 and @xmath140 are finite constants , and @xmath350 is a positive function of @xmath351 such that @xmath352 in particular , choosing @xmath353 gives @xmath138}\| u_\tau(s,0)p_0(0)-p_0(s)\| \le a g^{1/3 } + c(g^{1/6}).\label{c3}\ ] ] _ _ proof . _
let @xmath354 then @xmath355 where we have used the definition of the generator of the adiabatic evolution and the property that @xmath356 it follows that @xmath357 which is claim ( [ intertwining2 ] ) .
moreover , we are interested in estimating the difference between the true evolution and the adiabatic time evolution . for @xmath358 we have that @xmath359u_a^0(s',0)\psi \\ & = -\int_0^s ds ' u_\tau ( s , s')[\dot{p}_0(s'),p_0(s')]u_a^0(s',0 ) \psi .\end{aligned}\ ] ] since the domain of definition @xmath264 is dense in @xmath225 it follows that @xmath360u_a^0(s',0)\| .\ ] ] we will now use a variant of kato s commutator method to express the integrand as a total derivative plus a remainder term ; see @xcite .
let @xmath361 note that @xmath362 & = [ h_g(s)-\lambda_0(s)-i\epsilon , r_{g}(\lambda_0(s)+i\epsilon , s)\dot{p}_0(s)p_0(s ) ] \\ & + [ h_g(s)-\lambda_0(s)+i\epsilon , p_0(s ) \dot{p}_0(s)r_{g}(\lambda_0(s)-i\epsilon , s ) ] \\ & = [ \dot{p}_0(s),p_0(s ) ] + i\epsilon x_\epsilon ( s ) + o(g/\epsilon).\end{aligned}\ ] ] furthermore , @xmath363u_a^0(s',0 ) \\&+ u_\tau(s , s')x_{\epsilon}(s')[\dot{p}_0(s'),p_0(s')]u_a^0(s',0)\\ & + u_\tau ( s , s')\dot{x}_{\epsilon}(s')u_a^0 ( s',0).\end{aligned}\ ] ] therefore , @xmath364u_a^0(s',0)\| & \le & \sup_{s\in [ 0,1 ] } \ { \frac{1}{\tau } [ \| x_{\epsilon}(s)\| ( 1 + 2 \| \dot{p}_0(s)p_0(s)\| ) \nonumber \\ & + & \| \dot{x}_{\epsilon}(s)\| ] + \epsilon \| x_{\epsilon}(s)\| \ } + cg/\epsilon , \end{aligned}\ ] ] where @xmath44 is a finite constant independent of @xmath65.$ ] we claim that the following estimates are true for small enough @xmath62 and @xmath204 @xmath365 where @xmath366 and @xmath44 is a finite constant , uniformly in @xmath65.$ ] estimates @xmath367 and @xmath368 follow from our knowledge of the spectrum of @xmath320 and the resolvent identity . to prove estimate @xmath369 we compare the lhs of ( [ est3 ] ) to the case
when @xmath370 let @xmath371 then , by the second resolvent identity , @xmath372 uniformly in @xmath373 for some finite constant @xmath309 we claim that @xmath374 consider @xmath375 then @xmath376 using the spectral theorem for @xmath108 we have the following result .
@xmath377 and hence claim ( [ smalle ] ) .
therefore , @xmath138 } \| u_\tau(s,0)-u_a^0(s,0)\| \le \frac{c_1}{\tau \epsilon^2}+\frac{c_2g}{\epsilon}+c(\epsilon),\ ] ] where @xmath378 are finite constants , and @xmath379 choosing @xmath380 gives ( [ c2 ] ) . by choosing @xmath381 ( [ c3 ] )
follows from assumption ( d4 ) , ( [ intertwining2 ] ) and ( [ c2 ] ) .
@xmath301 _ remarks .
_ * we note that , using an argument due to kato , @xcite , the case of finitely many resonance crossings is already covered by theorem 4.1 , since the latter holds for @xmath328 twice differentiable as a bounded operator _ for almost all _
@xmath65 $ ] and continuous as a bounded operator for @xmath65.$ ] suppose that at time @xmath382 $ ] , a crossing of @xmath159 with an eigenvalue of @xmath147 happens .
it follows from continuity of @xmath328 that , for small @xmath383 @xmath384 and @xmath385 are close up to an error which is arbitrarily small in @xmath51 and hence our claim follows .
* further knowledge of the spectrum of @xmath147 will yield a better estimate of the convergence of @xmath386 to zero as @xmath387 for example , it is shown in @xcite that if the spectral measure @xmath388 is @xmath307-hlder continuous , for @xmath389,$ ] uniformly in @xmath65,$ ] then is @xmath307-hlder continuous , @xmath389,$ ] if there exists a finite constant @xmath44 such that , for every set @xmath62 with lebesgue measure @xmath390 @xmath391 , see , e.g. , @xcite . ]
@xmath262}\epsilon \| r_0(\lambda_0(s)+i\epsilon , s)\dot{p}_0(s)p_0(s)\| \le a \epsilon^{\alpha/2},\ ] ] for @xmath62 small enough , where @xmath139 is a finite constant , and hence estimate ( [ c3 ] ) becomes @xmath138}\| u_\tau(s,0)p_0(0)-p_0(s)\| = o(g^{\alpha/12})\ ] ] for @xmath189 small enough .
* proof of proposition 2.1 , section 2 * _ proof of proposition 2.1 .
_ this proposition effectively follows by integrating the liouville equation and applying the cauchy - schwarz inequality .
it is a special case of the generalized time - energy uncertainty relations derived in @xcite .
consider an orthogonal projection @xmath392 and selfadjoint operators @xmath139 and @xmath140 acting on a hilbert space @xmath393 then it follows from a direct application of the cauchy - schwarz inequality that @xmath394)^2 \le 4 tr(pa^2 -papa ) tr(pb^2-pbpb),\label{uncertainty1}\ ] ] with equality when there exist @xmath395 such that @xmath396p=0.\ ] ] we use inequality ( [ uncertainty1 ] ) to derive upper and lower bounds for @xmath397 let @xmath398 then @xmath399)|\\ & \le 2 \tau tr ( u_\tau(s , s')p_1^n(0)u_\tau ( s',s){{\mathsf p}}^2 - \\ & - u_\tau(s , s')p_1^n(0)u_\tau ( s',s){{\mathsf p}}u_\tau(s , s')p_1^n(0)u_\tau ( s',s){{\mathsf p}})^{1/2}\times\\ & \times tr(p_1^n(0)h(s')^2-p_1^n(0)h(s')p_1^n(0)h(s'))^{1/2 } \\ & \le 2 \tau \sqrt{p_{s , s'}^n - ( p_{s , s'}^{n})^2 } f(p_1^n(0),h(s')),\end{aligned}\ ] ] where @xmath400 it follows that @xmath401 and hence @xmath402 we note that @xmath403 together with ( [ genuncertainty1 ] ) , and the identification @xmath404 where @xmath104 is the generator of the auxiliary propagator @xmath405 , as defined in ( [ tildeh ] ) , we have @xmath406 @xmath301 it follows from assumption ( a3 ) that @xmath408 defined in ( [ omega ] ) is uniformly bounded by @xmath409 for @xmath23 therefore , the spectrum of @xmath410 for each _ fixed _ @xmath17 can be computed by applying analytic perturbation theory(see , e.g. , @xcite )
. also using analytic perturbation theory , one can show that the eigenstates of @xmath410 for each fixed @xmath373 decay like a gaussian away from the origin ; ( see @xcite ) . to prove the last claim , choose @xmath411 there exist finitely many sequences @xmath412 such that @xmath413 where @xmath414 @xmath139 is a finite geometrical constant , @xmath415 appears in assumption ( a2 ) , and @xmath416 is given in ( [ spech0 ] ) .
let @xmath417 then @xmath418 for @xmath419 choose a contour @xmath420 in the complex plane surrounding @xmath421 such that @xmath422=\frac{1}{2}\min_{s\in i}(e^s_{{{\mathbf l}}^{(k_e+1)}}-e^s_{{{\mathbf l}}^{(k_e)}})>0.\ ] ] for each fixed time @xmath38 , we define the spectral projection of @xmath59 , @xmath423 let @xmath424 be the orthogonal projection of @xmath147 onto the subspace @xmath425 spanned by the eigenfunctions @xmath426 and choose @xmath62 such that @xmath427 where @xmath41 is a finite constant appearing in assumption ( a3 ) .
it follows from analytic perturbation theory , with @xmath62 satisfying ( [ analpertbd ] ) , that @xmath428 and @xmath429 we have the following lemma .
_ suppose assumptions ( a2 ) and ( a3 ) hold .
choose @xmath62 satisfying ( [ analpertbd ] ) , and fix @xmath23 furthermore , suppose that @xmath430 then there exist finite constants @xmath431 and @xmath432 ( dependending on @xmath62 ) such that , for sufficiently small @xmath433 @xmath434 furthermore , @xmath435 for @xmath436 and @xmath307 small enough .
_ _ proof .
_ it follows from ( [ projdiff ] ) that there exists @xmath437 the subspace spanned by the eigenfunctions @xmath426 such that @xmath438 and hence @xmath439 for some finite constant @xmath309 moreover , it follows from ( [ proj1 ] ) and ( [ psis ] ) that @xmath440 for @xmath307 small enough , we know from ( [ hermite ] ) and ( [ hermitedecay ] ) that @xmath441 for some finite constant @xmath442 moreover , for @xmath443 it follows from analytic perturbation theory , @xcite , that @xmath444 for @xmath307 small enough ( depending on @xmath62 ) , where @xmath445 the claim ( [ claim1 ] ) follows from ( [ exppsi ] ) , ( [ expphi ] ) and ( [ bd1 ] ) .
now , @xmath446 where @xmath447 is the propagator generated by @xmath448 by applying analytic perturbation theory , it follows that @xmath449 where @xmath450 is a positive constant such that @xmath451 as @xmath452 together with ( [ claim1 ] ) , this implies ( [ claim2 ] ) for @xmath307 small enough .
_ proof of proposition 3.1 . _ fix @xmath453 with @xmath454 by assumptions ( b1 ) and ( b3 ) , there exists an open interval @xmath455 with @xmath456 choose @xmath457 then @xmath458 let @xmath459 where @xmath460 then @xmath461 the resolvent in @xmath462 can be decomposed into a singular and regular part , @xmath463 where @xmath464 is analytic in @xmath465 note that @xmath466 using ( [ regularres ] ) , the contribution of the regular part to @xmath467 defined in ( [ f ] ) is @xmath468 where @xmath469\psi_0(s,\theta),\ ] ] is of order @xmath470 since @xmath471 the last integral is bounded by @xmath472 for any @xmath473 and hence the contribution of the regular part is bounded by @xmath474
the contribution of the singular part of the resolvent to @xmath475 is @xmath476 using the fact that @xmath211 in some open interval @xmath477 one may deform the path @xmath18 into two contours , @xmath478 and @xmath479 in the lower complex half - plane , as shown in figure 2 . the term in ( [ singularcontribution ] ) corresponding to the path @xmath478 picks the residue @xmath480 it follows from the identity @xmath481[p_g(s,\theta)-1][p_g^n(s,\theta)-p_g(s,\theta ) ] , \ ] ] and from the fact that @xmath482 that @xmath483 using ( [ acoef ] ) , one may write the remainder term in ( [ singularcontribution ] ) due to the path @xmath484 as @xmath485 which is of order @xmath223 @xmath301 | we study the adiabatic time evolution of quantum resonances over time scales which are small compared to the lifetime of the resonances .
we consider three typical examples of resonances : the first one is that of shape resonances corresponding , for example , to the state of a quantum - mechanical particle in a potential well whose shape changes over time scales small compared to the escape time of the particle from the well .
our approach to studying the adiabatic evolution of shape resonances is based on a precise form of the time - energy uncertainty relation and the usual adiabatic theorem in quantum mechanics .
the second example concerns resonances that appear as isolated complex eigenvalues of spectrally deformed hamiltonians , such as those encountered in the n - body stark effect .
our approach to study such resonances is based on the balslev - combes theory of dilatation - analytic hamiltonians and an adiabatic theorem for nonnormal generators of time evolution .
our third example concerns resonances arising from eigenvalues embedded in the continuous spectrum when a perturbation is turned on , such as those encountered when a small system is coupled to an infinitely extended , dispersive medium .
our approach to this class of examples is based on an extension of adiabatic theorems without a spectral gap condition .
we finally comment on resonance crossings , which can be studied using the last approach . *
key words : * adiabatic theorem , quantum resonances . |
in this supplementary material we provide the details of some results in the main text .
k atoms . ] in this section we provide details of deriving the tight - binding hamiltonian for the @xmath42-band model .
as mentioned in the main text , only two lasers @xmath14 , with @xmath191 and @xmath192 used to induce the two - photon raman transition , are needed to generate simultaneously the periodic transverse zeeman term and the 1d optical lattice .
this further greatly simplifies the set - up for the experimental realization .
[ p - band ] shows the realistic transitions for @xmath178k atoms induced by @xmath14 .
it is noteworthy that besides the induced @xmath4-configuration , @xmath193 and @xmath194 can also couple respectively to the states @xmath16 and @xmath195 ( i.e. @xmath196 and @xmath197 ) ( fig .
[ p - band ] ) , while these couplings can not lead to additional raman transitions between the ground ( pseudo-)spin-@xmath80 subspace and other ground states since such transitions are associated with large two - photon detunings ( see the experiments ref . [
11 - 15 ] in the main text .
the typical two - photon detuning for such processes is @xmath198mhz , much larger than the raman transition strength which is about @xmath199mhz ) .
therefore the two couplings only lead to additional diagonal optical potentials @xmath200 and @xmath201 for @xmath195 and @xmath16 , respectively .
the light - atom coupling hamiltonian for the realistic @xmath178k system then reads @xmath202|g_\sigma\rangle\langle g_\sigma|+\hbar\delta|g_\downarrow\rangle\langle g_\downarrow|+\nonumber\\ & & + \hbar\delta|e\rangle\langle e|-\hbar\bigr(\omega_1|e\rangle\langle g_\uparrow|+\omega_2|e\rangle\langle g_\downarrow|+{\rm h.c.}\bigr).\end{aligned}\ ] ] for the large one - photon detuning condition @xmath203 , we can eliminate the excited state by @xmath19 and obtain @xmath204|g_\sigma\rangle\langle g_\sigma|-\nonumber\\ & & -\bigr[m_0\sin(k_0x)|g_\uparrow\rangle\langle g_\downarrow|+{\rm h.c.}\bigr],\end{aligned}\ ] ] with @xmath205 and @xmath206 note in realistic experiments the difference between @xmath207 and @xmath208 ( about @xmath198mhz ) is negligible relative to their magnitudes ( in the order of @xmath209ghz ) .
we can take that @xmath210 , and then get ( neglecting the constant terms ) @xmath211 the tight - binding hamiltonian for @xmath42-band model can be derived straightforwardly . by noticing the odd - parity of the local @xmath42-orbitals , the periodic term @xmath31 leads to the spin - flip hopping by @xmath212 , with @xmath213 .
the spin - conserved hopping reads @xmath214\phi^{(j+1)}_{p\sigma}(x)$ ] . with these results
we can finally get the tight - binding hamiltonian @xmath215,\end{aligned}\ ] ] which is the same as eq .
( 2 ) in the main text . more generally , with this configuration
all the bands with odd - parity local orbitals @xmath216 can be described by the tight - binding hamiltonian , while those bands with even - parity local orbitals ( @xmath217 ) are always topologically trivial .
accordingly , for another configuration with the lattice potentials @xmath27 considered in the main text , all the bands with even - parity local orbitals can be described by the above hamiltonian , while the other bands with odd - parity local orbitals are always topologically trivial .
with the inclusion of both @xmath124 and @xmath120 , the generic hamiltonian obtained in the main text is given by @xmath218+\sum_{j}\bigr[\gamma_z(\hat n_{j\uparrow}-\hat n_{j\downarrow})-\gamma_y(i\hat c^\dag_{j\uparrow}c_{j\downarrow}+{\rm h.c.})\bigr]\nonumber\\ & = & \sum_{k,\sigma}\hat c_{k,\sigma\sigma'}^{\dag}{\cal h}_{\sigma\sigma'}(k)\hat c_{k,\sigma'},\end{aligned}\ ] ] where @xmath219 with @xmath220 and @xmath221 . the full symmetry of the system is @xmath222 , where the phase transformation operator @xmath3 , charge conjugation operator @xmath223 and time reversal operator @xmath224 are defined as : @xmath225 the following commutation relations can be checked : @xmath226 , and @xmath227=0 $ ] .
the subgroup @xmath228 is anti - unitary and can be denoted as @xmath57 , and the complete symmetry group can also be written as @xmath56 . owning to this symmetry group
, the free fermion system belongs to chiral unitary ( aiii ) class and is characterized by a @xmath0 invariant in the noninteracting case .
if the zeeman field along @xmath55 axis vanishes , i.e. @xmath123 , alternatively the symmetry group can be chosen as @xmath229 where @xmath230 , @xmath231 , @xmath232 , with @xmath233 , and @xmath234=0 , \mathcal t u(\theta)=u(-\theta)\mathcal t , \mathcal c u(\theta)=u(-\theta)\mathcal c$ ] . in this case
both @xmath224 and @xmath223 are symmetries of the hamiltonian and the system then belongs to the bdi class which is also classified by a @xmath0 invariant in the non - interacting case . for a n - chain system under the half - filling condition ,
the total ground - state degeneracy is @xmath235 without interactions .
let @xmath236 and @xmath237 be the annihilation and creation operators of the left / right edge mode for the @xmath238-th chain , respectively , where @xmath239 and @xmath240 .
the edge states of the @xmath238-th chain can be written as @xmath241 and @xmath242 , where @xmath243 is the ground state for the bulk .
an interesting question is what happens if we turn on interactions .
it turns out that in the presence of interactions , the @xmath0 classification breaks down to @xmath1 . to confirm this result
, we will study the ground states of a n - chain system step by step .
first , for @xmath244 , a generic @xmath222-symmetric interaction between the edge zero modes reads @xmath245\nonumber\\ & = & \sum_{s}v_{12}^{(4)}(2\hat c_{1,s}^\dag \hat c_{1,s}\hat c_{2,s}^\dag \hat c_{2,s}-\hat c^\dag_{1,s}\hat c_{1,s}-\hat c^\dag_{2,s } \hat c_{2,s}+1).\end{aligned}\ ] ] with above interaction the minimum degeneracy of the 2-chain system is two - fold , which is obtained when @xmath246 .
it is straightforward to check that the states @xmath247 and @xmath248 have energy @xmath249 while the states @xmath250 , @xmath251,@xmath252 , and @xmath253 have a higher energy @xmath254 .
then the ground state is two - fold degenerate , as given by @xmath247 and @xmath248 .
second , for @xmath255 , the ground state degeneracy @xmath256 can be reduced to 2 by two - chain interactions according to ( [ 2chain ] ) : @xmath257 .
for instance , if @xmath258 , the two - fold degenerate ground states are @xmath259 and if @xmath260 , the ground states read @xmath261 where @xmath262 .
it is straightforward to see that above interaction is identical to a summation of two - body interactions and can not split the degeneracy of the ground states .
other possible 3-chain interactions include @xmath263\\ & = & v_{123}^{(4)}(2c_{2,s}^\dag c_{2,s}-1 ) ( c_{1,s}^\dag c_{3,s}+c_{3,s}^\dag c_{1,s}),\end{aligned}\ ] ] it can be easily checked that in the case ( [ gs3_1 ] ) the perturbation @xmath264 have zero matrix elements in the ground - state subspace and hence can not split the degeneracy .
the condition is similar in another case ( [ gs3_2 ] ) .
the difference is that , in case ( [ gs3_2 ] ) , @xmath264 mixes the ground states with higher energy states .
for instance , the state @xmath265 is mixed with @xmath266 to lower its energy . at the same time
the state @xmath267 is mixed with @xmath268 to lower the energy with the same amount . as a result
, the two new states , as the new ground states , are still degenerate .
therefore , the topological properties of 3 chains are stable against interactions respecting the symmetry , and we need to investigate the 4-chain system . finally , for @xmath269 , it turns out that we can find a path to smoothly reduce the ground state degeneracy to 1 .
it is easy to verify that degeneracy of the ground states can be reduced to 4 under two - body interaction @xmath270 with @xmath271 .
these four ground states are given by @xmath272 with the energy equal to @xmath273 .
however , the above ground states can be further gapped out by taking into account the following interactions : @xmath274 in the 4-dimensional hilbert space spanned by the ground states in eq . , the above interaction can be written in the matrix form @xmath275 whose eigenvalues are @xmath276 , and @xmath277 .
we therefore obtain the single non - degenerate ground state with the energy @xmath278 .
this implies that under interaction the 4-chain system can be smoothly connected to a trivial phase without closing the bulk gap , and we therefore complete the proof that the @xmath0 classification can be broken down to @xmath1 under interactions .
we prove in this section that each edge state leads to @xmath80-fractionalization .
a convenient way is to consider the semi - infinite geometry which has the open boundary at @xmath70 .
we then calculate the particle number of the zero mode localized on this boundary .
note the total number of quantum states in the system is given by @xmath279 where we denote by @xmath280 the state number operator and @xmath281 is the eigenstate with energy @xmath282 . since the hamiltonian satisfies @xmath283 ,
the energy spectrum is symmetric .
we have then @xmath284 .
it follows that @xmath285.\end{aligned}\ ] ] the particle number of the zero mode depends on its occupation . if the zero mode is unoccupied , the particle number of it is given by @xmath286.\end{aligned}\ ] ] here @xmath287 and @xmath288 represents the cases with one ( topological phase ) and zero ( trivial phase ) bound modes , respectively . using the eq .
one finds directly @xmath289 similarly , if the zero mode is occupied , the particle number is @xmath290 .
it is trivial to know that this result can be applied to the case with two boundaries located far away from each other , say respectively at @xmath70 and @xmath72 .
since the two zero modes are obtained independently , each of them carries @xmath86 ( @xmath87 ) particle if it is occupied ( unoccupied ) .
in this section we provide the derivation of the one - loop rg equations . we consider that in hamiltonian only one zeeman term , e.g. @xmath124 is nonzero ( the case with @xmath291 can be studied in the similar way ) .
we find it convenient to redefine that @xmath40 and rewrite the hamiltonian in the form : @xmath292 , with @xmath293 and @xmath294 .
the low - energy physics can be well captured by the continuum approximation ( @xmath295 ) : @xmath296 , \label{}\ ] ] with @xmath297 .
the continuum representation of the two mass terms are then given by : @xmath298 here @xmath299 .
using the standard bosonization formula @xmath300}$ ] with @xmath301 and @xmath302 , we reach the bosonized hamiltonian densities @xmath303 the allowed four - fermion interactions in the ( unperturbed ) hubbard model at half - filling is highly constrained by the @xmath304 symmetry .
it is convenient to define current operators : @xmath305 here @xmath306 is the fully anti - symmetric tensor .
the general form of four - fermion interaction is given by @xmath307 we then derive the rg flow equations to understand the fate of the topological phase transition driven by the competition of the staggered zeeman term and the spin - orbit coupling term .
the tree level term can be easily read off from the scaling dimensions of @xmath308 and @xmath309 in their bosonization form , both of which are @xmath310 .
so it is necessary to go to the next order in the perturbative expansion .
the one - loop order rg flow can be most easily derived by calculating the operator algebra of the various operators @xcite . to derive the rg equation , we consider the partition function in the euclidean functional integral representation @xmath311 the euclidean action @xmath312 where @xmath313 $ ] is the unperturbed gaussian part and the @xmath314 is the perturbation . to perform the rg ,
we expand the exponential to the second order in @xmath315 .
let us write @xmath316 .
then the second order term is given by @xmath317 we have introduced complex coordinates @xmath318 where @xmath319 . here
we assume the following operator product expansion ( ope ) @xmath320 which is sufficient for our purpose .
the opes are valid when two points @xmath17 and @xmath254 are brought close together , as replacement within correlation functions . at this point
the cutoff prescription needs to be carefully specified .
we will choose a short - distance cutoff @xmath39 in space , but none in imaginary time .
for a rescaling factor @xmath321 , we must then perform the integral @xmath322 eq . then becomes @xmath323 upon re - exponentiating we obtain the one - loop rg equation @xmath324 here @xmath325 .
now let us be more specifit .
besides the current operators defined in , we also need to define @xmath326 the fermionic field operators satisfy the following opes : @xmath327 the opes of the currents and the mass @xmath328 can be calculated from using wick s theorem .
those between the currents are standard and can be found in [ ] which we do not duplicate .
below are the needed ones : @xmath329(0)\sim -\frac{2}{4\pi^2|z|^2}m_i^\dag(z),\ \ m_i^\dag(z)[i_ri_l^\dag+\text{h.c.}](0)\sim -\frac{2}{4\pi^2|z|^2}m_i(z )
. \end{gathered } \label{}\ ] ] applying the above formalism to the model at hand , we find @xmath330 if we take the interaction in the lattice model to be an on - site hubbard form : @xmath331 , the bare values of the coupling coefficients are @xmath332 then for the repulsive interaction @xmath109 , the coupling parameter @xmath104 of the spin sector is marginally flows to zero , while @xmath103 is marginally relevant .
is the maximum value of @xmath333 and it has @xmath334 . ]
the appearance of the edge modes localized in the boundary between trivial insulating and topological insulating regions is protected by the nontrivial winding number in the bulk , and is independent of the details of the boundary .
the different configurations of the mass domain walls may at most quantitatively affect the coupling between the localized edge spins , as numerically shown in fig .
[ splitting_appendix ] , where we consider the mass domain with soft boundaries , described by [ fig .
[ splitting_appendix](a ) ] @xmath335+\tanh[\frac{x - l/2 - 5a}{\xi}\bigr]\biggr\}. \label{}\ ] ] here @xmath336 is the length of the lattice and @xmath94 represents the domain wall length .
the distance between the left and right domain walls is @xmath337 . in fig .
[ splitting_appendix](b ) we have shown that by varying the domain length from @xmath338 ( the profile is close to step functions ) to @xmath339 ( the profile is close to a gaussian function ) , the coupling slightly increases for fixed @xmath340 [ @xmath341 . the increasing coupling is because the wave - function overlapping of the localized spins is enhanced for wider domain walls , which can lead to a slightly larger spin rabi - oscillation frequency , and a shorter manipulation time for the tsqs . | we propose to observe and manipulate topological edge spins in 1d optical lattice based on currently available experimental platforms . coupling the atomic spin states to a laser - induced periodic zeeman field
, the lattice system can be driven into a symmetry protected topological ( spt ) phase , which belongs to the chiral unitary ( aiii ) class protected by particle number conservation and chiral symmetries .
in free - fermion case the spt phase is classified by a @xmath0 invariant which reduces to @xmath1 with interactions .
the zero edge modes of the spt phase are spin - polarized , with left and right edge spins polarized to opposite directions and forming a topological spin - qubit ( tsq ) .
we demonstrate a novel scheme to manipulate the zero modes and realize single spin control in optical lattice .
the manipulation of tsqs has potential applications to quantum computation .
_ introduction._@xmath2since the discovery of the quantum hall effect in two - dimensional ( 2d ) electron gas @xcite , the search for nontrivial topological states has become an exciting pursuit in condensed matter physics @xcite .
the recently observed time - reversal ( tr ) invariant topological insulators ( tis ) have opened a new chapter in the study of topological phases ( tps ) , attracting great efforts in both theory and experiments @xcite . depending on whether the ground states have long - range or short - range entanglement
, the tps can be classified into intrinsic or symmetry - protected topological ( spt ) orders @xcite .
being protected by the bulk gap , the intrinsic tps are robust against any local perturbations , and the spt phases are robust against those respecting given symmetries @xcite .
this property may be applied to the fault - tolerant quantum computation @xcite .
while in theory there are numerous types of tps , the existing topological orders in nature are rare .
the recent great advancement in realizing effective spin - orbit ( so ) interaction in cold atoms @xcite opens intriguing new possibilities to probe so effects @xcite and tps in a controllable fashion .
theoretical proposals have been introduced in cold atoms for the study of tis @xcite and topological superfluids @xcite .
experimental studies of these exotic phases are , however , a delicate issue due to stringent conditions such as complicated lattice configurations or so interactions . by far
the only experimentally realized so interaction @xcite is the equal rashba - dresselhaus - type so term as theoretically proposed by liu etal @xcite .
therefore , how to observe nontrivial topological states with currently available experimental platforms is a central issue in the field of cold atoms @xcite . in this letter
, we propose to observe and manipulate topological edge spins in 1d optical lattice with so interaction realizable in recent experiments @xcite .
the predicted spt phase belongs to aiii class and is protected by @xmath3 and chiral symmetries , with spin - polarized zero modes forming topological spin - qubits ( tsqs ) .
our results may open the way to observe topological states of all ten altand - zirnbauer symmetry classes @xcite with realistic cold atom systems , and have broad range of applications including realizing single spin control in optical lattice . _
model._@xmath2our model is based on quasi-1d cold fermions trapped in an optical lattice , with the internal three - level @xmath4-type configuration coupled to radiation , as shown in fig . [ lattice ] .
the transitions @xmath5 are driven by the laser fields with rabi - frequencies @xmath6 and @xmath7 , respectively . in the presence of a large one - photon detuning @xmath8 and a small two - photon detuning @xmath9 for the transitions [ fig .
[ lattice](a ) ] , the hamiltonian of the light - atom coupling system reads @xmath10 , with @xmath11|g_\sigma\rangle\langle g_\sigma|+\hbar\delta|g_\downarrow\rangle\langle g_\downarrow| , h_1=\hbar\delta|e\rangle\langle e|-\hbar\bigr(\omega_1|e\rangle\langle g_\uparrow|+\omega_2|e\rangle\langle g_\downarrow|+{\rm h.c.}\bigr)$ ] . here the diagonal potentials @xmath12 are used to construct the 1d optical lattice , and @xmath13 are the pauli matrices in spin space . for @xmath8 , the lasers @xmath14
induce a two - photon raman transition between @xmath15 and @xmath16 .
this configuration has been used to create the equal - rashba - dresselhaus so interaction @xcite .
the effect of the small two - photon detuning is equivalent to a zeeman field along @xmath17 axis @xmath18 , which in experiment can be precisely controlled with acoustic - optic modulator . eliminating the excited state by @xmath19 yields the effective hamiltonian @xmath20|g_\sigma\rangle\langle g_\sigma|-\nonumber\\ & & -\bigr[m(x)|g_\uparrow\rangle\langle g_\downarrow|+{\rm h.c.}\bigr],\end{aligned}\ ] ] where @xmath21 with @xmath22 represents a transverse zeeman field induced by the raman process .
-type configuration coupled to radiation .
( b ) energy spectra with open boundary condition in the topological ( diamond , @xmath23 ) and trivial ( circle , @xmath24 ) phases . the so coupled hopping @xmath25 . ]
we next derive the tight - binding model .
we consider first the @xmath26-band model in an optical lattice formed by the trapping potentials @xmath27 , with the lattice trapping frequency @xmath28 @xcite . from the even - parity of the local @xmath26-orbitals
@xmath29 ( @xmath30 ) , the periodic term @xmath31 does not couple the intrasite orbitals @xmath32 , but leads to a spin - flip hopping by @xmath33 [ see fig . [ lattice](a ) ] , representing the induced so interaction .
the spin - conserved hopping reads @xmath34\phi^{(j+1)}_{s\sigma}(x)$ ] .
bearing these results in mind we write down the effective hamiltonian in the tight - binding form : @xmath35 $ ] , with @xmath36 .
it can be verified that @xmath37 , where @xmath38 with @xmath39 the lattice constant . redefining the spin - down operator @xmath40 , we recast the hamiltonian into @xmath41.\end{aligned}\ ] ] the above model can also be realized with @xmath42-band fermions in a different configuration of the optical lattice that @xmath43 , which can be directly verified by noticing the odd - parity of @xmath42-orbitals .
remarkably , for @xmath42-band model the periodic zeeman term @xmath31 and the 1d lattice can be realized _
simultaneously _ by setting that @xmath44 and @xmath45 without applying additional lasers ( see supplementary material @xcite for details ) .
this further greatly simplifies the experimental set - up and we believe that our proposal can be realized with realistic experimental platforms .
we analyze the symmetry of the hamiltonian .
the tr and charge conjugation operators are respectively defined by @xmath46 with @xmath47 the complex conjugation , and @xmath48 .
one can check that while both @xmath49 and @xmath50 are broken in @xmath51 , the chiral symmetry , defined as their product , is respected and @xmath52 , with @xmath53 .
note the chiral symmetry is still preserved if a zeeman term @xmath54 along @xmath55 axis is included in @xmath51 .
the complete symmetry group then reads @xmath56 , where @xmath3 gives particle - number conservation and the anti - unitary group @xmath57 is formed by @xmath58 .
the spt phase of our free - fermion system belongs to the chiral unitary ( aiii ) class and is characterized by a @xmath0 invariant @xcite .
the @xmath51 can be rewritten in the @xmath59-space @xmath60_{\sigma,\sigma'}\hat c_{k,\sigma'}$ ] , with @xmath61 and @xmath62 .
this hamiltonian describes a nontrivial topological insulator for @xmath63 and otherwise a trivial insulator , with the bulk gap @xmath64 ( fig .
[ lattice ] ( b ) ) .
in particular , when @xmath65 and @xmath66 , our model gives rise to a _ flat band _ with nontrivial topology .
_ edge states._@xmath2the nontrivial topology can support degenerate boundary modes . considering hard wall boundaries located at @xmath67 , respectively @xcite and diagonalizing @xmath51 in position space @xmath68 with @xmath69
, we obtain the edge state localized on left boundary @xmath70 as @xmath71|\chi_+\rangle,\end{aligned}\ ] ] and accordingly the one on @xmath72 by @xmath73|\chi_-\rangle$ ] . here
@xmath74 is the normalization factor , the spin eigensates @xmath75 , and @xmath76 .
therefore the two edge modes are polarized to the opposite @xmath77 directions .
note @xmath78 and @xmath79 span the complete hilbert space of _ one single _
@xmath80-spin or spin - qubit .
each edge state equals _ one - half _ of a single spin , similar to the relation between a majorana fermion and a complex fermion in topological superconductors . as a result
, we expect the robustness of the zero modes to any local operations without breaking the @xmath3 and @xmath81 symmetries @xcite .
these properties of the tsq may be applicable to fault - tolerant quantum computation @xcite .
moreover , the @xmath0 classification implies that single - particle couplings respecting @xmath3 and @xmath81 can not gap out the edge modes in arbitrary @xmath82-chain system of 1d lattices .
interestingly , however , we have confirmed that with weak interactions a system with up to four chains of the 1d lattices can be adiabatically connected to a trivial phase without closing the bulk gap , implying that the @xmath0 classification breaks down to @xmath1 with interactions @xcite .
this result suggests an interesting platform to study the classification of spt phases with cold atoms . ;
( b ) @xmath80-particle fractionalization ( seen by @xmath83 ) for zero modes .
the parameters @xmath25 and @xmath84 , with which the localization length of bound modes @xmath85=2.36a . ]
existence of zero modes leads to particle fractionalization , which is another direct observable in experiment .
a zero mode is contributed half from the valence band and half from the conduction band .
therefore , an edge state carries @xmath86 ( @xmath87 ) particle if it is occupied ( unoccupied ) @xcite .
this result is confirmed by numerical simulation shown in fig .
[ zeromodes ] , where we calculate @xmath88 at half - filling , with @xmath89 the density of fermions when the left ( right ) edge mode is filled . the fermion number carried by an occupied ( unoccupied ) edge mode
is then given by @xmath90 with @xmath91 . here
@xmath92 is the localization length of @xmath93 .
the @xmath80-fractionalization is clearly seen when @xmath94 is several times greater than @xmath85 ( fig .
[ zeromodes ] ( b ) ) .
being a topological invariant , the @xmath80-fractionalization can be confirmed to be robust against weak disorder scatterings without breaking the given symmetries . _
correlation effects._@xmath2a particular advantage in cold atoms is that one can investigate correlation effects on the predicted spt phase by precisely controlling the interaction .
for spin-@xmath80 cold fermions the onsite hubbard interaction @xmath95 can be well controlled by feshbach resonance @xcite . for a single - chain system
, we expect that the topological phase is stable against weak interactions relative to the single - particle bulk gap , while the strong repulsive interaction can always drive the system into a mott insulating phase .
the correlation effects around the critical point can be probed by abelian bosonization approach combined with renormalization group ( rg ) analysis @xcite .
note in the non - interacting regime , the phase diagram of the single - chain system is determined by the so and zeeman terms which define two mass terms @xmath96 in the bosonized hamiltonian , where the masses @xmath97 , @xmath98 , and @xmath99 are boson representation of the fermion fields @xcite .
the fate of the system in the presence of the interaction depends on which mass term flows first to the strong coupling regime under rg .
a direct power counting shows the same rg flow for the masses @xmath100 and @xmath101 in the first - order perturbation .
therefore the next - order perturbation expansion is necessary to capture correctly the fate of the topological phase transition . by deriving the rg flow equations up to one - loop order @xcite
, we find the renormalization to @xmath102 , the umklapp scattering @xmath103 and spin backscattering @xmath104 by @xcite : @xmath105 where the bare values of the coupling constants @xmath106 , and @xmath107 is the logarithm of the length scale .
the renormalization of luttinger parameter @xmath108 has been neglected as it is a higher order correction . for @xmath109
, @xmath104 marginally flows to zero so we drop it off below .
this is consistent with the result that repulsive interaction can not gap out the spin sector in the 1d hubbard model .
@xmath103 is marginally relevant and can be solved by @xmath110 . substituting this result into rg equations of @xmath100 and @xmath101 yields after integration @xmath111^{\frac{1}{4}}e^{(3-k_\rho)l/2 } , w(l)=w(0)[1+\frac{g_\rho(0)l}{\pi v_f}]^{\frac{1}{4}}e^{(3-k_\rho)l/2}$ ] .
the physics is clear : the repulsive interaction ( @xmath112 ) suppresses so induced mass term @xmath100 while enhances the trivial mass term @xmath101 .
the fate of the system depends on which of @xmath100 and @xmath101 reaches the strong - coupling regime first .
assuming @xmath113 , we find the tp transition occurs at @xmath114^\gamma , \ : \gamma\approx 1-\frac{g_\rho(0)}{4\pi v_f(3-k_\rho)}. \label{}\ ] ] this gives the scaling law at the phase boundary with interaction .
note @xmath115 for @xmath109 .
the above scaling relation implies that a repulsive interaction suppresses the spt phase .
accordingly , if initially the noninteracting system is topologically nontrivial with @xmath116 , increasing @xmath117 to the regime @xmath118^\gamma$ ] drives the system into a trivial phase . _ single spin control._@xmath2now we study an interesting application of the present results to realizing single spin control .
besides the edge modes localized on the ends , tsqs can also be obtained in the middle areas by creating mass domains in the lattice .
this can be achieved by applying a local zeeman term @xmath119 or @xmath120 .
for example , we consider @xmath23 everywhere , but @xmath121 for @xmath122 and @xmath123 otherwise .
the local @xmath124 can be generated by applying another two lasers which cross with the 1d lattice and couple the atoms in the area @xmath122 to induce a local resonant raman coupling between @xmath15 and @xmath16 ( fig .
[ oscillation](a ) ) . employing a @xmath125-phase offset in the rabi - frequencies of the two lasers ,
the raman coupling takes the form @xmath126 , with @xmath127 controlled by the laser strength .
when @xmath128 a mass domain is created , associated with two midgap spin states @xmath129 respectively localized around @xmath130 ( fig .
[ oscillation](a ) ) . the width @xmath131 and height of the domain
are respectively adjusted by the waist size and strength of the two laser beams . due to the nonlocality of the tsq , creation of a single qubit here
is not restricted by the size of the laser beams .
this is a fundamental difference from creating conventional single qubit by optical dipole trapping which requires tiny - sized laser beams to reach a very small trapping volume @xcite .
note in realistic case the laser induced @xmath124 may vary fast but not in the form of step functions around @xmath130 , which , however , does not affect the main results presented here @xcite .
coupling between @xmath132 and @xmath133 results in an energy splitting @xmath134 , which is controlled by @xmath127 and @xmath135 ( fig .
[ oscillation](a ) , lower panel ) . in the limit @xmath136 ,
such coupling is negligible , and the two zero modes consist of a single spin qubit which is topologically stable .
let @xmath132 be initially occupied while @xmath133 be left vacancy .
reducing @xmath137 smoothly can open the coupling in @xmath129 and lead to spin state evolving as @xcite @xmath138 , with @xmath139 , and @xmath140 the spatial parts of the bound state wave - functions .
the spin - polarization densities are given by @xmath141 , and the spin expectation values @xmath142 .
it can be verified that @xmath143 , and @xmath144 .
\end{split } \label{}\ ] ] this phenomenon is analogous to spin precession with the rotating angle yielding @xmath145 .
we have then @xmath146 and @xmath147 .
the amplitude of @xmath148 is given by @xmath149 , which measures the overlapping integral of @xmath140 .
accordingly , if we apply the local zeeman field along @xmath17 rather than @xmath55 axis , we shall obtain that the spin evolves in the @xmath150-@xmath55 plane .
note the spin rabi - oscillation is induced by quantum tunneling .
therefore it is associated with a tunneling current given by @xmath151 between @xmath152 and @xmath153 . in experiment the internal states of a single atom
can be detected without energy transfer @xcite , which is applicable to observe the spin rabi - oscillations , while the oscillation of @xmath154 can be more conveniently observed by measuring the number of fermions @xmath155 localized around @xmath156 with single - site resolution technology @xcite , and @xmath157 can be detected by measuring the change rate with time of such fermion numbers .
khz , @xmath25 , and @xmath158 .
( a ) mass domain created by setting @xmath159 for @xmath122 which localizes a spin - qubit composed of two bound modes @xmath129 on @xmath160 , respectively ; ( b ) spin rabi - oscillation by smoothly reducing @xmath137 from @xmath161 to @xmath162 ; ( c ) the mass current @xmath157 and expectation values of particle numbers @xmath155 in states @xmath129 ; ( d ) spin - flip operation by controlling that @xmath163 with @xmath164 .
the initial spin state @xmath165 ( points a ) flips to be @xmath166 ( points b ) . ]
we show in fig .
[ oscillation ] ( b - d ) the numerical simulation for single spin control with the parameter regime that @xmath167khz , @xmath25 , and @xmath168
. for @xmath169 , @xmath170 and the coupling in @xmath129 is negligible .
reducing @xmath127 at @xmath171 leads to spin evolution and by fixing @xmath172 for @xmath173 the spin oscillates with a period of @xmath174ms ( b - c ) . note the quantum state of the spin can be precisely controlled by properly manipulating @xmath175 .
for example , in fig .
[ oscillation ] ( d ) we demonstrate the spin - flip operation @xmath176 by requiring @xmath163 . here
@xmath177 and in ( d ) we take @xmath164 .
note one may integrate multiple tsqs with e.g. atom - chip technology and individually control them by creating multiple mass domains in the 1d lattice .
the precise manipulation of such integrated tsqs may have interesting applications in developing scalable spin - based quantum computers .
before conclusion we estimate the parameter values for realistic experimental observations . for example , in @xmath178k atoms we have the recoil energy @xmath179khz using red - detuned lasers of wavelength @xmath180 nm to form the optical lattice .
taking that @xmath181 and @xmath182 , we have that the lattice trapping frequency @xmath183khz , and hopping coefficients @xmath184khz and @xmath185khz
. then the bulk gap equals @xmath186khz for @xmath23 , indicating a temperature @xmath187nk for the experimental observation .
also , under this parameter regime the life time of the atoms is over @xmath188s , which is long enough for the detection and manipulation of the topological edge spins . _
conclusions._@xmath2we have proposed to observe and manipulate spt phase of aiii class in 1d optical lattice , and demonstrated single spin control by manipulating spin - polarized zero modes which is applicable to spin - based quantum computation .
the minimum requirement for the proposed scheme is a regular 1d lattice and a transverse zeeman field , which can be realized simultaneously in a single two - photon raman transition as used in the recent experiments @xcite .
the present study may open the search for topological states of all ten altand - zirnbauer symmetry classes with realistic cold atom systems , and its remarkable feasibility will attract both theoretical and experimental efforts in future .
we thank tin - lun ho , x. g. wen , x. chen , v. w. liu , and cenke xu for helpful discussions .
we acknowledge the support from jqi - nsf - pfc , microsoft - q , and darpa - quest .
99 k.v .
klitzing , g. dorda , and m. pepper , phys .
rev .
lett .
* 45 * , 494 ( 1980 ) ; d. c. tsui , h. l. stormer , and a. c. gossard , phys .
rev .
lett . *
48 * , 1559 ( 1982 ) .
f. d. m. haldane , phys .
rev .
lett . * 50 * , 1153 ( 1983 ) ; phys . lett . * 93 * , 464 ( 1983 ) ; n. read and d. green , phys .
rev .
b * 61 * , 10267 ( 2000 ) ; a. y. kitaev , physics - uspekhi * 44 * , 131 ( 2001 ) . c. l. kane and e. j. mele , phys . rev .
lett . * 95 * , 146802 ( 2005 ) ; b. a. bernevig , t. l. hughes and s. -c .
zhang , science , * 314 * 1757 ( 2006 ) ; l. fu , c. l. kane , and e. j. mele , phys .
rev .
lett . * 98 * , 106803 ( 2007 ) ; j. e. moore and l. balents , phys .
rev .
b * 75 * , 121306 ( r ) ( 2007 ) ; r. roy , _ ibid _ * 79 * , 195322 ( 2009 ) .
m. z. hasan and c. l. kane , rev .
mod .
phys . * 82 * , 3045 ( 2010 ) ; x. -l .
qi and s. -c .
zhang , _ ibid _ * 83 * , 1057 ( 2011 ) .
a. altland and martin r. zirnbauer , phys .
rev .
b * 55 * , 1142 ( 1997 ) .
z. c. gu and x. g. wen , phys .
rev .
b * 80 * , 155131 ( 2009 ) ; x. chen , z. -c .
gu , and x. -g .
wen , _ ibid _ * 82 * , 155138 ( 2010 ) ; x. -g .
wen , _ ibid _ * 85 * , 085103 ( 2012 ) .
shinsei ryu , andreas schnyder , akira furusaki , andreas ludwig , new j. phys .
* 12 * , 065010 ( 2010 ) .
l. fidkowski and a. kitaev , phys .
rev .
b * 81 * , 134509 ( 2010 ) ; a. m. turner , f. pollmann , and e. berg , _ ibid _ * 83 * , 075102 ( 2011 ) ; l. fidkowski and a. kitaev , _ ibid _ * 83 * , 075103 ( 2011 ) .
y. kitaev , ann .
phys .
303 , 2 ( 2003 ) ; s. das sarma , m. freedman , and c. nayak , phys .
rev .
lett .
94 , 166802 ( 2005 ) ; c. nayak et al . , rev .
mod .
phys . * 80 * , 1083 ( 2008 ) .
x. -j .
liu , m. f. borunda , x. liu , and j. sinova , phys .
rev .
lett . * 102 * , 046402 ( 2009 ) .
y .- j .
lin , k. jimnez - garca , and i. b. spielman , nature * 471 * , 83 ( 2011 ) . m. chapman and c. s de melo , nature * 471 * , 41 ( 2011 ) . p. wang et al .
, phys .
rev .
lett . * 109 * , 095301 ( 2012 ) .
l. w. cheuk , a. t. sommer , z. hadzibabic , t. yefsah , w. s. bakr , and m. w. zwierlein , phys .
rev .
lett . * 109 * , 095302 ( 2012 ) .
jin - yi zhang et al . , phys .
rev .
lett . * 109 * , 115301 ( 2002 ) .
j. larson , j .-
p .
martikainen , a. collin , and e. sjoqvist , phys .
rev .
a * 82 * , 043620 ( 2010 ) ; d. sokolovski and e. ya .
sherman , _ ibid _ * 84 * , 030101(r ) ( 2011 ) ; j. d. sau , r. sensarma , s. powell , i. b. spielman , and s. das sarma , phys .
rev .
b * 83 * , 140510(r ) ( 2011 ) ; t. ozawa and g. baym , phys . rev . lett . * 109 * , 025301 ( 2012 ) ; g. i. martone , y. li , l. p. pitaevskii , and s. stringari , arxiv:1207.6804 ( 2012 ) .
x. -j .
liu , x. liu , l. c. kwek and c. h. oh , phys . rev .
lett . * 98 * , 026602 ( 2007 ) ; phys .
rev .
b * 79 * , 165301 ( 2009 ) . c. wu , phys . rev .
lett .
* 101 * , 186807 ( 2008 ) ; x .- j .
liu , x. liu , c. wu , and j. sinova , phys .
rev .
a * 81 * , 033622 ( 2010 ) ; y. yu and k. yang , phys .
rev .
lett . * 105 * , 150605 ( 2010 ) .
n. goldman , i. satija , p. nikolic , a. bermudez , m.a .
martin - delgado , m. lewenstein , and i. b. spielman , phys .
rev .
lett .
* 105 * , 255302 ( 2010 ) ; n. goldman , j. beugnon , and f. gerbier , _ ibid _ * 108 * , 255303 ( 2012 ) .
x. li , e. zhao , and v. w. liu , arxiv:1205.0254 ( 2012 ) .
g. liu , s. -l .
zhu , s. jiang , f. sun , and w. m. liu , phys .
rev .
a * 82 * , 053605 ( 2010 ) ; f. mei , s. -l .
zhu , z. -m .
zhang , c. h. oh , and n. goldman , _ ibid _ * 85 * , 013638 ( 2012 ) . c. zhang , s. tewari , r. lutchyn , and s. das sarma , phys .
rev .
lett .
* 101 * , 160401 ( 2008 ) ; y. zhang , l. mao , c. zhang , _ ibid _ * 108 * , 035302 ( 2012 ) .
m.sato , y. takahashi , and s. fujimoto , phys .
rev .
lett .
103 020401 ( 2009 ) .
s. -l .
zhu , l. b. shao , z. d. wang , and l. -m .
duan , phys .
rev .
lett . * 106 * , 100404 ( 2011 ) ; w. yi and g .- c .
guo , phys .
rev .
a * 84 * , 031608(r ) ( 2011 ) ; l. he and x. -g .
huang , phys .
rev .
lett . * 108 * , 145302 ( 2012 ) ; phys .
rev .
b * 86 * , 014511 ( 2012 ) . kangjun seo , li han , and c. a. r. s de melo , phys
. rev .
lett .
* 109 * , 105303 ( 2012 ) . c. v. kraus , s. diehl , p. zoller , and m. a. baranov , arxiv:1201.3253 ; s. nascimbe , arxiv:1210.0687 .
i. bloch , j. dalibard , and w. zwerger , rev .
mod .
phys . * 80 * , 885 ( 2008 ) .
see supplementary material for more details .
t. p. meyrath , f. schreck , j. l. hanssen , c .- s .
chuu , and m. g. raizen , phys .
rev .
a * 71 * , 041604(r ) ( 2005 ) .
note the degeneracy and spin - polarization of the edge modes obtained by f. mei et al . in ref .
@xcite are not symmetry protected and not stable against local perturbations .
therefore such edge modes can not form tsq .
t. giamarchi , quantum physics in one dimension ( oxford university press , 2004 ) .
l. balents and m. p. a. fisher , phys .
rev .
b * 53 * , 12133 ( 1996 ) .
n. schlosser , g. reymond , i. protsenko , and p. grangier , nature , * 411 * , 1024 ( 2001 ) .
we neglect the couplings between @xmath129 and bulk states @xmath189 , which is valid when @xmath190 during the manipulation .
j. volz , r. gehr , g. dubois , j. estve , and j. reichel , nature * 475 * , 210 ( 2011 ) .
w. s. bakr , j.i .
gillen , a. peng , s. foelling , and m. greiner , nature * 462 * , 74 ( 2009 ) ; m. karski , l. frster , j. m. choi , w. alt , a. widera , and d. meschede , phys . rev .
lett . * 102 * , 053001 ( 2009 ) . |
white dwarfs ( wds ) are the final stage for the the evolution of majority of low and medium mass stars with initial masses @xmath2 @xmath3 .
since there are no fusions reaction , the evolution of wds is primarily determined by a well understood cooling process ( fontaine et al . 2001 ; salaris et al . 2000 ) .
thus , they can be used for cosmochronology , an independent age - dating method .
also , the luminosity function of wds provides firm constraints on the local star formation rate and history of the galactic disk ( krzesinski et al .
2009 ) .
mccook @xmath4 sion ( 1999 ) present a catalog of 2249 wds which have been identified spectroscopically .
in addition , the sloan digital sky survey ( sdss ; york et al .
2000 ) has greatly expanded the number of spectroscopically confirmed wd stars ( harris et al .
2003 ; kleinman et al . 2004 ; eisenstein et al . 2006 ; kleinman et al .
the latter presented a catalog of 20,407 spectroscopically confirmed white dwarfs from the sdss data release 4 ( dr4 ) , roughly doubling the number of spectroscopically confirmed white dwarfs .
large sky area multi - object fiber spectroscopic telescope ( lamost , so called the guoshoujing telescope ) is a national major scientific project undertaken by the chinese academy of science ( wang et al .
1996 ; cui et al .
lamost has recently completed the pilot survey from october 2011 to may 2012 , which obtained several hundred thousand spectra ( luo et al .
2012 ) . from september of 2012
, lamost has undertaken the general survey and will observe about 1 million stars per year .
lamost has the capability to observe large , deep and dense regions in the milky way galaxy , which will enable a number of research topics to explore the evolution and the structure of the milky way .
therefore , it will definitely yield a large sample of wds .
wds whose primary spectral classification is da have hydrogen - dominated atmospheres .
they make up the majority ( approximately 75% ) of all observed wds ( fontaine @xmath4 wesemael 2001 ) .
such wds are easy to identify using optical spectra . here
we present a catalog of da wds from the lamost pilot survey ( luo et al .
* we do not expect the completeness of this sample . * in section 2
we describe the spectra obtained .
section 3 discusses how the @xmath5 , log @xmath6 , mass and distance of the wds were estimated .
the kinematics of these wds are illustrated in section 4 .
a summary of our pilot study results is given in section 5 .
the lamost spectra have a resolving power of r @xmath0 2000 spanning 3700@xmath7 9000@xmath8 .
two arms of each spectrograph cover this wavelength range with overlap of 200 @xmath9 .
the blue spectral coverage is 3700@xmath10 5900@xmath8 while that in the red is 5700@xmath10 9000 @xmath8 .
the raw data were reduced with lamost 2d and 1d pipelines ( luo et al .
these pipelines include bias subtraction , cosmic - ray removal , spectral trace and extraction , flat - fielding , wavelength calibration , sky subtraction , and combination .
the throughput in red is higher than the blue band .
the pilot survey obtained spectra of stars in the milky way , which included fainter objects on dark nights ( yang et al .
2012 ; carlin et al .
2012 ) , brighter objects on bright nights ( zhang et al .
2012 ) , objects in the disk of the galaxy with low latitude ( chen et al . 2012 ) and objects in the region of the galactic anti - center .
it also targets extragalactic objects located in two regions , i.e. , the south galactic cap and the north galactic cap .
we found twenty wd spectra in both sdss and lamost pilot survey catalogs .
1 shows a portion of a typical spectrum .
the top panel compares the sdss dr7 and lamost spectra for the object j100316.35 - 002336.95 .
the solid line is the sdss spectrum .
the dotted line is the lamost spectrum .
the bottom panel shows the residual between two spectra .
the mean difference between two spectra is less than 10 % .
the initial wd candidates we selected are from two sources .
one is the lamost pipeline ( luo et al .
2012 ) which yielded about 2000 candidates using the `` pcaz '' method . for stars with sdss photometry
, we used the formulas 1 - 4 of elsenstein et al .
( 2006 ) to idendify candidates .
next , each of these spectra was inspected by eye .
stars with signal - to - noise ratio ( s / n ) smaller than 10 were excluded . finally , if the balmer line profiles of the star were a little too narrow ( log @xmath6 @xmath2 7.0 ) , the spectrum was rejected even if selected by the pipeline . after these filters ,
72 da wds were left .
table 1 presents the physical data for these wds .
column 1 is an i d number . columns 2 - 5 list the name , ra and dec .
the estimated @xmath5 , log @xmath6 , mass and the cooling time are given in columns 5 - 8 .
columns 9 - 13 list the apparent magnitudes of each wds .
column 14 indicates the source of the magnitudes .
the last two columns are estimates of the color excess ( b - v ) and distance .
the e(b - v ) is estimated from schlegel et al .
for our da wd candidates , the @xmath11@xmath12 and log @xmath6 were derived via simultaneous fitting of the h@xmath13 to h8 balmer line profiles using the procedure outlined by bergeron et al .
the line profiles in both observed spectra and model spectra were first normalized using two points at the continuum level on either side of each absorption line .
thus , the fit should not be affected by the flux calibration .
model atmospheres used for this fitting were derived from model grids provided by koester ( 2010 ) .
details of the input physics and methods can be found in that reference .
fitting of the line profiles was carried out using the idl package mpfit ( markwardt 2008 ) , which is based on @xmath14 minimization using levenberg - marquardt method .
this package can be downloaded from the project website .
errors in the @xmath11@xmath12 and log @xmath6 were calculated by stepping the parameter in question away from their optimum values and redetermining minimum @xmath14 until the difference between this and the true minimum @xmath14 corresponded to 1@xmath15 for a given number of free model parameters .
2 - 3 show examples of @xmath11@xmath12 and log @xmath6 determinations for j150156.26 + 302300.13 .
fig.2 is the contour plot of the @xmath14 residual and the rough @xmath11@xmath12 and log @xmath6 implied by these error eclipses .
3 shows the actual fits of the observed balmer lines for j150156.26 + 302300.13 .
the black solid lines are the observed profiles of balmer lines from h@xmath13 to h8 .
the red dashed lines are the model spectra .
the derived @xmath11@xmath12 , log @xmath6 and uncertainties for all the wds are shown in columns 5 - 6 of table 1 . estimated @xmath11@xmath12 and log @xmath6 values for 14 das
were also available in the literature , allowing the comparisons shown in fig .
the solid line represents the unit slope relation . plus
( + ) symbols represent the wds with high s / n spectra while squares represent wds with low s / n spectra .
the three spectra of lowest s / n are outliers in the log @xmath6 comparison plot - suggesting the importance of s / n in determining this parameter . for most of other wds , the mean differences between our and the literature @xmath11@xmath12 values are less than 1000 k and the log @xmath6 difference is less than 0.2 dex . within this scatter ,
our results are consistent with those in the literatures .
one of our candidates , j104311.45 + 490224.35 has also been identified as da wd by mccook @xmath4 sion ( 1999 ) .
however , we were unable to determine its @xmath11@xmath12 and log @xmath6 because h@xmath13 was not included in the spectrum we obtained . from the @xmath11@xmath16 and log @xmath6 of each wd , its mass ( m@xmath17 ) and cooling time ( t@xmath18 )
were estimated from bergeron s cooling sequences .
for the model atmospheres above @xmath11@xmath16 = 30,000 k we used the carbon - core cooling models of wood ( 1995 ) , with thick hydrogen layers of q@xmath19 = m@xmath19/m@xmath20 = 10@xmath21 . for @xmath11@xmath16 below 30,000 k we used cooling models similar to those described in fontaine , brassard @xmath4 bergeron ( 2001 ) but with carbon - oxygen cores and q@xmath19 = 10@xmath21 ( see bergeron , leggett @xmath4 ruiz 2001 ) .
5 is the mass distribution of our sample resulting from the above procedure .
masses are found to range from 0.4 @xmath1 to 1.2 @xmath1 .
the curve is a gaussian fit with a peak at about 0.61 @xmath1 , which is consistent with the mean mass 0.613 @xmath1 from tremblay et al .
( 2011 ) derived from sdss da wds sample .
the determination of distances for wds is very difficult because of their low luminosity .
currently only about 300 wds have trigonometric parallaxes . in the absence of parallaxes , color - magnitude relations and empirical photometric methods
are often used .
holberg et al .
( 2008 ) provided improved distance estimates for da wds using multi - band synthetic photometry tied to spectroscopic temperatures and gravities .
this method was called synthetic spectral distances ( ssd ) .
the unique aspect of ssd is the systematic use of calibrated multi - channel synthetic absolute magnitudes , interpolated within the grid by the @xmath5 and log @xmath6 .
@xmath22 in this paper , the distances of wds in our sample were estimated using equation 1 . here
, @xmath23 are the photometric magnitudes of the wds .
most of our has _ u _ , _ g _ , _ r _ , _ i _ , _ z _ magnitudes . almost all have at least _ g _ ,
_ r _ and _ i _ magnitudes .
a few wds still only have _ v _ magnitude .
@xmath24 is the model absolute magnitudes calculated by interpolations in the atmospheric models provided by bergeron .
@xmath25 is the reddening and d is the distance in parsecs . in general for each magnitude a corresponding distance can be calculated . the final distance is estimated by using weighted average .
the weights adopted are the errors in the magnitude . here , we only calculated the distances for wds having _ u _ , _ g _ , _ r _ , _ i _ , _ z _ or _ v _ magnitude data .
distances for two wds in table 1 could not be estimated .
oppenheimer et al . ( 2001 ) suggested that halo wds could provide a significant contribution to the galactic dark matters component , which prompted much interest in wd kinematics . in a related study , silvestri et al .
( 2002 ) observed 116 common proper - motion binaries consisting of a wd plus m dwarf component .
they determined full space motions of their wds from the companion m dwarfs .
most of their wds were found to be members of the disk ; only one potential halo wd was identified . even the much larger samples of wds such as the pauli et al .
( 2003 , 2006 ) sn ia progenitor survey ( spy ) have found relatively few genuine halo and thick disk candidates . in their magnitude - limited sample of 398 wds
, they examined both the uvw space motions and the galactic orbits of their stars .
they found only 2% of their sample kinematically belonged to the halo and 7% to the thick disk .
sion et al .
( 2009 ) presented the kinematical properties of the wds within 20 pc of the sun . in their nearby sample
, they found no convincing evidence of halo members among 129 wds , nor was there convincing evidence of genuine thick disk subcomponent members within 20 parsecs .
the entire 20 pc sample likely belongs to the thin disk .
the proper motions of our sample were derived by the cross - correlating with ppmxl catalog ( roeser et al .
silvestri et al . ( 2002 ) ; pauli et al . ( 2003 , 2006 ) , sion et al .
( 2009 ) found relatively little kinematical difference among the samples whether they used radial velocity ( rv ) to compute full space motions or used the simple zero rv for simple wds .
we have assumed zero rvs in the analysis of our sample .
u is measured positive in the direction of the galactic anti - center , v is measured positive in the direction of the galactic rotation , and w is measured positive in the direction of the north galactic pole .
the u , v and w velocities were corrected for the peculiar solar motion ( u , v , w ) = ( -9 , + 12 , + 7 ) km s@xmath26 ( wielen 1982 ) . the space motions of 59 wds with sufficient kinematical information ( photometric or trigonometric parallax , proper motion ) in our sample were calculated .
the top panel of fig .
6 shows contours , centered at ( u , v ) = ( 0 , -220 ) km s@xmath26 , that represent 1@xmath15 and 2@xmath15 velocity ellipsoids for stars in the galactic stellar halo as defined by chiba @xmath4 beers ( 2000 ) .
only one of our candidate wds lies outside the 2@xmath15 velocity contour centered on ( u , v ) = ( 0 , -35 ) km s@xmath26 defined for disk stars ( chiba @xmath4 beers 2000 ) .
the bottom of fig .
6 shows a toomre diagram for our stars .
venn et al .
( 2004 ) suggested stars with v@xmath27 @xmath28 180 km s@xmath26 are possible halo members .
none of our stars meet this criterion .
we conclude that our sample consistes entirely of disk stars .
wegg @xmath4 phinney ( 2012 ) concluded that kinematical dispersion decreases with increasing wd mass among young wds whose cooling time is smaller than 3@xmath2910@xmath30 years .
progenitors of high mass wds have shorter lifetimes , hence they should be closer to the galactic plane and have small kinematical dispersion in accord with the disk ` heating ' theory . since most wds in our sample are relative young , we investigated the relation between mass and w , as well as mass and vertical distance of galactic plane @xmath31z@xmath31 ( see fig . 7 ) . in the top panel of fig .
7 , wds with mass larger than 0.8 @xmath32 are seen to have smaller w. also , the vertical distances from the galactic plane of wds with larger mass are relative small .
although there is no strict relation such as seen in wegg @xmath4 phinney ( 2012 ) , our sample support the general expectation that high mass wds tend to have lower w and @xmath31z@xmath31 .
from the lamost pilot survey data , 72 da wds were detected with s / n @xmath28 10 .
@xmath11@xmath12 , log @xmath6 , cooling time , mass and distance of these wds were determined from their spectra .
the @xmath5 of most wds range from 12000 k to 35000 k and the cooling times of all the wds are younger than 300 myr .
all these wds were found to be members of galactic disk .
wds with higher mass tend to have smaller vertical distance from the galactic disk , which partly supports the conclusions of wegg et al .
the da wd catalogue of the lamost pilot survey provides a first glimpse of how useful the survey will be to search for nearby wds .
the upcoming formal lamost survey will enlarge the sample of wds rapidly , perhaps providing the largest sample of wds available .
this large sample will open the door to much more detailed investigation of the physical @xmath4 kinematic properties of wds in the solar neighborhood as well as the local structure and evolution of the galaxy .
many thanks to d. koester for providing his wd models .
balmer / lyman lines in the models were calculated with the modified stark broadening profiles of tremblay @xmath4 bergeron ( 2009 ) , kindly made available by the authors .
this study is supported by the national natural science foundation of china under grant no .
11233004 , 11078019 and 10973021 .
t.d.o . acknowledges support from nsf grant ast-0807919 to florida institute of technology
. guoshoujing telescope ( the large sky area multi - object fiber spectroscopic telescope lamost ) is a national major scientific project built by the chinese academy of sciences .
funding for the project has been provided by the national development and reform commission .
lamost is operated and managed by the national astronomical observatories , chinese academy of sciences .
carlin , j. l. , lpine , s. , newberg , h. j. , et al .
2012 , raa ( research in astronomy and astrophysics ) , 12 , 755 chen , l. , hou , j. , yu , j. , et al .
2012 , raa(research in astronomy and astrophysics ) , 12 , 805 chiba , m. , @xmath4 beers , t. c. , 2000 , , 119 , 2843 cui , x. , zhao , y. , chu , y. , et al .
2012 , raa(research in astronomy and astrophysics ) , 12 , 1197 eisenstein , d. j. , liebert , j. , harris , h. c. , et al .
2006 , apjs , 167 , 40 fontaine , g. , brassard , p. , @xmath4 bergeron , p. 2001 , pasp , 113 , 409 harris , h. c. et al . , 2003 , , 126 , 1023 holberg , j. b. , bergeron , p. , @xmath4 gianninas , a. 2008 , , 135 , 1239 kleinman , s. j. , harris , h. c. , eisenstein , d. j. , et al . 2004 , , 607 , 426 kleinman , s. j. , kepler , s. o. , koester , d. et al .
2013 , , 204 , 5 koester , d. , 2010 , to appear in mem.s.a.i . , based on lectures given at the school of astrophysics `` f. lucchin '' , tarquinia , june 2008 eprint arxiv:0812.0482 krzesinski , j. , kleinman , s. j. , nitta , a. , hgelmeyer , s. , dreizler , s. , liebert , j. , @xmath4harris , h. 2009 , , 508 , 339 luo , a. l. , zhang , y. x. , @xmath4 zhao , y. h. 2004 , in society of photo - optical instrumentation engineers ( spie ) conference series , vol .
5496 , eds . h. lewis @xmath4 g. raffi , 756 luo , a. l. , zhang , h. t. , zhao , y. h , et al .
2012 , raa(research in astronomy and astrophysics ) , 12 , 9 markwardt , c. b. , 2008 , `` non - linear least squares fitting in idl with mpfit , '' in proc .
astronomical data analysis software and systems xviii , quebec , canada , asp conference series , vol .
411 , eds .
d. bohlender , p. dowler @xmath4 d. durand ( astronomical society of the pacific : san francisco ) , p. 251
- 254 ( isbn : 978 - 1 - 58381 - 702 - 5 ; ads bibcode : 2009aspc .. 411 .. 251 m ( click for bibtex and other citation formats ) arxiv preprint : arxiv:0902.2850v1 [ [ mccook @xmath4 sion ( 1999)]mcc99 mccook , e. p. , @xmath4 sion , e. m. 1999 , , 121 , 1 oppenheimer , b. r. , hambly , n. c. , digby , a. p. , hodgkin , s. t. , @xmath4 saumon , d. 2001 , science , 292 , 698 pauli , e .-
m . , napiwotzki , r. , heber , u. , altmann , m. , @xmath4 odenkirchen , m. , 2006 , , 447 , 173 roeser , s. , demleitner , m. , @xmath4schilbach e. 2010 , , 139 , 2440 salaris , m. , garca - berro , e. , hernanz , m. , isern , j. , @xmath4 saumon , d. 2000 , apj , 544 , 1036 schlegel , d. j. , finkbeiner , d. p. , @xmath4davis , m. 1998 , , 500 , 525 sion , e. m. , holberg , j. b. , oswalt , t. d. , mccook , g. p. , @xmath4 wasatonic , r. 2009 , , 138 , 1681 silvestri , n. , oswalt , t. d. , @xmath4 hawley , s. 2002 , aj , 124 , 1118 venn , k. a. , irwin , m. , shetrone , m. d. , tout , c. a. , hill , v. , @xmath4 tolstoy , e. , 2004 , , 128 , 1177 wang , s. g. , su , d. q. , chu , y. q. , cui , x. , @xmath4 wang , y. n. 1996 , appl .
, 35 , 5155 wegg , c. , @xmath4 phinney , e. s. 2012 , , 426 , 427 wielen , r. 1982 , landolt - brnstein tables , astrophysics , vol . 2c , sec .
8.4 ( berlin : springer ) , 202 yang , f. , carlin , j. l. , newberg , h. j. , et al .
2012 , raa ( research in astronomy and astrophysics ) , 12 , 781 york , d. g. , et al .
2000 , , 120 , 1579 zhang , y. , carlin , j. , liu , c. , et al .
2012 , raa ( research in astronomy and astrophysics ) , 12 , 792 ccccccccccccccccc 0 & j220522.86 + 021837.56 & 331.345250 & 2.310432 & 15377 @xmath33 493 & 8.02 @xmath33 0.10 & 0.63 @xmath33 0.06 & 190 @xmath33 43 & 17.35 & 17.00 & 17.25 & 17.45 & 17.71 & & 1 & 0.05 & 135 + 1&j025737.25 + 264047.89 & 44.405201 & 26.679970 & 19008 @xmath33 669 & 7.87 @xmath33 0.12 & 0.55 @xmath33 0.06 & 66 @xmath33 21 & & 16.91 & 17.00 & 17.16 & & & 2 & 0.16 & 139 + 2&j030214.72 + 285707.41 & 45.561340 & 28.952057 & 21894 @xmath33 1406 & 8.01 @xmath33 0.23 & 0.64 @xmath33 0.13 & 46 @xmath33 37 & & 17.21 & 17.60 & 17.80 & & & 2 & 0.18 & 173 + 3&j040449.34 + 280023.65 & 61.205600 & 28.006570 & 29302 @xmath33 2525 & 8.25 @xmath33 0.55 & 0.79 @xmath33 0.33 & 20 @xmath33 32 & & 15.96 & 15.84 & 16.00 & & & 2 & 0.21 & 87 + 4&j004036.79 + 413138.79 & 10.153296 & 41.527443 & 13000 @xmath33 651 & 7.75 @xmath33 0.08 & 0.48 @xmath33 0.04 & 216 @xmath33 47 & & 15.90 & 16.21 & 16.40 & & & 2 & 0.07 & 83 + 5&j003956.55 + 422929.55 & 9.985629 & 42.491542 & 18053 @xmath33 816 & 7.32 @xmath33 0.15 & 0.34 @xmath33 0.05 & 50 @xmath33 10 & & 16.43 & 16.58 & 16.72 & & & 2 & 0.06 & 181 + 6&j004128.67 + 402324.09 & 10.369458 & 40.390026 & 25996 @xmath33 733 .
& 7.92 @xmath33 0.10 & 0.59 @xmath33 0.05 & 15 .
@xmath33 3 . & & & & & & 17.14 & 3 & 0.08 & 68 + 7&j005340.53 + 360116.89 & 13.418857 & 36.021358 & 29772 @xmath33 158 & 7.96 @xmath33 0.04 & 0.63 @xmath33 0.02 & 9 @xmath33 0 & & 14.10 & 14.58 & 14.91 & & & 2 & 0.05 & 72 + 8&j100551.51 - 023417.87 & 151.464628 & -2.571630 & 22072 @xmath33 477 & 8.22 @xmath33 0.07 & 0.76 @xmath33 0.04 & 78 @xmath33 16 & 15.15 & 15.10 & 15.46 & 15.76 & 16.08 & & 1 & 0.05 & 68 + 9&j100316.35 - 002336.95 & 150.818141 & -0.393597 & 22249 @xmath33 330 & 7.92 @xmath33 0.05 & 0.59 @xmath33 0.03 & 33 @xmath33 4 & 15.97 & 15.93 & 16.25 & 16.56 & 16.85 & & 1 & 0.05 & 123 + 10 & j100941.45 - 004404.55 & 152.422705 & -0.734597 & 16489 @xmath33 601 & 7.98 @xmath33 0.13 & 0.60 @xmath33 0.08 & 140 @xmath33 45 & 17.36 & 16.98 & 17.24 & 17.44 & 17.74 & & 1 & 0.04 & 148 + 11 & j054613.53 + 255031.70 & 86.556364 & 25.842139 & 22935 @xmath33 498 & 7.99 @xmath33 0.07 & 0.63 @xmath33 0.04 & 34 @xmath33 10 & & 17.33 & 17.62 & 17.78 & & & 2 & 1.72 & 27 + 12 & j090734.26 + 273903.32 & 136.892757 & 27.650923 & 18619 @xmath33 386 & 8.56 @xmath33 0.07 & 0.97 @xmath33 0.04 & 272 @xmath33 39 & 16.31 & 16.08 & 16.37 & 16.64 & 16.89 & & 1 & 0.03 & 72 + 13 & j004628.31 + 343319.90 & 11.617971 & 34.555527 & 14644 @xmath33 808 & 7.60 @xmath33 0.18 & 0.41 @xmath33 0.08 & 120 @xmath33 47 & 16.83 & 16.33 & 16.40 & 16.53 & 16.75 & & 1 & 0.08 & 112 + 14 & j005340.53 + 360116.89 & 13.418857 & 36.021358 & 26534 @xmath33 394 & 7.88 @xmath33 0.06 & 0.58 @xmath33 0.03 & 13 @xmath33 1 & & 14.10 & 14.58 & 14.91 & & & 2 & 0.05 & 67 + 15 & j052038.36 + 304822.65 & 80.159836 & 30.806293 & 15924 @xmath33 348 & 8.00 @xmath33 0.07 & 0.61 @xmath33 0.04 & 164 @xmath33 28 & & 15.38 & 15.68 & 15.88 & & & 2 & 0.85 & 24 + 16 & j031236.50 + 515511.74 & 48.152099 & 51.919927 & 23558 @xmath33 1966 & 7.93 @xmath33 0.29 & 0.59 @xmath33 0.14 & 25 @xmath33 13 & & & & & & 15.44 & 3 & 0.84 & 103 + 17 & j055046.51 + 261220.27 & 87.693772 & 26.205631 & 28000 @xmath33 1916 & 8.34 @xmath33 0.39 & 0.84 @xmath33 0.24 & 37 @xmath33 57 & & 15.13 & 15.64 & 15.91 & & & 2 & 1.50 & 13 + 18 & j013938.94 + 291859.80 & 24.912266 & 29.316611 & 20934 @xmath33 515 & 8.13 @xmath33 0.08 & 0.70 @xmath33 0.05 & 77 @xmath33 22 & & 17.53 & 17.94 & 18.19 & & & 2 & 0.05 & 213 + 19&j105811.27 + 475752.75 & 164.546942 & 47.964653 & 29532 @xmath33 490 & 7.84 @xmath33 0.11 & 0.56 @xmath33 0.05 & 9 @xmath33 0 & 17.09 & 17.29 & 17.75 & 18.10 & 18.35 & & 1 & 0.01 & 353 + 20 & j104311.45 + 490224.35 & 160.797708 & 49.040097 & & & & & 15.47 & 15.84 & 16.40 & 16.76 & 17.18 & & 1 & 0.01 & + 21 & j053931.86 + 285456.66 & 84.882770 & 28.915740 & 23865 @xmath33 1774 & 8.63 @xmath33 0.26 & 1.01 @xmath33 0.15 & 147 @xmath33 97 & & 17.39 & 16.64 & 16.17 & & & 2 & 1.43 & 15 + 22 & j094104.43 + 282224.58 & 145.268457 & 28.373495 & 16713 @xmath33 438 & 7.86 @xmath33 0.09 & 0.54 @xmath33 0.05 & 109 @xmath33 24 & 15.70 & 15.42 & 15.70 & 15.94 & 16.25 & & 1 & 0.02 & 82 + 23 & j081845.28 + 121952.45 & 124.688667 & 12.331236 & 22271 @xmath33 531 & 8.34 @xmath33 0.08 & 0.83 @xmath33 0.05 & 100 @xmath33 22 & 16.32 & 16.18 & 16.57 & 16.88 & 17.19 & & 1 & 0.03 & 107 + 24 & j014147.59 + 302135.45 & 25.448307 & 30.359846 & 17520 @xmath33 367 & 8.17 @xmath33 0.07 & 0.72 @xmath33 0.04 & 162 @xmath33 27 & & 16.96 & 17.39 & 17.54 & & & 2 & 0.05 & 138 + 25 & j014933.76 + 285610.60 & 27.390679 & 28.936279 & 32200 @xmath33 631 & 8.33 @xmath33 0.12 & 0.84 @xmath33 0.07 & 17 @xmath33 10 & & 16.88 & 17.47 & & & & 2 & 0.06 & 140 + 26 & j074742.05 + 280945.57 & 116.925192 & 28.162658 & 15085 @xmath33 596 & 7.66 @xmath33 0.13 & 0.44 @xmath33 0.06 & 117 @xmath33 31 & 17.83 & 17.43 & 17.69 & 17.88 & 18.13 & & 1 & 0.04 & 209 + 27 & j075251.35 + 271513.85 & 118.213962 & 27.253847 & 25134 @xmath33 711 & 7.94 @xmath33 0.10 & 0.60 @xmath33 0.06 & 19 @xmath33 9 & 16.72 & 16.73 & 17.15 & 17.46 & 17.79 & & 1 & 0.03 & 206 + 28 & j075106.48 + 301726.96 & 117.776979 & 30.290822 & 34418 @xmath33 580 & 8.21 @xmath33 0.10 & 0.78 @xmath33 0.06 & 9 @xmath33 1 & 15.65 & 15.92 & 16.39 & 16.72 & 17.05 & & 1 & 0.05 & 156 + 29&j113614.04 + 290130.26 & 174.058504 & 29.025072 & 24106 @xmath33 255 & 7.80 @xmath33 0.03 & 0.53 @xmath33 0.02 & 20 @xmath33 1 & 14.64 & 14.68 & 15.13 & 15.44 & 15.75 & & 1 & 0.02 & 87 + 30 & j113705.17 + 294757.54 & 174.271529 & 29.799317 & 21786 @xmath33 160 & 8.58 @xmath33 0.03 & 0.98 @xmath33 0.02 & 174 @xmath33 11 & 12.29 & 12.31 & 12.69 & 12.99 & 13.31 & & 1 & 0.02 & 15 + 31 & j113423.35 + 314606.58 & 173.597300 & 31.768494 & 14683 @xmath33 832 & 8.02 @xmath33 0.14 & 0.62 @xmath33 0.08 & 219 @xmath33 73 & 15.53 & 15.17 & 15.44 & 15.68 & 15.95 & & 1 & 0.03 & 58 + 32 & j093903.33 + 114418.62 & 144.763879 & 11.738506 & 16673 @xmath33 815 & 8.75 @xmath33 0.09 & 1.08 @xmath33 0.05 & 513 @xmath33 116 & 17.37 & 17.01 & 17.21 & 17.41 & 17.67 & & 1 & 0.03 & 82 + 33 & j070755.01 + 265102.94 & 106.979210 & 26.850817 & 17854 @xmath33 893 & 8.87 @xmath33 0.12 & 1.14 @xmath33 0.06 & 554 @xmath33 202 & & 15.53 & 15.86 & 16.01 & & & 2 & 0.07 & 39 + 34 & j104946.47 + 003635.81 & 162.443625 & 0.609947 & 19832 @xmath33 550 & 8.08 @xmath33 0.10 & 0.67 @xmath33 0.06 & 87 @xmath33 23 & 17.25 & 17.27 & 17.67 & 17.97 & 18.30 & & 1 & 0.05 & 191 + 35 & j104623.28 + 024236.57 & 161.596987 & 2.710158 & 13000 @xmath33 728 & 7.73 @xmath33 0.08 & 0.47 @xmath33 0.04 & 211 @xmath33 52 & 16.43 & 16.03 & 16.26 & 16.48 & 16.72 & & 1 & 0.04 & 92 + 36 & j104928.89 + 275423.77 & 162.370375 & 27.906603 & 14212 @xmath33 681 & 7.68 @xmath33 0.15 & 0.45 @xmath33 0.07 & 148 @xmath33 44 & 15.74 & 15.32 & 15.51 & 15.75 & 15.98 & & 1 & 0.02 & 75 + 37 & j115506.22 + 264924.59 & 178.775929 & 26.823497 & 17291 @xmath33 679 & 8.47 @xmath33 0.13 & 0.91 @xmath33 0.08 & 285 @xmath33 88 & 17.03 & 16.69 & 16.97 & 17.23 & 17.50 & & 1 & 0.02 & 97 + 38 & j094627.81 + 313211.08 & 146.615867 & 31.536411 & 15000 @xmath33 2362 & 8.34 @xmath33 0.20 & 0.82 @xmath33 0.13 & 342 @xmath33 218 & 17.30 & 16.91 & 17.12 & 17.34 & 17.55 & & 1 & 0.02 & 103 + 39&j070057.53 + 284310.06 & 105.239692 & 28.719461 & 16000 @xmath33 735 & 8.15 @xmath33 0.13 & 0.70 @xmath33 0.08 & 207 @xmath33 66 & 17.37 & 16.98 & 17.23 & 17.45 & 17.73 & & 1 & 0.08 & 120 + 40 & j040613.25 + 465133.66 & 61.555205 & 46.859349 & 33026 @xmath33 436 & 7.50 @xmath33 0.10 & 0.45 @xmath33 0.03 & 6 @xmath33 1 & & & & & & 14.77 & 3 & 0.82 & 145 + 41 & j103535.22 + 395502.27 & 158.896764 & 39.917298 & 16652 @xmath33 550 & 8.05 @xmath33 0.11 & 0.65 @xmath33 0.07 & 155 @xmath33 39 & 17.43 & 17.12 & 17.33 & 17.55 & 17.84 & & 1 & 0.01 & 154 + 42 & j105443.36 + 270658.42 & 163.680650 & 27.116228 & 24915 @xmath33 131 & 8.38 @xmath33 0.02 & 0.86 @xmath33 0.02 & 74 @xmath33 4 & 13.86 & 13.98 & 14.34 & 14.64 & 14.97 & & 1 & 0.02 & 41 + 43 & j064452.84 + 260947.75 & 101.220170 & 26.163263 & 16835 @xmath33 598 & 7.78 @xmath33 0.13 & 0.50 @xmath33 0.06 & 90 @xmath33 20 & & 15.48 & 15.98 & 16.22 & & & 2 & 0.10 & 86 + 44 & j065601.55 + 115745.85 & 104.006460 & 11.962736 & 31347 @xmath33 603 & 7.42 @xmath33 0.15 & 0.41 @xmath33 0.05 & 8 @xmath33 1 & & 14.31 & 13.48 & 13.13 & 11.94 & & 1 & 0.22 & 63 + 45 & j013914.45 + 290057.61 & 24.810197 & 29.016003 & 16808 @xmath33 478 & 8.06 @xmath33 0.10 & 0.65 @xmath33 0.06 & 153 @xmath33 34 & & 16.20 & 16.53 & 16.68 & & & 2 & 0.06 & 97 + 46 & j094126.79 + 294503.39 & 145.361630 & 29.750942 & 21798 @xmath33 267 & 8.15 @xmath33 0.04 & 0.72 @xmath33 0.03 & 68 @xmath33 10 & 15.88 & 15.91 & 16.25 & 16.55 & 16.87 & & 1 & 0.02 & 106 + 47 & j100549.01 + 424804.68 & 151.454200 & 42.801300 & 23923 @xmath33 812 & 8.11 @xmath33 0.12 & 0.70 @xmath33 0.07 & 38 @xmath33 13 & 16.04 & 16.00 & 16.39 & 16.70 & 16.98 & & 1 & 0.01 & 127 + 48 & j093047.11 + 160012.98 & 142.696300 & 16.003606 & 32492 @xmath33 634 & 8.00 @xmath33 0.13 & 0.66 @xmath33 0.07 & 7 @xmath33 1 & 16.35 & 16.64 & 17.12 & 17.50 & 17.84 & & 1 & 0.04 & 249 + 49&j093451.69 + 171814.00 & 143.715358 & 17.303889 & 14645 @xmath33 566 & 7.80 @xmath33 0.12 & 0.50 @xmath33 0.06 & 156 @xmath33 45 & 17.10 & 16.79 & 17.09 & 17.36 & 17.66 & & 1 & 0.03 & 143 + 50 & j092518.36 + 180534.20 & 141.326500 & 18.092833 & 26274 @xmath33 324 & 8.29 @xmath33 0.06 & 0.81 @xmath33 0.04 & 43 @xmath33 11 & 16.07 & 16.17 & 16.61 & 16.94 & 17.23 & & 1 & 0.05 & 127 + 51 & j052147.24 + 283532.50 & 80.446823 & 28.592361 & 18917 @xmath33 466 & 7.81 @xmath33 0.09 & 0.52 @xmath33 0.05 & 59 @xmath33 16 & & 17.50 & 17.73 & 17.86 & & & 2 & 0.62 & 108 + 52 & j071223.81 + 260933.41 & 108.099190 & 26.159281 & 14278 @xmath33 632 & 7.75 @xmath33 0.14 & 0.48 @xmath33 0.06 & 159 @xmath33 40 & & 16.85 & 17.24 & 17.41 & & & 2 & 0.08 & 141 + 53 & j102521.36 + 455553.91 & 156.338987 & 45.931643 & 23547 @xmath33 908 & 7.51 @xmath33 0.13 & 0.41 @xmath33 0.05 & 19 @xmath33 3 & 18.12 & 18.27 & 18.60 & 18.89 & 19.26 & & 1 & 0.02 & 530 + 54 & j101806.60 + 455830.36 & 154.527482 & 45.975101 & 22475 @xmath33 1541 & 8.36 @xmath33 0.22 & 0.85 @xmath33 0.14 & 101 @xmath33 59 & 17.61 & 17.55 & 17.90 & 18.19 & 18.47 & & 1 & 0.01 & 201 + 55 & j033149.69 + 305944.92 & 52.957023 & 30.995811 & 19435 @xmath33 332 & 8.64 @xmath33 0.06 & 1.02 @xmath33 0.03 & 277 @xmath33 37 & & 16.89 & 17.41 & 17.67 & & & 2 & 1.18 & 25 + 56 & j033253.91 + 284006.91 & 53.224625 & 28.668586 & 19000 @xmath33 1400 & 9.84 @xmath33 0.09 & 1.20 @xmath33 0.15 & 155 @xmath33 0 & & 17.06 & 17.41 & 17.58 & & & 2 & 0.25 & 27 + 57 & j090918.99 + 292929.61 & 137.329125 & 29.491558 & 22588 @xmath33 346 & 8.04 @xmath33 0.05 & 0.65 @xmath33 0.03 & 43 @xmath33 8 & 15.74 & 15.68 & 16.04 & 16.34 & 16.59 & & 1 & 0.02 & 107 + 58 & j102155.50 + 405014.85 & 155.481261 & 40.837458 & 23364 @xmath33 999 & 7.96 @xmath33 0.15 & 0.61 @xmath33 0.09 & 28 @xmath33 18 & 16.19 & 16.25 & 16.61 & 16.91 & 17.23 & & 1 & 0.01 & 153 + 59&j121336.54 + 314808.77 & 183.402250 & 31.802436 & 13308 @xmath33 405 & 8.26 @xmath33 0.08 & 0.77 @xmath33 0.05 & 418 @xmath33 67 & 16.21 & 15.79 & 16.00 & 16.21 & 16.49 & & 1 & 0.01 & 60 + 60 & j134922.51 - 003503.15 & 207.343783 & -0.584208 & 16401 @xmath33 1151 & 8.52 @xmath33 0.19 & 0.94 @xmath33 0.12 & 363 @xmath33 153 & 17.24 & 16.91 & 17.25 & 17.46 & 17.73 & & 1 & 0.03 & 99 + 61 & j144433.83 - 005958.83 & 221.140967 & -0.999675 & 12165 @xmath33 856 & 8.01 @xmath33 0.21 & 0.61 @xmath33 0.13 & 367 @xmath33 154 & 16.58 & 16.20 & 16.38 & 16.58 & 16.85 & & 1 & 0.04 & 77 + 62 & j112518.85 + 541936.65 & 171.328550 & 54.326847 & 15272 @xmath33 209 & 7.85 @xmath33 0.04 & 0.53 @xmath33 0.02 & 147 @xmath33 14 & 15.62 & 15.28 & 15.57 & 15.83 & 16.08 & & 1 & 0.01 & 73 + 63 & j113203.47 + 065509.52 & 173.014441 & 6.919311 & 12455 @xmath33 1141 & 7.93 @xmath33 0.45 & 0.57 @xmath33 0.25 & 310 @xmath33 201 & 15.21 & 14.89 & 15.13 & 15.35 & 15.64 & & 1 & 0.04 & 46 + 64 & j084107.69 + 163221.71 & 130.282053 & 16.539363 & 16626 @xmath33 580 & 8.30 @xmath33 0.11 & 0.80 @xmath33 0.07 & 237
@xmath33 64 & 17.52 & 17.22 & 17.45 & 17.67 & 17.97 & & 1 & 0.02 & 134 + 65 & j150156.26 + 302300.13 & 225.484400 & 30.383369 & 27051 @xmath33 339 & 7.84 @xmath33 0.05 & 0.56 @xmath33 0.03 & 12 @xmath33 1 & 14.13 & 14.24 & 14.71 & 14.96 & 15.32 & & 1 & 0.02 & 77 + 66 & j063406.26 + 025401.30 & 98.526074 & 2.900361 & 31607 @xmath33 1274 & 7.82 @xmath33 0.31 & 0.56 @xmath33 0.14 & 7 @xmath33 1 & & & & & & & 4 & 1.60 & + 67 & j064438.16 + 030704.39 & 101.159020 & 3.117885 & 12067 @xmath33 2288 & 8.37 @xmath33 0.65 & 0.84 @xmath33 0.42 & 650 @xmath33 791 & & 14.10 & 13.66 & 13.51 & & & 2 & 1.02 & 5 + 68 & j063517.47 + 054917.94 & 98.822796 & 5.821650 & 22885 .
@xmath33 3739 & 7.48 @xmath33 0.54 & 0.40 @xmath33 0.14 & 21 @xmath33 13 & & 14.19 & 13.96 & 13.84 & & & 2 & 0.41 & 38 + 69&j152130.83 - 003055.70 & 230.378443 & -0.515472 & 13000 @xmath33 1056 & 7.63 @xmath33 0.10 & 0.42 @xmath33 0.05 & 186 .
@xmath33 71 .
& 15.52 & 15.24 & 15.53 & 15.78 & 16.09 & & 1 & 0.07 & 67 + 70 & j113705.14 + 294757.77 & 174.271408 & 29.799381 & 23829 @xmath33 127 & 8.49 @xmath33 0.02 & 0.93 @xmath33 0.02 & 111 @xmath33 6 & 13.54 & 12.45 & 12.88 & 13.16 & 13.69 & & 1 & 0.02 & 18 + 71 & j191927.67 + 395839.30 & 289.865292 & 39.977583 & 20376 @xmath33 345 & 7.93 @xmath33 0.06 & 0.59 @xmath33 0.03 & 54 @xmath33 8 & & & & & & & 4 & 0.14 & + | we present a spectroscopically identified catalogue of 72 da white dwarfs from the lamost pilot survey .
35 are found to be new identifications after cross - correlation with the eisenstein et al . and
villanova catalogues . the effective temperature and gravity of these white dwarfs are estimated by balmer lines fitting .
most of them are hot white dwarfs . the cooling times and masses of these white dwarfs
are estimated by interpolation in theoretical evolution tracks .
the peak of mass distribution is found to be @xmath0 0.6 @xmath1 which is consistent with prior work in the literature .
the distances of these white dwarfs are estimated using the method of synthetic spectral distances .
all of these wds are found to be in the galactic disk from our analysis of space motions .
our sample supports the expectation white dwarfs with high mass are concentrated near the plane of galactic disk . |
the papers in this series are devoted to a study of the period - colour ( pc ) and amplitude - colour ( ac ) relations for classical cepheid variables , the physics behind the connections of pc and ac relations , and the implication for cepheid period - luminosity ( pl ) relations .
the connection between the pc and ac relations was derived by @xcite , who applied the stefan - boltzmann law at the optical maximum and minimum light together with the fact that radial variations are small in the optical @xcite . specifically : @xmath1 where @xmath2 and @xmath3 are the effective temperature at the maximum and minimum light , respectively .
if @xmath2 , and hence the colour at maximum light , is independent of period , then equation ( 1 ) implies there is a relation between the temperature at minimum light and the optical amplitudes , and vice versa .
@xcite found that at the maximum light , the spectral type of galactic cepheids is independent of pulsating period .
convincing observational support for a flat ( i.e. zero slope ) maximum light pc relation has been documented by @xcite , ( * ? ? ?
* hereafter paper i ) , ( * ? ? ?
* hereafter paper ii ) and ( * ? ? ?
* hereafter paper iii ) for the galactic ( with @xmath0>0.8 $ ] , where @xmath4 is pulsation period in days ) and lmc ( with @xmath0>1.0 $ ] ) cepheids .
these authors imply that the flatness of the cepheid pc relation at maximum is due to the interaction between the hydrogen ionization front ( hif ) and the photosphere ( defined as optical depth @xmath5 ; see paper ii , iii and * ? ? ?
however , paper i found that the smc pc relation , in @xmath6 colour , at maximum light is not flat , but instead suffers a `` flattening '' of the slope from @xmath7 to @xmath8 for short and long period ( @xmath0 > 1.0 $ ] ) cepheids , respectively . this change of slope is marginally significant using the @xmath9-test described in paper i. one motivation for this paper is to investigate the physics behind pc relations at maximum light and in particular why galactic and lmc cepheids display flat maximum light pc relations and smc cepheids do not .
since the pc and pl relations are really projections of the period - luminosity - colour ( plc ) relation , pc and pl relations are closely connected : changes in one are generally reflected in changes in the other .
hence one way to look for changes in cepheid pl relations is to look for changes in pc and ac relations .
another important driving force behind this series of papers is to investigate the possible physics behind recent results suggesting that the lmc ( large magellanic cloud ) pl ( and pc ) relation is non - linear : in the sense that current lmc cepheid data are statistically consistent with two pl ( and pc ) relations of significantly differing slope with a break at / around a period of 10 days . the observational evidence for this is presented in @xcite , paper i & iii , @xcite , @xcite and ( * ? ? ?
* hereafter paper iv ) .
it is also interesting to note , as has been pointed out in previous work , that the `` break '' period of @xmath10 is also the location of the well known cepheid pulsation resonance .
paper iv use the ogle ( optical gravitational lensing experiment , * ? ? ?
* ) lmc data and phase all cepheids to a common epoch and then plot both pc and pl relations as a function of phase .
this approach presents clear and compelling evidence of a break at a phase around 0.8 and also attest to the dynamic nature of the pl relation .
the pc relation follows changes in the pl relation very closely .
however the standard pl relation used in the literature is at mean light : the average over phase of the multiphase relations presented in paper iv .
because the pl relation is indeed linear at certain phases ( e.g. , at maximum light ) , the net result is that the strength of the non - linearity is diluted at mean light .
we emphasize , though , that a battery of parametric and non - parametric statistical tests carried out on the lmc pl relation at mean light in optical bands confirm the existence of the non - linearity at high confidence levels - greater than @xmath11 . these tests include the @xmath9-test , robust methods like tukey s bi - weight function and non - parametric techniques ( paper i ; * ? ? ?
in contrast the galactic and small magellanic cloud ( smc ) pl and pc relations are found to be linear with current data @xcite .
since the galaxy , lmc and smc have different metallicity , it has been conjectured that metallicity is the key to understanding this behavior .
we now briefly discuss some criticisms of results implying a non - linear lmc pl relation .
one of these is that the ogle lmc sample lacks sufficient long period cepheids . despite the fact that the @xmath9-test takes into account both the number and nature of the observations , @xcite augmented the ogle sample with data taken from , for example , @xcite and @xcite .
they conclude on the basis of their rigorous statistical test ( the @xmath9-test ) that even with this augmented lmc cepheid data , the lmc pl relation is still found to be non - linear in the sense described above ( even though the difference in the slopes of the long and short period pl relations may be small ) .
another criticism is that extinction errors are influencing the result even though only published extinction values are used .
well established observations and theory have shown that the galactic and lmc cepheid pc relations are flat at maximum light , at least for the longer period ( @xmath0 > 1 $ ] ) cepheids ( see paper i , ii , iii and * ? ? ? * ) .
if the extinction values required to make the mean light lmc pc and pl relations linear are adopted , then the lmc pc relation at maximum light is no longer flat . in this scenario ,
lmc cepheids get hotter at maximum light as the period increases , which is in distinct disagreement with the theory and observations at maximum light @xcite .
the referee has pointed out that if the amplitude increases sufficiently fast with period , then it could produce hotter temperatures at maximum light with increasing period . but note that the amplitude increases fairly rapidly after @xmath12 days and this is precisely when the observed pc relation is flattest .
further if the lmc cepheids are such that they got hotter at maximum light as the period increases this is different to galactic cepheids at maximum light .
we therefore believe that extinction errors are not the cause of the observed non - linear lmc pl relation .
more detailed discussions regarding extinction errors are given in @xcite , @xcite and @xcite , and will not be repeated here .
since the cepheid pl relation is a cornerstone of modern astrophysics , a proper understanding of its properties and the physics behind it is crucial .
for example , it is of fundamental importance in establishing an extra - galactic distance scale that is independent of cmb studies .
an appropriate refinement of the cepheid pl distance scale , can provide a value of hubble s constant accurate to less than 3 - 4% .
though the current cepheid based estimate of @xmath13 is accurate to @xmath14 @xcite , a more accurate ( @xmath15 ) direct measurement of @xmath13 via a cepheid distance scale is still very relevant because it will help to break the degeneracy between @xmath16 and @xmath13 present from @xmath17 cmb estimates ( see , for example , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
cmb measurements can only estimate @xmath18 and estimate @xmath13 only if a simple flat @xmath19cdm model is assumed .
table 2 of @xcite points to the fact that an independent estimate of @xmath13 accurate to @xmath20 will result in a reduction of the @xmath21 confidence interval on @xmath16 by almost a factor of two over that with @xmath17 alone .
this is a greater reduction than if other data such as 2dfgrs are used ( table 5 of * ? ? ?
* ) . with the possibility of @xmath22 and @xmath23 satellites producing more accurate parallaxes for many cepheids a considerably more accurate cepheid
zero point is well within reach .
better refinement of the slope of pl relation will become vital .
equally as important , modeling the cepheid pl relation can yield important information about the mass - luminosity ( ml ) relation obeyed by cepheids . since these ml relations are sensitive functions of input physics , such as the amount of convective overshoot , such modeling is also important for theories of stellar evolution and pulsation . a working theoretical hypothesis to explain the non - linear lmc pl and pc relation is the interaction of the hif with the stellar photosphere at certain phases .
paper ii & iii published theoretical pulsation models of galactic and lmc cepheids , respectively , to model this behavior . in these papers we found that the interaction of the hif and stellar photosphere may play an important role in explaining the observed pc and pl behavior .
we briefly summarize this theory in what follows ( see paper ii & iii for more details ) . in a pulsating cepheid ,
the hif and stellar photosphere independently move in and out in the mass distribution as the star pulsates . in certain circumstances ,
the stellar photosphere can be located at the base of the hif . in this case , when the the density is low the temperature of the stellar photosphere and hence the effective temperature of the cepheid at that phase experiences a much reduced dependence on pulsation period .
thus the pc relation is flatter in this situation .
because the pl relation is intimately connected to changes in the pc relation , the pl relation slope will also change . because this hif - photosphere engagement happens suddenly ,
the changes in the pc and pl relation are sharp . the phase and period at which such a low density hif - photosphere interaction can occur depends on the ml relation since this controls the location of the hif in the mass distribution @xcite .
the connection to metallicity occurs because the ml relation , taken from stellar evolution calculations , is indeed a function of metallicity . in this paper
we extend the work of paper ii & iii by constructing a large grid of full amplitude hydrodynamic models of cepheids with an smc metallicity and study how the interaction of the photosphere and hif affects the pc ( and hence the pl relation ) in these models compared to previous work looking at galactic ( paper ii ) and lmc ( paper iii ) models .
section 2 outlines the methodology and pulsation codes we used in this paper and section 3 presents our results for the smc models .
the conclusion and discussion is given in section 4 .
ccccccc @xmath24 & @xmath25 & @xmath26 & @xmath27 & @xmath28 & @xmath29 & @xmath30 + + 12.0 & 4.612 & 5440 & 52.695 & 0.247 & 32.94 & @xmath31 + 10.8 & 4.458 & 5350 & 44.130 & 0.249 & 27.97 & @xmath32 + 9.50 & 4.272 & 5230 & 35.981 & 0.175 & 23.01 & @xmath33 + 8.70 & 4.144 & 5180 & 30.519 & 0.110 & 19.68 & @xmath34 + 8.00 & 4.022 & 5320 & 22.562 & 0.106 & 15.09 & @xmath35 + 7.10 & 3.848 & 5340 & 16.965 & 0.065 & 11.53 & @xmath36 + 6.50 & 3.720 & 5410 & 13.219 & 0.050 & 9.129 & @xmath37 + 5.90 & 3.579 & 5450 & 10.363 & 0.033 & 7.226 & @xmath38 + 5.80 & 3.554 & 5470 & 9.8372 & 0.032 & 6.879 & @xmath39 + 5.50 & 3.476 & 5500 & 8.5701 & 0.026 & 6.022 & @xmath39 + 5.20 & 3.395 & 5530 & 7.4236 & 0.020 & 5.238 & @xmath39 + 5.10 & 3.367 & 5530 & 7.1171 & 0.018 & 5.023 & @xmath40 + 5.00 & 3.338 & 5550 & 6.7222 & 0.017 & 4.755 & @xmath37 + 4.80 & 3.278 & 5570 & 6.0683 & 0.014 & 4.301 & @xmath40 + 4.60 & 3.217 & 5600 & 5.4218 & 0.012 & 3.854 & @xmath37 + + 8.50 & 4.504 & 5290 & 62.455 & 0.504 & 35.31 & @xmath41 + 7.80 & 4.384 & 5310 & 50.525 & 0.441 & 29.57 & @xmath42 + 7.00 & 4.232 & 5330 & 39.008 & 0.368 & 23.67 & @xmath31 + 6.50 & 4.129 & 5370 & 32.048 & 0.316 & 19.96 & @xmath34 + 5.70 & 3.945 & 5410 & 23.308 & 0.242 & 15.02 & @xmath43 + 4.70 & 3.675 & 5430 & 15.163 & 0.144 & 10.09 & @xmath44 + 4.30 & 3.551 & 5470 & 12.177 & 0.114 & 8.234 & @xmath40 + 3.95 & 3.432 & 5490 & 10.029 & 0.082 & 6.852 & @xmath35 + 3.80 & 3.378 & 5520 & 9.0437 & 0.074 & 6.221 & @xmath40 + 3.60 & 3.302 & 5550 & 7.9025 & 0.062 & 5.475 & @xmath37 + 3.45 & 3.243 & 5550 & 7.2318 & 0.050 & 5.022 & @xmath35 + 3.40 & 3.222 & 5570 & 6.9154 & 0.049 & 4.817 & @xmath38 + 3.30 & 3.181 & 5590 & 6.4078 & 0.045 & 4.479 & @xmath38 + 3.20 & 3.138 & 5610 & 5.9276 & 0.040 & 4.158 & @xmath40 + 3.00 & 3.047 & 5640 & 5.0804 & 0.031 & 3.581 & @xmath45 + ccccc @xmath4 & @xmath46 & @xmath2 & @xmath47 & @xmath3 + + 52.695 & 53454.42 & 5988.56 & 27021.41 & 5063.31 + 44.130 & 35895.77 & 5826.30 & 19377.45 & 5022.77 + 35.981 & 21806.85 & 5480.18 & 14111.82 & 5031.24 + 30.519 & 15662.90 & 5380.53 & 11737.37 & 5095.86 + 22.562 & 11969.05 & 5626.25 & 8755.998 & 5063.77 + 16.965 & 7880.140 & 5581.23 & 6253.112 & 5159.75 + 13.219 & 5731.252 & 5560.19 & 4789.290 & 5260.59 + 10.363 & 3934.212 & 5488.40 & 3546.536 & 5359.55 + 9.837 & 3699.569 & 5495.11 & 3336.452 & 5378.74 + 8.5701 & 3093.723 & 5609.36 & 2808.320 & 5421.22 + 7.4236 & 2555.649 & 5607.43 & 2357.080 & 5471.61 + 7.1171 & 2389.147 & 5605.87 & 2225.435 & 5481.90 + 6.7222 & 2234.535 & 5625.98 & 2089.173 & 5507.61 + 6.0683 & 1943.932 & 5642.11 & 1839.759 & 5538.16 + 5.4218 & 1681.461 & 5664.88 & 1604.783 & 5574.01 + + 62.455 & 37223.13 & 5497.13 & 20783.42 & 4867.51 + 50.525 & 28606.35 & 5590.89 & 15062.32 & 4869.70 + 39.008 & 20384.64 & 5630.58 & 10622.78 & 4916.65 + 32.048 & 16268.92 & 5710.44 & 8633.998 & 5001.55 + 23.308 & 10611.59 & 5727.19 & 6356.127 & 5179.16 + 15.163 & 5419.509 & 5562.85 & 3864.655 & 5147.97 + 12.177 & 4024.093 & 5771.01 & 2991.331 & 5226.61 + 10.029 & 3020.997 & 5703.97 & 2380.306 & 5291.96 + 9.0437 & 2653.422 & 5727.42 & 2129.332 & 5341.58 + 7.9025 & 2192.885 & 5700.61 & 1826.794 & 5404.38 + 7.2318 & 1872.266 & 5655.25 & 1622.456 & 5433.35 + 6.9154 & 1774.033 & 5653.59 & 1550.829 & 5465.08 + 6.4078 & 1592.429 & 5645.12 & 1409.870 &
5495.35 + 5.9276 & 1430.522 & 5645.05 & 1274.321 & 5516.67 + 5.0804 & 1148.483 & 5757.16 & 1041.914 & 5559.16 +
the codes for computational and numerical methods for doing the pulsation modelings are described in @xcite and @xcite , and are exactly the same as used in paper ii & iii . in brief , the input parameters to the pulsation codes include the mass ( @xmath24 ) , luminosity ( @xmath48 ) , effective temperature ( @xmath49 ) and chemical composition ( @xmath50 ) . in this paper , the chemical composition
is set to be @xmath51 to represent the smc hydrogen and metallicity abundance ( by mass ) . the input effective temperatures are chosen to ensure the models oscillate in the fundamental mode and located within the cepheid instability strip . to be consistent with paper ii & iii ,
we adopt the same ml relations as given below : 1 .
ml relation given in @xcite : @xmath52 2 .
ml relation given in @xcite for @xmath53 : @xmath54 the units for both @xmath24 and @xmath48 are in solar units . to convert the temperatures from the models to the @xmath6 colours , we use the basel atmosphere database @xcite to construct a fit giving temperature and effective gravity
( see paper ii ) as a function of @xmath6 colour .
the bolometric corrections ( @xmath55 ) are converted in a similar manner .
note that the effect of a micro - turbulence parameter that varies with phase has been discussed in paper iii : its variation with phase does not influence the results presented in this paper . to convert the observed colours to the temperatures appropriate for the smc data
, we use the prescriptions given in @xcite : @xmath56 more details of the methodology , the pulsation codes and the temperature - colour conversions can be found in paper ii & iii , and will not be repeated in detail here .
cccccccc @xmath4 & @xmath57 & @xmath58 & @xmath59 & @xmath60 & @xmath61 & @xmath62 ( asc ) & @xmath62 ( des ) + + 52.695 & 40955.18 & 41381.50 & 5794.90 & 40910.47 & 5191.18 & 5779.38 & 5192.37 + 44.130 & 29325.71 & 29139.10 & 5704.73 & 29292.77 & 5107.24 & 5713.09 & 5108.67 + 35.981 & 19241.07 & 19236.72 & 5557.14 & 19209.26 & 5025.52 & 5557.48 & 5027.67 + 30.519 & 13945.37 & 13879.68 & 5404.28 & 13960.74 & 4992.63 & 5411.22 & 4991.33 + 22.562 & 10490.84 & 10518.15 & 5535.73 & 10507.93 & 5138.62 & 5532.01 & 5136.74 + 16.965 & 7033.900 & 7014.821 & 5486.05 & 7035.167 & 5207.44 & 5489.46 & 5207.20 + 13.219 & 5237.352 & 5240.488 & 5502.74 & 5232.986 & 5324.74 & 5502.19 & 5326.06 + 10.363 & 3787.302 & 3787.677 & 5513.59 & 3791.380 & 5417.43 & 5513.40 & 5415.87 + 9.8372 & 3576.717 & 3579.774 & 5541.08 & 3572.195 & 5431.94 & 5539.47 & 5433.71 + 8.5701 & 2993.448 & 2996.177 & 5570.00 & 2991.997 & 5465.07 & 5568.28 & 5465.71 + 7.4236 & 2480.871 & 2483.940 & 5593.50 & 2482.250 & 5501.82 & 5591.18 & 5501.01 + 7.1171 & 2324.563 & 2325.461 & 5586.57 & 2324.964 & 5506.19 & 5585.84 & 5505.89 + 6.7222 & 2175.275 & 2175.738 & 5605.57 & 2175.519 & 5526.88 & 5605.17 & 5526.69 + 6.0683 & 1897.128 & 1899.297 & 5620.35 & 1897.321 & 5553.28 & 5618.19 & 5553.09 + 5.4218 & 1644.994 & 1643.720 & 5641.55 & 1644.522 & 5587.32 & 5642.97 & 5587.89 + + 62.455 & 31923.84 & 31934.40 & 5622.99 & 31788.30 & 5080.26 & 5622.53 & 5084.43 + 50.525 & 24193.87 & 24220.00 & 5657.21 & 24125.18 & 5092.35 & 5655.55 & 5094.78 + 39.008 & 17315.32 & 17452.68 & 5718.16 & 17393.96 & 5111.54 & 5706.24 & 5107.63 + 32.048 & 13480.61 & 13231.90 & 5702.29 & 13467.68 & 5118.54 & 5729.31 & 5119.08 + 23.308 & 9065.182 & 9050.282 & 5778.65 & 9095.283 & 5184.01 & 5780.89 & 5179.84 + 15.163 & 4886.975 & 4876.629 & 5719.32 & 4886.911 & 5268.85 & 5722.14 & 5268.87 + 12.177 & 3545.551 & 3517.459 & 5664.05 & 3535.748 & 5288.78 & 5674.57 & 5291.86 + 10.029 & 2700.579 & 2688.289 & 5642.62 & 2708.607 & 5350.09 & 5648.19 & 5346.47 + 9.0437 & 2383.037 & 2371.188 & 5644.71 & 2377.980 & 5391.28 & 5650.44 & 5394.08 + 7.9025 & 2003.229 & 1999.609 & 5639.31 & 2002.772 & 5459.24 & 5641.05 & 5459.58 + 7.2318 & 1746.108 & 1747.221 & 5620.58 & 1745.418 & 5640.07 & 5620.23 & 5640.65 + 6.9154 & 1666.125 & 1666.384 & 5630.74 & 1666.937 & 5651.07 & 5630.80 & 5649.58 + 6.4078 & 1514.393 & 1513.263 & 5663.32 & 1516.365 & 5549.10 & 5664.74 & 5547.18 + 5.9276 & 1371.416 & 1375.426 & 5695.05 & 1373.677 & 5570.67 & 5689.58 & 5568.37 + 5.0804 & 1113.786 & 1113.660 & 5717.14 & 1115.086 & 5603.50 & 5717.35 & 5602.00 + the results for our smc models are collectively summarized in table [ tabinput]-[tabmean ] .
table [ tabinput ] presents the input mass ( and hence the luminosity ) and effective temperature for our smc models and the resulting periods from linear non - adiabatic calculations .
table [ tabinput ] displays the fact that the models constructed in this study occupied a wide range of periods with an unstable and stable fundamental and first overtone mode respectively .
thus we avoid regions of the instability strip which are susceptible to first overtone or double mode pulsation .
table [ tabmaxmin ] gives the temperatures at the maximum and minimum luminosity for the models calculated in table [ tabinput ] .
table [ tabmean ] , with identical layout as in table 3 of paper ii , presents more detailed information regarding photospheric temperatures at mean luminosities during the ascending and descending branches of the light ( or luminosity ) curves ( see paper ii & iii for more details ) . [ cols="<,^,^,^,^,^,^,^,^,^ " , ] the results from the smc models were compared to the observations , also presented in figure [ pt]-[ac ] . in these figures , fourier fits to the smc cepheid data , taken from paper iv , were used to calculate the observed amplitudes and colours at maximum , mean , and minimum light using the definitions given in paper i & iii .
table [ tabpc ] & [ tabac ] summarizes the empirical pc and ac relations at maximum , mean and minimum light for the smc cepheid data plotted in these figures , respectively . in figure
[ pt ] , we compare the temperatures given in table [ tabmaxmin ] & [ tabmean ] to the observed smc cepheid data . similarly ,
figure [ pc ] displays the pc relations at various phases for the smc models superimposed on the smc data . from these figures
, we found that the smc pc relation at minimum light has the smallest scatter for all periods and does not suffer a statistically significant change of slope at @xmath10 , i.e. it is linear over the entire period range .
the pc relation at maximum light is significantly flatter for long ( @xmath0 > 1.0 $ ] ) period than short period cepheids and displays marginal evidence of a slope change at @xmath10 .
it also has the largest scatter over all phases .
the smc models do a reasonable job in matching the observations in the period - temperature and the period - colour plane , though the short period models at maximum light tend to be too cool compared to the observations ( as in figure [ pt ] & [ pc ] ) .
figure [ ac ] portrays the results of ac relations from the models as compared to the smc data .
several models , especially those using the @xcite ml relation , display a smaller amplitude as compared to the data .
this is seen in the galactic ( paper ii ) and lmc ( paper iii ) models as well , however we believe this does not affect our results on the hif - photosphere interactions ( paper iii ) . to the best of our knowledge
a comparison of models and theory on pc / ac planes at different phases , as described here , has not been carried out in such detail before : further work will investigate in detail some of the discrepancies between models and observations described in the text . from the full amplitude smc models , the temperature and the opacity profile
can be plotted at a given phase of pulsation .
similar to paper ii & iii , the locations of the hif ( sharp rise in the temperature profile ) and photosphere can be identified from the temperature profile .
figure [ profile ] displays the temperature and opacity profiles , with the locations of photosphere marked as filled circles , for a long period , a 10-days and a short period smc model . from the temperature profile ,
the `` distance '' @xmath63 between the hif and the photosphere from the temperature profile can be calculated ( see paper ii for the definition of @xmath63 ) .
a small @xmath63 implies there is a hif - photosphere interaction , and vice versa .
figure [ delta ] presents the @xmath63 as a function of pulsating period for the smc models with the two ml relations used in this paper . in papers ii & iii
, it is found that the distribution of @xmath63 as a function of period is almost independent of the adopted ml relations .
this is also seen in the smc models as depicted in figure [ delta ] .
furthermore , we see , in contrast to the behavior for galactic and lmc models , the distance is reasonably constant for all periods and at all phases .
figure 15 of paper iii shows that at minimum light this distance for the galactic models increases with period , but for the lmc models this distance only increases for periods greater than 10 days . at maximum light
, all models display a constant distance .
the smc models are different in that there is almost a constant distance between the hif and stellar photosphere , independent of pulsation period , pulsation phase and ml relation .
figure [ density ] depicts the photospheric density as a function of period for the smc models .
this figure needs to be compared with figure 16 of paper iii .
the difference between the smc models and the galactic / lmc models is in the photospheric density at minimum light .
the galactic models have a photospheric density at minimum light which is less than @xmath64 .
paper ii also found that for these galactic models , the photosphere and hif are not engaged . for the lmc models at minimum light
, paper iii found that the hif and photosphere are nearly always engaged with some indication that there is a separation for periods greater than 10 days .
we also see that the photospheric density for these longer period lmc models is less than @xmath64 .
the smc models always have an engaged photosphere and hif at all phases . but crucially , the density of the photosphere is higher than @xmath64 for models with @xmath65 .
the higher densities mean that no sharp change in the temperature of the photosphere independent of the global stellar parameters can occur because our premise is that a sharp change can occur in the slope of the pc relation when the photosphere and hif are engaged and this occurs at low densities .
thus the pc relation for smc cepheids should show a much reduced change in the pc slope in going from short to long period cepheids and this is what we see in the observations : at maximum light , galactic and lmc cepheids display a pc(max ) relation with a slope that decreases to zero in going from short to long period ( paper i & iii ) but smc cepheids only show a reduced pc slope at maximum light .
it may be possible to test this assertion spectroscopically and this will be the subject of a future investigation . because the pc relation at mean light is an average of the pc relation at different phases , if there is a discontinuity or not at a certain phase , this effect will be transmitted to the mean light pc relation and hence to the mean light pl relation .
in this paper we construct the full amplitude models appropriate for the smc cepheids to study the pc / ac relations and the hif - photosphere interaction . in short ,
the smc models constructed in this paper do a reasonable job of matching the observed pc relations at minimum and mean light .
the greatest discrepancies occur for short period cepheids at maximum light .
these smc models tend to not be driven to higher temperatures and hence bluer @xmath6 colours , though they do fall inside the outer envelope of the observed data points . lllll galaxy & photospheric density & hif - photosphere interaction & observed pc relation & observed ac relation + + gal & low for @xmath66 & yes for all periods & flat for @xmath67 with & higher amp
. @xmath68 hotter t. & + & & & significant break at @xmath69 & gets shallower for @xmath70 + + lmc & low for @xmath71 & yes for all periods & flat for @xmath72 with & higher amp . @xmath68
hotter t. & + & high for @xmath73 & & significant break at @xmath69 & gets shallower for @xmath70 + + smc & high for @xmath71 & yes for all periods & @xmath74flatter for @xmath72 with & higher amp . @xmath68 hotter t. & + & high for @xmath73 & & marginal break at @xmath69 & gets shallower for @xmath70 + + + gal & low for @xmath66 & no for @xmath66 & non - flat slope for all periods & higher amp . @xmath68
cooler t. & + & & & & no significant break & gets shallower for @xmath70 + + lmc & low for @xmath71 & no for @xmath71 & non - flat slope for all periods , & flat for @xmath75 & + & high for @xmath73 & yes for @xmath73 & significant break at @xmath69 & gets steeper for @xmath70
+ + smc & low for @xmath76 & yes for all periods & non - flat slope for all periods & @xmath74flat for all periods + & high for @xmath77 & & & no significant break & + from the previous papers in the series we note that the empirical lmc pc relation at minimum light has a statistically significant slope change or `` break '' at @xmath10 , but has the smallest scatter over all phases .
the lmc pc relation at maximum light has the largest scatter over all phases .
it also shows a statistically significant change of slope across @xmath10 but has a flat slope ( i.e. close to zero ) when only the long period cepheids are considered .
the empirical galactic pc relation displays a break at 10 days with a flat relation thereafter at the maximum light and a linear relation at minimum light .
the empirical smc pc relation also display a marginal break at maximum light but is linear at minimum light .
the pc relation is flatter for the smc cepheids with @xmath78 than the short period smc cepheids at the maximum light . in terms of the empirical ac relation , the ac relations at maximum / minimum light
are significantly broken ( at @xmath79=1.0 $ ] ) for all three galaxies ( see paper i & iii ) . at maximum light , higher amplitude cepheids in all three galaxies are driven to _
hotter _ temperature and hence bluer colours for all period ranges ( with negative slope in the ac relation ) .
nevertheless the slope of these empirical ac relations at maximum light becomes shallower for the long period cepheids .
however , the opposite behavior is found at the minimum light as the long period cepheids are driven to _
cooler _ temperature as the amplitude increases in all three galaxies . for short period cepheids
, the same behavior is found for the galactic cepheids but the lmc and smc short period ac relations are nearly flat at minimum light . in all cases
the behavior at mean light is intermediate between the properties at maximum / minimum light .
the saha ionization equation , which is used in most models of cepheid envelopes , is such that when the photosphere and hif interact at low densities , the photospheric temperature will be almost independent of pulsation period .
this is because when the hif and photosphere are interacting or engaged , the photospheric temperature is essentially the same as the temperature at which hydrogen ionizes , and not dependent on the pulsation period .
hence the colour , either in @xmath6 or in @xmath80 , is also almost independent of the pulsation period .
moreover , as the period increases , the @xmath81 ratio increases which , in turn , changes the location of the hif , in terms of the mass distribution , in the cepheid envelope . this can change the phase at which the hif and photosphere interact .
one possible difference between galactic , lmc and smc models is the photospheric density at minimum light . for galactic models
, the @xmath81 ratio causes the low density photosphere - hif interaction to only occur at / around the maximum light , leading to a stronger relation between period and colour at minimum light . at other phases ,
the hif and photosphere are well separated .
for lmc models , the @xmath81 ratio forces the hif - photosphere interaction to occur at all phases for short period cepheids at higher densities and at maximum light for longer period cepheids at low densities . at phases close to the minimum light ,
the hif - photosphere become disengaged for longer period cepheids .
this leads to change in the slope of pc relation at phases around minimum light for the long and short period lmc cepheids . for smc models ,
the hif - photosphere interaction occurs throughout the pulsation cycle , but only occurs at low density at phases close to , and including , the maximum light for the long period models .
this leads longer period smc cepheids to have a flatter pc relation slope than the short period smc cepheids at the maximum light .
the high density photosphere - hif interaction occurs at other phases and ensures that there is no significant change of slope in the pc relation as a function of period .
the low density hif - photosphere interaction at maximum light is the reason for the observed flat pc relation at maximum light for the galactic and lmc cepheids and the flatter pc(max ) relation for longer period smc cepheids . in table
[ tabsum ] , we summarize some of the properties discovered in this series of papers for the galactic , lmc and smc cepheids and models . because these changes depend on the @xmath81 ratio and because the @xmath82 relation is dependent on metallicity , the above argument outlines a mechanism whereby the observed changes in the pc relation between galactic , lmc and smc cepheids could occur .
moreover these changes do not have to be monotonic with metallicity .
since the mean light pc relation is the average of the pc relation at different phases , changes in the slope of the pc relation , can , in principle , affect the mean light pc relation ( see paper i to iv ) .
in addition to our interest in understanding the physics behind the multi - phase pc and pl relations as a function of metallicity ( in galaxy , lmc and smc ) , we are also interested in the changes of light curves shape in different galaxies .
this is because the ac relations are essentially equivalent to plotting the light curve shape against temperature ( or colour ) .
previous work has established that principal component analysis ( pca ) is a very efficient way of describing light curve structure with upwards of @xmath83 of structure variation being explained by just the first principal component @xcite .
since the first pca coefficient correlates well with amplitude , therefore there will be a correlation between the amplitude and light curve structure .
we remark in passing that @xcite mention a correlation between amplitude and effective temperature in their hydrodynamic models of cepheids .
it can be seen from this work and previous papers in this series that a correlation between amplitude and effective temperature is due to a more extreme correlation at maximum or minimum light : the physics of which has been explained in this paper and earlier papers in the series .
we thank an anonymous referee for useful suggestions .
we also thank t. barnes for helpful discussion .
smk acknowledges the support from hst - ar-10673.04-a and chretien international research award .
cn acknowledges support from nsf award opp-0130612 and a university of illinois seed funding award to the dark energy survey . | period - colour ( pc ) and amplitude - colour ( ac ) relations at maximum , mean and minimum light are constructed from a large grid of full amplitude hydrodynamic models of cepheids with a composition appropriate for the smc ( small magellanic cloud ) .
we compare these theoretical relations with those from observations .
the theoretical relations are in general good agreement with their observational counterparts though there exist some discrepancy for short period ( @xmath0 < 1 $ ] ) cepheids .
we outline a physical mechanism which can , in principle , be one factor to explain the observed pc / ac relations for the long and short period cepheids in the galaxy , lmc and smc .
our explanation relies on the hydrogen ionization front - photosphere interaction and the way this interaction changes with pulsation period , pulsation phase and metallicity .
since the pc relation is connected with the period - luminosity ( pl ) relation , it is postulated that such a mechanism can also explain the observed properties of the pl relation in these three galaxies .
cepheids stars : fundamental parameters |
the fock vacuum state with respect to a uniformly accelerating observer , behaves as a _
thermal _ state of a given temperature which is proportional to the magnitude , say @xmath0 , of the acceleration 4-vector .
this phenomena is referred as unruh effect @xcite .
the corresponding temperature is called _ unruh temperature _ given by @xmath1 where @xmath2 is the _ boltzmann constant_. this result follows from the application of standard quantum field theory techniques in a curved background @xcite .
the existence of unruh effect can be seen using many different methods .
firstly , one can use the method of _ bogoliubov transformations_. there one computes expectation value of number density operator for the accelerated observer in the fock vacuum state .
the corresponding expression turns out to be a blackbody distribution at unruh temperature . in this method
it appears explicitly that unruh effect depends on the contributions from trans - planckian modes as observed by an _
inertial _ observer .
this aspect makes the unruh effect to be a potentially interesting phenomena to explore and understand the implications of possible planck - scale physics @xcite . in other words ,
the unruh effect provides a probing window for many candidate theories of quantum gravity .
usually these theories predict large modifications for the trans - planckian modes .
while this approach is conceptually simpler but in order to arrive at the result one is required to use rather sophisticated regularization techniques to handle field theoretical divergences .
therefore , one is led to pursue other methods of derivation to ensure that the result is not an artifact of the employed regularization tools .
an alternative method which is often employed to verify the existence of unruh effect , is to compute the response function of the so - called unruh - dewitt detector that moves along the trajectory of the accelerated observer @xcite .
this method essentially relies on the concepts of einstein @xmath3 and @xmath4 coefficients that are used to describe the _ spontaneous _ and _ induced _ emission or absorption in statistical physics . in particular , one considers a two - level quantum mechanical system as a detector which weakly couples to the surrounding matter field .
the transition probability between the energy levels which is referred as the _ response function _ of the detector , is then used to conclude about the nature of the surrounding environment .
it turns out that the response function depends on the properties of two - point function of the matter field .
given that the unruh effect receives explicit contributions from trans - planckian modes in the method of bogoliubov transformations , it is compelling to check whether these contributions can also affect the corresponding response function . in this article , we compute the response function of a unruh - dewitt detector in the context of the so called polymer quantization of scalar matter field which couples weakly to the detector .
polymer quantization or loop quantization @xcite is used as a quantization method in loop quantum gravity @xcite .
it is known to differ from the schrdinger quantization method in multiple ways when applied to a mechanical system .
firstly , it comes with a new dimension - full parameter other than _ planck constant _ @xmath5 . in the context of full quantum gravity , this new scale essentially corresponds to _ planck length _ @xmath6 .
secondly , in the kinematical hilbert space which is non - separable , one can not define both position and momentum operators simultaneously but only one of them . due to the non - separability of kinematical hilbert space
, the stone - von neumann uniqueness theorem is also not applicable .
these aspects make polymer quantization unitarily _ inequivalent _ to schrdinger quantization @xcite , allowing a different set of results , in principle , from polymer quantization . in the section [ sec : rindlerobserver ]
, we briefly study the trajectory of a uniformly accelerated observer in minkowski spacetime and the properties of rindler metric as seen by the accelerated observer . in the section [ sec : einsteinabcoefficients ] , we review the basic ideas behind einstein @xmath3 and @xmath4 coefficients . in the section [ sec : masslessscalarfield ] , we consider a massless scalar field in the canonical approach which is convenient for application of polymer quantization .
we study the properties of the unruh - dewitt detector in the section [ sec : unruh - dewitt - detector ] .
subsequently , we study the short and long distance behaviour of the two - point function in fock and polymer quantization .
then we compute the corresponding response functions . in fock quantization , induced transition rate of the detector contains both transient and non - transient terms .
in particular , non - transient term is proportional to planck distribution and it contains a residue computed at the pole of the corresponding two - point function . on the other hand , we show that the two - point function has no pole in polymer quantization .
consequently , the corresponding residue which is non - zero in fock quantization vanishes in polymer quantization .
so the transition rate in polymer quantization contains only transient terms .
the result as shown here , provides an alternative evidence to the earlier reported results where it is shown using method of bogoliubov transformation @xcite as well as using kubo - martin - schwinger ( kms ) condition @xcite that unruh effect is absent in polymer quantization .
with respect to a uniformly accelerated observer , a section of minkowski spacetime can be described by the so - called rindler metric . in _ conformal _ rindler coordinates @xmath7 ,
the rindler metric can be expressed as @xcite @xmath8 where we have used the _
natural _ units ( @xmath9 ) .
the parameter @xmath0 is the magnitude of _ acceleration _ 4-vector .
minkowski metric with respect to an _ inertial _ observer with cartesian coordinates @xmath10 is @xmath11 .
the coordinates of rindler observer _
i.e. _ the uniformly accelerated observer , is related with the coordinates of the inertial observer as @xmath12 where we have assumed that rindler observer moves along @xmath13 x - axis with respect to the inertial observer
. therefore , @xmath14 and @xmath15 coordinates are the same for both observers .
it follows from the equation ( [ rindlerminkowskirelation ] ) that only a wedge - shaped section of minkowski spacetime is covered by rindler spacetime .
the wedge - shaped section is often referred as _ rindler wedge_.
the working principle behind unruh - dewitt detector can be understood in terms of the so - called _ einstein a and b coefficients _ of statistical physics .
let us consider a two - level atomic system of energy @xmath16 and @xmath17 which is in contact with a _ heat bath _ of spectral energy density @xmath18 such that @xmath19 .
the probability of transition from the _ level @xmath20 _ to the _ level @xmath21 _ _ per unit time _ , _ per unit atom _ is postulated to be @xmath22 where @xmath3 and @xmath4 are the einstein coefficients . the coefficient @xmath3 represents the transition rate for the _ spontaneous _ emission whereas the coefficient @xmath4 represents the same for the _ induced _ emission . on the other hand ,
transition probability _ per unit time _ , _ per unit atom _ from the _ level @xmath21 _ to the _ level @xmath20 _ is postulated to contain only the _ induced _ term as @xmath23 if the system is in equilibrium with the heat bath and number of particles in the energy _
levels @xmath21 , @xmath20 _ are @xmath24 , @xmath25 respectively , then the detailed balance relation implies @xmath26 on the other hand , particles numbers would satisfy boltzmann distribution law _
i.e. _ @xmath27 .
so if the ratio of the einstein coefficients is taken to be @xmath28 , then it precisely leads to planck distribution formula for spectral energy density of blackbody radiation @xmath29 clearly , the induced transition rate of the two - level system in equilibrium can be used to conclude about the thermal nature of the surrounding environment .
in particular , the equation ( [ twoleveltransition12 ] ) implies that if the induced transition rate from the _ level @xmath21 _ to the _ level @xmath20 _ is proportional to planck formula for _ mean energy _
@xmath30 of a given mode with angular frequency @xmath31 _ i.e. _
@xmath32 then surrounding environment must be a thermal radiation at a temperature @xmath33 .
we note that only the ratio of the einstein coefficients is relevant here but not their absolute values .
in order to study the response function of the unruh - dewitt detector , we consider here a massless scalar field @xmath34 in minkowski spacetime that weakly couples to the detector .
the corresponding scalar field dynamics is described by the action @xmath35 ~.\ ] ] in the canonical formulation , the dynamics of the scalar field can be equivalently described by the field hamiltonian @xmath36 ~,\ ] ] where @xmath37 is the metric on _ spatial _ hyper - surfaces labeled by @xmath38 .
poisson bracket between the field @xmath39 and its conjugate field momentum @xmath40 is @xmath41 where @xmath42 is the dirac delta .
it is convenient to use fourier space for studying dynamics of a massless free scalar field .
here we define fourier modes for the scalar field and its conjugate field momentum as @xmath43 where @xmath44 is the spatial volume .
the space being non - compact , the spatial volume would normally diverge for minkowski spacetime .
therefore , it is often convenient to use a fiducial box of finite volume so that one can avoid dealing with such divergent quantity .
kronecker delta and dirac delta then can be expressed as @xmath45 and @xmath46 .
the field hamiltonian ( [ sfhamgen ] ) can be expressed as @xmath47 , where hamiltonian density for the @xmath48th fourier mode is @xmath49 poisson brackets between these fourier modes and their conjugate momenta are given by @xmath50 in order to satisfy the _ reality condition _ of the scalar field @xmath51 , one usually redefines the _ complex - valued _
modes @xmath52 and momenta @xmath53 in terms of the real - valued functions @xmath54 and @xmath55 such that corresponding hamiltonian density and poisson brackets become @xmath56 this is the standard hamiltonian for a system of decoupled harmonic oscillators .
we may express the energy spectrum of these fourier modes as @xmath57 . using the vacuum state of each mode @xmath58
, we can express the vacuum state of the scalar field as @xmath59 .
unruh - dewitt detector is considered to be a point - like quantum mechanical system having two internal energy levels .
these system of detectors interact _ weakly _ with the scalar field through a _ linear _ coupling which is treated as perturbative interaction to the detectors .
if we denote the hamiltonian operator of an _ unperturbed _ detector by @xmath60 then we may express the energy eigenvalue equation of the detector as @xmath61 where @xmath62 and @xmath63 represent the _ ground state _ and the _ excited state _ respectively .
we may denote the energy gap between the levels as @xmath64 the interaction term in the hamiltonian of unruh - dewitt detector is taken to be of the form @xmath65 where @xmath66 denotes the _ coupling constant _ and @xmath67 is the monopole moment operator of the detector .
the world line @xmath68 of the detector is parametrized using the proper time @xmath69 .
therefore , the total hamiltonian of the detector becomes @xmath70 given the nature of the hamiltonian , it is convenient to work in the so - called ` interaction picture ' of quantum mechanics . in particular , by defining interaction state @xmath71 in terms of schrdinger state @xmath72 as @xmath73 one can express the time evolution of the detector state as @xmath74 where the evolution operator is given by @xmath75 we may denote the combined state of the detector and the scalar field at a given proper time @xmath69 as @xmath76 to the @xmath77 can be written as @xmath78 we are interested here in considering the induced transition of the detector which is initially at the ground state @xmath62 and the scalar field is in its vacuum state _ i.e. _ @xmath79 .
then transition probability of the detector being at the excited state @xmath63 at a later time @xmath69 for all possible field state is @xmath80 it is now straightforward to express the transition probability ( [ transitionprobabilitydetector ] ) in the form of @xmath81 where @xmath82 .
@xmath83 depends on the internal structure of the detector system .
the function @xmath84 is known as the _ response function _ of the detector and is defined as @xmath85 where @xmath86 is the _ two - point function _ of the scalar field and is given by @xmath87 we may recall that only the ratio of the einstein coefficients can be determined but not their absolute values .
we shall use this freedom to scale the response function of the detector such that proportionality constant between the induced transition rate and the mean energy per mode , as in equation ( [ twoleveltransition12numberdensity ] ) , becomes identity . using equation ( [ transitionprobabilitydetector2 ] )
we define the _ instantaneous _ transition rate , after being scaled by the chosen factor , as @xmath88 which can be further expressed as @xmath89 here @xmath90 represents time independent _ i.e. _ _ non - transient _ part of the induced transition rate and its expression is given by @xmath91 on the other hand @xmath92 represent time - dependent _ i.e. _ _ transient _ part of the induced transition rate and its expression is @xmath93 ~.\ ] ] clearly , @xmath94 as the time of observation increases _ i.e. _ @xmath95 .
we should emphasize here that the definition of instantaneous transition rate ( [ detectortransitionrate ] ) presumes the so - called _ sharp switching _ of the unruh - dewitt detector .
while sharp switching of the detector is used quite widely in the literature , it has certain key weaknesses .
this issue has been studied in @xcite and it is shown that one could overcome these weaknesses by considering either a spatially extended detector or a smooth switching function without affecting the main result regarding unruh effect .
it is evident from the equation ( [ detectortransitionrate ] ) that induced transition rate of the unruh - dewitt detector is fully determined once the properties of the two - point function is known . in terms of the fourier modes ( [ fouriermodesdef ] ) ,
the general form of such a two - point function can be written as @xmath96 where the matrix element @xmath97 is given by @xmath98 ) , the hamiltonians and the corresponding poisson brackets ( [ sfhamfourierminkowskireal ] ) are independent of the fiducial volume .
therefore , we can easily remove the fiducial box by taking the limit @xmath99 .
this essentially replaces the sum @xmath100 by an integration @xmath101 in the expression of the two - point function ( [ minkowskitwopointdef2 ] ) as @xmath102 using energy spectrum of the fourier modes and by expanding the state @xmath103 in the basis of energy eigenstates as @xmath104 , the matrix element @xmath97 formally becomes @xmath105 where @xmath106 and @xmath107 .
given the matrix element @xmath108 depends only on magnitude @xmath109 , one can carry out the angular integration using _ polar coordinates _
, to reduce the two - point function ( [ minkowskitwopointintegraldef ] ) as @xmath110 where @xmath111 with @xmath112 , @xmath113 and @xmath114 @xcite .
in fock quantization of the scalar field , fourier modes which are represented by the equation ( [ sfhamfourierminkowskireal ] ) , are essentially quantized using schrdinger quantization method . the corresponding energy spectrum and the coefficients @xmath115 ( [ dkfunctiongeneral ] )
are given by @xmath116 the two - point function ( [ kgpropagatordiffpm ] ) then reduces to its standard form for the fock space @xmath117}~,\ ] ] where @xmath118 is lorentz invariant spacetime interval and @xmath119 is a small , positive parameter that is introduced as the standard integral regulator . in order to illustrate the nature of the _ transient _ response of the unruh - dewitt detector , for simplicity
, we consider a detector which is located at a fixed position in an inertial frame _
i.e. _ its world line is given by @xmath120 where proper time @xmath69 is defined as @xmath121 .
along the world line of the detector the two - point function becomes @xmath122 one would get the same two - point function even for a detector which moves with a uniform velocity .
we can compute the non - transient part of the transition rate @xmath90 ( [ integralr0 ] ) by closing the contour in the lower half of the complex plane .
we note that the two - point function ( [ focktwopointfunctionintertial ] ) has a _ pole _ at @xmath123 and the contour does _ not _ enclose the pole .
therefore , the non - transient part of the transition rate vanishes _ i.e. _ @xmath124 the _ transient _ part @xmath92 can also be computed in a straightforward manner as @xmath125 ~,\ ] ] where @xmath126 is the _ sine integral _ function which goes to @xmath127 as @xmath95 .
clearly , the induced transition rate of an unruh - dewitt detector is _ non - vanishing _ even along an inertial trajectory if the observation is made for a finite period .
however , as expected , such a response of the detector decays out as the time of observation increases .
therefore , in order to conclude about _ thermal _ nature of the surrounding environment , one must look at the _ non - transient _ part of induced transition rate of the detector .
we now consider an unruh - dewitt detector which moves along a rindler trajectory given by @xmath128 . for convenience ,
we define a new variable @xmath129 .
it leads the time interval @xmath130 and spatial separation @xmath131 to become @xmath132 the non - transient part of the transition rate ( [ integralr0 ] ) then becomes @xmath133 which may also be expressed as @xmath134 we note that integrand @xmath135 has a _ pole of second order _ at @xmath136 and it is a _ multi - valued _ function of complex variable @xmath137 .
contour used to evaluate @xmath90 ] following the contour in the complex plane as shown in fig .
[ fig : contour ] , we express the contour integral as @xmath139_{|\eta=1+i\delta } ~,\ ] ] where @xmath140_{|\eta=1+i\delta}$ ] denotes the _ residue _ of the function @xmath135 evaluated at the pole @xmath141 . using the properties that @xmath142 and @xmath143 , non - transient part of the transition rate can be expressed as @xmath144_{|\eta=1+i\delta}}{e^{2\pi\omega / a } - 1 } ~.\ ] ] the evaluated residue at the pole of the two - point function in fock space @xmath145_{|\eta=1+i\delta } = \frac{i \omega}{2\pi } ~,\ ] ] leads the induced transition rate to become @xmath146 in other words , the induced transition rate @xmath90 is precisely equal to the standard expression for mean energy per mode of a system in thermal equilibrium at the unruh temperature @xmath1 .
the transient part @xmath147 can be computed easily by considering reasonably large but finite time of observation as @xmath148.\ ] ] the transient terms decays out exponentially as the time of observation increases .
we summarize two key inputs that are required to arrive at the equation ( [ rindlerfockr0final ] ) .
firstly , to ensure @xmath149 the function @xmath135 should have sufficient fall - off as @xmath150 so that _ jordan s lemma _ is applicable _
i.e. _ @xmath151 . secondly , in order to have a _
non - zero _ residue , the function @xmath135 should have at least one _ pole _ within the region enclosed by the contour .
it is evident from the equation ( [ integralr0fock3 ] ) , the _ short - distance _ singular nature of the fock space two - point function is a key requirement to ensure non - vanishing residue . in other words , the pole of two - point function @xmath152 at ( @xmath153 plays the crucial role for the existence of the unruh effect .
in order to perform the so - called polymer quantization of scalar field , we follow the approach as suggested here @xcite and was followed up later to study unruh effect @xcite . in this approach ,
one quantizes the system of fourier modes using polymer quantization instead of schrdinger quantization as used in fock space . as mentioned earlier ,
polymer quantization or _ loop quantization _ is a quantization technique that is used in _ loop quantum gravity _ and it comes with a new dimension - full parameter say @xmath154 along with planck constant @xmath5 . in full quantum gravity , the parameter @xmath154 would be analogous to planck length .
the energy spectrum of the @xmath48th oscillator mode in polymer quantization are given by @xcite @xmath155 where @xmath156 , @xmath157 , @xmath158 are _ mathieu characteristic value functions _ and @xmath159 is a _ dimension - less _ parameter .
the energy eigenstates are @xmath160 and @xmath161 where @xmath162 .
the functions @xmath163 and @xmath164 are solutions to _
mathieu equation_. they are referred as elliptic cosine and sine functions respectively @xcite .
in order to arrive at these @xmath165-periodic and @xmath165-antiperiodic states in @xmath166 , superselection rules are invoked .
this superselection leads to exact energy spectrum which in turns allows to study the system analytically .
besides , without imposition of superselection rules , certain key statistical notions for the system are known to become ill - defined @xcite . for low - energy modes _
i.e. _ for small @xmath167 , the energy spectrum ( [ eigenvaluemcfrelation ] ) reduces to regular harmonic oscillator energy spectrum along with perturbative corrections @xmath168 therefore , polymer quantization leads to expected results for low - energy modes .
however , we note that polymer energy spectrum has two - fold degeneracy as @xmath169 and it is lifted for finite values of @xmath167 . the coefficients @xmath170 are non - vanishing in polymer quantization for all @xmath171 , unlike in fock quantization where only _ one _ @xmath115 is non - vanishing ( [ fockspectrum ] ) . using asymptotic properties of mathieu functions , we can approximate the energy gaps @xmath172 between the levels and coefficients @xmath173 for low - energy modes or _ sub - planckian modes _ ( _ i.e. _ @xmath174 ) as @xmath175 for @xmath176 , and @xmath177 ~,~ \frac{c_{4n+3}}{c_{3 } } = \mathcal{o}\left(g^n\right),\ ] ] for @xmath178 . on the other hand , for high energy modes or _ super - planckian _ modes ( _ i.e. _ @xmath179 ) , we can approximate the energy gaps and coefficients @xmath173 as @xmath180 for @xmath176 , and @xmath181 , \frac{c_{4n+3}}{c_{3 } } = \mathcal{o}\left(\frac{1}{g^{2n}}\right),\ ] ] for @xmath182 . therefore , we can approximate matrix element @xmath183 in polymer quantization as @xmath184 for both the cases . as discussed earlier
, the singular nature of the short - distance two - point function plays the key role in providing non - vanishing residue to the expression of induced transition rate ( [ rindlerfockr0final ] ) in fock space .
therefore , in order to understand the effect of polymer quantization on the response function of unruh - dewitt detector it is crucial to determine the form of short - distance two - point function ( _ i.e. _ near @xmath185 ) in polymer quantization . given @xmath186 implies @xmath187 , we may express the two - point function @xmath188 , as a series of the form @xmath189 where we have used the identity @xmath190 . in fock quantization ,
the matrix element has exact expression @xmath191 whereas in polymer quantization we can approximate it as @xmath192 . in other words , we can represent the matrix element for both the cases in a general form @xmath193 .
further , in the domain where @xmath194 is very small , the temporal and the spatial separations satisfies @xmath195 and @xmath196 is at least @xmath197 or higher for all modes .
so by defining a new variable @xmath198 and keeping only the leading term , we may approximate the two - point function ( [ twopointfunctionseriesgeneral ] ) as @xmath199 where @xmath200 unlike in fock quantization , the expression for @xmath201 and @xmath202 are different for sub - planckian and super - planckian modes in polymer quantization due to the presence of the scale @xmath154 .
therefore , here we consider the sub - planckian and super - planckian contributions to the two - point function separately as @xmath203 the sub - planckian contributions to the two - point function , is defined as @xmath204 where @xmath205 and @xmath206 is some pivotal value of the wave - vector chosen such that @xmath207 is @xmath197 .
the limit @xmath208 implies @xmath209 which in turns leads @xmath210 . on the other hand ,
the super - planckian contributions to the two - point function , defined by @xmath211 are expected to dictate the short - distance ( @xmath186 ) behaviour of the two - point function . using the asymptotic expressions ( [ largegdeltae4n+3 ] ) and ( [ largegc4n+3 ] )
, one can compute the expression for @xmath212 for polymer quantization as @xmath213 ~.\ ] ] therefore , the short - distance two - point function with _
non - perturbative _ modifications from polymer quantization can be expressed as @xmath214 + \mathcal{o}(\lstar^{-2 } ) ~,\ ] ] where we have used _
analytic continuation _ to evaluate the integral as @xmath215 .
we note here that the short - distance two - point function ( [ gpolyfinal ] ) instead of diverging , it reaches a maximum value of @xmath216 .
this bounded from above behaviour of the two - point function is analogous to the behaviour of the spectrum of _ inverse scale factor _ operator as well as the effective hubble parameter in loop quantum cosmology ( lqc ) @xcite .
both of these can be associated with some inverse powers of the distance similar to the two - point function . in lqc , this crucial behaviour plays a key role in resolution of big bang singularity .
however unlike in lqc , here we have applied polymer quantization only for scalar matter field rather than for the geometry .
we may verify that the equation ( [ gsuper ] ) also reproduces the result of fock quantization up to @xmath217 term as @xmath218 ~.\ ] ] here we have used the fock space expression @xmath219 that follows from the equation ( [ fockspectrum ] ) .
we note that unlike in fock quantization , two - point function ( [ gpolyfinal ] ) _ does not _ have any pole as @xmath208 .
therefore , the corresponding residue @xmath140_{|\eta=1+i\delta}$ ] vanishes in polymer quantization .
this in turns , leads _
non - transient _ part of the induced transition rate ( [ rindlergeneralr0 ] ) to vanish _
i.e. _ @xmath220 thus , from the response of the unruh - dewitt detector which interacts with polymer quantized scalar field , one would conclude that unruh effect is not present in polymer quantization .
we should mention here that the large - distance two - point function including polymer corrections was computed in @xcite and it may be checked that _
jordan s lemma _ continues to be applicable even including polymer corrections .
as earlier , we may approximate the transient part of the transition rate by considering a reasonably large but finite observation time @xmath69 .
in particular , in the domain of perturbation where @xmath221 and @xmath222 , the transient part is @xmath223 } { \pi \left ( 1 + \omega^2/a^2 \right ) } \left[a \cos(\omega \tau ) -\omega \sin(\omega\tau)\right].\ ] ] clearly , the transient part of the induced transition rate decays exponentially in proper time @xmath69 , similar to the results of fock quantization .
we should mention here that the polymer quantization is known to cause a violation of lorentz invariance due to the presence of the length scale @xmath154 .
as one of the consequences , it is shown in @xcite that polymer vacuum state is not strictly invariant under the _ boost_. therefore , the specific nature of the transient terms as studied here are effectively tied to the observer in polymer quantization .
in summary , we have studied the properties of response function of an unruh - dewitt detector which moves along a rindler trajectory and interacts weakly to a polymer quantized massless scalar field . through a detailed calculation
we have shown that unlike in fock quantization , there are only transient terms present in the induced transition rate of the detector . in fock quantization ,
the induced transition rate of the detector contains also a non - transient term which is proportional to planck distribution .
this property of the induced transition rate signifies the existence of unruh effect in fock quantization . in polymer quantization of scalar field , however , the non - transient term is absent .
therefore , an unruh - dewitt detector would not perceive a flux of thermal particles along an uniformly accelerating trajectory in polymer quantization .
the result as shown here provides an alternative evidence for the earlier reported results by the authors where it is shown using method of bogoliubov transformation @xcite as well as using kubo - martin - schwinger ( kms ) condition @xcite that unruh effect is absent in polymer quantization .
we note the key ingredients that led to the main result of this paper .
as evident from the equation ( [ rindlergeneralr0 ] ) , the induced transition rate of an unruh - dewitt detector along a rindler trajectory is quite generically proportional to the planck distribution formula ( _ i.e. _ @xmath224 ) .
however , the proportionality constant contains the residue of the two - point function evaluated at its pole .
therefore , the _ short - distance _ singular nature of the fock space two - point function is a key input to ensure the non - vanishing residue which in turns plays the crucial role for the existence of unruh effect . on the other hand
, it is widely expected that the short distance behaviour of the two - point function would receive significant modifications from possible planck - scale physics . in this paper
, we have shown that this expectation is indeed borne out in polymer quantization where the short - distance two - point function receives _ non - perturbative _ modifications .
in particular , the short - distance singular two - point function is replaced by a regular function which has no divergence in short - distance .
finally , we may recall that several experiments have been proposed to detect possible signatures of unruh effect in laboratory @xcite .
therefore , the result as shown here using response function of an unruh - dewitt detector and also shown earlier using methods of bogoliubov transformation @xcite and kubo - martin - schwinger ( kms ) condition @xcite where criticism raised in @xcite was also addressed , clearly indicates that the experimental detection of unruh effect can be used as a probe of possible planck - scale physics .
in particular , such a detection can be used to either verify or rule out a candidate planck scale theory with a new dimension - full parameter , which affects quantization of matter fields as considered here .
we conclude by acknowledging few similar recent results in the context of polymer quantization where it is shown that some high energy modifications can indeed lead to the alteration of certain low energy phenomena @xcite .
in particular , it is shown in @xcite that polymer vacuum state is not strictly invariant under the _ boost_. therefore , the specific nature of the transient terms as studied here are effectively tied to the observer in polymer quantization . however , the transient terms are shown to decay out exponentially .
so the detector response after a reasonably long time , in the scale of @xmath225 , becomes essentially time independent .
clearly , the existence of unruh effect needs to be understood from the properties of the non - transient part of the detector response as studied here .
we would like to thank ritesh singh for discussions .
we thank subhajit barman and chiranjeeb singha for their comments on the manuscript .
gs would like to thank ugc for supporting this work through a doctoral fellowship . | if an unruh - dewitt detector operates for an infinite proper time along the trajectory of a uniformly accelerated observer in fock space then induced transition rate of the detector is proportional to planck distribution . for a realistic detector which operates only for a finite period , the instantaneous transition rate contains both transient and non - transient terms . in particular , the non - transient term contains a _ residue _ evaluated at the pole of the two - point function .
we show here by considering a massless scalar field that unlike in fock quantization , the short - distance two - point function contains no pole in polymer quantization , the quantization techniques used in loop quantum gravity .
consequently , corresponding transition rate of the unruh - dewitt detector contains only transient terms .
thus , the result presented here provides an alternative evidence for absence of unruh effect in polymer quantization which was shown earlier by the authors using methods of bogoliubov transformation and kubo - martin - schwinger condition . |
due to their short coherence length , long penetration depth and high anisotropy , the high - t@xmath4 superconductors ( hts ) , show very unusual behaviour . apart from a proposed vortex - glass phase @xcite , which has probably been observed in the highly anisotropic bi@xmath0sr@xmath0cacu@xmath0o@xmath5 ( bscco ) at high fields @xcite ,
there is also the exciting phenomenon of vortex - lattice melting in these compounds . unlike in bscco ,
the vortex lattice melting transition in yba@xmath0cu@xmath1o@xmath2 ( ybco ) takes place at sufficiently high fields to be studied using thermal measurements such as differential thermal analysis ( dta ) @xcite and specific heat @xcite .
this is mainly due to the much lower anisotropy of ybco ( @xmath6 5 ) .
apart from this phenomenological difference , there are however also more fundamental differences between ybco and bscco in the crystallographic symmetry , which is orthorhombic in the former , while it is tetragonal in the latter .
this leads to a fundamental difference in the growth of the samples , in that untreated ybco samples are usually heavily twinned ( having domains of interchanged @xmath7 and @xmath8 axes ) .
the boundaries between these domains ( twin - planes ) act as very strong pinning sites to the vortices @xcite , which may change some of the properties of the vortex lattice in the sample .
it has for instance been argued that such strong pinning may change the ( intrinsically ) first order transition to a second order transition @xcite .
this was also supported by the fact that the transitions observed ( for instance by muon - spin rotation ( @xmath9sr ) @xcite and neutron small - angle scattering ( sans ) @xcite ) in bscco were generally very sharp , not inconsistent with a first order transition , whereas twinned ybco samples gave a more gradual transition . with the advent of sufficiently sized untwinned and high purity twinned samples
, it became possible to study these effects in more detail using dta@xcite and specific heat@xcite to determine the true thermodynamic properties of the phase transition .
in untwinned samples , dta measurements have clearly shown the existence of a first order transition@xcite , even though the sample was nt pure but had some au impurities due to the growth process . in extremely pure samples ,
specific heat measurements have also shown a first order transition even though the sample was heavily twinned@xcite . however
, the twinned case is in general much more complicated than the untwinned one in that the order of the transition may also depend on the applied field@xcite .
the order of the transition is thus probably influenced by both kinds of defects ( impurities and twin planes ) , where an excess of pinning sites of any kind drives the transition to second order .
of further interest is the dependence of the melting transition on the angle of the field with respect to the @xmath3-axis @xcite .
there are detailed predictions on the field dependence of the melting temperature ( melting line ) on the angle from anisotropic london theory @xcite .
the nature of the transition , especially the change in entropy at melting has however only recently been investigated @xcite .
we have now performed @xmath9sr and sans measurements of the vortex lattice in a large untwinned crystal of ybco with the field at different angles to the crystallographic @xmath3-axis . as is the case in the dta and specific heat measurements , we probe the _ equilibrium state _ of the vortex lattice . from sans ,
we obtain a microscopic measurement of the global ( long range ) arrangement of the flux lines .
appropriately scaled , the sans intensity presents a phenomenological order parameter of the flux lattice ( see below ) .
this allows investigations similar to those of specific heat or dta , clarifying the order of the transitions .
it should however be noted that the samples used in sans are very much bigger than those used for thermal measurements .
thus a possibly present first order transition may already be smeared by slight inhomogeneities over the big sample volume .
in contrast , with @xmath9sr we observe the local magnetic field structure averaged over the sample .
therefore smearing effects may still be present , however a sharp change of the local equilibrium structure in itself may not be regarded as a sign of a first order transition . due to
the different time scales available to neutrons and muons one would expect to observe a difference in the melting line determined by the two methods if the transition is of a more glassy nature . for a ( first order )
phase transition , one would expect both methods of measurement to agree . as will be shown below
, our results differ quite strongly from those on a heavily twinned sample @xcite . there
the melting transition observed with sans is of second order , with the intensity decreasing to zero continuously at a temperature t@xmath10 t@xmath4 .
in contrast our results show a sharp decrease of the sans intensity with an infinite slope at t@xmath11 .
this is consistent with a first order transition considering the large size of the sample ( see below ) .
the sample consisted of a large ( 1.125 g ) ybco single crystal .
it was grown untwinned by application af uniaxial stress during the cooling from the melt @xcite . for the oxygenation process
, the sample was treated in flowing o@xmath0 at 490@xmath12c .
magnetisation measurements using a vibrating sample magnetometer ( vsm ) indicate a t@xmath4 of 93 k. the superconducting transition is estimated from these measurements to be @xmath130.5 k. this also gives a limit for the sharpness of the melting transition ( see below ) .
the @xmath9sr experiments were carried out on beamline @xmath14m3 at the paul scherrer institute ( psi ) , villigen switzerland , with a maximum field of 0.6 t and a cryostat capable of cooling to 2 k. in the experiments , a magnetic field was applied perpendicular to the incoming muon spin polarisation , while at a specified angle to the crystallographic @xmath3-direction of the sample . in this case , the spin of a stopped muon precesses with the frequency @xmath15b@xmath16 , where @xmath17 135 mhz / t is the gyromagnetic ratio of the muon in the transverse local field b@xmath16 .
the angular anisotropy of the muon decay , leading to a preferred emission of the decay - positron in the direction of the muon spin due to parity violation , allows the time evolution of the muon polarisation to be measured . in the mixed state of a type - ii superconductor ,
the magnetic field distribution is given by the structural properties of the vortex lattice ( i.e. the penetration depth @xmath18 , the coherence length @xmath19 and the lattice morphology ) .
the vortex lattice spacing is much bigger than the atomic spacing of the underlying crystal structure and is incommensurate to it .
therefore the muons sample the field distribution randomly , even though they are stopped at specific crystallographic sites .
this allows the field probability distribution p(b ) to be measured via the fourier transform of the oscillating time signal . as the different local fields cause the muon spin to precess at different frequencies the time signal consists of a sum of oscillations at different frequencies weighted by their relative occurrence . in order to obtain the fourier
transform from the observed time data , we use a maximum entropy algorithm @xcite . in this
procedure aberrations of the spectra due to finite time windows or large statistical errors at long times are avoided , as the data are not subjected to a fourier transform themselves . starting from an initially flat field distribution p(b ) , the most uniform distribution compatible with the given time spectrum is obtained by maximising its constrained entropy @xmath20 here @xmath21 is the entropy of the distribution according to shannon and @xmath22 is the usual measure giving the difference of the fourier transform of p(b ) ( @xmath23 ) to the data in the time domain ( @xmath24 ) , where @xmath25 is the error in the data .
the field distribution corresponding to a vortex lattice has very characteristic features arising from its two - dimensional periodicity . at extremal points in the spatial distribution the field probability distribution exhibits van hove singularities .
these singularities ideally are finite jumps at the minimum and maximum fields , and furthermore a logarithmic divergence at saddle points . due to the very high fields in the vortex cores ,
the field distribution is highly asymmetric with a long ` tail ' at fields higher than the average . in the entangled vortex liquid state
however , the core fields are smeared out over a large distance . therefore the high field tail is truncated leading to a symmetric lineshape .
we quantify the asymmetry of the lineshape by the dimensionless parameter @xmath26 a symmetric distribution corresponds to a value of @xmath27 , whereas a distribution with a high field tail has a positive @xmath28 .
the sans experiments were carried out on instruments d11 , d22 and d17 at the institut laue - langevin ( ill ) , grenoble france .
the neutrons typically had a wavelength of @xmath29 = 1 - 1.5 nm and the beam was collimated over a distance of 2.75 m in the experiments carried out at d17 , whereas the collimation length was 10 m in those at d11 and d22 .
the distance between the sample and the detector was then varied for optimal resolution .
the incoming neutron beam was aligned with the applied field to an accuracy of 0.1@xmath12 by observing the diffraction pattern from a nb single crystal .
the sample was mounted in a cryomagnet capable of a field of 5 t and a temperature of 1.5 k. the neutrons are scattered from the spatial variations of the magnetic field in the sample , due to the magnetic moment of the neutron ( @xmath30 ) , with the interaction energy @xmath31 . in the case of a type - ii superconductor in the mixed state ,
the field variation is periodic giving rise to bragg diffraction off the scattering planes of the vortex lattice @xcite , where the angle of refraction is given by @xmath32 here @xmath33 is the lattice spacing , which depends on the applied field as @xmath34 for a hexagonal lattice ( @xmath35 = h/2e is the magnetic flux quantum )
. thus the scattering is at very small angle , as typical vortex separations of @xmath1350 nm ( corresponding to a field of @xmath131 t ) result in a bragg angle ( from eq .
[ bragg ] ) of 2@xmath36 2@xmath12 .
the intensity of neutrons scattered into a bragg - peak , integrated over angle as the sample is rocked through the bragg condition , is then mainly given by the square of the form factor of a flux line , which in the london approximation is given by @xmath37 where @xmath18 is the penetration depth .
the london approximation is valid in the case of an extreme type - ii superconductor with @xmath38 , a condition well satisfied by the hts .
therefore , for fields much greater than b@xmath39 , corresponding to @xmath40 1 , the scattered intensity depends on the penetration depth as @xmath41 . for a classical type - ii superconductor ,
the temperature dependence of the scattered intensity therefore is a direct measure of that of the penetration depth . in the hts however , where thermal fluctuations may have a dominant contribution this is not necessarily the case . in order to discern effects arising from the temperature dependence of @xmath18 and those of the thermal fluctuations finally leading to the melting of the lattice we divide the observed intensities by the temperature dependence of @xmath42
this was obtained from the sans intensity at a low field ( 0.2 t ) , where the effects of thermal fluctuations are negligible ( see below ) .
this quantity then takes the role of an effective flux line lattice ( fll ) order parameter , which we use to observe the melting transition and its order .
because the scattered intensity depends on the penetration depth as @xmath41 the intensity in the hts is rather low , due to their long penetration depths ( @xmath18 = 150 - 200 nm ) .
furthermore , substantial background scattering from the cryostat and extended defects in the sample make it necessary to subtract a diffraction pattern taken above t@xmath4 to obtain a clear signal .
due to thermal contractions of the sample stick and cryostat , this background is shifted by fractions of a detector pixel in order to obtain a good subtraction . thanks to the high neutron intensity at the instruments d22 and d11 , a counting time of 30 minutes for each foreground and background was enough to obtain a reasonable signal .
a typical diffraction pattern can be seen in fig . [ pattern ] for an applied field of 0.2 t. there the field is applied at an angle of 33@xmath12 to the @xmath3-axis .
this is why the observed pattern is isotropic .
the distortion due to the @xmath43-anisotropy ( @xmath44 = 1.16 @xcite ) is compensated by the distortion due to the @xmath45-anisotropy at an angle of 33@xmath12 . furthermore , we note that second order diffraction spots are visible in the figure , indicating the high quality of the flux lattice .
the intensities measured from such a pattern are obtained from fitting a gaussian to each individual bragg - spot and taking the sum of all spots in order to reduce the statistical error .
all of these measurements of scattered intensity rely on the fact that the width of the rocking curve does not change as a function of temperature .
this is because the calculation relating the scattered intensity to the penetration depth is done for the total integrated intensity@xcite , which is given by the width of the rocking curve multiplied by the intensity when fulfilling the bragg condition .
the rocking curve width was checked to be constant for selected temperatures over a wide range .
we therefore only consider the scattered intensities at the bragg - angle . in order to determine an effective fll order parameter we have to divide the intensity by its temperature dependence intrinsically given by the penetration depth .
therefore we have measured the scattered intensity at a low field ( 0.2 t ) where the influence of thermal fluctuations should be small .
the result can be seen in fig .
[ tdep ] , where the intensity at 0.2 t is given divided by the expectation of the 3d xy model .
as the figure shows , the approximation of the 3d xy model , with @xmath46 ( 1 - t / t@xmath4)@xmath47 is valid for temperatures well below t@xmath4 in all of the region of interest . in the determination of the order parameter , the intensity was thus divided by ( 1 - t / t@xmath4)@xmath48 for simplicity .
furthermore the insert of fig .
[ tdep ] shows the temperature dependence over the full range .
there we have plotted the square root of the scattered intensity to present the familiar @xmath49 dependence @xcite . it can be seen that at lower temperatures the dependence is linear down to temperatures of @xmath134 k , below which it saturates . for a d - wave superconductor ,
as the hts are supposed to be , a linear temperature dependence would be expected down to the lowest temperatures from simple london theory .
however when non - local interactions are taken into account @xcite , a saturation at very low temperatures is predicted , not inconsistent with our data .
we will come back to the temperature dependence of @xmath18 below in the discussion of the melting line .
however , we note that such a big range of applicability ( down to @xmath1320 k below t@xmath4 ) for the 3d xy model is surprising , as this is thought only to be valid in the region of critical fluctuations in close vicinity of t@xmath4 . in fig .
[ sansmelt ] , we present the temperature dependence of the fll order parameter in the vicinity of t@xmath4 for two different fields . at both fields ,
the order parameter is constant at low temperature and decreases gradually towards the melting temperature .
when the order parameter finally drops to zero , the change is very sharp occurring over a temperature range of less than 1 k. the gradual decrease of the order parameter can be modeled by including a debye - waller factor into the theoretical description of the scattered intensity . the final ` jump ' however , is a sign of the transition being of first order .
the width of the transition of 1 k is not much bigger than that observed at t@xmath4 , which may be due to sample inhomogeneities .
such inhomogeneities leading to a distribution of t@xmath4 may also lead to a distribution of t@xmath11 which would probably be of the same order of magnitude .
furthermore , the distribution of fields inside the sample is inhomogenious @xcite , as we will also see in the discussion of the @xmath9sr results below , which will also lead to a further smearing of the melting temperatures .
however , as this macroscopic field distribution is very narrow compared to the applied fields , this should nt be a big factor .
it is therefore not unreasonable to expect the melting transition in such a big sample to be distributed over an interval of @xmath131 k. note also that the order parameter ( the scattered intensity ) decreases to zero above the melting transition , in agreement with the expectation for an entangled vortex liquid @xcite made up of two - dimensional ` pancake ' vortices .
the situation for the untwinned sample has to be compared to that in a heavily twinned sample @xcite .
this is done in fig .
[ comp ] , where we show the fll order parameter for both a twinned and our untwinned sample in a field of 4 t. as can be seen in the figure , the transition is much sharper in the untwinned one .
the order parameter decreases more or less linearly over a range of 5 k in the twinned sample , as would be expected from mean field theory . in the untwinned sample in contrast ,
the transition is at most 1 k wide .
takeing also into account the temperature dependence at lower temperature , where the untwinned sample varies much slower than the twinned one , the behaviour of the untwinned sample is consistent with a first order transition .
the width of the observed transition is furthermore not very much bigger than those determined from magnetisation@xcite and thermal measurements@xcite , which are done on much smaller samples where inhomogeneities and possible residual twin - planes must play a much smaller role .
we now turn to the discussion of the @xmath9sr results .
typical field probability distributions for temperatures above and below the melting transition can be seen in fig .
[ lineshape ] for a field of 0.6 t. the situation is similar to that of bscco @xcite , with a lineshape highly asymmetric towards high fields below the melting temperature .
as was discussed above , the lineshape for an ideal lattice made up of straight vortex lines is expected to show such a high field tail . above the melting temperature
however , the field distribution is skewed the opposite way .
this has been argued to arise from a macroscopic distribution of the local fields due to the sharp edges of the sample , with fields decreasing at the edges @xcite .
this kind of macroscopic distribution is also present below the melting temperature , however the large width of the field distribution due to the well formed lattice overcomes this effect , such that an almost ideal distribution is observed . above the melting temperature in contrast ,
the field distribution is smeared by the fast thermal fluctuations of the pancake vortices .
this leads to a strongly decreases the width of the microscopic field distribution ( ` motional narrowing ' ) and hence the observed distribution follows the macroscopic one . in fig .
[ musrmelt ] , we present the temperature dependence of the @xmath9sr lineshape asymmetry parameter @xmath28 , together with that of the mean internal field .
the mean field allows a direct measurement of t@xmath4 via the onset of diamagnetic behaviour .
as can be seen in the figure even at the very low field of 0.3 t , we observe a sharp melting transition in the temperature dependence of @xmath28 somewhat below t@xmath4 . the change in lineshape displayed in fig .
[ lineshape ] manifests itself by a change in sign of @xmath28 .
the transition is still very sharp , although thermal measurements are not sensitive enough to resolve a first order transition at such low fields .
there have however recently been magnetisation measurements by means of torque magnetometry , resolving a step in magnetisation and hence a first order transition at similarly low fields @xcite . the field dependence of the melting temperatures thus determined ( the melting line )
can be seen in fig .
[ meltline ] . in that figure
we have also included the line of onset of irreversible magnetisation ( irreversibility line , il ) as determined with a vsm . the microscopic melting line and the il are in good agreement with each other .
they follow a power law dependence of the form @xmath50 a least squares fit to both sets of data simultaneously yields the values of @xmath51 = 100(10 ) t and n = 1.30(5 ) .
this is in excellent agreement with many other studies @xcite on both twinned and untwinned samples .
it is thus only the order of the transition , which is affected by the presence of disorder ( e.g. in the form of twin - planes ) and not the meltingline itself .
the shape of the melting line has been the ground of some debate , as a dependence @xmath52 ( 1 - t / t@xmath4)@xmath48 is expected for the vortex - glass transition proposed by fisher _
this would indicate that the observed melting line is more likely to be a glass transition , which would however be in contradiction with the observation of a first order transition .
the temperature dependence of a melting line is usually calculated using the lindemann criterion that melting occurs when the mean square displacements of flux lines make up a certain fraction of the intervortex separation @xmath53 .
the lindemann number @xmath54 is supposed to be a universal constant in all kinds of melting processes and theoretical estimates indicate it to be 0.1 - 0.2 . using this criterion and modeling the thermal fluctuations in the vortex lattice one
can then derive an expression for the melting line @xcite @xmath55 due to the mean field assumption of the temperature dependence of the penetration depth @xmath49 ( t ) @xmath56 ( 1 - t / t@xmath4)@xmath57 this expression is usually taken to result in a value of n = 2 . as we have seen however in fig .
[ tdep ] , @xmath18 follows a powerlaw , such that the value of n = 1.33 is recovered over the range of temperatures where the meltingline can be determined .
furthermore , this expression allows a determination of @xmath51 , where @xmath54 is the only adjustable parameter . from the intensity in the bragg peaks
, we have an independent determination of @xmath18 = 138 nm @xcite and the anisotropy of our sample has been determined from the angular dependence of the @xmath9sr linewidth as well as from the distortion of the sans pattern at high angles to the @xmath3-direction
. therefore our value of @xmath51 = 100(10 ) t corresponds to a lindemann number of @xmath54 = 0.173(7 ) .
this is in excellent agreement with that found from the melting line in bscco using @xmath9sr @xcite .
we would like to note at this point that the lindemann number determined in bscco relied on a different expression for the thermal fluctuations taking into account the electromagnetic coupling between pancake vortices .
this indicates that the lindemann criterion may indeed be universally applicable .
finally , we discuss the dependence of the melting transition on the angle of the applied field to the @xmath3-direction . from the anisotropic scaling theory of blatter _
@xcite , a dependence of the melting line on angle is @xmath58 in the region of angles we investigated ( @xmath59 50@xmath12 ) , the influence of @xmath60 on @xmath61 is very small such that it can not be accurately determined this way . however using the value obtained from @xmath9sr and the distortion of the pattern @xcite
, we may calculate the expected melting lines at an angle from that determined parallel to @xmath3 .
this has been done in fig .
[ anglemelt ] , where we show the melting temperatures determined by sans at the three different angles of 0@xmath12 , 33@xmath12 and 55@xmath12 .
the lines correspond to the fit to the melting line discussed above , where the dependence on the angle has been taken into account as well .
as can be seen in the figure , the data are in good agreement with the expectations of the anisotropic scaling theory .
this is similar to the results of dta measurements of the melting transition as a function of angle @xcite .
in addition , that work also provided a determination of the entropy change at the transition as a function of angle . due to a cancellation of the angular dependencies in the melting line and the magnetisation jump
, the entropy - change turns out to be independent of the angle .
this is not inconsistent with the sans result displayed in fig .
[ angledep ] , where the temperature dependence of the fll order parameter is shown at the melting transition for different angles .
the temperatures are normalised to t@xmath11 for direct comparison and the applied field was 5 t. the decrease of the order parameter can be seen to be independent of the angle as it should be for a constant entropy change at the transition .
in conclusion , we have presented sans and @xmath9sr measurements of the melting of the vortex lattice in a sample of untwinned ybco .
the observed behaviour is consistent with dta measurements of the latent heat showing the existence of a first order transition @xcite .
this is inferred from a phenomenological fll order parameter in sans , which decreases to zero sharply at the melting transition .
these findings are compared to a previous sans study of melting in twinned samples @xcite .
the same fll order parameter vanishes continuously in that case , consistent with a second order phase transition .
we have furthermore studied the effects of anisotropy by applying the field at an angle to the @xmath3-direction .
the melting temperatures at constant field are consistent with the expectations of the scaling theory of blatter _
a determination of the anisotropy @xmath60 however is not possible , as the deviations of a cos ( @xmath62 ) dependence are negligible in the range of angles studied .
the nature of the transition is unchanged by having the field at an angle , consistent with the findings of schilling _ et al . _
@xcite , who studied the dependence of the entropy change on angle .
the temperature dependence of the melting line is consistent with that expected from a lindemann criterion @xcite , where the temperature dependence of the penetration depth follows that of the 3d xy model .
the observed melting line extends down to very low fields ( 0.2 t ) , where @xmath9sr still indicates a very sharp transition .
the range of applicability of this temperature dependence is limited , as we have shown from the temperature dependence of the scattered intensity at low fields , where thermal fluctuations do not play an important role . for temperatures below 75
k the approximation becomes invalid .
we would like to note that this is a rather unexpectedly big range of applicability of the critical exponent of the 3d xy model . using the independently measured penetration depth @xcite and anisotropy @xcite for our sample
, we have also determined the lindemann number from the melting line , which turns out to be c@xmath63 = 0.173(7 ) .
this is in very good agreement with the lindemann number previously obtained from a detailed @xmath9sr study of the melting line in bscco @xcite .
we would like to thank michel bonnaud , andreas polsak , pierre george ( ill ) and alex amato ( psi ) for technical support . financial support from the epsrc of the uk and the swiss national science foundation is gratefully acknowledged .
this work was funded in part by oak ridge national laboratory , which is managed by lockheed martin energy research corporation under contract no
. de - ac05 - 96or22464 for the us doe .
10 present address : division of physics and astronomy , faculty of sciences , vrije universiteit , de boelelaan 1081 , 1081hv amsterdam , the netherlands .
present address : laboratoire de physique des solides , btiment 510 , universit paris - sud , 91405 orsay , france .
d. s. fisher , m. p. a. fisher , and d. a. huse phys . rev .
b * 43 * , 130 ( 1991 ) .
s. l. lee , p. zimmermann , h. keller , m. warden , i. m. savi , r. schauwecker , d. zech , r. cubitt , e. m. forgan , p. h. kes , t. w. li , a. a. menovsky , z. tarnawski , phys .
71 * , 3862 ( 1993 ) .
a. schilling , r. a. fisher , n. e. phillips , u. welp , d. dasgupta , w. k. kwok , g. w. crabtree , nature * 382 * , 791 ( 1996 ) and phys .
. lett . * 78 * , 4833 ( 1997 ) .
m. roulin , a. junod , e. walker , science * 273 * , 1210 ( 1996 ) .
g. blatter , m. v. feigelman , v. b. geshkenbein , a. i. larkin , v. m. vinokur , rev .
* 66 * , 1125 ( 1995 ) .
r. cubitt , e. m. forgan , g. yang , s. l. lee , d. mck .
paul , h. a. mook , m. yethiraj , p. h. kes , t. w. li , a. a. menovsky , z. tarnawski , k. mortensen , nature * 365 * , 407 ( 1993 ) .
a. junod , m. roulin , j .- y .
genoud , b. revaz , a. erb , e. walker , physica c * 275 * , 245 ( 1997 ) .
a. schilling , r. a. fisher , n. e. phillips , u. welp , w. k. kwok , and g. w. crabtree , phys .
lett . * 78 * , 4833 ( 1997 ) .
g. blatter , v. b. geshkenbein , and a. i. larkin , phys .
lett . * 68 * , 875 ( 1992 ) .
m. j. w. dodgson , v. b. geshkenbein , h. nordborg , and g. blatter , phys .
lett . * 80 * , 837 ( 1998 ) ; phys .
b * 57 * , 14498 ( 1998 ) . c. m. aegerter , s. t. johnson , w. j. nuttall , s. h. lloyd , m. t. wylie , m. p. nutley , r. cubitt , s. l. lee , d. mck .
paul , m. yethiraj , and h. mook , phys .
b * 57 * , 14511 ( 1998 ) .
m. roulin , a. junod , e. walker , physica c * 296 * , 137 ( 1998 ) .
a. i. rykov _
et al . _ in advances in superconductivity viii ( proc .
iss95 ) , eds .
h. hayakawa and y enomoto ( springer , tokyo , 1996 ) p. 341 .
b. d. rainford , g. j. daniell , hyperfine interactions * 87 * , 1129 ( 1994 ) .
e. m. forgan , d. mck .
paul , h. a. mook , p. a. timmins , h. keller , s. sutton , j. s. abell , nature * 343 * , 735 ( 1990 ) .
d. cribier , b. jacrot , l. madhov rao , b. farnoux , phys .
* 9 * , 106 ( 1964 ) .
s. t. johnson , e. m. forgan , s. h. lloyd , s. l. lee , r. cubitt , p. g. kealey , c. ager , s. tajima , a. rykov , and d. mck .
paul , phys .
* 82 * , 2792 ( 1999 ) . c. ager , f. y. ogrin , s. l. lee , c. m. aegerter , s. romer , h. keller , i .m . savi , s. h. lloyd , s. t. johnson , e. m. forgan , t. m. riseman , p. g. kealey , s. tajima , a. rykov , phys .
b * 62 * , 3528 ( 2000 ) .
d. k. christen , f. tasset , s. spooner , h. a. mook , phys .
b * 15 * , 4506 ( 1977 ) .
s. kamal , d. a. bonn , n. goldenfeld , p. j. hirschfeld , r. liang , w. n. hardy , phys .
* 73 * , 1845 ( 1994 ) .
i. kosztin and a. j. leggett , phys .
lett . * 79 * , 135 ( 1997 ) .
m. v. indenbom , h. kronmller , t. w. li , p. h. kes , a. a. menovsky , physica c * 222 * , 203 ( 1994 ) ; m. v. indenbom , g. danna , m .- o .
andr , v. v. kabanov , w. benoit physica c * 235 - 240 * , 201 ( 1994 ) .
h safar , e. rodriuez , f. de la cruz , p. l. gammel , l. f. schneemeyer , d. j. bishop , phys .
b * 46 * , 14238 ( 1992 ) .
h. nordborg and g. blatter , phys .
lett . * 79 * , 1925 ( 1997 ) .
u. welp , w. k. kwok , g. w. crabtree , phys .
76 * , 4809 ( 1996 ) .
s. l. lee , c. m. aegerter , h. keller , m. willemin , b. stuble - pmpin , e. m. forgan , s. h. lloyd , g. blatter , r. cubitt , t. w. li , p. kes , phys .
b * 55 * , 5666 ( 1997 ) .
e. m. forgan , m. t. wylie , s. lloyd , m. p. nutley , s. l. lee , r. cubitt , c. aegerter , h. keller , and t. w. li , hyperfine interact . *
105 * , 61 - 69 ( 1997 ) .
m. willemin , a. schilling , h. keller , c. rossel , j. hofer , u. welp , w. k. kwok , r. j. olsson , and g. w. crabtree phys .
rev . lett . * 81 * 4236 ( 1998 ) .
a. houghton , r. a. pelcovits , a. sudb , phys .
b * 40 * , 6763 ( 1989 ) . | we present a study of the vortex lattice in untwinned yba@xmath0cu@xmath1o@xmath2 crystals , using a combination of muon spin rotation and neutron small angle scattering measurements .
both methods show a very sharp melting temperature consistent with a first order transition . the dependence of the melting temperature on the angle of the field with respect to the crystallographic @xmath3-axis is studied .
the results are compared to thermal measurements .
-1 cm 2 |
heavy quarkonium production remains a challenging problem for the understanding of quantum chromodynamics ( qcd ) . at the centre - of mass energies for proton - proton collisions at the large hadron collider ,
@xmath0 pairs are expected to be produced predominantly via leading order gluon gluon interactions , which can be computed using perturbative qcd , followed by the formation of the bound charmonium states described by non perturbative models .
recent approaches make use of non - relativistic qcd factorization ( nrqcd ) which assumes a combination of the colour - singlet and colour - octet @xmath0 as it evolves towards the final bound state via the exchange of soft gluons @xcite . next - to - leading order ( nlo ) qcd corrections in charmonium and bottomonium production
are also essential for the description of the experimental data @xcite .
studies of @xmath7 , @xmath8 and @xmath9 production cross sections are presented .
the @xmath4 meson is an exotic meson discovered in 2003 by the belle collaboration in the @xmath10 , @xmath11 decay chain @xcite .
its existence was confirmed by the cdf @xcite , d0 @xcite and babar @xcite collaborations . the @xmath4 particle , together with several other new states subsequently observed in the mass range 3.9 - 4.7 gev / c@xmath12 , has led to a resurgence of interest in exotic meson spectroscopy @xcite . several properties of the @xmath4
have been measured ; however , its nature is still uncertain and several models have been proposed .
first , it is not excluded that the @xmath4 is a conventional charmonium state with one candidate being the @xmath13 meson @xcite .
however , the mass of this state is predicted to be far below the observed @xmath4 mass .
given the proximity of the @xmath4 mass to the @xmath14 threshold , one possibility is that the @xmath4 is a loosely bound deuteron - like @xmath14 molecule , i.e. a ( ( uc ) - ( cu ) ) system @xcite .
another more exotic possibility is that the @xmath4 is a tetraquark state @xcite .
the cdf experiment has reported @xmath15 evidence for @xmath5 state ( also referred to as @xmath16 in the literature ) in 6.0 fb@xmath6 of @xmath17 data collected at tevatron @xcite .
the relative rate was measured to be @xmath18 .
this observation has triggered wide interest among model builders of exotic hadronic states .
it has been suggested that the @xmath5 could be a molecular state @xcite , a tetraquark state @xcite , a hybrid state @xcite or a re - scattering effect @xcite .
two decay modes of the @xmath19 meson have been studied @xcite : @xmath20 and @xmath21 .
the differential cross - section for the inclusive @xmath7 production is computed as @xmath22 where @xmath23 is the number of observed @xmath19 decays , @xmath24 is the integrated luminosity , @xmath25 is the total detection efficiency including acceptance effects , @xmath26 is @xmath27 for @xmath21 and the dielectron branching ratio ( @xmath28 ) for @xmath20 and @xmath29 1 gev / c is the bin size .
the dielectron branching ratio is used , assuming lepton universality , since it has a much smaller error than the dimuon one ( 2.2% and 10% respectively ) . in order to estimate the number of @xmath7 signal events ,
a fit is performed independently in each @xmath30 bin . with the predictions for prompt production by a nlo nrqcd model @xcite .
lhcb data include also @xmath7 from b.,width=302 ] in fig .
[ fig : psi2s_1 ] a comparison between the measurements presented of the differential @xmath7 production cross - section multiplied by the @xmath20 branching ratio and a recent theory prediction @xcite for prompt @xmath7 in the rapidity range @xmath31 is shown . here
@xmath32 and @xmath33 are the @xmath19 energy and momentum in the @xmath34 direction measured in the pp centre - of - mass frame ; the z - axis is defined along the beam axis in the lhcb frame , oriented from the velo to the muon detector .
the differential cross - section in @xcite has been evaluated for the @xmath7 prompt production at the lhc at next - to - leading order in non - relativistic qcd , including color - singlet and color - octet contributions .
the integrated cross - section in the full range of @xmath35 and @xmath36 respectively to @xmath37 and @xmath38 is found to be : @xmath39 where the first uncertainty is statistical , the second is systematic and the third is the uncertainty due to the unknown polarization .
@xmath8 meson has been studied in the decay mode @xmath40 @xcite .
the double differential cross - section for the inclusive @xmath8 production is computed as @xmath41 where @xmath42 is the number of observed @xmath40 decays , @xmath43 is the total detection efficiency including acceptance effects , @xmath24 is the integrated luminosity , @xmath44 is the branching fraction , and @xmath45 gev / c is the rapidity and @xmath35 bin sizes . in order to estimate the number of @xmath8 signal events , a fit is performed independently in each of the 15 @xmath35 times 5 @xmath36 bins .
the double differential cross - section as a function of @xmath35 and @xmath36 is shown in fig .
[ fig : y1s_1 ] .
the integrated cross - section in the full range of @xmath36 and @xmath35 is found to be @xmath46 where the first uncertainty is statistical , and the second systematic .
the latter includes @xmath47 nb from the unknown polarization , @xmath48 nb from the luminosity determination and @xmath49 nb from other sources .
the integrated cross - section is about a factor 100 smaller than the integrated @xmath50 cross - section in the identical @xmath36 and @xmath35 region @xcite , and a factor three smaller than the integrated @xmath8 cross - section in the central region @xmath51 as measured by cms @xcite .
production cross - section as a function of @xmath35 in bins of rapidity.,width=302 ] the cms and the lhcb measurements agree ; the difference in the integrated cross section results is due to the different rapidity ranges .
[ fig : y1s_2 ] shows the comparison of the lhcb cross - section measurement in bins of @xmath36 integrated over @xmath35 with the same measurement of cms , in the @xmath36 ranges covered by the two experiments .
production cross - section as a function of @xmath36 integrated over @xmath35 as measured by the cms @xcite and lhcb experiments.,width=302 ] the product of the inclusive production cross - section @xmath93 by the @xmath11 branching fraction is computed as @xcite @xmath94 where @xmath95 is the efficiency - corrected yield of @xmath96 signal decays , @xmath97 is a multiplicative factor to the efficiency that accounts for known differences between the data and simulation , @xmath24 is the integrated luminosity , and the notation @xmath98 is used for branching fractions .
the @xmath4 signal yield is determined from an extended , unbinned maximum likelihood fit of the reconstructed @xmath79 mass in the interval @xmath99 gev / c@xmath12 .
the fitting function is the same used for the @xmath4 mass measurement .
the mass resolution is fixed in this analysis to @xmath100 where @xmath101 mev / c@xmath12 is the mass resolution fitted in data for the kinematically similar decay @xmath75 , and where @xmath102 and @xmath103 are the mass resolutions determined from fully simulated monte carlo events . as in section [ sec_xmass_res ] , the natural width is fixed to 1.3 mev / c@xmath12 .
the effect of fixing parameters that are not perfectly known is investigated as part of the systematic studies .
the @xmath4 signal yield is approximately @xmath104 of that reported in @xcite due to the additional trigger requirements and fiducial cuts on @xmath36 and @xmath35 . applying the procedure discussed above , and using the @xmath105 and @xmath106dependent efficiency from simulation ,
the efficiency - corrected signal yield is @xmath107 , where the quoted uncertainty is statistical .
two factors enter into the determination of the correction factor @xmath97 .
the first is a factor @xmath108 which accounts for differences in the efficiency of the muon identification in the data and simulation .
the second is a factor of @xmath109 that accounts for observed differences in the efficiency of global event cuts applied in the trigger .
multiplying these values together gives @xmath110 .
the uncertainty on this number is taken into account in the estimation of the systematic uncertainty .
the preliminary measurement is @xmath111 where @xmath93 is the cross - section for producing an @xmath4 particle in @xmath2 collisions at @xmath3 tev ( either promptly or from the decay of other particles ) with a transverse momentum between 5 and 20 gev / c and a rapidity between 2.5 and 4.5 .
@xmath52 and @xmath53 have been reconstructed in the final state @xmath54
. the production cross - section ratio of the @xmath53 and @xmath52 states is measured @xcite using @xmath55 where @xmath56 and @xmath57 are the @xmath52 and @xmath53 branching ratios to the final state @xmath54 , @xmath58 ( @xmath59 ) is the efficiency to reconstruct and select a @xmath50 ( @xmath60 ) from @xmath61 decay and @xmath62 is the efficiency to select the @xmath61 candidate .
the measurement method consists of extracting the two @xmath63 yields ( for @xmath52 and @xmath53 ) from an unbinned maximum likelihood fit to @xmath64 mass difference distribution . with the ratio of the resolution parameters and the mass differences fixed
, a fit is then performed to the data in the full @xmath65 range @xmath66 $ ] gev / c in order to extract the resolution scale @xmath67 . here , the sample is subdivided into candidates with converted ( after the magnet ) and non - converted photons , in order to account for the different calorimeter resolution in the two cases .
converted photons are identified by the presence of activity in the scintillator pad detector at the entrance to the calorimeter system .
photons that convert before the magnet have a low probability to be reconstructed , because either one or both electrons are swept out of the detector acceptance by the magnetic field , and are not considered here .
the resolution scales , @xmath67 , are measured to be @xmath68 mev / c@xmath12 and @xmath69 mev / c@xmath12 for converted and non converted candidates , respectively .
the fit is then performed in bins of @xmath65 . for each bin
the value of @xmath67 is fixed to the value extracted from the fit to the full range @xmath65 .
the presence of @xmath61 polarized states would modify the efficiencies calculated from the monte carlo . in order to take into account possible polarization scenarios ,
the relevant combination of weights has to be taken into account .
the results from the not - converted and converted samples are combined by sampling the combination of the statistical and uncorrelated systematic uncertainties using a toy monte carlo .
finally , the correlated systematic uncertainty from the branching ratio is calculated using the combined central values for @xmath70 .
the preliminary result for the ratio of the prompt @xmath53 to @xmath52 production cross - sections as a function of @xmath65 is given in fig .
[ fig : chic_1 ] . comparisons to the theory predictions from the chigen mc generator @xcite and from the nlo nrqcd calculations @xcite , in the rapidity range [ 2 ; 4.5 ] , are also shown in the figure .
[ fig : chic_1 ] also shows the maximum effect of the unknown @xmath61 polarization on the result , shown as a black shaded area around the data points .
the upper limit of the shaded area corresponds to the spin state ( @xmath52 : @xmath71 = 1 ; @xmath53 : @xmath71 = 2 ) and the lower limit corresponds to the spin state ( @xmath52 : @xmath71 = 0 ; @xmath53 : @xmath71 = 0 ) .
the results are broadly in agreement at high @xmath65 with the colour singlet model , however , they are not yet precise enough to rule out sizeable colour octet terms
. there are indications of a discrepancy in the mid to low @xmath65 region
. this may be explained by a more complete modeling of the transition from the high to low @xmath65 regions , which is sensitive to non - perturbative effects and/or sizeable higher - order perturbative corrections . in bins of in the range of @xmath65 : [ 3 ;
15 ] gev / c .
the internal error bars correspond to the statistical error on the @xmath52 and @xmath53 yields ; the external error bars include the contribution from the systematic uncertainties ( apart from the polarization ) .
the shaded area around the data points ( black ) shows the maximum effect of the unknown @xmath9 polarization on the result .
the two other bands correspond to the chigen mc generator theoretical prediction @xcite ( in blue ) and nlo nrqcd @xcite ( in red).,width=302 ]
inclusively produced @xmath4 mesons are reconstructed and selected in the @xmath11 , @xmath72 decay mode @xcite .
a momentum scale calibration is performed to account for a mixture of effects related to imperfections in the knowledge of the magnetic field map and of the alignment of the tracking system .
the momentum scale is calibrated using a large sample of @xmath72 decays .
this calibration , gives an overall scale factor that is applied to all measurements of particle momenta .
the calibration is checked to be also valid for two - body decays of the @xmath8 , @xmath73 and @xmath74 .
in addition , its effect on @xmath75 , which have kinematics similar to the @xmath11 , is studied .
the uncalibrated value of the @xmath7 mass in data is @xmath76 mev / c@xmath12 .
after the momentum scale calibration it becomes @xmath77 mev / c@xmath12 , in good agreement with the pdg value of @xmath78 mev / c@xmath12 .
the measured @xmath50 mass after alignment and calibration is checked to be stable over the whole 2010 data - taking period .
the masses of the @xmath7 and @xmath4 are determined from an extended unbinned maximum likelihood fit of the reconstructed @xmath79 mass in the interval @xmath80 gev / c@xmath12 .
the functional form of the background is studied using the same - sign pion events .
the @xmath7 and @xmath4 signals are each described with a voigt function defined as the convolution of a non - relativistic breit - wigner with a gaussian function .
the intrinsic width of the @xmath7 is fixed to the pdg value , @xmath81 mev / c@xmath12 .
the @xmath4 intrinsic width is poorly known .
the babar @xcite and belle @xcite collaborations have published @xmath82 confidence level limits of @xmath83 mev / c@xmath12 and @xmath84 mev / c@xmath12 , respectively . the results of the belle and babar analyses have been combined by cdf to yield @xmath85 mev / c@xmath12 @xcite . for these studies , the procedure adopted in @xcite has been adopted and the natural width has been fixed to this value in the default fit .
[ fig : xmass_1 ] shows the invariant @xmath79 mass distributions for opposite - sign ( black points ) and same - sign ( blue filled histogram ) candidates .
clear signals for both the @xmath7 and the @xmath4 can be seen . from a comparison of fits with and without the @xmath4 component the statistical significance of the @xmath4 signal
is estimated to be @xmath86 .
the uncertainties on the parameters reported by the fit are in good agreement with expectations based on toy monte carlo studies .
the preliminary lhcb result @xmath87 is in good agreement with the published measurements @xcite and with their average , @xmath88 mev / c@xmath12 .
the new world average value including lhcb measurement , @xmath89 mev / c@xmath12 , is consistent within uncertainties with the sum of the @xmath73 and @xmath90 masses , @xmath91 mev / c@xmath12 , computed from the results of the global pdg fit of the charm meson masses .
( black points with statistical error bars ) and same - sign @xmath92 ( blue filled histogram ) candidates .
the red curve is the result of the fit described in the text .
the insert shows a zoom of the region around the @xmath4 mass.,width=302 ]
the @xmath5 has been studied using a data sample of approximately 0.376 fb@xmath6 @xcite selecting @xmath112 , @xmath113 .
no narrow structure has been seen near the threshold in @xmath114 distribution as shown in fig .
[ fig : x4140_1 ] . in the cdf analysis
, they fit their data with spin 0 relativistic breit - wigner function on top of three - body phase - space , all smeared with the detector resolution @xcite . to quantify the disagreement with cdf
the same function has been used .
the efficiency dependence is extracted from the mc simulations and applied as a correction to the three - body phase - space function .
mass and width of @xmath5 peak are fixed to the values obtained by the cdf collaboration .
the mass difference resolution is determined from the @xmath115 mc simulation .
the fit to @xmath114 distribution gives a @xmath5 amplitude of @xmath116 events ( fig .
[ fig : x4140_1]a ) and has a confidence level ( cl ) of @xmath117 . when using a quadratic polynomial instead of three - body phase - space function for the background , the preferred value of the @xmath5 amplitude is zero ( it is restricted not to go below zero ) , with a positive error of 3 events .
this fit is shown in fig .
[ fig : x4140_1]b and has a confidence level of @xmath118 . using the @xmath112 yield multiplied by this efficiency ratio ( @xmath115 , @xmath119)/(@xmath112 ) and multiplied by the cdf value for @xmath120 @xcite , leads to a prediction of observed @xmath121 events , where the first uncertainty is statistical from the cdf data and the second is systematic including both cdf and lhcb contributions .
the central value of this estimate is illustrated in fig .
[ fig : x4140_1 ] .
the cdf result disagrees by three standard deviations with the fit to lhcb data using the polynomial background .
the disagreement is 2.4 @xmath122 when the efficiency - corrected three - body phase - space background shape is used .
signal on top of smooth background to @xmath114 for the @xmath112 data .
the solid line represents the result of the fit to lhcb data .
the dashed line on top illustrates the expected signal amplitude from the cdf results as explained in the text .
the top and bottom plots differ by the type of the background function used in the fit : a ) efficiency - shaped three - body phase - spaced ; b ) quadratic polynomial.,width=302 ]
exclusive particle production in proton - proton collisions are elastic processes in which the protons remain intact , @xmath123 , and the additional particles are created through photon and/or gluon propagators . in the case of purely photon propagators , this is an electromagnetic process which can be theoretically calculated with high accuracy .
when gluons are involved , these processes constitute an important testing ground for qcd , since the object that couples to the proton must be colourless .
thus the pomeron ( two gluon states ) or possibly an odderon ( three gluons ) , predicted in qcd but never unambiguously observed , can be studied in a clean experimental environment .
the cleanest experimental manifestation of these occur in final states containing two protons and two muons or in events containing two protons , two muons and a photon .
the former can be produced in the diphoton process giving a continuous dimuon invariant mass spectrum , or in the photon - pomeron process which can produce @xmath124 , @xmath50 , @xmath7 , @xmath8 , @xmath125 or @xmath126 which decay to two muons .
the latter is a signal for double pomeron exchange which produces @xmath9(@xmath127 ) that decay to @xmath50 ( @xmath128 ) plus a photon .
the final state protons are only marginally deflected , go down the beam - pipe , and remain undetected .
the experimental signal therefore in lhcb is a completely empty event except for two muons and possibly a photon .
however , because lhcb is not hermetic , there will be sizeable backgrounds from non - elastic processes where the other particles travel outside the detector acceptance .
the cross - section @xmath122 , is calculated @xcite from the number , @xmath129 , of selected events having corrected for efficiency , @xmath130 , and purity , @xmath131 , and dividing by the luminosity , @xmath132 , via @xmath133 .
the efficiency for selecting the events has been determined from simulation .
the measured cross - sections are : @xmath134 where the first uncertainty is statistical , the second is systematic , and the third comes from the estimate of the luminosity .
note that these numbers are cross - section times the branching ratio into the final state of interest , and all final state particles are required to be between pseudorapidities of 2 and 4.5 .
the cross sections are quoted in a limited pseudorapidity range as the models under consideration have pseudorapidity dependence .
lhcb performed many analysis of the quarkonium states using 2010 collected data ( 35 pb@xmath6 ) .
the measurement of the production cross sections of charmonium and bottomonium states ( @xmath7 , @xmath8 , @xmath9 ) are useful to test theoretical models . for the exotic states ,
the measurement of the @xmath4 mass and cross section has been performed and the cdf narrow @xmath5 has been studied using a dataset of 376 pb@xmath6 and its existence not confirmed .
lhcb has a very high @xmath50 statistic in 2011 data that will allow to have a lot of new results in the near future .
99 g. t. bodwin , et al .
phys.rev . , d51:11251171 , 1995 .
j. m. campbell , et al .
phys.rev.lett . , 98:252002 , 2007 .
ma , et al .
phys.rev . ,
d83:111503 , 2011 .
choi et al .
[ belle collaboration ] , phys .
91 ( 2003 ) 262001 , arxiv : hep - ex/0309032 .
d. acosta et al .
[ cdf collaboration ] , phys .
( 2004 ) 072001 , arxiv : hep - ex/0312021 . v. m. abazov et al .
[ d0 collaboration ] , phys .
93 ( 2004 ) 162002 , arxiv : hep - ex/0405004 .
b. aubert et al .
[ babar collaboration ] , phys . rev .
d 71 ( 2005 ) 071103 , arxiv : hep - ex/0406022 .
e. swanson , physics reports 429 ( 2006 ) 243 , arxiv : hep - ph/0601110 .
l. maiani , f. piccinini , a. d. polosa and v. riquer , phys . rev .
d 71 ( 2005 ) 014028 , arxiv : hep - ph/0412098 .
the cdf collaboration , t. aaltonen et al .
, arxiv:1101.6058 .
liu and s .- l .
zhu , phys.rev .
d80 ( 2009 ) 017502 , [ arxiv:0903.2529 ] .
wang , z .- c .
liu , and x .- h .
zhang , eur .
j. c64 ( 2009 ) 373386 , [ arxiv:0907.1467 ] .
f. stancu , j. phys .
g37 ( 2010 ) 075017 , [ arxiv:0906.2485 ] .
n. v. drenska , r. faccini , and a. d. polosa , phys .
d79 ( 2009 ) 077502 , [ arxiv:0902.2803 ] .
n. 130 mahajan , phys .
b679 ( 2009 ) 228230 , [ arxiv:0903.3107 ] .
wang , eur . phys .
j. c63 ( 2009 ) 115122 , [ arxiv:0903.5200 ] .
x. liu , phys .
b680 ( 2009 ) 137140 , [ arxiv:0904.0136 ] .
d. v. bugg , arxiv:1103.5363 .
lhcb - conf-2011 - 021 .
b. aubert et al .
[ babar collaboration ] , phys .
d 77 ( 2008 ) 111101 , arxiv:0803.2838 [ hep - ex ] . t. aaltonen et al .
[ cdf collaboration ] , phys .
103 ( 2009 ) 152001 , arxiv:0906.5218 [ hep - ex ] .
lhcb - conf-2011 - 043
. m. pivk and f.r .
le diberder , nucl .
meth . a 555 ( 2005 ) 356369 , arxiv : physics/0402083v3 . s .-
k . choi et al .
[ belle collaboration ] , arxiv:1107.0163v1 [ hep - ex ] , submitted to phys . | the production of @xmath0 and @xmath1 is studied in @xmath2 collisions at @xmath3 tev with the lhcb detector .
the results of these measurements are compared with different theoretical models .
results and prospects are also shown for exotics : the mass measurement of the @xmath4 and the search of the @xmath5 .
most of the presented results are based on the 2010 lhcb dataset ( 35 pb@xmath6 ) . |
it is currently admitted that processes of structure formation in the universe are mainly dominated at large scale by an invisible component called dark matter .
although the microscopic origin of dark matter is still unclear , it is macroscopically described as a self - gravitating collisionless fluid following the collisionless boltzmann or vlasov equation in a cosmological background , @xmath0 f({\mbox{\boldmath$x$}},{\mbox{\boldmath$v$}},t ) = 0 , \label{eq : vp_3d}\end{aligned}\ ] ] supplemented with the poisson equation for the newton potential @xmath1 , @xmath2 , \label{eq : poisson_3d}\end{aligned}\ ] ] where @xmath3 is the phase - space density at comoving position @xmath4 , peculiar velocity @xmath5 and time @xmath6 , @xmath7 is the expansion factor of the universe , @xmath1 is the gravitational potential and @xmath8 is the average dark matter density . in the standard picture of structure formation ,
dark matter was initially cold i.e. with a virtually null local velocity dispersion , so the six - dimensional phase - space distribution is effectively reduced to a three - dimensional hyper - surface and this remains true at all times thanks to the hamiltonian nature of the system . at early times ,
dark matter thus follows the single - stream flow regime , with a velocity field @xmath9 uniquely determined as a function of position , and its evolution is that of a pressure - less fluid with a phase - space distribution function given by @xmath10 , \label{eq : df_single - stream}\end{aligned}\ ] ] where @xmath11 is the density contrast of the dark matter distribution , initially of very small magnitude , as well as @xmath9 . substituting eq .
( [ eq : df_single - stream ] ) into eqs .
( [ eq : vp_3d ] ) and ( [ eq : poisson_3d ] ) and taking the zeroth and first velocity moments yields the eulerian formulation of large - scale structure dynamics @xmath12=0 , \label{eq : continuity}\\ & \frac{\partial{\overline { \mbox{\boldmath$v$}}}}{\partial t}+\frac{1}{a}\,\bigl\{{\overline { \mbox{\boldmath$v$}}}\cdot\nabla\bigr\}\overline{\mbox{\boldmath$v$}}=-\frac{1}{a}\,\nabla\phi , \label{eq : euler}\\ & \frac{1}{a^2}\nabla^2\phi=4\pi\,g\,\overline{\rho}_{\rm m}\,\delta .
\label{eq : poisson_v2}\end{aligned}\ ] ] although the single - stream flow is , strictly speaking , valid only during the early phase of structure formation , the above equations have been shown in practice to describe accurately nonlinear mode - coupling in the weakly nonlinear regime and provide a solid basis for perturbation theory ( pt ) calculations to predict statistical quantities of large - scale structure , such as the power spectrum or the two - point correlation function of the matter distribution ( see , e.g. * ? ? ?
with the advent of cosmological observations aiming at precisely mapping galaxy clustering at large scales , techniques have been developed to improve on the slow convergence of the perturbative expansion as well as to accelerate higher - order calculations ( e.g. , * ? ? ? * ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
. improved calculations involving the next - to - next - to - leading order called two - loop give a remarkable agreement with cosmological @xmath13-body simulations at weakly nonlinear scales and have been applied to observations .
however , this does not ensure the convergence of perturbative calculations including higher - order corrections , e.g. , three - loop .
in fact , a direct calculation at three - loop order suggests a very large uv contribution to the large - scale modes through nonlinear mode - coupling @xcite and indicates a break down of higher - order perturbative expansions even at large scales .
deficiency of pt calculations has been also highlighted in recent numerical analyses .
@xcite directly measured the coupling between different scales in cosmological @xmath13-body simulations and found that the actual contribution from small scales to the large - scale modes is suppressed , as opposed to the prediction of pt based on the single - stream approximation .
these facts imply that the validity of the single - stream treatment is questionable even at large scales , and higher - order perturbative corrections need to be remedied with a proper account of small - scale dynamics , where the multi - stream flow contribution is important . one way to account for small - scale dynamics consists of using some ansatz to summarize the main physical effects in the multi - stream regime , such as burgers equation ( e.g. , * ?
* ; * ? ? ?
* ) and various alternatives ( e.g. , * ? ? ?
* ) , which roughly amount to adding a source term in the right hand side of equation ( [ eq : euler ] ) . in particular , it was proposed recently to use effective fluid equations giving account of the non - vanishing stress tensor arising in eq .
( [ eq : euler ] ) when calculating the first velocity moment of vlasov equation in the multi - stream regime .
this _ effective - field theory _ approach has attracted a lot of interest and has been studied in detail ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
the drawback of this approach is however that the parameters in the stress tensor characterizing the small - scale dynamics need to be calibrated with @xmath13-body simulations to keep predictions of perturbative calculations under control .
furthermore , these parameters generally vary with cosmology and with time and no prediction with pt is really possible in this framework independently of @xmath13-body simulations .
an alternative approach that we consider in this paper consists in going back to a more fundamental description , i.e. the vlasov - poisson system , eqs .
( [ eq : vp_3d ] ) and ( [ eq : poisson_3d ] ) , and trying to follow accurately the phase - space distribution function in the multi - stream regime , which is essential for describing the formation of dark matter halos . for this purpose
, it is useful to employ lagrangian pt , where the small parameter is the displacement field .
we shall consider one - dimensional gravitational dynamics in standard ( three - dimensional ) cosmology . in this case ,
large - scale structure dynamics is described by the gravitational interaction of massive parallel infinite planes moving left and right along a fixed axis , while hubble expansion is taking place as usual in all the directions following standard friedman - lematre equations . despite its simplicity , one dimensional dynamics defined as such still displays a rich physical content which somewhat shares the same features as 3d clustering .
this is partly the reason why the 1d model has recently attracted much attention ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
in particular , the zeldovich solution provides an exact solution for the dynamics of massive sheets before shell - crossing , and thus , starting with zeldovich solution , a tractable perturbative treatment of multi - stream flows is made possible based on a lagrangian description .
the analysis in the present paper is an extension of the method developed in @xcite to the cosmological setup .
we will describe perturbatively post - collapse dynamics around the shell - crossing region and apply it to several cases including random initial conditions .
in addition , we shall present a novel regularization scheme reducing the impact of small - scale clustering and improving greatly our post - collapse pt predictions of large - scale structure statistics .
the idea is to apply adaptive smoothing to initial density peaks and to better capture the bulk properties of phase - space structures in the post - collapse regions , where the interaction or merger of halos is supposed to be significant .
the idea is similar to the peak - patch treatment proposed by @xcite and subsequent works ( see , e.g. * ? ? ?
* ) , but we here implement it in the pt prescription in order to better describe the late - time post - collapse dynamics . indeed , our perturbative approach will not allow us to follow post - collapse dynamics beyond next - crossing time , although an iterative prescription such as proposed by @xcite could make this possible but is out of the scope of the present work . here , we will show instead that the predictions with adaptive smoothing reproduce simulations remarkably well even at nonlinear scales , where zeldovich solution significantly deviates from simulations .
note finally that an analytical study in 1d of course represents only a first step toward a proper description of 6d phase - space dynamics .
apart from a few examples including self - similar solutions @xcite , little is indeed known analytically for the phase - space dynamics of vlasov - poisson systems , particularly in cosmology .
therefore , further development of analytical treatment in this framework seems indispensable , complementary to simulations , and even helpful to cross check simulation codes .
this is all the more motivated by the fact that simulations in 6d phase space have now become available thanks to recent efforts on the development of pure vlasov codes @xcite .
this paper is organized as follows . in
[ sec:1dcosmology ] , we begin by describing the basic setup of our calculations in one - dimensional cosmology .
we then discuss in
[ sec : postcollapse ] the analytic treatment beyond shell - crossing and develop post - collapse pt .
[ sec : improvement ] introduces a novel filtering scheme to improve the perturbative description of post - collapse dynamics , suited for a system with interacting clusters . in [ sec : comparison_nbody ] , analytic calculations in the post - collapse perturbative framework are tested against controlled @xmath13-body experiments .
several cases are considered , including the single initial sine wave and random initial conditions . finally ,
[ sec : conclusion ] is devoted to discussion and conclusion .
we consider an ensemble of massive parallel infinite planes moving along @xmath14 axis and interacting through gravitational force in the expanding universe .
the lagrangian equations of motion of the planes are given by @xmath15 where @xmath16 and @xmath17 are respectively the comoving position and peculiar velocity of each plane , @xmath18 the gravitational potential , @xmath8 the average matter density , @xmath11 the density contrast and @xmath7 the expansion factor of the universe .
to simplify the equations , it is useful to introduce the _ super - conformal time _
@xmath19 defined by ( e.g. , * ? ? ?
* ; * ? ? ?
* ) , @xmath20 the new velocity @xmath21 and potential @xmath1 : @xmath22 then , eqs .
( [ eq : eom0_x])([eq : eom0_delta ] ) are simplified : @xmath23 where @xmath24 is the matter density parameter of the universe and @xmath25 is the hubble constant . to deal with this system , especially for the dynamics after shell - crossing
, it is useful to introduce in the cold case considered here the lagrangian coordinate @xmath26 defining the initial position of the planes in the absence of perturbation @xmath27 and to express the subsequent position and velocity of each mass element as @xmath28 and @xmath29 . in particular
, mass conservation implies @xmath30\,{\rm d}x \,\,\longrightarrow\,\ , \delta(x)=\left(\frac{\partial x}{\partial q}\right)^{-1}-1 .
\label{eq : jacobian_delta}\end{aligned}\ ] ] this equation is valid in the single - stream regime , i.e. as long as @xmath31 remains monotonic as a function of @xmath26 . with the new expressions above
, the solution of the equations of motion can be formally written as : @xmath32 , \label{eq : sol1_p}\end{aligned}\ ] ] where @xmath33 and @xmath34 are the initial positions and velocities given at a starting time @xmath35 , which will be specified below . in
what follows , we consider the dynamics of the cosmological system in a finite - size box with periodic boundaries , @xmath36 .
the solution of poisson equation , eq .
( [ eq : eom1_delta ] ) , can be expressed in an integral form as : @xmath37\,\delta(x ' ) .
\label{eq : phi_poisson}\end{aligned}\ ] ] the derivation of this expression is presented in appendix [ sec : green_func ] .
then , the force exerted on a mass element at position @xmath14 is given by : @xmath38 , \label{eq : force_poisson}\end{aligned}\ ] ] where function @xmath39 represents the heaviside step function . in the above
, we used the fact that fluctuations averaged over space vanish , @xmath40 . in the limit @xmath41 , equation ( [ eq : force_poisson ] ) of course converges to the well - known result in the infinite space . in one - dimension and in the cold case
, zeldovich approximation is known to provide the exact solution for the dynamics of mass elements before shell - crossing . it can be explicitly written as @xmath42 here , function @xmath43 corresponds to the linear growth factor satisfying the following equation : @xmath44\,d_+(\tau)=0 .
\label{eq : evolv_lin}\end{aligned}\ ] ] note that in terms of cosmic time @xmath6 , eq . ( [ eq : evolv_lin ] ) reduces to the standard form of linear evolution equation : @xmath45\,d_+(t)=0.\end{aligned}\ ] ] the zeldovich solution in eq .
( [ eq : zeldovich_sol ] ) contains an arbitrary function @xmath46 that we call displacement field .
it is related to the linear density field @xmath47 given at a very early time ( @xmath48 or @xmath49 ) through @xmath50 since zeldovich solution is exact until shell - crossing , we do not necessarily assume that the evolved density field @xmath51 is small .
one may thus consider the situation where in a region around a lagrangian coordinate @xmath52 , the density field becomes large and the region will undergo shell - crossing at time @xmath53 .
recalling the fact that in 1d , a shell - crossing point corresponds to an inflection point of the mapping from lagrangian to eulerian space , the actual conditions for shell - crossing are given by @xmath54 at the time of shell - crossing , @xmath53 , the solution ( [ eq : zeldovich_sol ] ) around the shell - crossing region can be expanded as follows : @xmath55 using eq .
( [ eq : def_deltal ] ) , the conditions ( [ eq : condition_shellcrossing ] ) for shell - crossing can be rewritten as @xmath56 in other words , shell - crossing takes place at local density peaks .
hence , in the 1d case , the conditions for shell - crossing are equivalent to peak constraints and collapse take places exactly when the eulerian linear density contrast at the peak positions becomes equal to unity .
the structure of the system around the density peak is symmetric with respect to the lagrangian coordinate @xmath57 , where @xmath52 is the lagrangian position of the shell - crossing point ( see also * ? ? ?
[ fig : shellcrossing],width=321 ] we are interested in the dynamics of mass elements after shell - crossing , when the zeldovich solution is no longer valid . in this section , extending the work of @xcite , we develop perturbative calculations to deal with the multi - stream motion around the shell - crossing region .
the basic formalism to treat post - collapse dynamics can be described as follows . starting with the cold initial conditions in sec .
[ sec : initial ] , we first follow pre - collapse dynamics with the exact zeldovich solution . then , in regions undergoing shell - crossing , we switch to a perturbative treatment and compute the backreaction to the zeldovich flow based on an explicit functional form of the displacement field in the shell - crossing region .
to be precise , we compute the force from eq .
( [ eq : force_poisson ] ) exerted at each position using the extrapolation of the zeldovich flow . integrating the force over time , we obtain the correction for the velocity to the zeldovich motion from eq .
( [ eq : sol1_p ] ) . further integrating the corrected velocity over time ,
a correction for the position is obtained from eq .
( [ eq : sol1_x ] ) . throughout these calculations , we assume that the collapsing region is small , which allows us to express the phase - space configuration in terms of polynomial forms of low order of the lagrangian coordinate . let us focus on a collapse point centered on lagrangian position @xmath52 and perform a perturbative description of the multi - stream flow around @xmath58 .
as we mentioned earlier , the conditions for shell - crossing are given by eq .
( [ eq : condition_shellcrossing ] ) . shortly after shell - crossing time @xmath53 ,
the multi - stream flow has just started growing in a small lagrangian region around @xmath52 and the displacement field @xmath46 can be approximately described by a third - order polynomial in @xmath59 .
since the deviation from zeldovich flow is small , the motion of a mass element around @xmath58 may be expanded at third - order in @xmath26 for @xmath60 as @xmath61 with the time - dependent coefficients @xmath62 , @xmath63 and @xmath64 defined by @xmath65 where the prime denotes the derivative with respect to @xmath26 . here , we used eqs .
( [ eq : def_deltal ] ) and ( [ eq : shellcrossing_peak ] ) .
the above expressions imply that the shell - crossing point slightly moves from @xmath66 to @xmath67 and the shell - crossing structure develops as an `` @xmath68 '' shape , as shown in fig .
[ fig : shellcrossing ] .
we now introduce the new lagrangian coordinate , @xmath57 , for which the structure of the local density peak is symmetric . since we expand the displacement field at third - order , the equation @xmath69 inside the shell - crossing region has three solutions .
an example is shown in fig .
[ fig : shellcrossing ] for the equation @xmath70 with solutions @xmath71 . in general , the three ordered roots of the equation @xmath72 , when they exist , are related to each other through the following expressions : @xmath73 here , the lagrangian extent @xmath74 of the boundary of the multi - valued region , also related to the lagrangian position @xmath75 of the left caustic in fig .
[ fig : shellcrossing ] through @xmath76 , is expressed as @xmath77 because of the time dependence of the coefficients ( see eqs .
[ eq : def_b ] and [ eq : def_c ] ) , the position of the boundary gradually changes in time . the leading - order expression in powers of time
@xmath78 reads @xmath79 with @xmath80 in other words , it also means that a given element of fluid of lagrangian position @xmath81 will enter the multi - valued region at some time @xmath82 defined by @xmath83 and then subsequently coincide exactly with the lagrangian position of the caustic at a time @xmath84 with @xmath85 . by inverting the relations @xmath86 and @xmath87
, we obtain @xmath88 and @xmath89 the quantities defined above thus play a crucial role to disentangle the single - stream regime from post - collapse dynamics , as well as the inner part of the `` @xmath90 '' shape of fig .
[ fig : shellcrossing ] from its outer part , the lagrangian position of the caustic acting as a separator in the latter case .
the expression of the force will indeed be different between the zeldovich single - flow regime @xmath91 and the multi - valued regime composed of an outer caustic region , @xmath92 , and an inner caustic region , @xmath93 .
one important remark for the subsequent calculations is that the boundaries coordinates @xmath74 and @xmath94 are assumed to be small and that the backreaction to the position and velocity inside the multi - valued region can be described perturbatively in a polynomial form of @xmath95 [ partly allowing the fractional power of @xmath96 .
thus , the resultant expressions are , rigorously speaking , only valid for a short period after collapse time , but we shall see in practice , as also shown by @xcite in the non - cosmological case , that they remain impressively accurate even up to next - crossing time ( see sec . [
sec : comparison_nbody ] ) , which will allow us to set up the framework for a powerful self - adaptive scheme .
to derive the corrections to the motion , we first compute the force exerted on a mass element inside the multi - valued region , @xmath97 ( see fig . [ fig : shellcrossing ] ) .
note again that the outer regions @xmath98 and @xmath99 are described by the zeldovich solution .
the force in the multi - valued region is computed using eq .
( [ eq : force_poisson ] ) , dividing each integral of the right - hand - side into three contributions : @xmath100 assuming that the collapse region , @xmath101 , is small enough , the contributions to the integrals from each domain can be computed analytically , based on the geometrical setup in fig .
[ fig : shellcrossing ] .
the detailed calculations are presented in appendix [ sec : derivation ] .
summing up all the contributions given in eqs .
( [ eq : integral_inner1 ] ) , ( [ eq : integral_inner2 ] ) , ( [ eq : integral_outer1 ] ) and ( [ eq : integral_outer2 ] ) , the force exerted on the mass element at @xmath102 inside the multi - valued region becomes @xmath103 \label{eq : force_inner}\end{aligned}\ ] ] with the functions @xmath104 and @xmath105 respectively defined by @xmath106 and @xmath107 where the quantities @xmath62 , @xmath63 , and @xmath64 are defined by eqs .
( [ eq : def_a])([eq : def_c ] ) .
note that in deriving eq .
( [ eq : force_inner ] ) , we have assumed that the system follows zeldovich solution if @xmath108 . since the resultant expressions are written in terms of the local quantities characterizing the density peak at position @xmath52 and the shell - crossing time @xmath53 , eq .
( [ eq : force_inner ] ) is in fact still applicable to other shell - crossing regions possibly appearing elsewhere in the region @xmath108 , allowing one to generalize the result to a smooth random initial field with multiple peaks .
one issue , discussed later , is then to treat mergers , i.e. the case when for instance two multi - stream regions start overlapping .
given the explicit expression for the force in the multi - stream region and using the formal solution given by eqs .
( [ eq : sol1_x ] ) and ( [ eq : sol1_p ] ) , we now compute corrections to the zeldovich flow that we write as follows : @xmath109 we noticed in previous section that , depending on the lagrangian position of interest , the expression for the force is different and we have to divide the domain of the integrals in eqs .
( [ eq : delta_p_formal ] ) and ( [ eq : delta_x_formal ] ) into several pieces : 1 .
@xmath110 : the position @xmath81 is located in the single - valued region ( i.e. , @xmath111 ) and the motion is still described by the zeldovich solution .
we have @xmath112 1 .
@xmath113 : the position @xmath81 is in the multi - valued region and satisfies @xmath114 , i.e. lies in the outer part of the caustic .
in addition to the zeldovich flow , the corrections arising from the multi - stream flow need to be added and we note them as follows : @xmath115 in this region , which correspond to the tails of the @xmath90 shape in fig .
[ fig : shellcrossing ] , the system is globally expanding in phase space and gaining energy , at variance with the central part , which on the contrary , contracts .
this can be understood from computing the variation of energy of a typical test particle during a fraction of orbit .
this phenomenon was studied in the non - cosmological case by @xcite and the results should not be fundamentally different in the cosmological case studied here . 1 .
@xmath116 : this corresponds to @xmath117 , i.e. the position @xmath81 now lies in the inner part of the multi - stream region with respect to the caustic .
similarly to the above case , the backreaction to zeldovich flow needs to be computed , including at present both the multi - stream dynamics of the inner part and the incoming flow from the outer part .
we may write @xmath118 note that the corrections @xmath119 and @xmath120 partly come from outer part contributions , @xmath121 and @xmath122 . as mentioned in the previous point ( ii ) , in this region , which corresponds to the central part of the @xmath90 on fig .
[ fig : shellcrossing ] , the system is globally contracting and losing energy in favor of the tails of the @xmath90 . in
what follows , we compute the backreaction to zeldovich flow and derive the expressions for @xmath123 and @xmath124 in each domain .
the calculation for the corrected motion is rather straightforward but needs several steps . here , we give a brief sketch of the calculation , deferring details to appendix [ sec : pcpt_basic ] .
the final results are given in eqs .
( [ eq : p_for_tau_c > tau > hattau_c ] ) and ( [ eq : x_for_tau_c > tau > hattau_c
] ) for the outer part , eqs . ( [ eq : p_for_tau > tau_c ] ) and ( [ eq : x_for_tau > tau_c ] ) for the inner part , together with the coefficients in tables [ tab : xv_pcpt1 ] and [ tab : xv_pcpt2 ] .
note that although our expansion is rigorously valid only at third - order in @xmath81 , higher - order contributions will appear in the corrections to provide a continuous solution , up to fifth- and seventh - order for the velocity and the position , respectively .
cll coefficients & velocity @xmath125 [ eq . ( [ eq : p_for_tau_c > tau > hattau_c ] ) ] & position @xmath126 [ eq .
( [ eq : x_for_tau_c > tau > hattau_c ] ) ] + + @xmath127 & @xmath128 & @xmath129 + + @xmath130 & @xmath131 & @xmath132 + + @xmath133 & @xmath134 & @xmath135 + + @xmath136 & @xmath137 & @xmath138 + + @xmath139 & ' '' '' & @xmath140 + + @xmath141 & @xmath142^\tau_{\widehat{\tau}_{\rm c}(q)}}$ ] & @xmath143 + in the outer part of the multi - valued region , ( ii ) , the correction to the velocity can be expressed as @xmath145 in the above , while the first integral is performed with the help of formulae in appendix [ sec : integral ] , the second integral is computed exactly .
the whole derivation is provided in appendix [ subsec : delta_v_out ] .
the resultant expression can be summarized as @xmath146 + \ , \widetilde{\epsilon}_1(\tau,\widehat{\tau}_{\rm c } ) , \label{eq : p_for_tau_c > tau > hattau_c}\end{aligned}\ ] ] with the time - dependent coefficients given in table [ tab : xv_pcpt1 ] .
note that the coefficient @xmath147 implicitly depends on the lagrangian position @xmath81 through @xmath148 but is not taylor expanded with respect to @xmath81 , for simplicity .
a fully analytical theory , in particular to predict the power spectrum of the projected density field , would in principle require such a taylor expansion . performing
it should not change significantly the performances of post - collapse pt as presented in this article .
once @xmath125 is obtained , the expression for the correction @xmath126 is evaluated by further integrating eq .
( [ eq : p_for_tau_c > tau > hattau_c ] ) over time : @xmath149 the resultant expression becomes ( see appendix [ subsec : delta_x_out ] for derivation ) : @xmath150 \nonumber\\ & \quad\quad + \,\widetilde{\epsilon}_2(\tau,\widehat{\tau}_{\rm c } ) , \label{eq : x_for_tau_c >
tau > hattau_c}\end{aligned}\ ] ] with the time - dependent coefficients presented in table [ tab : xv_pcpt1 ] .
note again the dependence on @xmath81 of parameter @xmath151 through @xmath82 and the fact that we did not taylor expand it in polynomials of @xmath81 , as it would be required for a fully analytical theory .
let us consider now the inner part of the multi - valued region ( iii ) . for the correction to the velocity , @xmath153 , the expression to be evaluated
is divided into three contributions : @xmath154 .
\label{eq : delta_p_2 } \end{aligned}\ ] ] evaluating each contribution in appendix [ subsec : delta_v_in ] , we obtain @xmath155\,+\ , \widetilde{\epsilon}_1(\tau , \widehat{\tau}_{\rm c } ) , \label{eq : p_for_tau > tau_c}\end{aligned}\ ] ] with the time - dependent coefficients given in table [ tab : xv_pcpt2 ] .
note that the coefficient @xmath147 in the above equation is the same one as in eq .
( [ eq : p_for_tau_c > tau > hattau_c ] ) .
cll coefficients & velocity @xmath153 [ eq . ( [ eq : p_for_tau > tau_c ] ) ] & position @xmath119 [ eq . ( [ eq : x_for_tau > tau_c ] ) ] + + @xmath127 & @xmath156 & @xmath157 + + @xmath130 & @xmath158 & @xmath159 + + @xmath136 & @xmath160 & @xmath161 + + @xmath139 & ' '' '' & @xmath162 + + @xmath141 & @xmath142^\tau_{\widehat{\tau}_{\rm c}(q)}}$ ] & @xmath143 + similarly , the correction to the position , @xmath119 , is divided into three pieces : @xmath163 where the function @xmath164 represents the correction of the velocity given in either eq .
( [ eq : p_for_tau_c > tau > hattau_c ] ) or eq .
( [ eq : p_for_tau > tau_c ] ) , but without the contribution corresponding to the zeldovich motion , i.e. , the last term in their expressions .
the term @xmath165 is the same quantity as in eq .
( [ eq : x_for_tau_c > tau > hattau_c ] ) ( see table [ tab : xv_pcpt1 ] ) . based on the calculation presented in appendix [ subsec : delta_x_in ]
, we obtain the expression for @xmath119 : @xmath166\ , + \,\widetilde{\epsilon}_2(\tau,\widehat{\tau}_{\rm c } ) , \label{eq : x_for_tau > tau_c}\end{aligned}\ ] ] with the time dependent coefficients given in table [ tab : xv_pcpt2 ] .
the perturbative description of post - collapse dynamics introduces two kinds of approximations around the singularities ( shell - crossing points ) .
firstly , we assumed the position and velocity to be described by third - order polynomials in @xmath81 in the multi - stream regions . to preserve continuity between the multi - stream and the single - stream regime
, we added higher - order terms in @xmath81 to the solutions for @xmath14 and @xmath167 , but strictly speaking , these solutions are valid locally only at third - order in @xmath81 around the singularities . secondly
, these solutions are correct only very shortly after collapse time @xmath53 up to second - order in @xmath168 .
our post - collapse dynamics can thus be improved at a twofold level : ( a ) : : _ improvement at the spatial level : _ we could try , starting from initial conditions , to use a description of the @xmath90 shape of fig .
[ fig : shellcrossing ] at higher - order in @xmath81 , e.g. fourth- or fifth- instead of third - order . in this latter case ,
deriving the solution of equation @xmath70 comes down to find the roots of a third or fourth - order polynomial instead of a second - order one , which makes , in addition to the symmetry breaking of the @xmath90 shape , the calculations much more cumbersome , even though a series expansion of the roots at fourth or fifth - order in @xmath81 remains tractable and might bring interesting improvements on the description of the tails of the @xmath90 shape .
( b ) : : _ higher - order in time : _ we can introduce higher - order corrections in time to follow more accurately the evolution of the central part of the singularity after collapse .
note that after integration of the equations of motion , such corrections bring out higher - order terms in @xmath81 .
strictly speaking , the right way to proceed would consist of performing `` post - post - collapse '' dynamics , i.e. of computing a counter - term to the zeldovich+leading - order post - collapse solution .
this procedure is cumbersome , and seems to make sense only if performed simultaneously with a description at higher - order in @xmath81 , in particular because the tails of the @xmath90 structure in fig .
[ fig : shellcrossing ] also influence the dynamics of the central part of the system ( but this remains to be proved ) .
going beyond leading - order in post - collapse dynamics thus seems rather challenging even in the simple one - dimensional case considered in this article .
however , we can still try to investigate some potential improvements of the theory while maintaining the complexity of the calculations at an acceptable level . for this purpose
, we consider here mainly two approaches .
the first one just consists in incorporating next - to - leading order terms into the expressions for @xmath74 and @xmath169 ( and @xmath75 and @xmath170 ) .
these quantities , which determine the boundary between the single - stream and the multi - stream regime , are indeed crucial because they correspond to the main scale / timescale of the system .
the second approach consists , in addition , of improving the evaluation of the time integrals in eqs .
( [ eq : delta_p_formal ] ) and ( [ eq : delta_x_formal ] ) . in sec .
[ subsec : corrections_pcpt ] , when evaluating these time integrals , a part of the integrands is taylor - expanded and the results are presented at leading - order in time .
it is thus possible to include higher - order corrections from the taylor - expanded integrands .
again , one has to be aware that the two corrections we propose might not improve the results , since a correct higher - order time approach , as discussed in point ( b ) above , would actually require to compute counter terms to the standard post - collapse result , which we do not do here .
in fact , our higher - order corrections might even _ worsen _ the results since the corrections brought by such additional terms might have opposite sign compared to the iterated counter - term approach .
the derivation of the higher - order corrections we propose here and summarize just below is rather straightforward compared to its rigorous alternatives ( a ) and ( b ) above , but still cumbersome .
the resultant expressions now include higher - order polynomials of @xmath81 as well as new contributions , which are given in appendix [ sec : pcpt_higher - order ] . to be more specific , we consider the following cases : 1 . higher - order for the expressions of critical time @xmath171 and position @xmath172 ( _ hc _ ) : based on the basic results in sec .
[ subsec : corrections_pcpt ] , we include the contributions arising from the next - to - leading order expressions for @xmath74 and @xmath169 ( as well as @xmath75 and @xmath173 ) .
the calculation to derive higher - order terms is almost the same as presented in sec .
[ subsec : corrections_pcpt ] , but we replace the leading - order expressions for @xmath74 and @xmath169 in eqs .
( [ eq : qchat_approx ] ) and ( [ eq : hattau_c ] ) with those including the next - to - leading order , given in eqs .
( [ eq : higher_order_qhat ] ) and ( [ eq : higher_order_hattau_c ] ) .
the higher - order results , denoted by @xmath174 and @xmath175 , correspond to the expressions summarized in appendix [ sec : pcpt_higher - order ] , but where the terms involving the quantities @xmath176 or @xmath177 are all set to zero . 1 .
higher - order in time ( _ ho _ ) : in addition to the above treatment , we also include higher - order corrections to the time - integrals , based on the formulas given in appendix [ sec : integral ] . the resultant expressions , @xmath178 and @xmath179 , are summarized in appendix [ sec : pcpt_higher - order ] . finally , we also consider a semi - analytic approach in order to smoothly connect the inner part of the multi - stream region to the zeldovich solution : 1 .
third - order spline ( _ spl _ ) : basic post - collapse pt is applied to the inner part , @xmath152 , but we take the higher - order expression for @xmath75 . in the outer part , @xmath180
, on the other hand , a third - order spline interpolation is used to smoothly connect the inner post - collapse region to the zeldovich solution , with again , the higher - order expression for @xmath74 .
post - collapse pt as described in
[ sec : postcollapse ] provides a way to describe the dynamics around an initial density peak shortly beyond collapse time . in practice ,
the quality of the description is expected to be good until next - crossing time , at least this is what is known in the non - cosmological case @xcite . using the leading - order expression given in eq .
( [ eq : x_for_tau > tau_c ] ) , the duration of time spent between collapse and next shell - crossing can be estimated from eq .
( [ eq : x_post - collapse_inner ] ) , @xmath181 which is derived by solving the equation @xmath182 . beyond time
@xmath183 , post - collapse pt is expected to fail , similarly as zeldovich approximation fails beyond @xmath184 .
the only way to fix this problem at the fine grain level would be to use an iterative procedure similar to what was proposed by @xcite to reconstruct the spiral structure built up during the course of the dynamics .
in addition , post - collapse pt assumes the initial density distribution to be sufficiently smooth .
while this is certainly expected at very small scales in the standard cold dark matter scenario , the issue of mergers arises . indeed , post - collapse pt is valid only if the forming halo is isolated .
if there are two adjacent collapsing regions , they might , at some point , overlap in lagrangian space , hence merge with each other in eulerian space .
after the merger , a separate description of each halo becomes inadequate . in this case , the prediction of post - collapse pt , which is in principle able to handle multiple but non - overlapping structures in lagrangian space , will break down much earlier than expected .
one simple way to fix both limitation in time and mergers issues , already applied in the past to the zeldovich approximation ( see , e.g. * ? ? ?
* ; * ? ? ?
* and references therein ) , consists in trying to describe the dynamics at the coarse - grained level by introducing a smoothing procedure in lagrangian space , so that post - collapse pt remains applicable , in particular that the condition @xmath185 remains valid for the system with coarse - grained initial conditions .
this means that single or composite halos as well as mergers are locally summarized by a simplified @xmath90 shape ( fig .
[ fig : shellcrossing ] ) corresponding to a single peak of the coarse - grained initial density field and such that @xmath185 , while retaining the correct large - scale dynamics .
the hope is that this simplified structure still captures the most important features of the nonlinear system , e.g. its mass and its typical size . of course , the coarsening scale should depend on environment , hence on lagrangian position @xmath81 , so for best results , _ adaptive smoothing _ of initial conditions has to be introduced . in this sense
our method is very similar to the `` peak - patch '' treatment by @xcite ( see also * ? ? ?
* ) , but we implemented it in the post - collapse pt framework . to be precise
, the practical set up of post - collapse dynamics with adaptive smoothing is summarized as follows :
the initial density field is smoothed at various scales , employing the sharp-@xmath186 filter function in fourier space : @xmath187 with @xmath39 being the heaviside step function .
the cutoff wavenumber @xmath188 ranges from the fundamental mode @xmath189 to a certain value @xmath190 , where @xmath191 is an integer corresponding to the effective resolution of the simulation , as discussed more in details in
[ subsec : simulation ] . in our investigations
, we covered all the possible values of @xmath188 , i.e. @xmath192 with @xmath193 , but this is costly and probably not required , in practice . 1 .
zeldovich solutions are constructed from the smoothed density fields .
then we identify the critical points @xmath52 in lagrangian space that have undergone shell - crossing at time @xmath194 , where @xmath19 is the time of interest for the analyses .
for each critical point , various quantities needed for perturbative calculations , such as @xmath195 and @xmath196 , are also computed . 1 .
starting from the largest smoothing scale ( i.e. @xmath197 ) , we compute the next - crossing time @xmath198 associated to each critical point . for the points satisfying @xmath199 with @xmath200 a constant parameter ,
the post - collapse correction to the zeldovich flow is obtained in the interval @xmath201 $ ] , which is labeled . 1
. the procedure ( iii ) is repeated , but with a smaller smoothing scale , i.e. , larger @xmath188 , and only for peaks lying in the unlabeled regions , until the cutoff wavenumber reaches @xmath202 . note that if two lagrangian regions overlap , e.g. @xmath203 and @xmath204 , corresponding to two different values of the smoothing scale , @xmath205 , the interval @xmath204 corresponding to the smallest smoothing scale , i.e. largest wavenumber @xmath206 , prevails .
this is the way we solve the cloud - in - cloud problem , but not the only possible choice . 1 . at the end of the iterative process
, post - collapse dynamics is applied as well for the rest of the critical points corresponding to the highest resolution scale @xmath207 that still require treatment , i.e. for those with @xmath208 .
the procedure is the same as described in point ( iv ) above , i.e. only critical points which do not fall in a labeled region are treated , and if it is the case , then their lagrangian interval @xmath209 prevails for applying post - collapse dynamics . for the rest of the unlabeled regions ,
standard zeldovich dynamics is applied and the mapping uses the initial density corresponding , of course , to the highest resolution scale .
note that the approximation of the phase - space density distribution constructed this way is generally discontinuous , because we collect perturbative solutions for the displacement field in phase space coming from different coarse - graining scales without imposing smoothness at the transition between these solutions . in this respect ,
the procedure given above is not entirely consistent and remains to be improved in the future .
nevertheless , the discontinuities only affect the results at small scales and in regions where the projected density is small . indeed , the discontinuities correspond to transitions between tails of the @xmath90 shape representing a halo and a non collapsed region or transitions between tails of two @xmath90 shapes in the case of a merger .
as illustrated below through numerical examples , these regions of phase space have small density contrast and thus contribute weakly to second- or higher - order statistics , hence to the power spectrum which is the statistics under focus in this work . with an appropriate choice of @xmath210
, we shall show that the adaptive smoothing procedure provides a substantial improvement on the post - collapse prediction for the overall phase - space structure of the system and the power spectrum of the projected density .
we are now in position to compare the predictions of post - collapse pt with simulation results . in [ subsec : simulation ] , we briefly describe the one - dimensional cosmological @xmath13-body code we wrote to perform the numerical analyses of this article and discuss various important parameters settings .
we then present the results of the comparison between theory and numerical experiments in various cases with cold initial conditions : formation of an isolated single halo ( [ subsec : single_cluster ] ) , merging of two neighboring halos ( [ subsec : merging_clusters ] ) , as well as random initial conditions with cold dark matter - like ( cdm ) and power - law power spectra ( [ subsec : cdm_initial ] and [ subsec : power - law_initial ] ) . in the later case , we shall discuss more in details highly nonlinear dynamics , in particular predictions from stable clustering , the expected outcome on the power spectrum from halos with power - law density profiles , as well as the statistical properties of the phase - space sheet resulting from a caustic distribution .
to solve numerically vlasov - poisson equations in one dimension , we wrote a fortran90 particle - mesh code , vlafroid . in this code , publicly available through the following web page , www.vlasix.org , poisson equation is solved on a periodic grid of fixed resolution @xmath211 using fast fourier transform , after projecting the @xmath212 particles on the grid using simple cloud - in - cell interpolation ( see , e.g. * ? ? ? * ) .
the particles positions and velocities are updated at each time step using a standard predictor - corrector integration scheme .
after the predictor step , @xmath213 poisson equation is solved and the force is estimated on the grid by direct derivation in fourier space .
the acceleration @xmath214 of each particle is computed by linear interpolation of the force .
this is followed by the corrector step : @xmath215 the value of the time step @xmath216 is slowly varying with time and is bounded by 3 constraints : 1 . a limit on the relative variations of the expansion factor during a time step to have a good description of the linear growing modes at early stages of the simulation , @xmath217 2 . a dynamical condition which states that the time step should be a small fraction of the smallest harmonic oscillator time scale ( see , e.g. * ? ? ?
* ) , @xmath218 where @xmath219 is the maximum value of the projected density ( normalized to be unity in average ) , 3 .
the standard courant - friedrichs - lewy ( cfl ) condition @xmath220 with @xmath221 the maximum magnitude of the velocity , stating that a particle should not travel more than a fraction of a grid cell size during a time step . for all the runs performed in this article ,
our practical and safe choice of the time step parameters was : @xmath222 the evolution of single or composite objects as studied in
[ subsec : single_cluster ] and [ subsec : merging_clusters ] does not pose any problem from the numerical point of view . indeed , we consider here only a small number of dynamical times , so we do not expect significant defects due to the discrete nature of the particle distribution ( for instance , such as described in * ? ? ?
the case of a random field , as studied in
[ subsec : cdm_initial ] for cdm and in [ subsec : power - law_initial ] for power - law power spectra , is less trivial .
indeed , one needs accurate numerical modeling of the dynamics of each single structure , which has to be sampled with many particles and be resolved with a sufficient number of grid elements .
in particular , to avoid contamination by the discrete nature of the particle distribution , we require , in average , many particles per grid element : @xmath223 to avoid the formation of small structures that would be insufficiently resolved by the computational grid , there must be a coarse - graining scale below which the initial conditions are smooth .
this also facilitates , of course , the calculation of ( post - collapse ) pt predictions .
here we impose a cut - off on the power spectrum of initial conditions using the sharp-@xmath186 filter introduced in eq .
( [ eq : sharp - k ] ) , with @xmath224 in practice , a choice which works well and adopted in this paper is @xmath225 which means that @xmath202 is five time smaller than the nyquist frequency of the grid .
since the cut - off scale on the power spectrum can affect the dynamics compared to the true , un - smooth system , we shall perform some convergence studies to determine the available dynamic range in which the results are not influenced by @xmath202 .
let us first present the simplest case of the formation of a single structure seeded with the following sinusoidal density perturbation : @xmath226 we adopt the einstein - de sitter cosmology ( @xmath227 , @xmath228 with @xmath229 ) and the box size of the simulation is arbitrarily set to @xmath230 .
the simulation was started at expansion factor @xmath231 with initial amplitude @xmath232 and was run using @xmath233 .
pt analyses were performed on the fly as part of routines of the code , when needed .
[ fig : single - cluster_xv_basic ] shows snapshots of the phase - space structure ( upper inserts ) and projected density profiles ( lower inserts ) obtained from the simulation ( red ) and post - collapse pt ( blue ) based on the basic prescription described in sec .
[ subsec : corrections_pcpt ] .
also , for reference , the zeldovich solution is plotted as a green dashed line on each panel . although the three curves are identical at early time , differences show up , as expected , after collapse , which occurs at @xmath234 .
a closer look at the central part of the system reveals however that post - collapse pt captures the trends of the @xmath13-body simulation quite well and succeeds to reproduce the structure of the inner part of the system until next - crossing time , which occurs at @xmath235 .
this is in marked contrast with zeldovich solution , which , without the back reaction contribution present in the post - collapse correction , overshoots the displacement of the phase - space sheet after first crossing time , in particular of the caustics .
after next - crossing time , @xmath236 , the phase - space structure given by post - collapse pt goes away from the @xmath13-body result .
nevertheless , the predicted projected density profile still provides a reasonably good approximation of the evolved density structure , which is clearly not the case for the zeldovich solution .
[ fig : single - cluster_xv_variant ] presents results of post - collapse pt when including the higher - order corrections discussed in
[ subsec : higher - order ] . while the three prescriptions examined here , namely _ hc _ , _ ho _ and _ spl _ , account for higher - order terms in different manners , they are all based on the force expression in eq .
( [ eq : force_inner ] ) , which is valid only shortly after collapse . indeed , as discussed in
[ subsec : higher - order ] , our implementation of higher - order corrections is not performed in a fully rigorous way , that would require to compute a back reaction to post - collapse pt , similarly as post - collapse pt introduces a back reaction correction to zeldovich solution .
it also does not propose an iteration procedure that would allow us , similarly as in @xcite , to follow the system behind next - crossing time .
hence , all of the treatments break down at the end of the simulation .
nevertheless , taking the higher - order corrections to @xmath74 and @xmath169 into account improves the prediction of the outer boundary of the multi - valued region , and the phase - space description becomes visually better than that of the basic pt prescription in fig .
[ fig : single - cluster_xv_basic ] . even after next - crossing time ( @xmath236 ) ,
the predicted phase - space structure still remains consistent , although the size of the halo is prone to be over - estimated , leading to an extended density profile , the best result being obtained with the _ spl _ prescription .
still , one can notice that if only the projected density is under consideration , basic post - collapse pt does better than any of its higher - order counter - part , except for _ spl _ which , not surprisingly , provides nearly equivalent results , since it is exactly the same as post - collapse pt in the center of the system .
as far as statistics depending on the projected density are concerned , such as the power spectrum that will be studied later , we thus expect from the analyses of this single halo that the simplest prescription of post - collapse pt is going to perform best .
we examine now the second simplest case of two merging structures in einstein - de sitter cosmology ( @xmath227 , @xmath237 and @xmath229 ) .
the initial density contrast is given by @xmath238 .
\label{eq : delta_init_two_clusters}\end{aligned}\ ] ] the constant @xmath239 is determined so as to satisfy the condition @xmath240 . with @xmath241 , @xmath242 and @xmath243 , we have @xmath244 .
the overall amplitude is set to @xmath245 and the simulation , of box size unity ( @xmath230 ) , was started at @xmath231 and was run using @xmath233 .
the results are shown in fig .
[ fig : merging - cluster_xv_basic ] for basic post - collapse pt and fig .
[ fig : merging - cluster_xv_variant ] for the higher - order variants of post - collapse pt .
here , we focus on the merger phase which takes place after collapse of each individual structure . figs .
[ fig : merging - cluster_xv_basic ] and [ fig : merging - cluster_xv_variant ] demonstrate how the adaptive smoothing introduced in sec .
[ sec : improvement ] improves pt prescriptions to describe the overall phase - space structure after the merger . in each figure
, the upper and lower panels show the results without and with adaptive smoothing , respectively .
the free parameter controlling adaptive smoothing , @xmath210 , is set here to @xmath246 for post - collapse pt and @xmath247 for the zeldovich solution . without adaptive
smoothing , as soon as the two clusters cross each other ( @xmath248 ) , both post - collapse pt and zeldovich solution start failing to describe the real dynamics in the @xmath13-body simulation . indeed , while post - collapse pt tries to provide a local perturbative correction for each single cluster ,
it is clearly unable to account for nonlinear couplings taking place at larger scales , where it simply follows zeldovich motion . as a result ,
the location of multi - valued regions predicted by post - collapse pt , as well as their shape , largely deviates from the correct one .
on the other hand , implementing adaptive smoothing in the pt calculations significantly improves the predicted phase - space structure , starting already from the beginning of the merger phase ( @xmath248 ) . after merging ,
dynamics of the central part of the system is now described by the displacement field derived from the smoothed initial density field . by construction
, it summarizes the composite structure made of two sub - halos into a single cluster .
while this is a rough approximation of the real dynamics , the description of the outer part of the system is substantially improved as well as the prediction for projected density profiles . introducing both adaptive smoothing and higher - order corrections to post - collapse pt further improves the results from the visual point of view ( fig .
[ fig : merging - cluster_xv_variant ] ) , although the predicted size of the halo tends to be slightly overestimated . as already noticed in
[ subsec : single_cluster ] , the best results are obtained for _ spl _ , which also gives , after adaptive smoothing , comparable projected density profiles to basic post - collapse pt .
the above results demonstrate that post - collapse pt with adaptive smoothing is effective in capturing the main features of phase - space structures .
strictly speaking , it does not give an accurate prescription for the fine structure of high - density regions , but it provides a way of regularizing or mitigating the impact of highly nonlinear dynamics , keeping the location and size of halos reasonably accurate . as we will see next , adaptive smoothing brings a drastic improvement on the prediction of power spectra for systems with random initial conditions .
furthermore , the introduction of adaptive smoothing makes the analytic calculations much less sensitive to the choice of the small - scale cutoff in the initial conditions so that pt predictions become more robust . in these respects ,
the determination of the extension of multi - streaming regions for a given smoothing scale of initial conditions , as performed in step ( iii ) of [ sec : improvement ] , is essential , making the choice of @xmath210 crucial .
concerning the particular merger experiment studied here , @xmath249 and @xmath247 seem to be the best choices for tuning post - collapse pt and zeldovich solutions , respectively , and we shall adopt these values in subsequent section . we shall however discuss again about the choice of @xmath210 later .
let us now consider random cdm initial conditions .
although there is no realistic setup in 1d , a way of mimicking as well as possible the 3d case may be to consider initial conditions given by a random gaussian field with the following power spectrum ( see , e.g. * ? ? ?
* ) : @xmath250 with @xmath251 being the 3d matter power spectrum of the initial density fluctuations .
here we use for @xmath251 the transfer function of @xcite who wrote the corresponding fortran module that we included in vlafroid .
we set the cosmological parameters to those of the base @xmath252cdm model determined by planck @xcite : total matter density parameter @xmath253 , cosmological constant @xmath254 , baryonic matter density parameter @xmath255 , hubble constant @xmath256kms@xmath257mpc@xmath257 , effective slope @xmath258 and root mean square deviation @xmath259 . through @xmath260 with @xmath261mpc , @xmath262 and @xmath263 .
[ foot : toto ] ] the simulations were performed with a box size @xmath264mpc and started at initial redshift @xmath265 .
convergence of simulations results was tested by varying the number of particles @xmath266 , the resolution @xmath211 of the grid used to compute the force and the cutoff wavenumber of the initial power spectrum , @xmath202 , introduced in [ subsec : simulation ] . ,
@xmath267 and @xmath268 . while the highest resolution runs with @xmath269 only allowed us to test the @xmath13-body results and zeldovich solution without adaptive smoothing , lower resolution runs included full pt predictions with and without adaptive smoothing .
] here , we present the results of the analyses for @xmath270 , @xmath271 , and @xmath272 mpc@xmath257 . our convergence study shows , unless specified otherwise below , that the results can be trusted for @xmath273 mpc@xmath257 , i.e. in the available dynamic range shown in left panels of figs .
[ fig : pk_snap_cdm_midres ] and [ fig : pk_snap_cdm_midres2 ] . to have sufficient statistics for measuring power spectra , we ran @xmath274 simulations with different random seeds . figs .
[ fig : pk_snap_cdm_midres ] and [ fig : pk_snap_cdm_midres2 ] show the measured power spectra at different times ( left panel of each figure ) as well as snapshots of phase space for a given realization ( right panel ) .
right panel of fig .
[ fig : pk_snap_cdm_midres ] is supplemented with fig .
[ fig : snap_qxv_midres ] , which displays in approximately the same scale interval , eulerian positions and velocities as functions of lagrangian coordinates .
one can first notice that , in contrast to the 3d case , the amplitude of the power spectrum at small scales is not strongly enhanced in 1d , but damped .
for the cdm like initial conditions considered here , function @xmath275 becomes nearly flat in the highly nonlinear regime , which is probably a consequence of stable clustering as discussed further in details in next section , where the case of scale - free initial conditions is considered . in fig .
[ fig : pk_snap_cdm_midres ] , the simulations results are compared to basic pt predictions ( solid blue ) and zeldovich solution ( dotted green ) , while , in fig .
[ fig : pk_snap_cdm_midres2 ] , they are compared to various higher - order variants of post - collapse pt .
note , that except for the solid line that we explain below , all power spectra predicted by post - collapse pt or zeldovich solution are computed directly from the particle distribution evolved according to the displacement field predicted by each theoretical model under scrutiny .
so it is important to mention that we did not actually compute analytically power spectra in these cases , although it would be in fact possible .
the thin black solid lines , on the other hand , provide the analytic results for the zeldovich solution without adaptive smoothing , @xmath276 ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) : @xmath277\ , ; \nonumber\\ & \qquad\qquad\ i(q)=\int_0^\infty \frac{{\rm d}p}{\pi}\cos(p\,q)\frac{p_{\rm 1d}(p)}{p^2 } , \label{eq : pk_za_analytic}\end{aligned}\ ] ] where @xmath278 is the _ un - truncated _ power spectrum , i.e. without suppression of power at @xmath279 . the black solid curves therefore differ slightly from the green dotted ones on left panel of fig .
[ fig : pk_snap_cdm_midres ] , because of the effect of the cut - off at @xmath202 ( and not because of discreteness effects or finite grid resolution ) .
while the cut - off wavenumber @xmath202 has , as expected , a noticeable effect on the un - smoothed zeldovich solution as well as the un - smoothed post - collapse pt prediction and its variants ( a property that we do not show but that we verified ) , its influence becomes negligible in the range of values of @xmath186 under consideration , both for the simulations and the theoretical models with adaptive smoothing .
as expected , without adaptive smoothing , both post - collapse pt prescriptions and zeldovich solution deviate from the simulations as soon as the nonlinear scale becomes larger than the cut - off scale @xmath280 , which is already the case in upper insert of left panel of figs .
[ fig : pk_snap_cdm_midres ] and [ fig : pk_snap_cdm_midres2 ] , i.e. at @xmath281 . still , at this redshift , post - collapse pt ( or its variants ) without adaptive smoothing is able to capture the main features of phase space better than zeldovich solution , although this is not obvious to decipher on upper left insert of right panel of figs .
[ fig : pk_snap_cdm_midres ] and [ fig : pk_snap_cdm_midres2 ] . at lower redshift ,
halos relax to a highly nonlinear state and can merge together , which can not be described well by zeldovich solution that gives elongated structures in phase space ( middle - left and lower - left inserts of right panels of fig .
[ fig : pk_snap_cdm_midres ] .
as already argued in previous section , post - collapse pt provides only local corrections to zeldovich flow and therefore can not account for nonlinear dynamics at large scales , hence presents the same defect .
this elongation effect obviously leads to a strongly underestimated power spectrum .
note however that this depends also on the choice of the cut - off scale @xmath207 : if this scale would be kept , for the theoretical predictions , variable so that it remains close to the typical transition scale to the nonlinear regime , the results would improve greatly both for zeldovich solution and for post - collapse pt , as it is the case for upper insert of left panels of figs .
[ fig : pk_snap_cdm_midres ] and [ fig : pk_snap_cdm_midres2 ] . but here , instead of varying the effective smoothing scale @xmath207 for the pt predictions , we now focus on adaptive smoothing , which should be even better because it provides locally the `` optimal '' softening of initial conditions .
when adaptive smoothing is employed , agreement between pt predictions and simulation improves strikingly both from the visual and statistical points of view , even for zeldovich solution which performs comparably well to post - collapse pt at @xmath282 , reproducing simulations results in the whole available dynamic range , including the highly nonlinear regime , with a slight underestimation of the power spectrum , though .
of course , superiority of post - collapse pt shows up at @xmath283 but zeldovich solution remains quite good .
keep in mind , however , that zeldovich solution with adaptive smoothing as presented here uses the calculation of next - crossing time to estimate the local smoothing scale , hence relies as well partly on post - collapse pt .
these results confirm the conclusions of
[ subsec : merging_clusters ] : adaptive smoothing allows one to summarize composite structures into a single s shape like halo with the right size so that two - point statistics matches the true one even in the highly nonlinear regime .
in fact , pt predictions reproduce , at least partly , the plateau that can be observed at small scales in middle and lower insert of left panels of figs .
[ fig : pk_snap_cdm_midres ] and [ fig : pk_snap_cdm_midres2 ] . in this regime
that we will discuss more in details in next section , the system is most probably following stable clustering .
of course , because of the way we implement our adaptive smoothing , continuity of the phase - space sheet is not preserved anymore , as illustrated by right panel of figs [ fig : pk_snap_cdm_midres ] and [ fig : pk_snap_cdm_midres2 ] , but the overall description of phase - space structures is nevertheless improved tremendously .
this is illustrated even more clearly by fig .
[ fig : snap_qxv_midres ] , which also shows , as expected , that positions predicted by post - collapse pt or zeldovich solution match the simulation significantly better than predicted velocities .
this is a trivial consequence of the fact that positions correspond to velocities integrated over time : by construction , velocities are better approximated by theoretical models at early than at late times , hence their integrated counterpart , which corresponds to some averaged behavior , will compare better to the simulations .
note thus on right panel of fig .
[ fig : snap_qxv_midres ] the excellent agreement with the simulation for the predicted position from basic post - collapse pt , even when significant mergers take place : one just needs to compare bottom right insert of right panel of fig .
[ fig : pk_snap_cdm_midres ] to bottom insert of right panel of fig .
[ fig : snap_qxv_midres ] to be convinced by this state of fact . as a consequence
, we obtain an excellent match between basic post - collapse pt with adaptive smoothing and simulations measurements for the power spectrum , which is a two - point statistics not depending directly on the quality of representation of the velocity .
these arguments also apply , to a lesser extent , obviously , to the zeldovich solution . when focusing again on the power spectrum ( left panel of figs .
[ fig : pk_snap_cdm_midres ] and [ fig : pk_snap_cdm_midres2 ] ) , a detailed comparison between various variant of post - collapse pt models reveals that the basic prescription remains the best among all of them , although _ spl _ performs nearly as well .
this confirms the results of
[ subsec : single_cluster ] and [ subsec : merging_clusters ] , where we noticed that higher - order prescriptions for post - collapse pt tended to over - predict the size of halos and mergers ( although this is not obvious at first sight when examining phase - space diagrams of right panels of figs .
[ fig : pk_snap_cdm_midres ] and [ fig : pk_snap_cdm_midres2 ] ) , leading to a smaller power spectrum than the basic post - collapse pt prediction at small scales . note that , in agreement with intuition , the results of pt with adaptive smoothing are rather insensitive to the choice of cutoff wavenumber @xmath202 ( as long as it is kept large enough compared to the wavenumber corresponding to transition towards nonlinearity ) , but the quality of the agreement with simulations depends , on the other hand , on the value of the parameter @xmath210 introduced in sec . [ sec : improvement ] . here
, we adopt the intuitive choices @xmath249 for post - collapse pt and @xmath247 for zeldovich solution , but , as illustrated in appendix [ sec : choice_of_f_cross ] , it is possible to optimize the value of @xmath210 .
for instance at @xmath283 , we find for the analyses performed in this section that @xmath284 and @xmath285 would provide better agreement with simulation measurements .
however we also noticed that the optimal value of @xmath210 depends slightly on redshift as well as on initial conditions , this is why we decided for the analyses presented in this article to stick to our fiducial , slightly suboptimal choice .
we now turn to random initial conditions with a power - law power spectrum : @xmath286 this allows us to test post - collapse pt for different kinds of initial conditions and to understand more in details small - scale dynamics . assuming einstein - de sitter universe ( @xmath227 , @xmath237 and @xmath287 ) , we performed simulations for @xmath288 and @xmath246 , with the same resolution as our main cdm runs , i.e. @xmath270 , @xmath271 and @xmath289 mpc@xmath257 .
although the initial conditions are scale free , we set an actual size for the simulation box , @xmath290 mpc , which allows us to normalize the power spectrum using @xmath291 ( see footnote [ foot : toto ] ) .
with such a normalization , the level of nonlinearity of @xmath292 at the end of the simulations is roughly the same as in the cdm runs studied in previous section . to measure power spectra with sufficient statistics , we performed sets of 50 simulations with different initial random seed for each value of @xmath293 .
[ fig : pk_snap_powerlaw ] displays the power spectra ( left panel ) and phase - space diagrams ( right panel ) measured at @xmath283 in the scale - free simulations .
the numerical results are compared to basic post - collapse pt ( blue solid ) and to zeldovich solution ( green dotted ) with adaptive smoothing . for reference ,
linear theory ( black dashed ) and the analytic prediction given by eq .
( [ eq : pk_za_analytic ] ) for the zeldovich solution without smoothing ( black solid ) are also plotted in left panel . in eq .
( [ eq : pk_za_analytic ] ) , we introduced the actual cut - offs in the initial conditions of the simulations , i.e. below @xmath189 and above @xmath289mpc@xmath257 . ] in addition , predictions from stable clustering @xcite are displayed as a thin solid line and will be discussed more in details below .
to complete visual inspection of phase space , fig .
[ fig : snap_qxv_powerlaw ] plots velocities and positions as functions of lagrangian coordinate @xmath26 in the same region of phase space as in right panel of fig .
[ fig : pk_snap_powerlaw ] .
note that the simulations considered in right panel of fig .
[ fig : pk_snap_powerlaw ] and in fig .
[ fig : snap_qxv_powerlaw ] have the same random initial seed as for right panel of fig . [
fig : pk_snap_cdm_midres ] and fig .
[ fig : snap_qxv_midres ] , so one can compare directly phase - space features between scale - free and cdm - like simulations .
examination of figs .
[ fig : pk_snap_powerlaw ] and [ fig : snap_qxv_powerlaw ] demonstrates again the success of post - collapse pt with adaptive smoothing . while the results are nearly as good for @xmath288 as for the cdm - like case , post - collapse pt does slightly less well for @xmath294 .
the same trend can be seen for zeldovich solution with adaptive smoothing .
this result is not surprising , for two reasons .
firstly , the level of nonlinearity in the @xmath294 simulations is slightly larger than for @xmath288 and the cdm - like initial conditions . indeed ,
deviation from linear theory or pure zeldovich solution happens at smaller values of @xmath186 for @xmath294 than for other cases . secondly and more importantly , adaptive smoothing using a sharp low - pass filter is less efficient , from the dynamical point of view , when considering a power spectrum with a high slope .
the results would look even worse in the case @xmath295 that we simulated but do not show here . to explain this , one can just take the example of the single @xmath296 mode system , e.g. initial conditions with a single sine wave as shown in fig .
[ fig : single - cluster_xv_basic ] . in this case , with a sharp-@xmath186 filter , the fluctuation at @xmath297 is either kept , hence post - collapse pt becomes invalid after the system reaches next - crossing time , or it is completely suppressed , which means that the procedure with adaptive smoothing is at the end equivalent to no smoothing at all . increasing the power spectrum slope
@xmath293 basically makes us approaching the single mode case with @xmath298 , since we always introduce a cut - off at some value @xmath202 to have a sufficiently smooth representation of the phase - space sheet at very small scales .
of course , this situation is unrealistic in practice , because no such problem occurs when considering cdm - like power spectra , but one has to stay aware of the fact that adaptive smoothing does not always work .
a possible way to improve it could consist in changing the filtering window so that all the wavenumbers are affected by the softening procedure , even if it is performed predominantly at a given scale : to this respect , adaptive smoothing with a gaussian window might represent an interesting alternative to our sharp-@xmath186 filter .
another interesting result of our measurements is the excellent agreement between the simulations and the stable clustering prediction from @xcite and @xcite in the nonlinear regime at very high-@xmath186 , irrespective of the initial power - law slope @xcite .
] , who predict that the slope of the power spectrum should be equal to @xmath299 in fact , even in the cdm - like case , which roughly corresponds , in one dimension , to an asymptotic value of @xmath293 equal to @xmath300 at large @xmath186 , stable clustering prediction works , since it predicts a plateau , @xmath301 , which is indeed what we observe on left panel of fig .
[ fig : pk_snap_cdm_midres ] .
note that what @xcite and @xcite call `` stable clustering '' is very specific . in the 1d case considered here ,
it is equivalent to assuming that relaxed objects become of constant size in the coordinate @xmath302 instead the physical coordinate @xmath303 as normally supposed in the three - dimensional case ( see , e.g. * ? ? ?
* ; * ? ? ?
it is important to notice here that post - collapse pt seems to be able to reproduce correctly the measured power spectrum in the beginning of the stable clustering regime .
this suggests that stable clustering regime is established only in a few dynamical times , since post - collapse pt is only able to follow the dynamics of a single halo until next - crossing time . on the other hand , if one considers a `` halo model '' approach ( see , e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* and references therein ) and assumes that at large @xmath186 the power spectrum is dominated by the `` one - halo term '' , interesting properties are expected , at least when initial density fluctuations are sufficiently smooth at small scale . indeed , one dimensional simulations of single cold halos suggest that their central density profile nearly behaves like a power - law @xmath304 with @xmath305 .
the actual value of @xmath306 is controversial . while @xcite postulated it to be equal to @xmath247 ,
detailed @xmath13-body simulations suggest instead @xmath307 @xcite . on the other hand
, pure vlasov simulations results using the waterbag method suggest @xmath308 @xcite .
other values to keep in mind are @xmath309 from the singularity occurring exactly at crossing times and @xmath310 from the caustics ( see , e.g. * ? ? ? * ) .
these singularities imply , as discussed below , some expected behavior for @xmath311 at large @xmath186 @xcite .
a slope @xmath312 for the density profile corresponds , after some trivial algebraic calculations , to a power - law slope @xmath313 for the power spectrum .
for instance , one expects the zeldovich approximation to be dominated by the caustics at very small scales , i.e. @xmath314 constant at large @xmath186 , which is indeed the case on left panels of figs .
[ fig : pk_snap_cdm_midres ] and [ fig : pk_snap_powerlaw ] in both analytic ( black ) and numerical ( green ) cases , irrespective of adaptive smoothing . turning to post - collapse pt with adaptive smoothing , since , by using @xmath249 , we push the ( smoothed ) system exactly to next - crossing time , we are dominated by the singularity with @xmath309 . hence , in this case one expects @xmath315 close to @xmath316 when @xmath186 becomes very large , as can be observed on all the figures . with a smaller value of @xmath210 , the system is dominated , similarly as for zeldovich solution , by the caustics at very small scales and then @xmath317 reaches a plateau at large @xmath186 as can be seen in fig .
[ fig : pk_xv_fcross ] of appendix [ sec : choice_of_f_cross ] . without adaptive
smoothing , one also intuitively expect post - collapse pt power spectrum to be dominated by caustics at very small scales , however the objects have a more complex structure than zeldovich solution as shown by lower right panel of fig .
2 , hence the plateau at large @xmath186 is not clearly visible on left panel of figs .
[ fig : pk_snap_cdm_midres ] , [ fig : pk_snap_cdm_midres2 ] and [ fig : pk_snap_powerlaw ] probably because not reached as quickly as for the pure zeldovich solution when increasing @xmath186 . turning to the actual nonlinear evolution of single objects , assuming as found in the literature @xmath308 and @xmath247 for the density profile of a halo would respectively give @xmath318 and @xmath319 for the `` one halo '' power spectrum .
this of course supposes that statistical averaging over many halos of different masses does not affect the slope of the power spectrum predicted by the single halo .
again , we have to be aware of the fact that at very large @xmath186 , one must be dominated by the caustics , since we introduce a cut - off at large @xmath186 .
such a cut - off enforces some level of smoothness of the curve representing the system in phase space and thus actually imposes the existence of well - defined caustic structures in the nonlinear regime , hence a plateau for @xmath317 at very large @xmath186 , even for the simulations .
however this plateau is not visible in the simulations , because it lies outside the range of @xmath186 shown in the figures . on the other hand
, there must be , according to our `` halo model '' prescription , a regime where @xmath317 might decrease ( @xmath308 ) or present another plateau ( @xmath320 ) or at least some kind of inflection at some intermediate values of @xmath186 .
this is what can indeed be observed e.g. on lower insert of left panel of fig .
[ fig : pk_snap_cdm_midres ] for @xmath321 mpc@xmath257 for the cdm - like case and on lower insert of left panel of fig .
[ fig : pk_snap_powerlaw ] for @xmath322 mpc@xmath257 for the @xmath294 simulation .
there is no such signature obviously visible in the @xmath288 simulation , although the measurements are too noisy to conclude in this case .
interestingly , when the inflection is visible , its extension in @xmath186 space is correctly predicted by basic post - collapse pt with adaptive smoothing .
this just shows that basic post - collapse pt is able to predict approximately the correct size for the halos , which we already knew .
note that here , we did not try to position our single halo term discussion in the context of stable clustering predictions .
these latter can be considered as the outcome of some statistical averaging after weighting the one - halo term by the mass function .
they are also expected to be valid only in a finite range of values of @xmath186 if one assumes a cut - off at @xmath202 : what we mean here is that our `` one - halo term '' discussion is in fact not incompatible with the concept of stable clustering although the link between the halo model and the stable clustering predictions remain to be performed in details in the present case . clearly , however , the small inflection discussed above marks a small deviation from stable clustering .
in this paper , we have developed a lagrangian perturbation theory method for solving , beyond crossing time , vlasov - poisson equations in the expanding universe for initially cold systems in the infinite parallel planes geometric set up .
the proposed approach captures post - collapse dynamics by computing at leading - order a counter term to the zeldovich solution just after first crossing time .
it extends earlier work of @xcite to the cosmological case and , thanks to adaptive smoothing , to random initial conditions instead of a single halo . by performing a local taylor expansion of the velocity and position of each mass element as functions of lagrangian
coordinate around each initial density peak , we are able to compute the force field in the multi - valued region . a correction to the single - stream flow motion , described by the zeldovich solution , is systematically obtained .
the results are given as polynomial functions of lagrangian position and the coefficient of each power is expressed in terms of quantities associated to density peaks .
a simple analysis of the formation of a single cluster shows that post - collapse pt reproduces well the phase - space structure up to next - crossing time and can describe the projected density profile quite well even at later time . to deal with the bulk properties of phase - space structures , we propose an adaptive filtering scheme
the idea is very close to the peak - patch treatment by @xcite . in particular ,
the local smoothing scale of initial conditions is determined by the calculation of a dynamical time , but in post - collapse pt this dynamical time is given by next - crossing time instead of collapse time .
this allows us to produce a realistic coarse - grained version of the phase - space structure , even in the presence of mergers . after demonstrating that adaptive smoothing indeed works well in a simple setup with coalescing clusters
, we applied it to more general , random initial conditions that include cdm - like as well as scale - free power spectra .
the results of our analyses show that post - collapse pt predictions with adaptive smoothing reproduce the power spectrum of our numerical simulations remarkably well even at scales sufficiently nonlinear to probe the beginning of the stable clustering regime . on important thing
to keep in mind for further work is that the critical quantity in post - collapse pt is the time of next crossing , which can in fact be used to also create a spectacularly good pure zeldovich prescription with adaptive smoothing .
while the calculations of post - collapse pt can be cumbersome , the expression ( [ eq : tau_cross ] ) of this time is relatively simple and this can be very interesting when trying to generalize results of post - collapse pt to the three - dimensional case . when analyzing equation ( [ eq : tau_cross ] ) , one indeed notes that @xmath198 depends both on collapse time @xmath53 and on the height @xmath323 of the initial density peak . since @xmath53 depends solely on @xmath323 , we can see that basically the dynamical time we consider for defining the local smoothing scale is a single function of @xmath323 , larger than @xmath53 .
in fact , things are very simple : when performing adaptive smoothing one enforces @xmath324 to be equal to present time , or in other words , adaptive smoothing of initial conditions is performed so that the local peak height given by linear theory at present time , @xmath323 , is a number larger than unity ( instead of exactly unity for the collapse time ) . for instance , in the einstein - de sitter case , @xmath325 and @xmath237 , we have @xmath326 with @xmath210 chosen in this article to be equal to unity for post - collapse pt and one - half for the zeldovich solution .
note that the optimal value of @xmath210 is smaller than these intuitive choices ( see appendix [ sec : choice_of_f_cross ] ) .
the present post - collapse pt treatment together with the adaptive smoothing scheme seems a very promising tool .
a successful extension to the 3d case should lead to a major breakthrough in the precision pt calculations of large - scale structure statistics beyond the single - stream approximation .
our 1d study is however only a first step . toward a practical application to the 3d case
, there are several issues to be addressed .
one is the analytic calculation of statistical quantities such as the power spectrum .
indeed , in the present work , we measured directly the power spectrum from a particle distribution following the dynamics prescribed by the various pt prescriptions under scrutiny .
in fact , analytic calculations of the power spectrum are tractable , at least in the presence of fixed smoothing .
adaptive smoothing itself can in principle be implemented in the analytic framework by inspiring for instance from the theory of excursion sets discussed in e.g. @xcite , although we can foresee that actual calculations will probably be very cumbersome .
another issue , when generalizing to the 3d case , is that the zeldovich solution no longer provides a sufficiently accurate description of the dynamics before collapse .
at least third - order lagrangian pt is required to estimate correctly collapse times as well as the power spectrum beyond tree - level approximation , which clearly adds some significant level of complexity .
furthermore , the variety and complexity of topological configurations of singularities appearing in the 3d case might in fact represent a nearly impassable obstacle in three dimensions ( see , e.g. * ? ? ?
we will tackle these issues in future work . finally ,
while we have focused in this article on the impact of post - collapse dynamics in 1d on the phase - space structure and on the power spectrum of the matter distribution , it would be interesting to see quantitatively how the small - scale modes responsible for the multi - stream flows couple to large - scale fluctuations described by the single - stream flows .
this can be presumably clarified by measuring the response of the evolved power spectrum to a small initial perturbation of high-@xmath186 modes , through the comparison between a pair of particle distributions modeled with post - collapse pt that slightly differ in their initial conditions @xcite . in the 3d case
, a strong damping is found for the mode transfer from small to large scales , but the physical origin of it is still unclear .
a detailed analysis in the 1d cosmological setup would provide important clues on the nature of mode transfers .
at thanks patrick valageas for references and useful comments on the adhesion model .
sc greatly acknowledges hospitality of yukawa institute for theoretical physics ( yitp ) , where this work was initiated .
we also acknowledge the support of yitp in organizing the workshop `` vlasov - poisson : towards numerical methods without particles '' in kyoto , funded by grant yitp - t-15 - 02 , anr grant anr-13-monu-0003 and by institut lagrange de paris ( anr-10-labx-63 and anr-11-idex-0004 - 02 ) . this work was also supported in part by mext / jsps kakenhi grant numbers jp15h05889 and jp16h03977 ( at ) as well as anr grant anr-13-monu-0003 ( sc ) .
in this appendix , we derive the expression in eq .
( [ eq : phi_poisson ] ) . to do this ,
we first consider the 1d green function which satisfies the following equation : @xmath327 with this green function , the solution of the poisson equation is formally expressed as @xmath328 to solve eq .
( [ eq : green_func ] ) under periodic boundaries condition , we move to harmonic space and express the green function as : @xmath329 recalling that the fourier transform of dirac s delta function reads @xmath330 $ ] , . ]
substitution of eq .
( [ eq : g_fourier ] ) into eq .
( [ eq : green_func ] ) gives @xmath331 thus , we obtain @xmath332 note that there is a subtlety in the above equation . in deriving it ,
we ignore the contribution from @xmath288 mode in @xmath333 , while the @xmath288 mode does not vanish in dirac s delta function .
the result remains however correct as long as the density field does not contain a @xmath288 mode , or equivalently , @xmath334 .
keeping this point in mind , we can further simplify the expression .
we use the following formula ( see , e.g. chap.1.44 of * ? ? ?
* ; * ? ? ?
* ) : @xmath335 the final expression for the green function is then given by @xmath336,\qquad |x - x'|\leq l.\end{aligned}\ ] ] substituting this into eq .
( [ eq : integral ] ) , we obtain the expression in eq .
( [ eq : phi_poisson ] ) .
in this appendix , we derive the expression for the force in the multi - stream region , given in eqs .
( [ eq : force_inner ] ) , ( [ eq : force_inner_q ] ) and ( [ eq : force_inner_const ] ) . to start with
, we divide the integral in eq .
( [ eq : force_poisson ] ) into several pieces according to eq .
( [ eq : part_by_part_integrals ] ) : @xmath337 .
\label{eq : force_pieces}\end{aligned}\ ] ] we shall below evaluate these integrals one by one . consider the first integral in right - hand side of eq .
( [ eq : force_pieces ] ) .
we evaluate it at position @xmath338 in fig .
[ fig : shellcrossing ] . using eq .
( [ eq : jacobian_delta ] ) , the integral is rewritten @xmath339.\end{aligned}\ ] ] exploiting the geometric properties of the configuration in fig .
[ fig : shellcrossing ] , substitution of the explicit form of @xmath340 ( eq . [ eq : ballistic_shellcrossing ] ) into the above leads to @xmath341 here we used the fact that @xmath71 .
recalling that @xmath342 and @xmath343 are related with each other through eqs .
( [ eq : relation_three - values_1 ] ) and ( [ eq : relation_three - values_2 ] ) , eq .
( [ eq : integral_inner1_mid ] ) can be recast as a single function of either @xmath342 or @xmath343 , which respectively provides the general expression valid for @xmath180 and @xmath344 . omitting the subscript @xmath345 or @xmath346 in either cases , we finally obtain @xmath347 next , we consider the second integral in eq .
( [ eq : force_pieces ] ) . rewriting the integrand in terms of the variable @xmath81 with the help of eqs .
( [ eq : jacobian_delta ] ) and ( [ eq : ballistic_shellcrossing ] ) , it is straightforward to obtain @xmath348 the remaining integrals to be computed , i.e. , the third and fourth terms in eq .
( [ eq : force_pieces ] ) , are performed over intervals outside the multi - valued region , which allows us to use zeldovich solution to evaluate the integrands . for the third integral in eq .
( [ eq : force_pieces ] ) , we obtain @xmath349 although this expression is exact , our perturbative treatment implicitly assumes that the multi - stream region is small and the quantities inside this region are written in polynomial forms of @xmath81 .
we may expand the last two terms of eq .
( [ eq : integral_outer1_mid ] ) as @xmath350 where we used the fact that @xmath351 ( see eq . [ eq : shellcrossing_peak ] ) .
additionally , periodic boundaries imply @xmath352 .
therefore , eq . ( [ eq : integral_outer1_mid ] ) can be recast as @xmath353_{q_0}^{q(x = l ) } d_+(\tau ) .
\label{eq : integral_outer1}\end{aligned}\ ] ] in similar manner , we evaluate the fourth integral in eq .
( [ eq : force_pieces ] ) as follows : @xmath354 the first of these two integrals is performed exactly to give @xmath355_{q(x=0)}^{q(x = l ) } -\frac{1}{2 } \bigl[\psi(q')^2\bigr]_{q(x=0)}^{q(x = l)}\ , \nonumber\\ & \quad\quad + \int_{q(x=0)}^{q(x = l ) } { \rm d}q'\,\psi(q')\biggr\}\,d_+(\tau ) \nonumber\\ & \quad = -\psi(q(x=0))\,d_+(\tau ) , \label{eq : outer2_part1}\end{aligned}\ ] ] where we exploited properties of periodic boundaries , @xmath356 and @xmath357 , as well as momentum conservation , which implies @xmath358 . turning to the second integral in eq .
( [ eq : outer2_step1 ] ) , we use again the fact that the domain of integration is assumed to be narrow to taylor expand the integrand using the formula @xmath359 to obtain the following expression , valid at third - order in @xmath74 : @xmath360 here , eqs .
( [ eq : def_a])([eq : def_c ] ) are used to rewrite the expression in the last line .
summing up the results in eqs .
( [ eq : outer2_part1 ] ) and ( [ eq : outer2_part2 ] ) , we obtain , at the end , @xmath361 plugging eqs .
( [ eq : integral_inner1 ] ) , ( [ eq : integral_inner2 ] ) , ( [ eq : integral_outer1 ] ) and ( [ eq : integral_outer2 ] ) into eq .
( [ eq : force_pieces ] ) , we finally obtain the expression of the force in the multi - valued region ( eqs .
[ eq : force_inner ] , [ eq : force_inner_q ] and [ eq : force_inner_const ] ) .
in this appendix , we detail the calculations of basic post - collapse pt predictions summarized in eqs .
( [ eq : p_for_tau_c > tau > hattau_c ] ) , ( [ eq : x_for_tau_c > tau > hattau_c ] ) , ( [ eq : p_for_tau > tau_c ] ) and ( [ eq : x_for_tau > tau_c ] ) , together with the coefficients in table [ tab : xv_pcpt1 ] and [ tab : xv_pcpt2 ] .
our starting point is eq .
( [ eq : v_ii ] ) , which involves two integrals . to evaluate the first integral , the approximate formula given by eq .
( [ eq : integ_formula_j ] ) is applied . recalling that we are considering the outer part of the multi - valued region , @xmath180 , we have @xmath363 , \label{eq : integ_j1}\end{aligned}\ ] ] with @xmath364 . here , we used the leading - order expression for critical time @xmath82 , @xmath365 , and ignore in eq .
( [ eq : integ_formula_j ] ) the higher - order terms involving the quantities @xmath176 or @xmath177 . on the other hand , the second integral in eq .
( [ eq : v_ii ] ) is analytically performed . using the evolution equation in eq .
( [ eq : evolv_lin ] ) , we obtain @xmath366_{\widehat{\tau}_{\rm c}(q)}^\tau .
\label{eq : integ_f}\end{aligned}\ ] ] summing up the above two contributions , we obtain the expression for @xmath362 summarized in eq .
( [ eq : p_for_tau_c > tau > hattau_c ] )
. to derive the expression for @xmath126 , we just need to integrate over time eq .
( [ eq : p_for_tau_c > tau >
hattau_c ] ) , as written in eq .
( [ eq : x_ii ] ) : @xmath367 \nonumber \\ & ~\quad + \int_{\widehat{\tau}_{\rm c}(q)}^\tau { \rm d}\tau'\,\,\widehat{\epsilon}_1 \,(\tau',\widehat{\tau}_{\rm c } ) .
\label{eq : x_ii_prime}\end{aligned}\ ] ] making use of the expressions summarized in table [ tab : xv_pcpt1 ] , we evaluate the first integral .
a straightforward calculation leads to @xmath368 \nonumber\\ & \simeq \frac{t^2}{2}\,q+ \bigl\{-\frac{\kappa}{8}\,t+\frac{\delta_{\rm l}''(q_0)}{6}d_+(\tau_0 ) \ , \bigr\}q^3 \nonumber\\ & \quad -\mbox{sgn}(q)\,\frac{\kappa^2}{80\sqrt{3}}\ , \left\{{\widehat{q}_{\rm c}}^2(\tau)-q^2\right\}^{5/2 } \nonumber\\ & \quad + \biggl[\frac{1}{2}\left(\frac{\kappa}{8}\right)^2-\left(\frac{\kappa}{8}\right)\,\frac{\delta_{\rm l}''(q_0)}{6}d_+(\tau_0)\,t\,\biggr]q^5 \nonumber\\ & \quad+ \frac{1}{2}\left(\frac{\kappa}{8}\right)^2 \frac{\delta_{\rm l}''(q_0)}{6}d_+(\tau_0)q^7 .
\label{eq : tilde_x_i}\end{aligned}\ ] ] note again that we used the leading - order expression for @xmath82 ( see eq . [ eq : hattau_c ] ) .
the second integral in eq .
( [ eq : x_ii_prime ] ) is analytically performed to give @xmath369^{\tau'}_{{\widehat{\tau}_{\rm c}}(q ) } \nonumber\\ & \quad = \psi(q_0)\,\bigg\ { d_+(\tau)-d_+({\widehat{\tau}_{\rm c}}(q ) ) \nonumber\\ & \quad \quad \quad \quad -\left.\left.\frac{{\rm
d}d_+}{{\rm d}\tau}\right|_{{\widehat{\tau}_{\rm c}}(q)}(\tau-{\widehat{\tau}_{\rm c}}(q))\right\}. \label{eq : integ_eps_1}\end{aligned}\ ] ] collecting eqs . ( [ eq : tilde_x_i ] ) and ( [ eq : integ_eps_1 ] ) , we obtain eq .
( [ eq : x_for_tau_c > tau > hattau_c ] ) . for @xmath153 ,
the expression to be evaluated is given by eq .
( [ eq : delta_p_2 ] ) , which involves three integrals .
the first integral is obtained from eq .
( [ eq : integ_j1 ] ) by setting @xmath370 ( eq . [ eq : tau_c ] ) .
recalling the fact that @xmath371 ( eq . [ eq : qchat_qc ] ) , we have @xmath372.\end{aligned}\ ] ] for the second integral in eq .
( [ eq : delta_p_2 ] ) , we use the approximate formula given by eq .
( [ eq : integ_formula_j ] ) in the regime @xmath373 . ignoring higher - order terms involving the quantities @xmath176 or @xmath177
, we obtain @xmath374.\end{aligned}\ ] ] finally , the last integral in eq .
( [ eq : delta_p_2 ] ) is evaluated using the formula given in eq .
( [ eq : integ_f ] ) . collecting the three contributions above
, we obtain the expression for @xmath153 ( eq .
[ eq : p_for_tau > tau_c ] ) .
the starting expression for @xmath119 is given in eq .
( [ eq : x_iii ] ) , which involves two integrals . to evaluate the first integral in eq .
( [ eq : x_iii ] ) , we use eq .
( [ eq : tilde_x_i ] ) . setting @xmath375 , an expression at seventh - order in @xmath81
is obtained : @xmath376 to compute the second integral in eq .
( [ eq : x_iii ] ) , we substitute eq .
( [ eq : p_for_tau > tau_c ] ) into the integrand , but without the term @xmath377 . using the explicit expression for the coefficients in table [ tab : xv_pcpt2 ] , a straightforward calculation gives @xmath378 \nonumber\\ & ~=-\frac{3}{2}h_0 ^ 2\,\omega_{\rm m,0}\,a(\tau_0)\ , \nonumber\\ & \quad \times\biggl(\,-t^2\,q+ \left\{\frac{5\kappa}{8}\,t+\frac{\delta_{\rm l}''(q_0)}{12}\,d_+(\tau_0 ) t^2\right\}\,q^3 \nonumber\\ & \qquad\quad- \bigl\{\left(\frac{\kappa}{4}\right)^2\ , + \left(\frac{\kappa}{8}\right)\frac{\delta_{\rm d}''(q_0)}{6}\,d_+(\tau_0)\,t\ , \bigr\}\,q^5 \nonumber\\ & \qquad\quad -4\left(\frac{\kappa}{8}\right)^2\frac{\delta_{\rm l}''(q_0)}{6}\,d_+(\tau_0)\,q^7 \biggr ) .
\label{eq : integ_x_ii_part2}\end{aligned}\ ] ] combining the above two results , we obtain eq .
( [ eq : x_for_tau > tau_c ] ) , which gives the final expression for @xmath119 along with the coefficients in table [ tab : xv_pcpt2 ] .
in this appendix , we detail calculations of the various higher - order corrections to post - collapse pt we propose in [ subsec : higher - order ] . one possible way to improve on basic post - collapse pt is to taylor expand at next - to - leading order the expression for @xmath74 ( as well as @xmath75 ) that determines the boundary of the post - collapse region ( eqs .
[ eq : qchat_qc ] and [ eq : qchat_approx ] ) : @xmath379 with @xmath380 inverting the relation @xmath381 [ @xmath382 , we obtain the corresponding critical times at next - to - leading order , @xmath383 and @xmath384 using these expressions , we repeat below the same calculations as in sec . [ subsec : corrections_pcpt ] .
higher - order expressions for the correction to zeldovich flow are derived on top of basic post - collapse pt results and are denoted by @xmath385 and @xmath386 . in sec .
[ subsec : higher - order ] , a semi - analytic treatment designed by _ spl _ is also discussed . in this case
, we adopt the basic post - collapse pt results in the inner part of the multi - stream region , but the extension of this latter is calculated using the higher - order expression ( [ eq : higher_order_qhat ] ) for @xmath75 .
then , a straightforward third - order spline interpolation , that we do not detail here , is used to connect the inner part to the zeldovich solution , assuming again that the lagrangian boundary @xmath74 of the multi - stream region is given by the higher - order expansion ( [ eq : higher_order_qhat ] ) .
another improvement may come from the calculation of the integrals in eqs .
( [ eq : delta_p_formal ] ) and ( [ eq : delta_x_formal ] ) . in sec .
[ subsec : corrections_pcpt ] , a part of the integrands is taylor - expanded in time and the integration is performed for the leading - order terms .
the resultant expressions include time - dependent terms up to @xmath387 for the velocity and @xmath388 for the position . here , using the approximate integral formulae for the taylor - expanded integrands in appendix [ sec : integral ] , we keep terms up to @xmath388 for the velocity , @xmath389 for the position , and derive the corresponding higher - order corrections to the zeldovich flow , @xmath390 and @xmath391 .
these corrections assume , of course , higher - order expansions for critical times @xmath82 and @xmath170 .
below , we present the expressions for @xmath390 and @xmath391 .
the expressions for @xmath385 and @xmath386 are obtained by simply setting @xmath176 and @xmath177 to zero .
the expression for the higher - order correction to the velocity , @xmath393 , is @xmath394 , \label{eq : p_higher_tau_c > tau > hattau_c_higher}\end{aligned}\ ] ] with @xmath395 .
\label{eq : def_delta_zeta_1_higher}\end{aligned}\ ] ] here , we have introduced the following function : @xmath396 where @xmath177 is the value of the hubble parameter at time @xmath53 .
for the higher - order correction to the position , @xmath397 , we have @xmath398 , \label{eq : x_higher_tau_c > tau > hattau_c_higher}\end{aligned}\ ] ] with @xmath399 note the term proportional to @xmath400 , which is necessary to enforce continuity of the solution .
the expression for the higher - order corrections to the velocity , @xmath402 , becomes @xmath403 , \label{eq : p_higher_tau > tau_c}\end{aligned}\ ] ] with @xmath404 , \\ & \widetilde{\beta}_3(\tau)= \frac{g(\tau_0)}{2}t^2\,\frac{\delta_{\rm l}''(q_0)}{6}\,d_+(\tau_0 ) , \\ & \widetilde{\delta}_3(\tau_0)=\frac{47}{64}\kappa^2\,\eta -\frac{\kappa^2}{128 } \bigl[\,g(\tau_0)-48\{a(\tau_0)\}^2\,h(\tau_0)\bigr ] , \\
& \widetilde{\zeta}_3(\tau_0)=-\left(\frac{\kappa}{8}\right)^2\ , \left\{\,\frac{g(\tau_0)}{2}+\eta\,\right\ } \frac{\delta_{\rm l}''(q_0)}{6}\,d_+(\tau_0 ) \nonumber\\ & \qquad\qquad-\left(\frac{\kappa}{8}\right)^3\,\eta\ \bigl[\,g(\tau_0)-192\{a(\tau_0)\}^2\,h(\tau_0)\bigr].\end{aligned}\ ] ] finally , the higher - order corrections to the position , @xmath405 , are given by @xmath406 , \label{eq : x_higher - order_tau > tau_c}\end{aligned}\ ] ] with @xmath407 , \label{eq : def_delta_alpha_4 } \\ & \widetilde{\beta}_4(\tau)= \frac{\delta_{\rm l}''(q_0)}{6}\,d_+(\tau_0 ) \frac{g(\tau_0)}{6}\,t^3 , \label{eq : def_delta_beta_4 } \\
& \widetilde{\delta}_4(\tau)= -\left(\frac{\kappa}{4}\right)^2\,\left(\frac{g(\tau_0)}{8}\,t-6\{a(\tau_0)\}^2h(\tau_0)\,t\,\right ) \nonumber\\ & \qquad\quad+\,47\left(\frac{\kappa}{8}\right)^2\,\eta\,t , \label{eq : def_delta_delta_4 } \\ & \widetilde{\zeta}_4(\tau)= -\left(\frac{\kappa}{8}\right)^2 t \left(\frac{g(\tau_0)}{2}+\eta\right)\frac{\delta_{\rm l}''(q_0)}{6}\,d_+(\tau_0 ) \nonumber\\ & \qquad\quad + \left(\frac{\kappa}{8}\right)^3\biggl\{\,\frac{g(\tau_0)}{3}-64\,\{a(\tau_0)\}^2h(\tau_0 ) \nonumber\\ & \qquad\quad -\eta\,\bigl(95+g(\tau_0)\,t-192\,\{a(\tau_0)\}^2h(\tau_0)\,t\bigr)\,\biggr\ } , \label{eq : def_delta_zeta_4 } \\ & \widetilde{\mu}_4(\tau)=\kappa^3\,\delta_{\rm l}''(q_0)\,d_+(\tau_0 ) \left(\frac{g(\tau_0)}{9216}+\frac{1}{3072}\eta-\frac{g(\tau_0)\,\eta\,t}{3072}\right ) \nonumber\\ & \qquad\quad + \kappa^4\,\eta\,\biggl\{\frac{g(\tau_0)}{4096}-\frac{3}{16}\,\{a(\tau_0)\}^2h(\tau_0)-\frac{767\eta}{8092 } \nonumber\\ & \qquad\quad -\frac{g(\tau_0)\,\eta\,t}{8192}+\frac{3}{32}\eta\,\{a(\tau_0)\}^2h(\tau_0)\,t\biggr\ } , \nonumber\\ & \widetilde{\nu}_4(\tau)=\frac{1}{49152}\,\kappa^4\eta^2\delta_{\rm l}''(q_0)\,d_+(\tau_0 ) .
\label{eq : def_delta_nu_3}\end{aligned}\ ] ]
in this appendix , we provide useful formulae for primitive integrals over time , which are used to derive the post - collapse pt results in appendices [ sec : pcpt_basic ] and [ sec : pcpt_higher - order ] .
our aim here is to evaluate the time - integral of the force in the multi - valued region , given in eq .
( [ eq : force_inner ] ) .
the time - integral of the second term , involving function @xmath408 , is performed exactly ( see eq .
[ eq : integ_f ] ) . on the other hand ,
the integral of the first term needs an approximate treatment which will be valid as long as the multi - valued region is sufficiently small , or equivalently , shortly after collapse time @xmath53 .
below , we evaluate the integral by taylor - expanding the integrand .
we first give the results term by term : @xmath409 where @xmath176 is given by eq .
( [ eq : def_t_g ] ) . for the term with fractional power of time ,
we obtain @xmath410 where we used the leading - order approximation for @xmath74 , i.e. , @xmath411 . summing up the above formulas
, we obtain the primitive of the time integrals of the force involving function @xmath412 .
the result including polynomials of time up to @xmath413 is @xmath414_{\tau_{\rm i}}^{\tau_{\rm f}}\,q \\ \\ \qquad+\frac{1}{6}\delta_{\rm l}''(q_0)\,d_+(\tau_0)\ , \\ \\
\qquad\times\ , \bigl[\ , ( \tau-\tau_0)+ \frac{1}{2}g(\tau_0)\,(\tau-\tau_0)^2 \bigr]_{\tau_{\rm i}}^{\tau_{\rm f}}\,q^3 \\ \qquad - \mbox{sgn}(q)\,{\displaystyle \frac{\kappa}{4\sqrt{3}}}\,\bigl[\ , \left({\widehat{q}_{\rm c}}(\tau)^2-q^2\right)^{3/2 } \bigr]_{\tau_{\rm i}}^{\tau_{\rm f } } \,\biggr\ } \\ \qquad\qquad\qquad\qquad\qquad\qquad;\quad q_{\rm c}<|q|<{\widehat{q}_{\rm c } } , \\ \\
a(\tau_0)\,\biggl\{\ , \bigl[\,-2 ( \tau-\tau_0)+ \frac{1}{2}\bigl\{-3\,h(\tau_0)\,a(\tau_0)^2 \\ \qquad + g(\tau_0)\bigr\}(\tau-\tau_0)^2\ , \bigr]_{\tau_{\rm i}}^{\tau_{\rm f}}\,q + \frac{1}{6}\delta_{\rm l}''(q_0)\,d_+(\tau_0 ) \\
\qquad \times \bigl[\,(\tau-\tau_0)+ \frac{1}{2}g(\tau_0)\,(\tau-\tau_0)^2\ , \bigr]_{\tau_{\rm i}}^{\tau_{\rm f}}\,q^3 \biggr\ } \\ \qquad\qquad\qquad\qquad\qquad\qquad ; \quad |q|<q_{\rm c}. \end{array } \right . \label{eq : integ_formula_j}\end{aligned}\ ] ]
in this appendix , we study the effect of varying the parameter @xmath210 on the performances of pt predictions with adaptive smoothing ( see sec . [ sec : improvement ] for definition ) . in the main text
, we adopt the intuitive setup @xmath249 for post - collapse pt and @xmath247 for zeldovich solution .
however , pt predictions are prone , after shell - crossing , to get worse over time , so we can naively expect that choosing a smaller value of @xmath210 will provide a better agreement with simulations .
[ fig : pk_xv_fcross ] shows the results obtained in the case of the cdm - like power spectrum studied in [ subsec : cdm_initial ] when varying the value of @xmath210 . here
, the calculations are performed using @xmath415 , @xmath416 and @xmath417mpc@xmath257 for a boxsize @xmath290 mpc , which is a relatively low resolution set up , but this will not have any consequence on the discussion that follows .
left panel shows the power spectrum at @xmath283 and the resultant predictions for different values of @xmath210 are plotted as solid lines with different colors .
as anticipated , decreasing @xmath210 results in an enhancement of small - scale power and a better agreement between post - collapse pt and the simulation is obtained . but
a too small value of @xmath210 overshoots the simulation and an optimal choice of @xmath210 roughly corresponds to @xmath418 .
the same trend can be also seen for the zeldovich solution , but the improvement of the power spectrum amplitude is rather mild : while @xmath419 seems to provide the best overall behavior for the zeldovich solution , the result is clearly still not as good as what post - collapse pt can provide , in particular in the regime @xmath420 mpc@xmath257 .
right panel of fig .
[ fig : pk_xv_fcross ] shows the phase - space portraits and compares the simulation results ( red ) to predictions obtained with our `` standard '' value of @xmath249 for post - collapse pt ( upper insert ) and @xmath421 for zeldovich solution ( lower insert ) and to predictions obtained with respective nearly optimal values of @xmath284 and @xmath285 .
it seems difficult to judge by eye whether the prediction with a smaller value of @xmath210 really improves the description at small scales , although more structures seems to be captured for post - collapse pt with @xmath249 , which contradict the results obtained for the power spectrum .
however the simple examination of this figure corresponding to a single realization of the random initial conditions is not conclusive , obviously : it does not preclude the fact that a smaller value of @xmath210 can give a better result for the power spectrum obtained from averaging over many realizations .
finally , to conclude this section , although we do not show the results here , we noticed as well that the best choice of @xmath210 could vary slightly according to redshift or initial condition : for example , for the cdm cosmology considered here , a better choice of @xmath210 at redshift @xmath282 is @xmath422 and @xmath423 respectively for post - collapse pt and zeldovich solution .
this finally explains why we decided to keep , for simplicity , the generic values @xmath249 and @xmath421 respectively for post - collapse pt and zeldovich solution , even if they are sub - optimal . | we develop a new perturbation theory ( pt ) treatment that can describe gravitational dynamics of large - scale structure after shell - crossing in the one - dimensional cosmological case . starting with cold initial conditions ,
the motion of matter distribution follows at early stages the single - stream regime , which can , in one dimension , be described exactly by the first - order lagrangian perturbation , i.e. the zeldovich solution .
however , the single - stream flow no longer holds after shell - crossing and a proper account of the multi - stream flow is essential for post - collapse dynamics . in this paper , extending previous work by ( * ? ? ?
* mnras 446 , 2902 ) , we present a perturbative description for the multi - stream flow after shell - crossing in a cosmological setup . in addition , we introduce an adaptive smoothing scheme to deal with the bulk properties of phase - space structures .
the filtering scales in this scheme are linked to the next - crossing time in the post - collapse region , estimated from our pt calculations .
our pt treatment combined with adaptive smoothing is illustrated in several cases .
predictions are compared to simulations and we find that post - collapse pt with adaptive smoothing reproduces the power spectrum and phase - space structures remarkably well even at small scales , where zeldovich solution substantially deviates from simulations .
[ firstpage ] large - scale structure cosmology vlasov - poisson equation |
it is a well known and intriguing fact that , in the genetic code , 64 codons code the biosynthesis of 20 amino - acids ( a.a . ) with a structure in multiplets reported in table [ tablerep ] .
+ it is also a well known and , at our knowledge , unexplained fact that the frequency rate of usage ( codon usage ) of the different codons inside a multiplet is not the same .
+ it is the aim of this paper to emphasize for the vertebrate series a correlation in the codon usage , in the quartets and sextets , which is naturally explained in the framework of the mathematical model of the genetic code recently proposed by the authors @xcite .
moreover we put in evidence an universal function behaviour connected with the codon usage , which also finds a justification in the model .
2 we recall the essential features of the model and in sec .
3 we present the analysis of the codon usage for a set of biological sequences in the vertebrate series @xcite .
let us briefly recall some properties of the crystal basis @xcite .
we limit ourselves to the case of @xmath0 , but such basis exists for any finite dimensional representation of @xmath1 , @xmath2 being any ( semi)-simple classical lie algebra .
the crystal basis has the nice property , for @xmath3 , that in the limit @xmath4 @xmath5 and @xmath6 where the operators @xmath7 are a redefinition , using an element of the center , of the generators @xmath8 of @xmath3 , @xmath9 , and @xmath10 are the basis vectors of the irreducible representation labelled by @xmath11 ( @xmath11 being an integer or half - integer ) .
the labels of the irreducible representation are connected to the eigenvalues of the `` casimir '' operator @xmath12 : @xmath13 its eigenvalue on any vector basis of the irreducible @xmath11-representation is @xmath14 .
+ moreover any state in the tensor product of two irreducible representations @xmath15 is written in the crystal basis as one and only one tensor product of a @xmath16 state by a @xmath17 state @xcite . for example , taking for @xmath16 and @xmath17 the two - dimensional representation @xmath18 of @xmath19 with states @xmath20 and @xmath21 , one will get in @xmath15 the @xmath22 representation displayed by @xmath23 , @xmath24 and @xmath25 , and the @xmath26 representation with the state @xmath27 .
now let us state the main hypothesis of our model @xcite .
* assumption i * - the four nucleotides containing the bases : adenine @xmath28 and guanine @xmath29 , deriving from purine , and cytosine @xmath30 and thymine @xmath31 , coming from pyrimidine , are the basis vectors of a crystal basis of the @xmath32 irreducible representation of the quantum algebra @xmath33 in the limit @xmath4 . in the following ,
we denote with @xmath34 the basis vector corresponding to the eigenvalue @xmath35 of @xmath36 , where @xmath37 specifies the generator of the first ( second ) @xmath38 .
we assume the following `` spin '' structure ( we remind that the thymine @xmath39 in the dna is replaced by the uracile @xmath40 in the rna ) : @xmath41 let us remark that the @xmath42-symmetry is associated to the purine - pyrimidine structure , while the @xmath43-symmetry reflects the complementarity rule ( that is @xmath44 and @xmath45 interactions )
. + * assumption ii * - the _ codons _ are the basis vectors , in the crystal basis , of the irreducible representations build up by the tensor product of three 4-dimensional @xmath46 fundamental representations describing the nucleotides .
+ we have reported in table [ tablerep ] the assignment of the codons classified in the representations which appear in the r.h.s . of the following relation : @xmath47 in @xcite , an operator ( called the _ reading or ribosome operator _ )
@xmath48 has been constructed out of the algebra @xmath49 , which describes the multiplet structure of the the genetic code in the following way : _ two codons have the same eigenvalue under @xmath48 if and only if they are associated to the same amino - acid_. moreover an `` hamiltonian '' depending on 4 parameters has been build up which gives a very satisfactory fit of the 16 values of the free energy released in the folding of a rna sequence into a base paired double helix .
let us close this section by drawing the reader s attention to fig .
[ figa ] where is specified for each codon its position in the appropriate representation .
the diagram of states for each representation is supposed to lie in a separate parallel plane .
thick lines connect codons associated to the same amino - acid .
one remarks that each segment relates a couple of codons belonging to the same representation or to two different representations .
this last case occurs for quadruplets or sextets of codons associated to the same amino - acid .
it is the purpose of this letter to show a remarkable relation between such multiplets of codons ( or amino - acids ) involving the same subset of representation and ( branching ratios of ) the probabilities of presence of codons in the amino - acid biosynthesis .
we define the codon usage as the frequency of use of a given codon in the process of biosynthesis of all the amino - acids .
we define the probability of usage of the codon @xmath50 of a given amino - acid as the ratio between the occurrence of the codon @xmath50 and the occurrence @xmath51 of the corresponding amino - acid , i.e. as the relative codon frequency , in the limit of very large @xmath51 . here and in the following the labels
@xmath52 represent the bases @xmath53 .
the frequency rate of usage of a codon in a multiplet is connected to its probability of usage @xmath54 .
it is reasonable to assume that @xmath54 depends on : the biological organism ( b.o . )
from which the sequence considered has been extracted + the sequence analyzed + the nature of the neighboring codons in the sequence + the amino - acid ( a.a . ) + the nature and structure of the multiplet associated to the amino - acid + the biological environment + the properties of the codon itself ( @xmath50 ) .
we neglect the time in which the biosynthesis process takes place as we assume that the biosynthesis processes are considered at the same time , at least compared to the time scale of evolution of the genetic code .
we define the _ branching ratio _ @xmath55 as @xmath56 we argue that in the limit of very large number of codons , for a fixed biological organism and amino - acid , the branching ratio depends essentially on the properties of the codon . in our model
this means that in this limit @xmath55 is a function , depending on the type of the multiplet , on the _ quantum numbers _ of the codons @xmath50 and @xmath57 , i.e. on the labels @xmath58 , where @xmath59 or @xmath43 , and on an other set of quantum labels leaving out the degeneracy on @xmath60 ; in table [ tablerep ] different irreducible representations with the same values of @xmath60 are distinguished by an upper label
. moreover we assume that @xmath55 , in the limit above specified , depends only on the irreducible representation ( ir ) of the codons , i.e. : @xmath61 let us point out that the branching ratio has a meaning only if the codons @xmath50 and @xmath62 are in the same multiplet , i.e. if they code the same amino - acid . + we consider the quartets and sextets .
there are five quartets and three sextets in the eukariotic code : that will allow a rather detailed analysis .
moreover the 3 sextets appear as the sum of a quartet and a doublet , see table [ tablerep ] . in the following we consider only the quartet sub - part of the sextets .
we recall that the 5 amino - acids coded by the quartets are : [ pro , ala , thr , gly , val ] and the 3 amino - acids coded by the sextets are : [ leu , arg , ser ] .
there are , for the quartets , 6 branching ratios , of which only 3 are independent .
we choose as fundamental ones the ratios @xmath63 , @xmath64 and @xmath65 .
it happens that we can define several functions @xmath55 , considering ratios of probability of codons differing for the first two nucleotides @xmath66 , i.e. @xmath67 the analysis was performed on a set of data retrieved from the data bank of `` codon usage tabulated from genbank '' @xcite .
in particular we analyzed two different data set : the first one comprises all the data of at least 2 000 codons , while the second set represents all the data with at least 30 000 codons .
the referring organism for the analysis was _ homo sapiens _ , whose codon usage table derives from the analysis of more than 12 500 coding sequences , and corresponds to about 6 000 000 codons .
three quartets , coding the amino - acids pro , ala and thr , have exactly the same content in irreducible representations , see table [ tablerep ] . in table [ tabledata ] we report the 16 biological organisms with highest statistics . in figs .
[ fig1 ] , [ fig2 ] and [ fig3 ] the @xmath63 , @xmath65 and @xmath64 are reported for the 8 amino - acids coded by the quartets and sextets showing : * a clear correlation between the four amino - acids pro , ala , thr and ser . from table
[ tablerep ] we see that for these amino - acids the irreducible representation involved in the numerator of the branching ratios ( see ( [ eq8 ] ) ) is always the same : @xmath71 for @xmath63 , @xmath72 for @xmath65 , @xmath73 for @xmath64 , while the irreducible representation in the denominator is @xmath71 for the whole set .
the relative position of each of these quartets of codons can be more easily visualized in fig . [ figa ] where pro , ala , thr and ser ( quartet part ) constitute the four edges of a vertical column linking the representation @xmath71 , sitting at the ground floor , first to the representation @xmath74 , then to the @xmath72 one and finally to the representation @xmath73 , this last one located at the top floor . * a clear correlation between the two amino - acids val and leu . from table [ tablerep ]
we see that also for these two amino - acids the irreducible representation in the numerator of ( [ eq8 ] ) is the same : @xmath75 for @xmath63 , @xmath76 for @xmath65 , @xmath76 for @xmath64 , and the irreducible representation in the denominator is @xmath75 . considering fig .
[ figa ] , it is now the two representations @xmath77 and @xmath76 which are brought together , the codons associated to val and leu ( quartet part ) determining the vertices of two parallel and vertical plaquettes . * no correlation of the arg and also of the gly with the others amino - acids , in agreement with the irreducible representation assignment of table [ tablerep ] .
indeed we can note in fig .
[ figa ] that the representations @xmath78 and @xmath79 are connected by the codon quartet relative to arg and ( but ) only by this multiplet .
we also remark the gly quartet in the representations @xmath72 and @xmath80 : its position is completely different from the above discussed quartets which show up in these representations .
then in figs . [ fig4 ] , [ fig5 ] and [ fig6 ]
we have drawn the normalized branching ratios @xmath81 , @xmath82 , defined by : @xmath83 where the sum @xmath84 is extended to the eight amino - acids above listed .
the mean value and the standard deviation are : | the ratios of the codon usage in the quartets and sextets for the vertebrate series exhibit a correlated behaviour which fits naturally in the framework of the crystal basis model of the genetic code @xcite . moreover the observed universal behaviour of these suitably normalized ratios can be easily explained . * symmetry and codon usage correlations * * in the genetic code * l. frappat , p. sorba _ laboratoire de physique thorique lapth , ura 1436 , _ + _ chemin de bellevue , bp 110 , + f-74941 annecy - le - vieux , france _
+ _ e - mail : _ ` frappat(sorba)@lapp.in2p3.fr ` a. sciarrino _ dipartimento di scienze fisiche , universit di napoli `` federico ii '' _ + _ and i.n.f.n .
, sezione di napoli , italy _ + _ mostra doltremare pad .
20 , 80125 napoli , italy _
+ _ e - mail : _ ` [email protected] ` |
in 1962 , little and parks@xcite measured a mesoscopic superconducting cylinder in an axial magnetic field .
the superconducting critical temperature @xmath3 showed oscillations periodic in the normalized flux , with the period corresponding to the superconducting flux quantum @xmath4 .
these oscillations in @xmath3 are a straightforward consequence of the fluxoid quantization constraint , which was introduced by f. london@xcite .
fluxoid quantization can be easily understood by integrating the second ginzburg - landau ( gl ) equation for the supercurrent@xcite @xmath5 along a closed contour . here , @xmath6 is the supercurrent density , @xmath7 is the superfluid velocity , @xmath8 is the phase of the complex order parameter @xmath9 , and @xmath10 is the magnetic vector potential .
integration along an arbitrary closed contour yields the following equation : @xmath11 where the fluxoid @xmath12 is quantized in units of @xmath13 and @xmath14 is the applied flux threading the area inside the contour . the integer number @xmath15 is the phase winding number , or also called the fluxoid quantum number , counting the number of flux quanta @xmath16 penetrating the enclosed area . when the applied flux @xmath14 is not equal to integer times the flux quantum @xmath16 , a supercurrent @xmath17 has to be generated in order to fulfill eq .
( [ eq : flux ] ) . for a superconducting ring of radius @xmath18 , made of wires of vanishing width ( @xmath19 )
, a one - dimensional ( 1d ) gl model can be used to describe the onset of superconductivity .
the relative @xmath0 variations can be written as : @xmath20 where @xmath21 is the coherence length at zero temperature , @xmath22 is the critical temperature in zero field , and the integer number @xmath15 is chosen to maximize the critical temperature @xmath23 . for each fluxoid quantum number @xmath15 , the critical temperature @xmath3 has a parabolic shape .
the little - parks ( lp ) oscillations appear due to the transitions from the integer value @xmath15 to @xmath24 , at a half integer value of @xmath25 . for @xmath26 ,
no supercurrent flows in the ring , while for @xmath27 , the supercurrent reaches a maximal value and changes sign .
the maximum normalized variation of the critical temperature @xmath28 is @xmath29 , as can be evaluated from eq .
( [ eq : wire ] ) . to observe these quantization effects experimentally , structures with the radius of the loop of the order of the coherence length @xmath21 should be used . since the pioneering work of little and parks , single mesoscopic superconducting loops and cylinders has been largely studied .
recently , zhu @xmath30 @xmath31@xcite studied the flux state in two magnetically coupled mesoscopic normal loops .
the magnetic coupling of an array of normal@xcite and superconducting@xcite loops has also been studied . in those cases the loops are electrically isolated from each other and
can only interact magnetically .
our work will focus on the magnetic coupling of two concentric superconducting loops .
the modification of the phase boundary @xmath2 of the outer loop due to magnetic coupling with the inner loop will be studied .
magnetic coupling effects between two loops are potentially very important since for the two loops made from different materials , a new unusual effect of enhancing @xmath32 due to a higher @xmath33 can be expected .
the gl free energy for such system can be written as follows : @xmath34 with @xmath35 the total free energy in the normal state , @xmath36 and @xmath37 the expansion coefficients , @xmath38 the mass of a cooper pair and @xmath39 and @xmath40 the self- and mutual inductance .
@xmath41 is the volume of the loop , @xmath42 is the order parameter , @xmath43 is the velocity of the superfluid and @xmath44 is the supercurrent
. the indexes @xmath45 and @xmath46 refer to the inner and outer loop , respectively .
the superfluid velocities of the two loops are determined from the fluxoid quantization constraint : @xmath47 to solve this equation , the free energy must be minimized with respect to variations in @xmath48 , @xmath49 , @xmath50 and @xmath51 .
the free energy @xmath52 contains in this case the term @xmath53 responsible for the mixing of the two individual order parameters @xmath48 and @xmath49 .
we present the results of transport measurements carried out on two different types of mesoscopic al structures ( fig .
[ fig:1 ] ) . the first type of sample is composed of two concentric loops , with the outer loop being electrically connected to the experimental set - up in order to perform four - points resistance measurements .
three samples of this type have been studied : sample a and sample b with the same thickness and the same dimensions of the loops and sample c with a smaller thickness and slightly different dimensions of the loops .
the reference sample ( sample r ) is analogous to the first structure , but without the inner loop .
all the samples discussed , except sample c , are evaporated in the same run .
all samples have been prepared by thermal evaporation of 99.9999% pure al on a sio@xmath54 substrate .
the patterns are defined using electron beam lithography on a bilayer of pmma / pmma - copmma resist before the deposition of an aluminum film with a thickness @xmath55=50 nm and @xmath55=28 nm for sample a , b and r and for sample c , respectively .
after the evaporation , the lift - off was performed using dichloromethane . in fig .
[ fig:1]a , the geometry and the dimensions of the different structures are shown .
the thickness and the lateral dimensions of the samples have been characterized by the x - ray diffraction on a co - evaporated plane film and afm ( see fig .
[ fig:1]b ) , respectively .
the wire width of the loop and of the connecting wires has been determined from sem investigations .
the superconducting / normal phase boundaries are obtained from transport measurements , carried out with a transport current @xmath56 flowing through the outer loop .
the phase boundary is measured holding the resistance at a fixed resistive criterion ( we used the criterion @xmath57 , with @xmath58 the resistance in the normal state ) .
this is achieved using an electronic feedback circuit .
once a required temperature stability is obtained , the magnetic field is swept at a very slow rate ( with a typical frequency of 20 @xmath59hz ) . to improve the signal - to - noise ratio a par 124a
lock - in amplifier has been used , operating at 27.7 hz .
the width of the wires of the studied superconducting loops determines the parabolic background of the @xmath2 phase boundary@xcite : @xmath60 where @xmath61 is the width of the wires .
( [ eq : phaseb ] ) also describes the @xmath2 line for a superconducting thin film of thickness @xmath61 subjected to a magnetic field parallel to the film plane . in a practical situation
, the phase boundary of a superconducting loop of finite wire width will show a parabolic background ( see eq .
( [ eq : phaseb ] ) ) .
the suppression of @xmath0 can be written as the sum of two components : an oscillatory term as described by eq .
( [ eq : wire ] ) , and a monotonic term ( eq . ( [ eq : phaseb ] ) ) .
the coherence length @xmath21 of the samples can be determined from the parabolic background of the phase boundary , since the width of the wires @xmath61 is a known parameter .
a second method to evaluate the coherence length is to determine @xmath21 from the slope of @xmath2 of a co - evaporated macroscopic reference film .
a third method is based on the dirty limit of the gl theory@xcite , where from the known value of the elastic mean free path @xmath62 , the coherence length @xmath63 is obtained , using the clean limit value @xmath64=1.6 @xmath65 for al .
the results of the three methods are summarized in table [ table : param ] .
the difference between the @xmath21 values calculated with the three different methods can be partially explained by the rather broad error margins on the wire width @xmath61 .
ccccc sample & r & a & b & c + @xmath55 ( nm ) & 50 & 50 & 50 & 28 + @xmath61 ( nm ) & 180 & 180 & 180 & 140 + @xmath66 ( @xmath59 m ) & 2 & 2 & 2 & 1.9 + @xmath67 ( @xmath59 m ) & & 0.9 & 0.9 & 1 + @xmath22 ( k ) & 1.324 & 1.326 & 1.327 & 1.363 + r@xmath68 ( @xmath69)=@xmath70 ( 4.2 k ) & 0.65 & 0.72 & 0.76 & 1.36 + @xmath71 ( nm ) & 12.3 & 11 & 10.6 & 10.5 + @xmath72(0 ) ( eq .
( [ eq : phaseb ] ) ) ( nm ) & 103 & 105 & 102 & 117 + @xmath72(0 ) ref . sample ( nm ) & 128 & 128 & 128 &
+ @xmath72(0 ) dirty limit ( nm ) & 114 & 112 & 120 & 112 + @xmath73 ( ph ) & & 1.6 & 1.6 & 1.9 + @xmath74 ( ph ) & 4.5 & 4.5 & 4.5 & 4.6 + @xmath40 ( ph ) & & 22 & 22 & 21 + in fig .
[ fig:2 ] , the theoretical phase boundaries are shown for the two single loops when they are not coupled magnetically : one using the dimensions of the inner loop to calculate the phase boundary ( solid line ) , and the other using the size of the outer loop ( lowest dashed line ) , thus corresponding to the reference sample ; this has been calculated from eq .
( [ eq : wire ] ) using the dimensions of the loops summarized in table [ table : param ] .
it should be mentioned that an increase along the vertical axis corresponds to a decreasing temperature .
the amplitude of the lp oscillations is the larger the smaller the loop size ( see eq .
( [ eq : wire ] ) ) .
the period of the @xmath2 oscillations for the inner and outer loop is given by @xmath75 , and @xmath76 , respectively with @xmath77 and @xmath78 the enclosed areas ( see fig .
[ fig:1 ] ) .
points x and y in fig .
[ fig:2 ] are situated at two fixed applied magnetic field values on the @xmath79 phase boundary for the outer loop . in the case of point x ,
the inner loop is in the normal state , thus the inner loop carries no supercurrent .
in this situation , no flux is coupled to the outer loop . for point
y , on the contrary , the inner loop is in the superconducting state , and a supercurrent will flow in both loops in order to satisfy the respective fluxoid quantization constrains ( eq .
( [ eq : flux ] ) ) . under these conditions , due to the mutual inductance between the two current loops ( see eqs .
( [ eq : freeenergy ] ) and ( [ eq : velocity ] ) ) , an influence of the fluxoid quantization in the inner loop on the measured @xmath2 phase boundary of the outer loop is expected . to extend the flux interval for which the inner loop remains superconducting ,
a higher transport current @xmath56 can be applied to the outer loop .
the phase boundary @xmath23 of the outer loop is schematically presented in fig .
[ fig:2 ] by the dotted ( @xmath56=0.3 @xmath59a ) and the dashed - dotted ( @xmath56=0.9 @xmath59a ) lines .
we now follow through the fixed magnetic fields lines following the points x@xmath80x and y@xmath80y by increasing the transport current @xmath56 . the point on the phase boundary of the outer loop , with the same magnetic field value as x for @xmath56=0 , will cross the phase boundary of the inner loop in point x by increasing @xmath56 . for a highest transport current
, the inner loop will be superconducting for each point on the phase boundary of the outer loop ( dashed - dotted line )
. the inner loop will be deeper in the superconducting state , following the shift from y to y while increasing the transport current . as a result , an increase of @xmath56
will not only broaden the interval in which the inner loop is superconducting , but also increase the supercurrent in the inner loop .
zhang @xmath30 @xmath31@xcite have calculated the self - flux for a typical mesoscopic ring . from these calculations
, we can find the self - flux and the additional flux in the outer loop due to the presence of the inner loop .
the self - flux for the inner and outer loop will be not larger than 0.4% of the applied flux for @xmath81 .
the additional flux @xmath82 in the outer loop will be less than 4% in this temperature interval but can be higher than 20% for @xmath83 .
all the measurement presented in this paper are in the region @xmath81 .
a possibility to further increase the supercurrent in the inner loop @xmath50 and thus the magnetic coupling between the two loops would be the use of a material with a higher @xmath0 for the inner loop . in fig .
[ fig:3 ] , the phase boundaries of the samples a , b , c and r ( @xmath84 : sample a , @xmath85 : sample b , @xmath86 : sample c , @xmath87 : sample r ) are shown for a ac transport current @xmath56=0.3 @xmath59a rms after subtraction of the monotonic parabolic background due to the finite width of the strips ( eq .
further on , the flux @xmath88 will refer to the flux threading the surface of the outer loop .
the @xmath23 part below 5 @xmath16 is not shown because the experimental data were rather noisy in the low field region for some measurements .
the curves corresponding to different samples are arbitrarily shifted in fig .
[ fig:3 ] .
the variation of the critical temperature with increasing transport current can be estimated from ref@xcite for zero field : @xmath89 , with @xmath90550 @xmath59a and @xmath90240 @xmath59a for sample a , b and r and for sample c , respectively .
the critical temperatures of the different samples after extrapolation of the experimental results to @xmath56=0 are given in table [ table : param ] .
for @xmath56=0.3 @xmath59a , the shift of the zero field critical temperature @xmath91 ( used in fig .
[ fig:2 ] ) is 7 mk ( 12 mk for sample c ) , for 0.5 @xmath59a : 9 mk and for 0.7 @xmath59a : 12 mk . within each oscillation period , a parabolic function
is fitted through the data points in between the transition to a different fluxoid quantum number @xmath92 ( eq . ( [ eq : wire ] ) ) .
the value for the self- and mutual inductance@xcite for the different samples are summarized in table [ table : param ] . in fig .
[ fig:3 ] , a smaller amplitude of the oscillations is observed for sample c , compared to the other samples .
this smaller amplitude is to be expected because of a smaller thickness of the film , what would lead to a reduced mean free path and a smaller coherence length value for sample c ( see eq .
( [ eq : wire ] ) ) .
but the coherence lengths are comparable for all four different samples ( see table [ table : param ] ) .
the phase boundaries of the double loop sample a and for the reference sample r are also measured for different transport currents @xmath56 .
the results for three different ac transport currents @xmath56 ( @xmath85 : 0.3 @xmath59a , @xmath86 : 0.5 @xmath59a , @xmath87 : 0.7 @xmath59a rms ) are shown in fig .
[ fig:4]a and fig .
[ fig:4]b for sample a and sample r , respectively . from these experimental data ,
clear differences between sample a and sample r are seen .
first of all , the amplitude of the @xmath23 oscillations is stronger in sample r , secondly the phase boundary of the single loop ( fig .
[ fig:4]b ) matches very well with the fitted parabolic curves ( solid curves ) , while the oscillations of sample a ( fig .
[ fig:4]a ) are not parabolic at all , since the cusps in @xmath23 are always rounded .
the rounding of the cusps in @xmath23 , observed for sample a , are not reproduced in samples b and c ( see fig . [ fig:3 ] ) .
these rounded cusps can not be attributed with all certainty to the magnetic interactions with the inner loop , but may be also related to the presence of scattering imperfections in the outer loop of sample a , as has been shown theoretically in ref@xcite .
a possible source of such imperfections could be the variation of the strip width along the loops written by e - beam lithography . using the micronet approach@xcite ,
the gl equation can also be solved when imperfections are present in a loop .
a single period of the phase boundary calculated from this micronet approach@xcite is presented in fig .
[ fig:5 ] for different magnitudes of imperfections .
notice that the stronger the imperfection the more the oscillation amplitude is damped .
it is clear from our experiments and from@xcite that the influence of the inner loop on the measured phase boundary is quite small . to detect traces of the periodicity coming from the @xmath23 oscillations of the inner loop ,
a fast fourier transform analysis of the phase boundaries is carried out using @xmath93 equally spaced points in an interval between 5 and 24 @xmath25 for sample a , b and r and between -36 and 49 @xmath25 for sample c , but with approximately the same number of data points . in order to remove most of the contribution arising from the fluxoid quantization in the outer loop , the parabolas fitted in fig .
[ fig:3 ] , [ fig:4]a and [ fig:4]b ( solid line ) are subtracted from the experimental data , prior to taking the fourier transform .
the resulting spectra are presented in fig .
[ fig:6]a for sample a , [ fig:6]b for sample b , [ fig:6]c for sample c and [ fig:6]d for the single loop ( sample r ) , for an external transport current of @xmath56=0.3 @xmath59a .
the fourier spectra of sample a and sample r for @xmath56=0.5 @xmath59a are shown in fig .
[ fig:7]a and [ fig:7]b , respectively . in fig .
[ fig:6]a , [ fig:6]c and [ fig:7]a , a peak is clearly seen at @xmath94 corresponding to the frequency of the lp oscillations for the outer loop ; the second to the sixth harmonics of this base frequency are also seen . in the case of the sample r ( fig .
[ fig:6]d and [ fig:7]b ) and of sample b ( fig . [ fig:6]b ) , the first harmonic is strongly reduced .
this reduction is quite reasonable because of the subtraction procedure applied before the fourier transformation .
the lower scale for the fourier spectrum for sample c ( fig .
[ fig:6]c ) is probably due to a smaller amplitude of the oscillations in the phase boundary . for the double loop ( sample a , b and c )
, supplementary peaks are clearly distinguished in between the harmonics ( see inset in fig . [ fig:6]a , [ fig:6]b , [ fig:6]c and [ fig:7]a ) , which are considerably weaker in the reference sample r. these peaks can be due to the coupling with the supercurrent in the inner loop . comparing the insets of fig .
[ fig:6]b and fig .
[ fig:6]d , we note that the supplementary peaks are substantially sharper and higher for sample b than for sample r. the period of the lp oscillations for the outer loop is @xmath95=0.525 mt and @xmath96=0.550 mt for sample a , b and r and for sample c , respectively , corresponding to the first harmonic peak at @xmath94 .
the periodicity of the @xmath23 oscillations for the inner loop ( samples a and b ) is @xmath97=2.6 mt , which is approximately 5 times larger than @xmath98 , since the surface of the inner loop is approximately 5 times smaller than the one of the outer loop .
this periodicity is in good agreement with the measurement on sample b ( fig . [ fig:6]b ) , where four peaks are indeed observed between the first and the second harmonic , thus indicating a five times smaller frequency compared to the frequency of the @xmath23 oscillations of the outer loop . the @xmath23 measurements on sample a show two pronounced peaks between each harmonic ( fig .
[ fig:6]a and [ fig:7]a ) .
this suggests a period for the inner loop @xmath99 which is in disagreement with the dimensions of sample a. for sample c , which has a slightly different size , the periodicity of the inner loop oscillations in @xmath23 is @xmath75=2.1 mt@xmath100 .
this periodicity is in agreement with the measurements where three peaks are clearly observed before the first harmonic ( see fig .
[ fig:6]c ) .
the reason of the disagreement between the periodicity of the oscillations and the peaks in the fourier spectrum of sample a can be a different effective surface . to calculate the periodicity of the oscillations of the inner and outer loop , the average size of the loops ( through the middle of the wires ) has been used .
it is possible that we have to take a slightly smaller or larger effective surface for sample a. the dimensions of samples a and b may also be not exactly the same .
but the effective surface has to be taken larger than the largest dimensions of the inner loop to obtain a value of @xmath99 .
we therefore think that the smaller peaks might be hidden , and not clearly seen however in the fourier spectrum . in that case , a value for the periodicity of 6 times the periodicity of the inner loop is more realistic for sample a. coming back to fig .
[ fig:2 ] , we can see that the interval ( between 5 and 24 @xmath25 ) , where a fourier transform was performed for sample a , b and r , corresponds only to 3 or 4 periods @xmath101 .
the resolution of the fourier spectrum in the low frequency regime will therefore be low , thus presenting an additional difficulty in interpreting the intermediate peaks in the fourier spectra .
the phase boundary @xmath23 of sample c is measured over a broader interval ( between -36 to 49 @xmath25 ) .
this interval corresponds to more than 20 periods @xmath101 , what results in a better resolution of the fourier spectrum in the low frequency regime . in a first approximation , with the currents @xmath50 and @xmath51 independent from each other , the mutual inductance @xmath40 can be evaluated from the amplitude of the additional peaks in the fourier spectrum .
the additional energy due to the mutual inductance in eq .
[ eq : freeenergy ] ( @xmath102 ) has to be of the same order of magnitude as the amplitude of the oscillations @xmath103 due to coupling in the phase boundary of the outer loop .
the value for the mutual inductance can then be evaluated with the formula @xmath104 with @xmath105 the boltzmann constant .
this gives @xmath106 17 ph , 13 ph and 14 ph for sample a , b and c , respectively .
the average of the supercurrents @xmath50 and @xmath107 is calculated from ref@xcite , for @xmath56=0.3 @xmath59a and with the criterion for @xmath108 at 90% of @xmath58 , what correspond to a typical temperature interval of 3 mk between this criterion and the 50% criterion used for the measurements of the phase boundary .
it is quite clear that the supercurrents , and thus also the mutual inductance are strongly dependent of this chosen criterion .
the values calculated from eq . [ eq : compare ] are comparable to the calculated mutual inductances from table [ table : param ] . due to the strong temperature dependent currents
, only the order of magnitude of the mutual inductance can be evaluated from our measurements . for high ( @xmath56=0.9 @xmath59a for sample
a and 1.0 @xmath59a for sample c ) transport current , the peaks between the harmonics completely disappear ( not shown )
. this vanishing may be due to increasing shift of @xmath23 with the applied transport current .
the highest current corresponds to the case of the upper curve ( dashed - dotted line ) in fig .
[ fig:2 ] .
hence , for all flux values , the inner loop is in the superconducting state at @xmath23 of the outer loop , and the phase boundaries of the two loops are not intersecting each other .
there will be no sharp interruption of the supercurrent of the inner loop going from the superconducting to the normal state .
therefore , the shape of the phase boundary of the outer loop can be less sensitive to the presence of the inner loop .
on the other hand , once the inner loop is deep in the superconducting state , a higher supercurrent in the inner loop would be present , and this higher supercurrent would require a higher coupling . if the transport current @xmath56 is high enough , the inner loop will be always superconducting at @xmath23 of the outer loop . a discontinuity in the measured phase boundary of the outer loop
is expected when the inner loop changes from a fluxoid quantum number @xmath109 to @xmath110 .
however , we could not see a discontinuity in the measured phase boundaries , corresponding to a sign reversal of the supercurrent in the inner loop .
we have measured the normal / superconducting phase boundary of a superconducting system consisting of two concentric mesoscopic loops , to study magnetic interactions between the two loops .
the modification of the @xmath23 oscillations of the outer loop is seen in the fourier spectrum of the @xmath23 line due to the coupling between the outer and the inner loops . to interpret these observations ,
we have used two different models .
the first model assumes the presence of scattering imperfections in the outer loop .
this model can not explain the observed evolution of the fourier spectrum with the current , although it might be applicable for a fixed weak current .
the second model explains the extra peaks in the fourier spectrum by the magnetic coupling of the two loops .
the systematic shift of the @xmath23 phase boundary of the outer loop with the applied current @xmath56 induces a well defined evolution of the fourier spectrum which was indeed found in our experiments .
this evolution of the extra peaks in the fourier spectrum with the applied current gives an experimental evidence for the presence of the magnetic interaction between the two superconducting loops .
future magnetic measurements on huge arrays of magnetically coupled loops , deeper in the superconducting state , could be helpful to reveal an enhanced magnetic coupling of both loops at lower temperatures .
a inner loop made from a different superconductor with a higher critical temperature would certainly increase the magnetic coupling between the two loops .
an enhanced critical field is expected in this case for the lower @xmath0 loop , at the expense of sharing the fluxoid quantization `` burden '' with a loop where superconductivity is stronger .
the authors wish to thank m. cannaerts , e. seynaeve and k. temst for the afm , sem and x - ray measurements , and h.j .
fink and l. van look for useful discussions .
this work has been supported by the belgian iuap , the fwo and goa - programs , and by the esf programme vortex .
d. davidovic , s. kummar , d.h .
reich , j. siegel , s.b .
field , r.c .
tiberio , , r. hey , and k. ploog , phys .
lett . * 76 * , 815 ( 1996 ) ; d. davidovic , s. kummar , d.h .
reich , j. siegel , s.b .
field , r.c .
tiberio , , r. hey , and k. ploog , phys
b. * 55 * , 6518 ( 1997 ) | we study the magnetic interaction between two superconducting concentric mesoscopic al loops , close to the superconducting / normal phase transition . the phase boundary is measured resistively for the two - loop structure as well as for a reference single loop . in both systems little - parks oscillations ,
periodic in field are observed in the critical temperature @xmath0 versus applied magnetic field @xmath1 . in the fourier spectrum of the @xmath2 oscillations ,
a weak low frequency response shows up , which can be attributed to the inner loop supercurrent magnetic coupling to the flux of the outer loop .
the amplitude of this effect can be tuned by varying the applied transport current . |
the tendency toward for the formation of the charge ordering is a common characteristic of the transition metal oxides with a perovskite structure including high - temperature superconducting cuprates and colossal magnetoresistive manganites .
@xcite in particular , various kind of the manganites with the doping level @xmath4 of 1/2 show the charge - exchange ( ce)-type charge ordering , @xcite in which mn@xmath5 and mn@xmath6 ions by the ratio of 1:1 regularly distribute in the underlying lattice below the charge and orbital ordering temperature @xmath7 [ fig .
[ fig1](a ) ] .
such a real - space charge ordering is characterized by a single - particle excitation across the charge gap @xmath8 . using the light of frequency @xmath9 , the single - particle excitation spectrum of the charge - ordered ( co ) manganites
have been extensively investigated in recent years ; @xcite for example , okimoto _
et al_. identified @xmath10 mev of pr@xmath11ca@xmath12mno@xmath2 by using polarized reflectivity measurements and clarified that the electronic structure is dramatically reconstructed in the order of ev by varying temperature @xmath13 and magnetic field @xmath14 ( refs . and ) . using transmission measurements , calvani _
et al_. found that @xmath15 of la@xmath16ca@xmath17mno@xmath2 with @xmath18 and 0.67 can be well described by the bardeen - cooper - schrieffer ( bcs ) relation .
another important viewpoint is the collective nature of @xmath19 carriers including the orbital degree of freedom , which is thought to have a great impact on extraordinary properties of the manganites , i.e. , colossal magnetoresistance ( cmr ) .
@xcite the conspicuous examples are recent findings of the `` orbiton '' , a new elementary excitation of the orbital - ordered state observed in lamno@xmath2 @xcite and of the `` charge stripes '' , a new periodic form composed of a pairing of diagonal mn@xmath5 with mn@xmath6 observed in la@xmath16ca@xmath17mno@xmath2 @xmath20 .
@xcite therefore , the understandings of the role of @xmath19 carriers with variation of @xmath4thus the collective phenomena of @xmath19 carriers in connection with the specific kind of the charge ordering depending on @xmath4have attracted interests and may be of central topics in co manganites .
more recently , asaka _ et al_. proposed the model of the ce - type structure in pr@xmath21ca@xmath22mno@xmath2 by using the transmission electron spectroscopy as schematically shown in fig .
[ fig1](b ) .
@xcite it may be viewed as a quasi - one dimensional electronic structure with the reduced dimensionality as compared to the well - known ce - type structure [ fig .
[ fig1](a ) ] : @xmath23 or @xmath24 orbital ordering of mn@xmath5 occurs along @xmath25-axis . in another angle ,
@xmath26 orbital of mn@xmath5 was cliped by three mn@xmath6 ions .
such a regular pattern of the distinct charges modifies the uniformity of the charge density , leading to the development of the charge - density - wave ( cdw ) condensate .
@xcite this situation is depicted in fig .
[ fig1](c ) ; the charge density ( shared area ) periodically varies with the position and also time ( dashed line ) , so the new modulation occurs in the underlying lattice ( the period of this modulation is larger than the lattice constant ) .
such a cdw formation is recently deduced from the one - dimensional fermi - surface ( fs ) topology in a layered manganite la@xmath27sr@xmath28mn@xmath29o@xmath30 by chuang _ et al_. ; @xcite they determined the detailed map of the fs by using the angle - resolved photoemission spectroscopy and concluded that fs is subject to cdw instabilities even in the ferromagnetic metallic state . in the sight of these considerations ,
the charge ordering is also characterized by a collective excitation well below @xmath8 .
however , no clear evidence for the collective excitation of the co state was reported so far ( see , _ note added _ ) .
the collective excitation is associated with the mode depending on both the position and the time ; the modes arising from the former and the later are referred to as amplitudons and phasons , respectively .
@xcite according to lee _
et al_. , @xcite the dispersion relation of the phasons is given by @xmath31 , where @xmath32 is the wavenumber . by the illumination of light ,
the collective excitation can make a coupled oscillation , which yields the change of the dielectric constant at @xmath33 , being in proportional to @xmath34 , where @xmath35 and @xmath36 are the longitudinal and transverse optical frequencies , respectively .
therefore , this kind of the coupled mode comes to the surface as a collective excitation mode in the complex optical spectrum .
it is also important that the impurity and/or the lattice imperfection has striking influence on the cdw dynamics , because the cdw condensate is pinned by the impurity and therefore the peak position of the collective excitation mode shifts to a finite frequency .
@xcite in this article we provide the spectroscopic evidence for the cdw condensate in the typical co manganite pr@xmath0ca@xmath1mno@xmath2 .
the basis of this conclusion is derived from the first observation of the collective excitation mode in the optical conductivity spectrum .
as well - known cdw system like a quasi - one dimensional k@xmath1moo@xmath2 , the collective excitation mode frequently appears in the millimeter @xmath9 range @xmath37 mev , @xcite where it is difficult to get the probing light by the conventional optical spectroscopy .
we overcome this limitation by using thz time - domain spectroscopy ( tds ) , @xcite which is a powerful tool to unveil the low - energy charge dynamics of the cmr manganites .
@xcite in addition to this advantage , thz - tds is especially suited to capture the signal of the fluctuation phenomenon as the high @xmath9 thz pulse is sensitive and matches on the typical time scale of the charge fluctuation in correlated electron system , which is instantiated by thz - tds experiments on underdoped bi@xmath29sr@xmath29cacu@xmath29o@xmath38 .
@xcite _ material_material we used in this work is the typical co manganite ; pr@xmath0ca@xmath1mno@xmath2 with a orthorhombically disordered three dimensional structure .
this material exhibits the insulating behavior and undergoes the charge @xcite and orbital @xcite ordering below @xmath39 k and the antiferromagnetic spin ordering below @xmath40 k. due to the deviation of @xmath4 from 1/2 , at which the charge ordering is most stable , the extra electrons @xmath41 occupy at the mn@xmath6 site with maintaining the charge ordering [ fig .
[ fig1](b ) ] .
the remarkable characteristics compared to other co manganites are such an admissibility of the charge ordering and the cmr effect , in which the dramatic variance of the resistance by more than 10 orders of magnitude was attained @xcite therefore , various kind of the experimental results are accumulated to date .
@xcite it should be mentioned that there are variant interpretaions about the low - temperature phase below 110 k : one is due to the coexistence of the ferromagnetic and antiferromagnetic phases proposed by jirk _ et al_. ; @xcite another is the canted antiferromagnetic phase ( or the spin - glass phase ) reported by yoshizawa _
et al_. @xcite however , the intimate identification is beyond the scope of this study , but we can say that magnetic properties of our samples are consistent with the results by deac
_ et al . _
@xcite dependence of the magnetization @xmath42 in a unit of bohr magnetron @xmath43 at the mn - site of the sample a. circles and squares represent @xmath44 taken from the field cooling ( fc ) and zero field cooling ( zfc ) scans under the magnetic field of 500 oe or 10 koe , respectively.,width=264 ] [ cols="^,^,^,^,^,^,^ " , ] dependence of the scattering rate @xmath45 of the sample a deduced from eq .
( [ eqn1 ] ) .
the data are taken from the @xmath13-warming run .
the solid line denotes a fitting result to the data ( open circles ) using eq .
( [ eqn3 ] ) .
the dashed line indicates the charge and orbital ordering temperature @xmath7.,width=245 ] one may consider the phase separation ( ps ) picture @xcite the dynamical coexistence of the ferromagnetic metallic and co insulating states @xcite , which is thought to be a key feature of the cmr phenomena and indeed experimentally discussed in pr@xmath0ca@xmath1mno@xmath2 .
@xcite such ps as well as the presence of the disorder commonly seen in thin films using mgo substrates , produces a finite peak structure as a result of the shifting of the drude peak centered at @xmath46 .
our samples show the ferromagnetic order at low - temperature ; there is a significant hysteresis loop in @xmath47 curve at 30 k with a coercive force of @xmath48 oe ( figs .
[ fig3 ] ) , which may imply the presence of the ferromagnetic domain in the co insulating phase .
therefore , we compare our experimental results to the typical characteristics of the ps in co manganites .
@xcite best example in the ps picture is the slow relaxation effect observed in pr@xmath49ca@xmath50mno@xmath2 by anane _ et al_. ( ref . ) and cr - doped manganites by kimura _
et al_. , @xcite in which the ferromagnetic fraction can be artificially controlled by @xmath14-annealing process and also the cr ( impurity)-doping at the mn - site . in these materials , the physical quantities [ i.e. , @xmath51 and @xmath42 ] strongly depend on the history of the external perturbations . for example , anane _
et al_. measured the time dependence of @xmath51 in the ferromagnetic phase of pr@xmath49ca@xmath50mno@xmath2 after @xmath14 of 100 koe is applied at low - temperature .
@xcite they found that the co insulating phase is restored in the time range of @xmath53@xmath54 second ( the slow relaxation effect ) .
very recently , mathieu _
et al_. reported that @xmath55 under 20 oe in nd@xmath56ca@xmath56mno@xmath2 exhibits the different relaxation effect in the typical characteristic time of @xmath57 second by changing only the cooling rate .
@xcite moreover , the difference of @xmath55 in the history of the cooling rate increases with time . based on these studies
, we have done following thz - tds experiments : we rapidly cooled the another sample used in sec .
[ thz ] ( the growth condition is same as samples a and b ) from room temperature to 4 k with a cooling rate of @xmath58 k per minute and measured the time dependence of @xmath59 while keeping @xmath60 k constant .
we observed no slow relaxation effect in the measured @xmath9 range ; the spectral shape of the peak structure and the peak frequency @xmath61 mev exhibit no variation per @xmath62 second , which is a contrarious evidence for the ps picture .
as described in sec . [ intro ] , we infer the possibility of the cdw condensate due to the quasi - one dimensional charge and orbital ordering in pr@xmath0ca@xmath1mno@xmath2 .
we discuss here that the observed peak structure arises from the cdw condensate as compared to the low - energy optical properties of the well - known cdw system .
the typical characteristics of the collective excitation mode observed in low - dimensional cdw materials are as follows : @xcite ( i ) as mentioned earlier , the pinning frequency is in general of the order of mev .
@xcite ( ii ) the spectral weight of the collective excitation mode is typically 2 orders of magnitude smaller than that of the single - particle excitation .
@xcite this is due to the relatively large @xmath63 mev , which is in strong contrast to the small spin gap @xmath64 mev observed in the spin - density - wave ( sdw ) system . in the sdw system , the spectral weight of the collective excitation mode should be comparable to that of the single - particle excitation due to the electron - electron interaction.@xcite ( iii ) kim _ et al_. found that the pinning frequency in ( ta@xmath65nb@xmath66se@xmath67)@xmath29i alloys linearly shifts to higher @xmath9 and @xmath59 at the pinning frequency does not change with increasing the impurity concentration @xmath68 .
@xcite same tendencies are also observed in k@xmath1mo@xmath65w@xmath66o@xmath2 alloys .
@xcite ( iv ) mihly _ et al_. identified the swelling structure in the low - energy side of the collective excitation mode in k@xmath1moo@xmath2 as the internal deformation on the basis of the generalized debye analysis .
@xcite our present findings ( i)-(iv ) described in sec .
[ thz ] and various works ( i)-(iv ) described above in the cdw system , both of which have apparent similar properties and have no contrariety .
moreover , @xmath69 dependence of @xmath45 ( fig .
[ fig7 ] ) has been also found in the cdw system @xcite and eqs .
( [ eqn1 ] ) and ( [ eqn2 ] ) are same expressions derived from the equation of motion for the cdw dynamics .
@xcite therefore , we arrive at the conclusion that the observed structure in pr@xmath0ca@xmath1mno@xmath2 is assigned to the collective excitation mode arising from the cdw condensate . to the best of our knowledge ,
this is a first spectroscopic evidence for the cdw condensate in co manganites .
the description of the cdw condensate in pr@xmath0ca@xmath1mno@xmath2 is consistent and may explain following previous studies .
the cdw easily couples with the lattice and affects phonon modes .
therefore , new infrared modes are expected to be active below @xmath7 .
et al_. found that in addition to phonon modes due to the orthorhombic distortion , many new modes emerge as sharp excitations in pr@xmath11ca@xmath12mno@xmath2 at 10 k. @xcite asamitsu _
et al_. have measured current - voltage characteristics of pr@xmath0ca@xmath1mno@xmath2 and found the beginning of the nonlinear conduction above the threshold electric field , @xcite which reminds us of the sudden motion of the cdw ( cdw depinning ) as observed in cdw materials .
@xcite we should note the previous work about the impurity effect on @xmath59 in a copper oxide superconductor .
et al_. reported that the zn ( impurity)-doping of the cu - site in yba@xmath29cu@xmath67o@xmath38 distracts the superconductivity and induced the finite peak structure around 10 mev ( ref . ) .
they concluded that this structure arises from the localization of the charge carriers produced by the randam distribution of the zn on the cu - site ( disorder ) on the basis that the integrated area of @xmath59 up to 1 ev conserves as far as 3.5% zn - doping .
however , same authors also pointed out that if the drude - like response in yba@xmath29cu@xmath67o@xmath38 appears as a result of the collective excitation which is prospective by theory , the peak structure in yba@xmath29(cu@xmath65zn@xmath66)@xmath67o@xmath38 can be regard as the collective excitation mode as in both cases of the cdw and sdw . @xcite it is well known that the collective excitation mode originates in two different states ; one is a pinned " state due to the pinning of the cdw condensate by the impurity and/or the lattice imperfection and another is the bound " state , which is created by the coupling of the pinned state with the optical phonon or the impurity near the pinned state .
@xcite despite the fact that the swelling of @xmath59 below the low - energy side of the finite structure , which is usually ascribed to a internal deformation of the pinned collective excitation mode , @xcite can be seen in figs .
[ fig4 ] , we can not clearly claim whether the observed mode is assigned to the pinned or the bound collective excitation mode of the cdw .
one pregnant result is that @xmath70 below the peak structure crosses zero around 8 mev and reaches the minus value with decreasing @xmath9 [ fig . [ fig4](a ) ] .
this implies the presence of another @xmath71 and of the another peak structure below the measured @xmath9 range .
further experiments on @xmath72 in the ghz @xmath9 range are necessary to perform the detailed discussion and are now in progress using the cavity perturbation technique . as shown in fig .
[ fig5](d ) , the collective excitation mode can be seen even above @xmath7 , while its spectral shape is blurred due to the large proportion of the background contribution above @xmath7 .
@xmath45 continues to follow @xmath69 dependence above @xmath73 as shown in fig .
[ fig7 ] [ the fitting procedure by eq .
( [ eqn3 ] ) was performed using the data below @xmath73 ] .
this excludes the thermal fluctuation as the origin of this broadening , leading us to conclude that the dynamical short - range co ( charge fluctuation ) is still subsistent above @xmath7 instead of the long - range co below @xmath7 .
such a short - range co has been already reported by other experiments : radaeli _ et al_. and shimomura _ et al_. have revealed that the short - range co of pr@xmath16ca@xmath17mno@xmath2 with @xmath75 , @xcite 0.35 , 0.4 , and 0.5 , @xcite exists even above @xmath7 by neutron and x - ray scattering experiments , respectively . as in the case of the cdw system such as k@xmath1mo@xmath65w@xmath66o@xmath2 and ( tase@xmath67)@xmath29i , the collective excitation modes
are also observed above the cdw transition temperature @xmath76 due to the presence of the fluctuating cdw segments , which are systematically investigated by schwartz _
et al_. @xcite for example , the collective excitation mode of k@xmath1moo@xmath2 is visible even at 300 k ( @xmath77 k ) . finally , we comment on thz radiation from pr@xmath0ca@xmath1mno@xmath2 thin films excited by the femtosecond optical pulses , which we found recently ; @xcite the radiated spectrum decreases rapidly in intensity with increasing @xmath9 and seems to show the depletion around 2.4 mev . as
compared to figs .
[ fig4 ] , such a thz response is due to the absorption of the collective excitation mode during the propagation of the generated thz pulse inside the material .
summarizing , by using thz - tds , for the first time , we reported the presence of the finite peak structure around 23 mev in pr@xmath0ca@xmath1mno@xmath2 and assigned it to the collective excitation mode arising from the cdw condensate
. the measurements on the polarization dependence of @xmath59 with the grazing incidence of light using the single - domain single crystal will provide the direct evidence and more detailed information for the collective excitation mode .
in especially , the measurements of @xmath59 in the ghz @xmath9 range are needed to perform further quantitative discussions .
further interests are to study the melting process under external perturbations and @xmath4 dependence of the collective excitation mode ; for example , @xmath8 decreases linearly with @xmath4 from 0.3 to 0.5 , @xcite whereas @xmath7 gradually increases and behaves the less-@xmath4 dependent .
@xcite this indicates the breaking of the bcs relation given by @xmath78 .
so , it is indispensable to clarify how the collective excitation mode manifesting well below @xmath8 changes with @xmath4 using thz - tds .
_ note added._after the submission of this article , we became an aware of some reports related in the present study .
et al_. [ europhys . lett . *
56 * , 434 ( 2001 ) ] and gorshunov _
et al_. [ cond - mat/0201413 ( unpublished ) ] found peak structures around 0.11 mev in spin - ladder materials sr@xmath79ca@xmath17cu@xmath80o@xmath81 by means of optical spectroscopy and assigned them to the collective excitation modes due to the cdw formation . about manganites , campbell _ et al_. [ phys .
b * 65 * , 014427 ( 2001 ) ] interpreted that cdw fluctuations inhere in a layered ferromagnetic manganite la@xmath27sr@xmath28mn@xmath29o@xmath30 based on their diffuse x - ray scattering data . recently ,
et al_. [ phys . rev .
b * 65 * , 060405(r ) ( 2002 ) ] reported that the electron microscopic data of a layered co manganite nd@xmath16sr@xmath82mno@xmath67 are consistent with assumptive images of the cdw formation . very recently
, the finite peak structure around 4 mev was also found in 1/8 hole - doped la@xmath83nd@xmath11sr@xmath84cuo@xmath67 by dumm _
et al_. [ phys .
lett . * 88 * , 147003 ( 2002 ) ] as in the case of yba@xmath29(cu@xmath65zn@xmath66)@xmath67o@xmath38 .
they also ascribe this structure to the localization of the charge carriers in the static stripe phase due to the lack of @xmath8 in this material .
however , fujita _
et al_. reported the evidence for cdw and sdw formations in 1/8 hole - doped la@xmath85ba@xmath86sr@xmath17cuo@xmath67 by elastic neutron scattering experiments [ phys .
lett . * 88 * , 167008 ( 2002 ) ] .
we would like to thank in particular y. okimoto and also k. miyano , y. ogimoto , and y. tokura for giving fruitful comments and discussions .
we are also grateful to t. kawai for giving an opportunity of using squid apparatus , t. kanki , i. kawayama , k. takahashi , and h. tanaka for their help to squid measurements , and m. misra for reading the manuscript .
it is worth noticing that the appearance of the finite peak structure is a common feature in the disordered system but only in _ conductors _ near the compositional insulator - metal phase boundary . in general , the characteristic energy scale of the localized state in the impurity - doped _ conducting _ manganites , i.e. , al ( impurity)-doping at the mn - site of la@xmath87sr@xmath88mno@xmath2 ( ref . ) is about 100 mev , which is considerably different from the peak position of the observed structure . | thz time - domain spectroscopy was used to directly probe the low - energy ( 0.55 mev ) electrodynamics of the charge - ordered manganite pr@xmath0ca@xmath1mno@xmath2 .
we revealed the existence of a finite peak structure around 23 mev well below the charge gap @xmath3 mev .
in analogy to the low - energy optical properties of the well - studied low - dimensional materials , we attributed this observed structure to the collective excitation mode arising from the charge - density - wave condensate .
this finding provides the importance role of the quasi - one dimensional nature of the charge and orbital ordering in pr@xmath0ca@xmath1mno@xmath2 . |
the field of ultracold atom physics has explored a wide variety of phenomena since its relatively recent accessibility , with a major feature being the tunability of experiments across wide parameter regimes @xcite , to easily access and probe phase transitions @xcite , as well as excitation spectra and dynamics of systems analog to condensed matter @xcite .
even features such as artificial gauge fields can be implemented for neutral atomic particles , allowing for the investigation of topological phases @xcite . within this ultracold toolbox
, one ingredient is becoming of increasing interest in recent years which is of vital importance to real solid - state systems : phonons and atom - phonon coupling @xcite .
such a coupling provides many interesting possibilities @xcite . for one , it can lead to effective hamiltonians , such as extended hubbard models or the holstein model @xcite , as well as dissipative two - level system @xcite .
polaronic effects from electron - phonon interactions have also long been suggested to be the proponent behind high-@xmath0 superconductivity in one- and two - band solid - state systems @xcite . in ultracold quantum gases ,
evidence of polarons has been found in systems with trapped ions @xcite or systems with a single ion immersed in a degenerate quantum gas @xcite . on the other hand , the atomic polaron has also been studied both experimentally and theoretically in systems of imbalanced bose - fermi mixtures @xcite and fermi - fermi mixtures @xcite . in the particular case of a system with impurities immersed in a bosonic bath , these impurities couple to bosonic excitations . for suitable parameters ,
polaronic phenomena arise generically in such systems @xcite .
there are also other proposals for realizing atom - phonon couplings and polaronic effects , including crystals of dipolar molecules @xcite , nanoparticles @xcite and hybrid atom - ion coupled systems @xcite .
none of these works consider polaronic phenomena of multi - band systems with ultracold quantum gases . in such systems ,
the inter - band dynamics of polarons , as well as intra - band dynamics , lead to new effects which have no analogue in single band systems .
motivated byrecent experiments @xcite , we consider a system of a few impurities in an optical lattice , populating the lowest two bloch bands , immersed in a bose - einstein condensate .
these impurities are coupled to bogoliubov phonons ( of the bec ) via both intra- and inter - band transitions . in order to decouple this frhlich - like term
, we derive a generalized two - band lang - firsov polaron transformation .
the transformed effective hamiltonian still contains two bands , where the impurity is now dressed by phonons as a quasi - particle ( polaron ) .
we use a variational approach to connect between the weak and strong coupling limits and calculate the dressing parameters .
polaronic effects modify both intra - band coherent transport and polaron energy shifts , and also induce a long - range interaction between different polarons .
we then focus on inter - band relaxation effects and specify our system as a single impurity trapped in a quasi-1d system .
we study the residual incoherent coupling between polaron and bath by using a lindblad master equation .
the impurity inter - band relaxation process under this polaronic treatment is beyond a fermi s golden rule description .
these polaronic renormalization effects of the inter - band relaxation rate should be accessible in current experiments .
on the other hand , for large impurity - phonon coupling , the polaron is tightly dressed in each band and can not hop between different bands . in this limit ,
inter - band self - trapping _ effect is expected .
this work is organized as follow : in section [ sec : effective - hamiltonian ] , we introduce the effective two - band hamiltonian of a realistic experiment setup , with a few impurities in a lattice , immersed in a bose - einstein condensate ( bec ) of a different atomic species . in section [ sec : variational - polaron - transformati ] , we describe in detail the generalized lang - firsov polaron transformation for the two - band system .
the transformed hamiltonian with two ( polaron ) bands can be separated into a coherent part and an incoherent part .
the coherent part , arising from the thermal average over the phonon bath , is discussed in section [ sec : coherent - part ] . the inter - band relaxation and decoherence effects , which are all included by the incoherent part of this hamiltonian , are discussed in section [ sec : lindblad - equation ] .
we derive the lindblad equation and correlation functions for the residual incoherent impurity - phonon coupling .
this polaronic impurity dynamics is closely related to recent experiments .
the polaronic inter - band relaxation rate is compared to fermi s golden rule .
we give concluding remarks and an outlook in section [ sec : conclusion ] .
here we consider a few neutral impurities with mass @xmath1 interacting with a bose - einstein condensate of another neutral species .
the impurities are trapped by a 3d optical lattice and their hamiltonian is denoted by @xmath2 .
the homogeneous bec system @xmath3 is formed by another atomic species with mass @xmath4 and a weak repulsive interaction @xmath5 .
the impurity - bec interaction @xmath6 is caused by @xmath7-wave interactions between the different species , which can be tuned by standard feshbach resonance techniques . the total hamiltonian is hence @xmath8 we describe these different terms in detail in the following parts of this section .
we consider a quasi-1d system with impurities trapped in an anisotropic 3d optical lattice : @xmath9 with lattice constant @xmath10 and laser wavelength @xmath11 .
we use the lattice constant @xmath12 as the unit of length throughout this paper .
the single impurity recoil energy is @xmath13 .
the trapping strength in the transverse ( @xmath14 ) direction is assumed to be much stronger than in longitudinal ( @xmath15 ) direction with @xmath16 .
the impurities are therefore tightly trapped in the transverse direction , and remain in the ground state of the associated harmonic oscillator potential : @xmath17 with transverse characteristic length @xmath18 and frequency @xmath19 . in the longitudinal direction , on the other hand , the impurities populate both the lowest and the first excited band . in principle , exact bloch wave functions need to be calculated numerically , which can then also be represented as linear combinations of wannier states @xmath20 . in a deep lattice ,
the wannier functions can be approximated by the lowest two eigenstates of a harmonic oscillator @xmath21 with longitudinal characteristic length @xmath22 and oscillation frequency @xmath23 . in this highly anisotropic system
, impurities can hop to their nearest - neighbor sites only along the longitudinal direction . a particle localized on lattice site @xmath24
will be described by the wave function @xmath25 due to the low density of impurities , their dynamics can be modeled by a non - interacting two - band hubbard hamiltonian : @xmath26 where @xmath27 and @xmath28 are the hopping parameters and on - site energy for each band with index @xmath29 . in a deep lattice ,
the band gap @xmath30 can be approximated as the longitudinal oscillator frequency @xmath31 . in this work
we only consider inter - band dynamics between the lowest two bloch bands and ignore higher band effects . ) .
the impurity moves in the longitudinal direction of the system with two bloch bands @xmath32 ( blue line ) and @xmath33 ( green line ) .
the dashed blue lines at @xmath34 indicate the first brillouin zone in longitudinal direction .
the phonon dispersion @xmath35 ( red line ) is described by bogoliubov theory where @xmath36 is the phonon momentum .
the dashed red lines at @xmath37 indicate that the phonon energy matches with the impurity band gap @xmath38 with @xmath30.[fig : dispersion],scaledwidth=45.0% ] the impurities are immersed in a homogeneous bec with weakly repulsive boson - boson interaction @xmath5 between the atoms . in a dilute system
, this weak interaction can be described by boson - boson scattering length @xmath39 as @xmath40 . for vanishing inter - species interaction @xmath41 between impurity and bath ,
the bec can be described by standard bogoliubov theory and treated as a phonon bath ( see fig .
[ fig : dispersion ] ) .
once @xmath41 is introduced , the bec becomes deformed due to the presence of impurities .
this interaction is closely related to the impurity - boson scattering length and other system parameters such as the impurity - boson mass ratio and the impurity confinement strength .
the relation can be determined by making use of scattering theory in the low - energy limit , such as the lippmann - schwinger equation or effective field theory @xcite . for an unconfined impurity
, the inter - species interaction @xmath41 can be derived as @xmath42 with the reduced mass @xmath43 and 3d impurity - boson scattering length @xmath44 . on the other hand , for the confined impurity
, @xmath41 needs to be treated carefully due to lattice effects such as confinement induced resonances @xcite . in the specific system we considered here ,
the impurity is confined in one - dimensional tube ( quasi-1d ) by an anisotropic optical lattice while the bosonic atoms are free in three - dimensional space ( 3d ) .
when the transverse characteristic length @xmath45 is much smaller than any other length scales , the resulting system is mixed - dimensional . at low energies ,
the inter - species interaction is solely characterized by a single parameter , the effective scattering length @xmath46 , whose value can be obtained numerically @xcite .
this fact also allow us to arbitrarily tune the value of @xmath46 by tuning the transverse confinement strength , independently of tuning the feshbach resonance position .
here we use the approach in @xcite , where the deformation is treated as a perturbation around the bec ground state .
the bosonic field operator is expanded as @xmath47 , where @xmath48 is the order parameter in the absence of inter species interaction and @xmath49 represents the perturbation itself , which consists of a correction to the order parameter , @xmath50 , and bogoliubov excitation operators @xmath51 .
the modified bec ground state is described as @xmath52 , which is the gross - pitaevskii solution including the presence of impurities .
this modification @xmath50 shifts the equilibrium positions in order to minimize the total energy of the system .
the small excitations @xmath51 , around the static gp ground state @xmath52 of the condensate , can be described in the terms of bogoliubov modes @xmath53,\label{eq : phonon_beta}\ ] ] with bosonic operators @xmath54 ( @xmath55 ) creating ( annihilating ) a bogoliubov quasi - particle with momenta @xmath56 .
the coefficients @xmath57 and @xmath58 can be determined by bogoliubov - de gennes equations : @xmath59}\\ v_{\mathbf{q}}= & \sqrt{1/\left(2\omega\right)\left[\left(\epsilon_{\mathbf{q}}+g_{b}n_{0}\right)/\left(\hbar\omega_{\mathbf{q}}\right)-1\right]}\end{aligned}\ ] ] with the system _ quantization volume _ @xmath60 and the bogoliubov phonon dispersion @xmath61 where @xmath62 is the condensate density .
as pointed out in @xcite , the coefficients @xmath57 , @xmath58 and bogoliubov phonon spectrum @xmath63 are not changed by the presence of the impurity , although the equilibrium positions of the modes are modified by @xmath50 . in @xcite
, they express @xmath49 in terms of bogoliubov modes around the state @xmath48 as @xmath64.\ ] ] in contrast to eq .
( [ eq : phonon_beta ] ) , the bosonic operators @xmath65 ( @xmath66 ) create ( annihilate ) a quasi - particle around the ground state @xmath48 in absence of impurity rather than around state @xmath52 .
the hamiltonian ( up to constant terms ) for this effective phonon bath can be simply expressed as @xmath67 if the impurity - bec coupling @xmath41 satisfies condition @xmath68 , which implies the dimensionless relation : @xmath69 where @xmath70 is the condensate healing length and @xmath71 for a 3d homogeneous bosonic bath @xcite . strictly speaking , even when the coupling @xmath41 is stronger than this limit , the perturbation theory still qualitatively applies .
for very strong interactions , however , the bath can not be properly described by bogoliubov quasi - particles any more .
the above perturbative method can also be used in a system where the impurities and bec are both trapped by optical lattices @xcite .
the impurity - bec interaction of the hamiltonian can be written as @xmath72 with @xmath73 indicating the impurity bloch bands and @xmath74 .
this term describes the impurity coupling to the phonon bath by creating or annihilating phonons .
the non - local coupling terms with @xmath75 are highly suppressed due to the local form of the interaction which requires overlap between localized wannier functions . for these reasons ,
the above integrals @xmath76 can be well approximated by @xmath77 such that the value depends only on @xmath73 and @xmath56 .
the dimensionless impurity - phonon coupling @xmath78 can also be written as @xmath79 with @xmath80 note that the impurity - phonon coupling @xmath81 obeys : @xmath82 the integral factors @xmath83 take the form @xmath84 in a deep optical lattice , where the impurity wannier functions in eqs .
( [ eq : wannier_y]-[eq : wannier_x1 ] ) can be approximated as harmonic oscillator states , we can explicitly evaluate these factors : @xmath85 by using the identity for hermite polynomial integrals @xcite .
here @xmath86 indicates the transverse phonon momentum . in order to describe polaronic effects resulting from the impurity - bec coupling ,
we introduce a dimensionless coupling constant as in @xcite and name it @xmath87 .
this constant depends on the impurity - boson and boson - boson interactions @xmath41 , @xmath88 and the condensate parameters @xmath89 and @xmath4 as @xmath90 as described in @xcite , this constant is the ratio @xmath91 between the characteristic impurity - boson interaction @xmath92 and the typical phonon energy @xmath93 , where @xmath94 is the condensate speed of sound . in @xcite
an alternative dimensionless coupling constant is used as @xmath95 or equivalently @xmath96 .
these two coupling constants are related by @xmath97 , where @xmath98 is the reduced mass .
there are also other coupling constants used @xcite , which are slightly different from @xmath99 or @xmath87 . by tuning the impurity - boson scattering length @xmath46 ,
the coupling constant @xmath87 can be tuned continuously . however , the condition in eq .
( [ eq : couling_g_c ] ) requires that the coupling constant satisfies the relation @xmath100 this upper limit of the coupling constant @xmath101 depends only on the boson - boson scattering length @xmath39 and the condensate density @xmath102 instead of the specific mass ratio @xmath103 . in a typical bec system
, this maximum coupling constant is relatively small , such as @xmath104 for @xmath105 with @xmath106 and @xmath107 . in order to reach larger values of @xmath101 in realistic systems ,
one need to reduce the condensate density @xmath102 or the bose - bose scattering length @xmath39 .
finally the effective hamiltonian with two - band frhlich impurity - phonon coupling is @xmath108 here we used the relations in eq .
( [ eq : m_relation ] ) .
this hamiltonian describes a general two - band system with impurity - phonon coupling , which can also be realized by other experimental setups such as hybrid atom - ion systems @xcite .
this two - band hamiltonian in eq .
( [ eq : hamil_0 ] ) , with frhlich - type impurity - phonon coupling , can not be solved analytically even for the case of a single impurity .
the goal of this paper is to find a simple but non - trivial variational method which can deal with the two - band system ( [ eq : hamil_0 ] ) in general .
we choose the lang - firsov polaron transformation approach and generalize it to two - band system .
this canonical transformation is exact and decouples the impurity - bec interaction term in a new quasi - particle basis . in this basis ,
the kinetic part in hamiltonian ( [ eq : hamil_0 ] ) contains the dynamics of quasiparticle and its interactions between the transformed phonon bath .
we solve the coherent part of this transformed hamiltonian with a variational treatment and take into account the remaining incoherent parts by a master equation .
firstly we introduce the basic concept of the transformation for a single band before we extend it to the two - band case .
when a single impurity moves in a lattice and couples to a phonon bath , there are exact solutions in both the weak and strong coupling limit @xcite .
when the impurity - phonon interaction is much weaker than the impurity kinetic energy , the impurity behaves as a free particle in a lattice . on the other hand ,
when the interaction is much larger than the kinetic part , the impurity will be tightly dressed by a `` cloud '' of phonons , forming a quasi - particle .
the phonons are tied to the impurity such that the impurity can not move on its own but must drag around a phonon cloud .
this increases the effective mass of the quasiparticle . in the intermediate coupling region ,
the phonon dressing competes with the impurity dynamics . in order to describe this competition ,
a variational ground state can be used to connect between the weak and strong coupling limits @xcite .
this variational ansatz is equivalent to a canonical transformation @xmath109 with @xmath110 where @xmath111 are the variational parameters .
the transformed hamiltonian @xmath112 still can not be solved analytically , but can be separated into a coherent part @xmath113 and an incoherent part @xmath114 where @xmath115 indicates a thermal average over the phonon bath .
the coherent part , which is decoupled from the phonon bath , is of the form of an extended ( polaronic ) hubbard model .
the incoherent part describes the residual coupling between polaron quasi - particle and phonon bath .
compared to the initial `` bare '' impurity - phonon coupling , this incoherent part is significantly reduced by the polaron transformation .
we first focus on the coherent part and neglect the incoherent terms .
the variational parameters @xmath111 are determined by minimizing the coherent hamiltonian energy and approach @xmath116 in the strong coupling limit . finally , the residual incoherent part can be included by a perturbative approach such as the lindblad master equation . in order to find a suitable variational transformation for the two - band system , we modify the lang - firsov polaron transformation @xmath117 by extending the impurity - phonon interaction to the two - band form .
it takes the form : @xmath118,\label{eq : transformation_s}\ ] ] with @xmath119 .
a similar method was also applied by sibley and munn in @xcite and stojanovi@xmath120 _ et al . _ in @xcite for a single - band system with non - local impurity - phonon coupling .
our initial guess for the variational parameters is the coupling element itself , @xmath121 , and we constrain the variational parameters to obey the same symmetry properties as @xmath121 , namely that of eq .
( [ eq : m_relation ] ) .
by using the baker - campbell - hausdorff formula , @xmath122+\frac{1}{2!}\left[\hat{s},\left[\hat{s},\;\hat{a}\right]\right]+\cdots$ ] , we canderive the transformed hamiltonian with exponential quadratic operators as outlined in appendix [ sub : app_polaron - transformation ] . for convenience , these expressions can be written in @xmath123 matrix form via @xmath124 and @xmath125 after the transformation , the impurity annihilation and creation operators can be expressed through @xmath126 and similarly for the creation operators .
the matrix operators @xmath127 are found to be @xmath128 in the same fashion as was described above for the single band system , this canonical transformation is equivalent to defining a new quasi - particle , which represents an impurity dressed by the phonon cloud forming a polaron .
this transformation also shifts the equilibrium position of the phonon bath by @xmath129 but it does not modify the phonon dispersion relation . in appendix
[ sub : app_polaron - transformation ] , we show the derivations for eq .
( [ eq : transform_a]-[eq : transform_b ] ) .
the transformed polaronic hamiltonian can be written as : @xmath130_{\alpha\beta}\hat{a}_{i}^{\alpha\dagger}\hat{a}_{i}^{\beta}\nonumber \\ + & \sum_{i,\alpha,\beta}\sum_{j,\alpha',\beta'}\sum_{\mathbf{q}}\frac{\hbar\omega_{\mathbf{q}}}{2}\left[\left(\hat{\mathbf{x}}_{i}^{\dagger}\hat{\mathbf{b}}_{\mathbf{q}}^{\dagger}\hat{\mathbf{x}}_{i}-\hat{\mathbf{b}}_{\mathbf{\mathbf{q}}}^{\dagger}\right)_{\alpha\beta}\left(2\hat{\mathbf{x}}_{j}^{\dagger}\mathbf{m}_{j,\mathbf{q}}^{\dagger}\hat{\mathbf{x}}_{j}+\hat{\mathbf{x}}_{j}^{\dagger}\hat{\mathbf{b}}_{\mathbf{q}}\hat{\mathbf{x}}_{j}-\hat{\mathbf{b}}_{\mathbf{q}}\right)_{\alpha'\beta'}+h.c.\right]\hat{a}_{i}^{\alpha\dagger}\hat{a}_{j}^{\alpha'\dagger}\hat{a}_{j}^{\beta'}\hat{a}_{i}^{\beta},\label{eq : transformed - hamil_0}\end{aligned}\ ] ] where @xmath131 and @xmath132 is the ( diagonal ) matrix form of @xmath27 and @xmath28 .
here we combine all single impurity contributions in the second line .
the last term , which is zero when only a single impurity is considered , describes induced polaron - polaron interactions due to the coupling with the phonon bath .
compared to the original hamiltonian in eq .
( [ eq : hamil_0 ] ) , the hopping and on - site energy have been modified . as for the single band system
, this hamiltonian contains all interactions exactly and is hard to solve analytically .
motivated by the single - band system , we separate the hamiltonian into coherent and incoherent parts , @xmath113 and @xmath133 respectively .
the explicit form of the coherent terms , which conserve the number of phonons , is determined in appendix [ sub : app_polaron - transformation ] . in eqs .
( [ eq : thermal_1_end ] , [ eq : thermal_2 ] , [ eq : thermal_3 ] , [ eq : thermal_4 ] ) we calculate all possible coherent terms in eq .
( [ eq : transformed - hamil_0 ] ) . after determining the variational parameters @xmath134 by minimizing the free energy of the coherent part
, we treat the residual polaron - bath coupling in the incoherent part as a perturbation and solve it by a lindblad master equation . in appendix [ sub : app_polaron - transformation ] , we calculate the coherent part @xmath113 by averaging over the phonon bath and assuming it is thermal .
in contrast to the single - band case , these calculations are demanding when both intra- and inter - band phonon couplings are included . in order to determine the variational parameters @xmath135
, we finally need to minimize the free energy of whole coherent hamiltonian . until now
, we did not make any assumptions for our variational parameters @xmath135 except for the symmetry relations in eq .
( [ eq : m_relation ] ) . unfortunately , the general result of the transformed hamiltonian in eq .
( [ eq : transformed - hamil_0 ] ) and its corresponding coherent part are still quite complicated .
it can be further simplified by making some approximations suitable to our specific system . due to conservation of energy ,
the phonon - induced inter - band dynamics requires the phonon energy to match the impurity band gap , i.e. @xmath136 .
this energy scale involves a phonon with particle - like dispersion and momentum @xmath137 significantly far from zero .
the inter - band coupling @xmath138 for this large phonon momentum is highly reduced due to the gaussian decay of @xmath139 in eq .
( [ eq : coupling_factor ] ) . on the other hand ,
intra - band dynamics requires a phonon energy @xmath140 with phonon - like dispersion and small momentum @xmath141 .
the intra - band coupling @xmath142 is not reduced too much at this smaller momentum . in the polaron transformation ,
the parameters @xmath143 reflect the dressing of the impurity by phonons and are closely related to @xmath121 . for this reason ,
we treat the inter - band coupling as small and approximate the matrices @xmath135 as diagonal : @xmath144,\ ] ] with variational parameters @xmath145 and @xmath146 . since the intra - band couplings @xmath142 are purely real numbers in our system , we assume @xmath147 are also real numbers .
after the transformation with these diagonal matrices @xmath148 , we have then decoupled the intra - band impurity - phonon coupling and leave the ( relatively ) small inter - band coupling in the new polaronic two - band hamiltonian .
the coherent part , with phonons eliminated by thermal averaging , is a many - body hamiltonian : @xmath149 with renormalized polaronic hopping terms @xmath150 and on - site energy @xmath151 including the polaron energy shift : @xmath152 in the above formula , we use the fact that operators @xmath127 have only diagonal terms . there are also induced interactions @xmath153 between multiple polarons , which can lead to strong correlations in the system and will be discussed in the next section . in the above calculations , we need to sum over all possible phonon momenta @xmath56 . in the thermodynamic limit of the phonon bath
, we use the relation @xmath154 with quantization volume @xmath60 for the phonons , and write this explicitly in cylindrical coordinates : @xmath155 for the polaronic intra - band hopping @xmath150 , we only need to calculate nearest - neighbor terms with @xmath156 regardless of the specific value of @xmath157 . by noting that @xmath158 where @xmath159 is the thermally averaged phonon occupation number , we define polaronic renormalization factors @xmath160|\lambda_{\mathbf{q}}^{\alpha\alpha}|^{2}\ ] ] for each band and thus @xmath161 .
we would like to note that the choice of diagonal transformation matrices @xmath148 works well when the inter - band coupling is less important that the intra - band coupling . by choosing diagonal matrices
, we treat each of the two bands independently .
this approach is similar as the authors in ref .
@xcite did for single band problem , except a variational treatment was used . on the other hand , when the off - diagonal terms in the coupling matrix @xmath162 are comparable or larger than the diagonal terms , one must instead deal with the general transformed hamiltonian in eq .
( [ eq : transformed - hamil_0 ] ) . for completeness
, we describe a more general approximate technique in appendix [ sub : app_polaron - transformation ] , which may be useful for other types of systems .
we intend to address the approximation introduced in this section more explicitly and quantitatively in a future publication .
in the previous section , we derived a general form of the two - band polaron transformation and calculated the resulting coherent part of the hamiltonian in eq .
( [ eq : coherent ] ) . for the system with a single polaron
, there are only intra - band terms in the coherent hamiltonian ( [ eq : coherent ] ) .
the single impurity is dressed by a coherent phonon cloud in each band .
the residual inter - band polaron - bath coupling will appear only in the incoherent hamiltonian .
it is easy to diagonalize the single - polaron coherent part in the momentum representation and minimize the free energy @xmath163 for this two - band system , with polaron dispersion : @xmath164 here @xmath165 is quasi - momentum in longitudinal direction . these variational parameters , which are real numbers , can then be determined by the self - consistent equations : @xmath166\exp\left(-e_{k}^{\alpha'}/k_{\text{b}}t\right)}\label{eq : lambda}\ ] ] with @xmath167 $ ] . before numerically calculating the variational parameters @xmath147 , we firstly discuss some properties of this self - consistent equation . considering a simplified model with momentum - independent variational parameters @xmath11 in a single band system ,
we choose to minimize only the ground state energy .
the self - consistent equation ( [ eq : lambda ] ) will be modified as @xmath168^{-1}.\ ] ] from the analysis in @xcite , this equation has two locally stable solutions @xmath169 once the _ adiabatic _ regime is achieved when @xmath170 these two solutions @xmath169 , corresponding to two local minima of ground state energy , indicate the impurity is respectively loosely or tightly dressed by phonons . at a critical impurity - phonon coupling with @xmath171 , when the two minima become equal
, the lowest energy state solution abruptly switches from @xmath172 to @xmath173 , indicating a first - order _ polaronic transition_. on the other hand , the solution of @xmath11 is a smooth and continuous crossover when the adiabatic condition is broken with small value of @xmath174 this transition - crossover behavior also appears later when solving the self - consistent equation ( [ eq : lambda ] ) numerically , although we have considered a much more simplified model here . strictly speaking
, this sharp polaronic transition in the adiabatic regime is due to the mean - field approximation by thermal averaging of the phonon degrees of freedom .
this drawback could be improved if we were to treat the incoherent dynamics properly , by taking into account fluctuations or using a master equation method .
we now compare the variational parameters @xmath175 from eq .
( [ eq : lambda ] ) and polaron dressing effects for different system parameters . in fig .
[ fig : variational lambda](a ) we show the coupling factor @xmath147 in each band for the temperatures @xmath176 , where it can be seen that the variational parameters @xmath147 are always smaller than 1 .
this shows the competition between intra - band dynamics and phonon dressing effects .
these variational parameters are also different for the two bands , since the polaron is dressed differently in each band .
we have focused on the results for momentum in the longitudinal direction with @xmath177 and find that the low - momentum phonons are less dressed at higher temperatures @xmath178 . here
@xmath179 corresponds to a temperature of about @xmath180 for a @xmath181 impurity trapped by lasers with wavelength @xmath182 . in fig .
[ fig : variational lambda](b , c ) we also show the the factors @xmath183 . these factors @xmath184 thus indicate the differences between our initial guess for the variational parameters @xmath185 , as is sometimes used for the transformation , and the full minimization of free energy for the variational parameters . as shown in fig . [
fig : variational lambda ] , although these factors are different for each band and different temperatures , they always approach @xmath186 in the limit of strong interaction . in order to reach larger values of the coupling constant , in fig .
[ fig : variational lambda](c ) we show results with @xmath187 by assuming condensate density @xmath188 . in the upper plot of fig .
[ fig : variational lambda](c ) , we notice that the parameter @xmath11 shows a polaronic transition in the adiabatic regime when the condition eq .
( [ eq : adiabatic ] ) is satisfied at finite temperature . on the other hand
, @xmath11 shows a smooth crossover behavior in the non - adiabatic regime at zero temperature . with @xmath177 and @xmath189 in higher band ( upper ) and lower band ( lower ) at different temperatures @xmath176 .
( b)(c ) factors @xmath190 in each band at different temperatures .
maximum coupling constant @xmath191 for condensate density @xmath192 ( b ) and @xmath187 for condensate density @xmath188(c ) .
other parameters are : @xmath193 ; @xmath194 and @xmath106 .
[ fig : variational lambda],scaledwidth=50.0% ] in the coherent part of the hamiltonian eq .
( [ eq : coherent ] ) , the single polaron band structure is modified by phonon dressing effects in eq .
( [ eq : j_eff ] ) .
effectively , the polaron is trapped in a deeper lattice , with larger mass .
the effective mass of a single polaron ( at @xmath195 ) in an optical lattice can be defined as @xmath196 if the impurity - bec coupling @xmath41 increases , the polaron effective mass will increase exponentially as @xmath197 , with @xmath198 indicating the impurity effective mass at @xmath199 . in fig .
[ fig:7 e_k ] we compare energy spectrum @xmath200 , renormalization of intra - band hopping @xmath201 , polaron effective mass @xmath202 and renormalization factor @xmath203 for each band at different temperatures .
due to phonon dressing effects , the polaronic band gap @xmath204 is also increased .
this will affect the inter - band relaxation dynamics . in fig .
we show the on - site polaron energy and band gap renormalization versus impurity - bec coupling constant . in both fig .
[ fig:7 e_k](a ) and fig .
[ fig:8 e_p-1](a ) , without loss of generality , we set the initial lower band on - site energy @xmath205 to zero . as shown in fig .
[ fig:8 e_p-1](b ) , the band gap renormalization is almost temperature independent .
these quantities are only slightly affected by temperature due to different variational transformation matrices @xmath206 , as predicted in eq .
( [ eq : j_eff ] ) .
although the band gap is not significantly changed , this renormalization effect is important for inter - band resonance conditions , which are required for landau - zener tunneling to take place in a tilted lattice @xcite . including polaron energy shift at zero temperature ; ( b ) renormalized band gap @xmath204 at different temperatures @xmath207 .
the energy renormalization is almost temperature - independent .
other parameters are @xmath193 ; @xmath194 ; @xmath208 and @xmath191 .
[ fig:8 e_p-1],scaledwidth=50.0% ] between different band at @xmath189 .
( b ) ratio between induced on - site and nearest neighbor interaction terms .
the healing length @xmath89 is increased by reducing the condensate density @xmath102 from @xmath209 to @xmath210 while keeping the coupling constant @xmath211 .
( c ) effective polaronic interactions in a multiple tube setup with two bloch bands .
other parameters are @xmath193 ; @xmath194 and @xmath212 .
[ fig:9 effect_v_ij-1],scaledwidth=50.0% ] here we briefly discuss interactions between polarons , by considering the additional interactions between impurities beyond the original hamiltonian ( [ eq : hamil_0 ] ) , which are given by @xmath213 with the impurity on - site intra-/inter - band hubbard - type interaction @xmath214 . after the polaron transformation , the coherent hamiltonian eq .
( [ eq : coherent ] ) for the multi - polaron system contains effective interactions @xmath153 with both intra- and inter - band terms . in principle , the variational parameters @xmath215 should be determined by minimizing the total free energy .
this is hard for the many - impurity system .
here we assume that the polaron parameters are not modified from the single polaron case , due to the low density of impurities in our dilute system .
the effective interaction terms @xmath153 in hamiltonian eq .
( [ eq : coherent ] ) take the form : @xmath216 with long - range bath induced interaction terms : @xmath217\nonumber \\ & \:\quad\times\left(\lambda_{\mathbf{q}}^{\alpha\alpha}\right)\left(2m_{\mathbf{q}}^{\beta\beta}-\lambda_{\mathbf{q}}^{\beta\beta}\right)\end{aligned}\ ] ] and @xmath218\nonumber \\ & \;\times\left(\lambda_{\mathbf{q}}^{\alpha\alpha}\right)\left(2i\cdot m_{\mathbf{q}}^{\beta\beta'}\right)\langle\hat{k}_{j , j}^{\beta\beta'}\rangle_{t},\end{aligned}\ ] ] with @xmath219 as it was shown in @xcite , these induced density - density interactions @xmath220 are always attractive with @xmath221 and take the form of yukawa - type interactions . the first line in eq .
( [ eq : interaction ] ) contains on - site effective interactions between polarons , also within different bands . due to these attractive interactions @xmath220 , the final on - site interactions @xmath222 must be repulsive for bosonic impurities in order to keep the system stable . the second part in eq .
( [ eq : interaction ] ) describes the long - range part of these induced interactions .
there are also density - induced inter - band transitions , which are due to the inter - band polaron - bath coupling . in fig .
[ fig:9 effect_v_ij-1](a ) we show the induced density - density interactions @xmath220 versus distance @xmath223 between polarons between each band . since these interactions are induced by the impurity - bec coupling , the long - range behavior is related to the condensate healing length @xmath89 . here
we assume we can obtain large values for @xmath89 by tuning the bose - bose scattering length @xmath39 to a small positive value or keeping the condensate density @xmath102 small . in fig .
[ fig:9 effect_v_ij-1](b ) we also show the ratio of @xmath220 between on - site and nearest neighbor terms versus @xmath89 .
these ratios increase with the bec healing length , indicating a longer effective range of the interactions .
the effects of these interactions , which are beyond this paper , might include new ordered polaron phases in our system with hamiltonian eq .
( [ eq : coherent ] ) . as shown in fig .
[ fig:9 effect_v_ij-1](c ) , if we consider that the transverse confinement is due to a very deep optical lattice instead of a single well , different 1d tubes could then interact due to the bath - induced long range ( attractive ) interactions .
a more rich phase diagram is expected in such mixed - dimensional system coupled to a bosonic bath @xcite .
the incoherent part of the hamiltonian @xmath224 includes the residual coupling between polarons and phonon bath : @xmath225\hat{a}_{i}^{\alpha\dagger}\hat{a}_{i}^{\alpha}\nonumber \\ + & \sum_{i}\sum_{\alpha\neq\beta}\left\ { \sum_{\mathbf{q}}\hbar\omega_{\mathbf{q}}\left[\hat{b}_{\mathbf{q}}^{\dagger}\left(m_{i,\mathbf{q}}^{\alpha\beta}\right)^{*}\left(\hat{k}_{i , i}^{\alpha\beta}\right)-\left(\lambda_{i,\mathbf{q}}^{\alpha\alpha}\right)\left(m_{i,\mathbf{q}}^{\alpha\beta}\right)^{*}\left(\hat{t}_{i , i}^{\alpha\beta}\right)\right]+h.c.\right\ } \hat{a}_{i}^{\alpha\dagger}\hat{a}_{i}^{\beta}\nonumber \\ - & \sum_{i\neq j}\sum_{\alpha}\sum_{\alpha'\neq\beta'}\left\ { \sum_{\mathbf{q}}\hbar\omega_{\mathbf{q}}\left[\left(\lambda_{i,\mathbf{q}}^{\alpha\alpha}\right)\left(m_{j,\mathbf{q}}^{\alpha'\beta'}\right)^{*}\left(\hat{t}_{j , j}^{\alpha'\beta'}\right)\right]+h.c.\right\ } \hat{a}_{i}^{\alpha\dagger}\hat{a}_{i}^{\alpha}\hat{a}_{j}^{\alpha'\dagger}\hat{a}_{j}^{\beta'}.\label{eq : incoherent}\end{aligned}\ ] ] with @xmath226 .
there are three different coupling terms in eq .
( [ eq : incoherent ] ) : 1 ) an intra - band part due to hopping of the polaron to nearest neighbor sites and residual phonon dressing effects ; 2 ) an inter - band part due to polaron dynamics between two bands ; and 3 ) a mixed many - body term due to effective interactions between polarons . the coherent part of the hamiltonian eq .
( [ eq : coherent ] ) , as calculated before , is a two - band hubbard hamiltonian . in order to focus on inter - band relaxation effects ,
we restrict our investigations to the case of a single polaron and ignore the interaction part @xmath227 . the incoherent hamiltonian part eq .
( [ eq : incoherent ] ) can be treated by a lindblad master equation . after applying the born - markov approximation , the lindblad equation has the form : @xmath228+\mathcal{l}_{i}\left[\rho\left(t\right)\right]\ ] ] with reduced density operator @xmath229 of the polaron system .
all decoherence effects in the system are described by the dissipator : @xmath230= & \sum_{\alpha\beta}\sum_{\alpha'\beta'}\sum_{i , j}\sum_{i',j'}\gamma_{ij;i'j'}^{\alpha\beta;\alpha'\beta'}\nonumber \\ & \times\left(c_{i'j'}^{\alpha'\beta'}\rho c_{ij}^{\alpha\beta\dagger}-\frac{1}{2}\left\ { c_{ij}^{\alpha\beta\dagger}c_{i'j'}^{\alpha'\beta'},\;\rho\right\ } \right)\label{eq : lindblad}\end{aligned}\ ] ] with the quantum jump operators @xmath231 .
these jump operators describe the dynamics of polarons instead of bare impurities , since they include the creation and annihilation operators of the polaron . in the interaction picture with @xmath232 ,
the decoherence rates are defined as : @xmath233 and the correlation functions @xmath234 are defined in appendix [ sub : app_lindblad_polaron ] . in this section
, we focus our study on inter - band spontaneous relaxation of the polaron . due to longitudinal
trapping , the energy scales for intra- and inter - band dynamics are mismatched as @xmath235 .
this allows us to use the rotating wave approximation ( rwa ) for inter - band dynamics and decouple it from intra - band dynamics in eq .
( [ eq : lindblad_rwa ] ) .
we then use the short - hand notation @xmath236 and @xmath237 to describe the polaron relaxation * * processes . due to the coupling with the bath
, a polaron in the upper band can spontaneously relax to the lower band with rate @xmath238 and emit a phonon . in our system ,
the corresponding phonon energy @xmath35 for inter - band dynamics is much larger than the bec temperature @xmath239 .
for inter - band dynamics , the phonon bath temperature is thus effectively zero . in the interaction picture ,
this process is described by the master equation @xmath240 the single polaron spontaneous relaxation rate @xmath238 is calculated in eq .
( [ eq : decay_rate_1 ] ) .
it is convenient to write the master equation for relaxation processes in momentum space , as long as @xmath238 only depends on the value of @xmath241 : @xmath242 with @xmath243 and @xmath244 is the polaron annihilation operator in momentum space @xcite . here , @xmath245 are the quasi - momenta in the first brillouin zone and @xmath246 . the relaxation rate is written as a function of longitudinal phonon momentum : @xmath247 where @xmath248 is lattice number .
when @xmath249 , the sum over with @xmath75 , involving polaron relaxation effects over different sites , decays rapidly with @xmath223 . the total relaxation rate @xmath250 is a sum over longitudinal phonon momenta , @xmath251 @xcite .
from the expression of the correlation function @xmath252 in eq .
( [ eq : correlation_g-1 ] ) , the polaronic inter - band relaxation contains two terms .
the first term in ( [ eq : correlation_g-1 ] ) is a single - phonon process , with the polaron absorbing ( emitting ) one phonon from ( to ) the bath .
this term is similar to fermi s golden rule except a renormalization factor @xmath253 .
the second term describes higher order processes involving two phonons begin absorbed ( emitted ) , which are absent in fermi s golden rule . in the total relaxation rate , we will only consider the leading - order single - phonon process in eq .
( [ eq : decay_rate_1 ] ) and neglect the higher - order processes . at zero temperature , only spontaneous emission of phonons is allowed with the single polaron spontaneous relaxation rate @xmath254 where @xmath255 is defined in eq .
( [ eq : kikj ] ) as a renormalization factor for inter - band relaxation dynamics . in order to investigate
how polaron effects affect the inter - band dynamics , we also derive a lindblad equation for the original `` bare '' hamiltonian in eq .
( [ eq : hamil_0 ] ) , before the lang - firsov transformation had been applied .
the corresponding single - impurity spontaneous relaxation rate in this case is @xmath256 and reduces to fermi s golden rule formula : @xmath257 comparing eq .
( [ eq : gamma_polaron ] ) and eq .
( [ eq : gamma_bare ] ) , we observe that the renormalization of the band gap from @xmath258 to @xmath259 increases the energy of the phonon which is created , due to energy conservation .
the polaron relaxation rate @xmath260 is also reduced by an additional exponential factor , since the coupling between impurity and bath is also renormalized .
before we discuss the numerical results for the relaxation rate , let us first look at the behavior of @xmath261 as a function of phonon momentum in the longitudinal direction .
for the bare particle relaxation rate , the prediction of fermi s golden rule gives : @xmath262 in a deep lattice , the impurity - phonon coupling matrices @xmath121 can be approximated by gaussian functions in eq .
( [ eq : coupling_factor ] ) . then the value of fermi s golden rule for @xmath263 in eq .
( [ eq : gamma_bare_q ] ) is found to be : @xmath264 where @xmath265 is the phonon quantization length ( analogous to the quantization volume @xmath60 ) in the longitudinal direction .
the phonon momentum in the transverse direction @xmath266 is fixed by energy conservation @xmath267 , where @xmath268 and @xmath269 is the longitudinal oscillation frequency .
an additional consequence of energy conservation , is that @xmath261 is cut off when @xmath270 .
the cut - off is @xmath271 which only depends on the mass ratio @xmath103 between impurity and bec particles . on the other hand , the polaron relaxation rate @xmath272 versus longitudinal phonon momentum needs to be calculated numerically from eq .
( [ eq : gamma_q ] ) and eq .
( [ eq : decay_rate_1 ] ) . as we discussed before for the total relaxation rate eq .
( [ eq : gamma_polaron ] ) , we consider only the single - phonon emitting processes and calculate @xmath272 as : @xmath273 comparing with eq .
( [ eq : gamma_bare_q ] ) , the polaron relaxation rate has a similar behavior except for an additional renormalization factor . due to energy conservation
, @xmath272 also has a cut - off when @xmath274 .
the renormalized polaron band gap @xmath259 , which is larger than the bare gap @xmath258 , will shift this cut - off position to higher phonon momentum . at the same time , the renormalization factor in eq .
( [ eq : gamma_polaron_q ] ) will reduce the value of @xmath272 . in fig .
[ fig : decay - q ] we show the behavior of the relaxation rate @xmath272 and @xmath263 as functions of longitudinal momentum @xmath275 with different impurity - bec mass ratio . in order to compare the differences between @xmath272 and @xmath263 at different coupling ,
we divide both @xmath272 and @xmath263 by @xmath276 and then normalize them . in this way
, the fermi s golden rule value @xmath263 always stays the same for different coupling constants @xmath87 . as shown in fig .
[ fig : decay - q ] , the dashed black lines indicate normalized @xmath263 from eq .
( [ eq : gamma_bare_q_app ] ) , while the cut - off of @xmath277 in eq .
( [ eq : cut_off ] ) is indicated by black lines .
the total relaxation rate @xmath278 from fermi s golden rule is the gray area below the black curve . at @xmath199 , the polaronic relaxation rate @xmath272 is identical to the results from the golden rule . on the other hand , as the impurity - bec coupling is increased , the polaron relaxation rate @xmath272 from eq .
( [ eq : gamma_polaron_q ] ) is renormalized by the polaron band gap @xmath259 and renormalization factor . the increased polaron band gap will involve more phonons by shifting the cut - off momentum ( [ eq : cut_off ] ) , while the renormalization factor reduces the whole momentum range . in fig .
[ fig : decay - q ] we also show the polaron relaxation rate @xmath272 with blue ( red ) curve at @xmath279 and total relaxation rate @xmath260 with blue ( red ) area . as shown in fig .
[ fig : decay - q ] , the renormalization of the relaxation rate is different for various impurity - bec mass ratios . for larger mass ratio ( heavy impurity ) such as a system with single @xmath181 impurity coupled with @xmath105 bec
, the polaron relaxation processes will be enhanced by the shift of the cut - off momentum and increase the total rate . on the other hand , for smaller mass ratio ( light impurity ) such as a @xmath280 impurity coupled with @xmath281 bec ,
the higher momentum cut - off for @xmath277 is not so important due to gaussian decay of @xmath261 . in this case
, the renormalization factor will reduce @xmath272 as well as the total relaxation rate @xmath260 .
we can also expect that , for extremely strong impurity - bec coupling , the total relaxation rate will be reduced due to this renormalization factor with any mass ratio . and
@xmath263 as functions of longitudinal momentum @xmath275 with different impurity - bec mass ratio .
the dashed black lines indicate fermi s golden rule results @xmath263 from eq .
( [ eq : gamma_bare_q_app ] ) with momentum cut - off given by the solid black line ( @xmath199 ) .
the blue ( red ) curves indicate normalized polaron relaxation rate @xmath272 at coupling @xmath282 .
the total relaxation rates @xmath283 are shown as the area below the corresponding curves .
( a ) system with @xmath284 , i.e. @xmath181 impurity in @xmath105 bec .
( b ) system with @xmath285 , i.e. @xmath280 impurity in @xmath281 bec .
other parameters are @xmath194 and @xmath208 .
[ fig : decay - q],scaledwidth=50.0% ] and fermi golden rule results @xmath286 at different impurity - bec mass ratio . ( a ) in the weak coupling regime , this ratio is increased with the coupling constant @xmath87 for a heavy impurity coupled to a light bec bath , and decreased for a light impurity coupled to a heavy bec bath .
the boson - boson scattering length here is chosen as @xmath106 and @xmath191 .
( b ) in the strong coupling regime , the polaron inter - band relaxation processes for any impurity - bec mass ratio are suppressed .
this leads to an inter - band self - trapping effect .
the boson - boson scattering length here is @xmath287 and @xmath288 .
we use @xmath193 for a @xmath181 impurity and @xmath105 bec ( blue line ) , @xmath289 for a @xmath290 impurity and @xmath291 bec ( green line ) , and @xmath292 for a @xmath280 impurity and @xmath281 bec ( red line ) .
other parameters are @xmath194 and @xmath192.[fig_decay],scaledwidth=50.0% ] finally we compare the polaronic inter - band relaxation rate @xmath293 and fermi golden rule results @xmath286 in fig.@xmath294[fig_decay ] as a function of coupling constant and for different impurity - bec mass ratios . as we have already seen in fig.@xmath294[fig : decay - q ] , the polaron formation will renormalize the inter - band relaxation rate differently , depending on the mass ratio .
we first consider the weak impurity - bec coupling regime with small @xmath87 in fig.@xmath294[fig_decay](a ) . for a heavy impurity coupled to a light bec bath ,
the inter - band relaxation process involves more phonon modes and will be enhanced .
the difference in the ratio @xmath295 between polaronic relaxation and the fermi golden rule result increases with coupling constant @xmath87 . on the other hand , for a light impurity
coupled to a heavy bec , the impurity is dressed by heavy phonons in each band and tends to localize in the same band . although the inter - band relaxation process does involve more phonon modes than the fermi golden rule result , this effect is highly suppressed due to gaussian decay of @xmath261 with high momentum .
the ratio @xmath295 will be reduced in this system .
we also notice that the ratio @xmath295 is not sensitive to the longitudinal trapping potential . on the other hand
, we can access the strong coupling regime by tuning the bose - bose scattering @xmath39 to a small positive value , since the allowed maximum coupling constant in eq .
( [ eq : kappa_c ] ) goes as @xmath296 . for the strong impurity - bec coupling region with large @xmath87 in fig.@xmath294[fig_decay](b ) , though the lindblad master equation might not be accurate enough for such strong inter - band coupling , we can still obtain some qualitative insight . in each band , the polaron is tightly dressed by phonons with different coupling strength , such that the polaron behaves as a quasi - particle with rather different properties in each band . also , the band gap @xmath259 is enhanced in comparison to the bare case , so the polaron can not hop between bands by creating or annihilating phonons . as shown in fig.@xmath294[fig_decay](b ) , we indeed obtain a suppressed inter - band relaxation process
inter - band self - trapping _ effect is expected in a strongly coupled impurity - bec system . in a realistic impurity - bec system , this effect might be also observed together with the well known _
self - trapping _ effect due to deformation of bec @xcite .
in conclusion , we have studied a two - band hamiltonian with frhlich impurity - phonon coupling with both intra- and inter - band terms .
such a hamiltonian can be realized in experiments where few impurities are immersed in a bose - einstein condensate of another species .
the impurities are trapped by an anisotropic optical lattice behave as quasi-1d particles .
based on the lang - firsov transformation , we have derived and applied a variational two - band polaron transformation .
we have calculated the coherent part of the resulting effective hamiltonian with two ( polaron ) bands . in each band
the impurity is dressed by phonons as a quasi - particle ( polaron ) with different properties .
the polaronic intra - band coherent transport and polaron effective mass are both renormalized . due to the coupling with bath ,
there are also induced on - site polaron energy shifts , and long - range interactions between different polarons . in order to account for the residual incoherent coupling between polaron and bath , we have derived a lindblad master equation , and focused on the single - polaron inter - band relaxation dynamics . comparing to fermi s golden rule calculations for the bare impurities
, we found the renormalization of the relaxation rate to depend on the mass ratio of impurity and bec particles .
these polaronic effects in the inter - band relaxation dynamics might be observed in ongoing experiments . in the strong coupling limit of the two - band frhlich hamiltonian
, the impurity is _ inter - band self - trapped _ and can not tunnel between bands by creating or annihilating phonons .
the authors thank a. daley , e. demler , r. gerritsma , f. grusdt , g. pupillo , t. rentrop , r. schmidt , t. shi and a. widera for fruitful discussions .
support by the deutsche forschungsgemeinschaft ( dfg ) via sonderforschungsbereich sfb / tr 49 , forschergruppe for 801 and the high - performance computing center loewe - csc is gratefully acknowledged .
in this section , we derive the two - band polaron transformation in detail . the transformation operator @xmath297 in eq.@xmath294([eq : transformation_s ] ) can be written as @xmath298 with @xmath299 or in matrix form as @xmath300 . by using the relation @xmath301=-\delta_{ij}\delta_{\alpha\gamma}\hat{a}_{i}^{\beta}$ ] , which is valid for both bosons and fermions ,
we obtain the transformed impurity annihilation operator as : @xmath302= & -\sum_{\beta}\left(\hat{\mathbf{c}}_{i}\right)_{\alpha\beta}\hat{a}_{i}^{\beta};\quad\left[\hat{s},\;\left[\hat{s},\;\hat{a}_{i}^{\alpha}\right]\right]=+\sum_{\beta}\left(\hat{\mathbf{c}}_{i}\hat{\mathbf{c}_{i}}\right)_{\alpha\beta}\hat{a}_{i}^{\beta};\quad\cdots;\\ e^{\hat{s}}\hat{a}_{i}^{\alpha}e^{-\hat{s}}= & \sum_{\beta}\left(e^{-\hat{\mathbf{c}}_{i}}\right)_{\alpha\beta}\hat{a}_{i}^{\beta}=\sum_{\beta}\left(\hat{\mathbf{x}}_{i}\right)_{\alpha\beta}\hat{a}_{i}^{\beta},\end{aligned}\ ] ] with @xmath303 is defined in eq .
( [ eq : transform_a ] ) .
the phonon annihilation operators can also be transformed in a similar way as : @xmath304= & \sum_{i,\alpha,\beta}\left(\left[\hat{\mathbf{c}}_{i},\;\hat{\mathbf{b}}_{\mathbf{q}}\right]\right)_{\alpha,\beta}\hat{a}_{i}^{\alpha\dagger}\hat{a}_{i}^{\beta};\\ \left[\hat{s},\;\left[\hat{s},\;\hat{b}_{\mathbf{q}}\right]\right]= & \left[\sum_{i,\alpha,\beta}\hat{c}_{i}^{\alpha,\beta}\hat{a}_{i}^{\alpha\dagger}\hat{a}_{i}^{\beta},\;\sum_{i',\alpha',\beta'}\left(\left[\hat{\mathbf{c}}_{i'},\hat{\mathbf{b}}_{\mathbf{q}}\right]\right)_{\alpha',\beta'}\hat{a}_{i'}^{\alpha'\dagger}\hat{a}_{i'}^{\beta'}\right]=\sum_{i,\alpha,\beta}\left(\left[\hat{\mathbf{c}}_{i},\;\left[\hat{\mathbf{c}}_{i},\hat{\mathbf{b}}_{\mathbf{q}}\right]\right]\right)_{\alpha\beta}\hat{a}_{i}^{\alpha\dagger}\hat{a}_{i}^{\beta};\\ \vdots\nonumber \end{aligned}\ ] ] and thus @xmath305 different from the single band case , all higher commutators in eq .
( [ eq : trans_b ] ) are nonzero except when @xmath306=0 $ ] .
the hamiltonian is transformed as : @xmath307 the resulting polaron hamiltonian is presented in eq .
( [ eq : transformed - hamil_0 ] ) . for completeness , we describe a general calculation of coherent hamiltonian after two - band transformation .
we apply a _ shorthand notation _
technique , which was used by sibley and munn @xcite , to our two - band system .
the transformed hamiltonian in eq .
( [ eq : hamil_0 ] ) is complicated since it has no assumptions for the variational parameters @xmath308 except the symmetry relations in eq .
( [ eq : m_relation ] ) . on the other hand , for our specific choice of assuming diagonal matrices for the variational parameters ,
all of the expressions in this section can be shown to be straightforward and simple .
the coherent part @xmath113 of eq .
( [ eq : hamil_0 ] ) can be calculated by averaging over the phonon bath . here
we assume the phonon bath is thermally distributed at the bec temperature . from eq .
( [ eq : hamil_0 ] ) we find these calculations require several types of thermal average values such as ( i ) @xmath309 , ( ii ) @xmath310 , ( iii ) @xmath311 and ( iv ) @xmath312 .
note that some terms inside @xmath313 are @xmath123 matrices . here
we will discuss these terms in details , without making any assumption for the parameters @xmath135 except that they obey the same symmetry relations as for @xmath162 as given in eq .
( [ eq : m_relation ] ) .
the first coherent part describes a matrix @xmath131 or @xmath132 transformed by operator @xmath127 .
we use a general @xmath123 matrix @xmath315 instead of @xmath316 or @xmath317 as : @xmath318 in the last step , we apply the shorthand notation technique and imply index @xmath319 for all functions of @xmath320 and the index @xmath321 for all functions of @xmath322 @xcite .
this shorthand notation ignores the commutation of matrices and collects all terms of @xmath320 together and all @xmath322 together , before applying the indices . for an arbitrary operator @xmath323 and an arbitrary type of averaging involved
, the expansion up to the second - order cumulant reads @xmath324 this expansion is exact if applied without truncation directly to equation ( [ eq : first - shorthand ] ) when the matrices @xmath323 commute , i.e. @xmath325=0 $ ] .
however , we make the approximation @xcite , that the exponential in the short - hand notation also follows such a cumulant expansion . for small deviations from community
, this will be a negligible approximation . since the operators @xmath326 contain only creation and annihilation operators , the coherent part @xmath327 . in this case
, we have @xmath328 . for the single band calculation
, this relation reduces to the _ bloch identity _ and is valid exactly @xcite .
however , in the two band system this relation is only an approximation because higher order terms are also present .
the accuracy depends on the commutator @xmath329 $ ] .
here we first apply this approximation with general parameters @xmath135 : @xmath330 the exponents above contains matrices @xmath331 , which do not commute with each other .
fortunately , by using the shorthand notation @xmath332 , those matrix elements with index @xmath333 do commute between each other .
they can be separated into independent exponents .
thermal averages such as @xmath334 can be calculated easily : @xmath335 where @xmath159 is the thermally averaged phonon occupation number .
the result of @xmath334 is guaranteed to be a real number because of the symmetry relations @xmath336 .
the thermal averages in eq .
( [ eq : thermal_1 ] ) can be calculated as : @xmath337 the first and third term are easy to calculate .
then we expand the middle exponential under the shorthand notation @xmath338 : @xmath339 in each term , we combine elements with the same momentum @xmath56 : @xmath340 in order to simplify last part in eq .
( [ eq : expg ] ) , we define 4th rank tensors * @xmath341 * with elements : @xmath342 and @xmath343 then the eq .
( [ eq : expg ] ) can be written as : @xmath344 where @xmath345 are also 4th rank tensors .
the tensors * @xmath341 * involve only numbers , and thus @xmath345 can be calculated exactly .
finally , the first thermal average term can be calculated as : @xmath346 here we calculate the second type of thermal averages by the relation @xmath347 : @xmath348 where we have used the relation @xmath349=-\mathbf{\lambda}_{i,\mathbf{q}}$ ] .
the first term can be calculated by separating out the bath terms : @xmath350 this element can be calculated by the same method as we applied for @xmath351 in the previous section . under the same shorthand notation : @xmath352 with @xmath353 applied separately for @xmath354 , @xmath355 and @xmath356 .
we have introduced the different indices @xmath157 , @xmath24 and @xmath165 so that we can keep track of the terms required in applying the short - hand notation . in the end , we will set @xmath357 to recover the desired result . using the results from previous section
, these thermal averages will be : @xmath358 with @xmath341 from eq .
( [ eq : matrices_g ] ) . on the other hand , the second term @xmath359 in eq .
( [ eq : thermal_2 ] ) can be calculated as : @xmath360 from the results in @xcite , the elements such as @xmath361 can be obtained by differentiating @xmath362 before taking the thermal average . on the other hand ,
since the thermal average only involves the phonon degrees of freedom , we can switch the order and differentiate @xmath363 after taking the thermal average . by comparing the two steps , we obtain the relation : @xmath364 which can be calculated by the results in eq .
( [ eq : thermal_1_end ] ) . as an example
, we show a simplified calculation in eq.([eq:<b_k > ] ) , where the transformation matrices @xmath365 are diagonal .
we finally obtain the second term in eq .
( [ eq : thermal_2 ] ) : @xmath366.\label{eq : thermal_2_3}\end{aligned}\ ] ] this term can be calculated by a similar integral method as in the previous section : @xmath367 here the thermal average @xmath368 can be obtained by changing @xmath320 to @xmath369 and @xmath370 to @xmath365 in eq .
( [ eq : thermal_2 ] ) : @xmath371\nonumber \\ & + n_{\mathbf{q}}\sum_{\alpha'\beta'}\left(\mathbf{\lambda}_{i,\mathbf{q}}^{\dagger}\right)_{\alpha'\beta'}y\cdot\left[\sum_{\alpha''}\left(-\mathbf{\lambda}_{i,\mathbf{q}}\right)_{\alpha\alpha''}\langle\left(e^{y\hat{\mathbf{c}}_{i}-y\hat{\mathbf{c}}_{i}}\right)_{\alpha''\alpha';\beta'\beta}\rangle_{t}+\sum_{\beta''}\langle\left(e^{y\hat{\mathbf{c}}_{i}-y\hat{\mathbf{c}}_{i}}\right)_{\alpha\alpha';\beta'\beta''}\rangle_{t}\left(\mathbf{\lambda}_{i,\mathbf{q}}\right)_{\beta''\beta}\right].\end{aligned}\ ] ] this last term describes bath - induced interactions between polarons . by using a relation similar to eq .
( [ eq : thermal_2 ] ) and applying it only to the @xmath372 term , we obtain @xmath373 and @xmath374 this can thus be calculated by the same method as in eqs .
( [ eq : thermal_2_1 ] , [ eq : thermal_2_2 ] ) .
in addition , we would like to point out that the thermal average calculation in eq .
( [ eq : thermal_1 ] ) is exact only when the matrices @xmath308 commute with each other : @xmath375=0 $ ] , even though the other derivations are exact for an arbitrary choice of @xmath308 . in order to maintain good control over our two - band polaron transformation method , we can also limit the variational parameters @xmath308 to be a set of _ commuting matrices_. though the ground state energy and the contribution from the incoherent part might thus increase , this transformation has the great advantage of an exactly solvable coherent part . another benefit of using commuting matrices
is that the transformed creation / annihilation operators take a simplified form @xmath376 .
after the polaron transformation , the hamiltonian can be separated into a coherent part eq .
( [ eq : coherent ] ) and an incoherent part eq .
( [ eq : incoherent ] ) .
this hamiltonian describes an quantum system ( coherent part ) coupled to a bosonic bath via interactions ( incoherent part ) .
we apply the lindblad master equation to include the incoherent part . in the interaction picture , due to @xmath377 for a deep lattice
, the polaron operators can be approximated as @xmath378 where @xmath151 is the renormalized polaron on - site energy .
the interaction ( incoherent part ) in eq .
( [ eq : incoherent ] ) is given by : @xmath379\nonumber \\ + & \sum_{\mathbf{q}}\hbar\omega_{\mathbf{q}}\left[\left(\lambda_{\mathbf{q}}^{\alpha}\right)\left(\mathbf{m}_{\mathbf{q}}^{\dagger}\right)_{\alpha\beta}\left(\hat{t}_{i , i}^{\alpha\beta}\left(t\right)\right)+\left(\lambda_{\mathbf{q}}^{\beta}\right)^{*}\left(\mathbf{m}_{\mathbf{q}}\right)_{\beta\alpha}\left(\hat{t}_{i , i}^{\beta\alpha}\left(t\right)\right)^{*}\right],\end{aligned}\ ] ] where @xmath380 and @xmath381 are also in the interaction picture .
the first term in @xmath382 describes intra - band dynamics , while the second term describes inter - band dynamics . in both processes ,
the polaron will emit ( absorb ) phonons to ( from ) the bath . using the general form of the lindblad equation in eq .
( [ eq : lindblad ] ) , we can define the decoherence factor @xmath250 as : @xmath383 by using the rotating wave approximation ( rwa ) , we decouple intra- and inter - band dynamics in the lindblad equation : @xmath230= & \sum_{\alpha,\alpha'}\sum_{ij}\sum_{i'j'}\gamma_{ij;i'j'}^{\alpha\alpha;\alpha'\alpha'}\left(c_{i'j'}^{\alpha'\alpha'}\rho c_{ij}^{\alpha\alpha\dagger}-\frac{1}{2}\left\ { c_{ij}^{\alpha\alpha\dagger}c_{i'j'}^{\alpha'\alpha'},\;\rho\right\ } \right)\nonumber \\ + & \sum_{i}\sum_{i'}\gamma_{ii;i'i'}^{01;01}\left(c_{i'i'}^{01}\rho c_{ii}^{01\dagger}-\frac{1}{2}\left\ { c_{ii}^{01\dagger}c_{i'i'}^{01},\;\rho\right\ } \right)\nonumber \\ + & \sum_{i}\sum_{i'}\gamma_{ii;i'i'}^{10;10}\left(c_{i'i'}^{10}\rho c_{ii}^{10\dagger}-\frac{1}{2}\left\ { c_{ii}^{10\dagger}c_{i'i'}^{10},\;\rho\right\ } \right).\label{eq : lindblad_rwa}\end{aligned}\ ] ] the first part @xmath384 describes intra - band decoherence * * effects .
the second part @xmath385 describes polaron relaxation process from upper to lower band , while the third part @xmath386 describes an excitation from lower to upper band .
we use the short notation @xmath236 and @xmath237 for the second term .
@xmath389\nonumber \\ & \qquad\qquad\qquad\cdot\left[\hat{b}_{\mathbf{\mathbf{q}'}}^{\dagger}\left(0\right)\left(\hat{k}_{j , j}^{01}\left(0\right)\right)+\left(\hat{k}_{j , j}^{01}\left(0\right)\right)\hat{b}_{-\mathbf{\mathbf{q}'}}\left(0\right)+\left(\lambda_{j,\mathbf{\mathbf{q}'}}^{0}-\lambda_{j,\mathbf{q}'}^{1}\right)\left(\hat{k}_{j , j}^{01}\left(0\right)-\langle\hat{k}_{j , j}^{01}\left(0\right)\rangle_{t}\right)\right]\rangle_{t}.\label{eq : correlation_g}\end{aligned}\ ] ] from the definition of @xmath390 , we have @xmath391;\nonumber \\ \left(\mu_{ij,\mathbf{q}}^{\alpha\beta}\right)\equiv & -\left(\lambda_{i,\mathbf{q}}^{\alpha\alpha}\right)+\left(\lambda_{j,\mathbf{q}}^{\beta\beta}\right),\label{eq : kab}\end{aligned}\ ] ] and the thermal averaged results from bloch identity : @xmath392,\ ] ] as well as : @xmath393\exp\left[-\sum_{\mathbf{q}}\left(n_{\mathbf{q}}+\frac{1}{2}\right)\left(\mu_{i'j',\mathbf{q}}^{\alpha'\beta'}\right)^{*}\left(\mu_{i'j',\mathbf{q}}^{\alpha'\beta'}\right)\right]\nonumber \\ & \quad\cdot\exp\left[-\sum_{\mathbf{q}}\left(n_{\mathbf{q}}\right)\left(\mu_{ij,\mathbf{q}}^{\alpha\beta}\right)^{*}\left(\mu_{i'j',\mathbf{q}}^{\alpha'\beta'}\right)e^{i\omega_{\mathbf{q}}t}-\sum_{\mathbf{q}}\left(n_{\mathbf{q}}+1\right)\left(\mu_{ij,\mathbf{q}}^{\alpha\beta}\right)\left(\mu_{i'j',\mathbf{q}}^{\alpha'\beta'}\right)^{*}e^{-i\omega_{\mathbf{q}}t}\right].\end{aligned}\ ] ] the elements such as @xmath394 can be obtained by using the glauber formula @xmath395}$ ] ( with the condition @xmath396=[\hat{b},[\hat{a},\hat{b}]]=0 $ ] ) and differentiating @xmath397 in eq .
( [ eq : kab ] ) as : @xmath398\hat{k}_{i , j}^{\alpha\beta}\left(t\right);\nonumber \\
\frac{\partial}{\partial\left(\mu_{ij,\mathbf{q}}^{\alpha\beta}\right)}\hat{k}_{i , j}^{\alpha\beta}\left(t\right)= & \hat{k}_{i , j}^{\alpha\beta}\left(t\right)\left[-\hat{b}_{\mathbf{q}}\left(t\right)-\frac{1}{2}\left(\mu_{ij,\mathbf{q}}^{\alpha\beta}\right)^{*}\right].\label{eq : b_k}\end{aligned}\ ] ] as an example , averages of @xmath394 in eq .
( [ eq : correlation_g ] ) can then be obtained by using eq .
( [ eq : b_k ] ) and switching the order of differentiation and thermal averaging : @xmath399.\label{eq:<b_k>}\end{aligned}\ ] ] finally the correlation function eq .
( [ eq : correlation_g ] ) for inter - band relaxation is : @xmath400\left[\left(n_{\mathbf{q}}+1\right)e^{-i\omega_{\mathbf{q}}\tau}+\left(n_{\mathbf{q}}\right)e^{\text{+}i\omega_{\mathbf{q}}\tau}\right]\langle\hat{k}_{i , i}^{10}\left(\tau\right)\hat{k}_{j , j}^{01}\left(0\right)\rangle_{t}\nonumber \\ & + \sum_{\mathbf{q}\mathbf{\mathbf{q}'}}\hbar^{2}\omega_{\mathbf{q}}\omega_{\mathbf{\mathbf{q}'}}\left(m_{\mathbf{q}}^{01}\right)\left(m_{\mathbf{\mathbf{q}'}}^{01}\right)\left(\lambda_{\mathbf{q}}^{0}-\lambda_{\mathbf{q}}^{1}\right)\left(\lambda_{\mathbf{\mathbf{q}}'}^{0}-\lambda_{\mathbf{q}'}^{1}\right)\sin\left[q_{x}(i - j)d\right]\sin\left[q'_{x}(i - j)d\right]\nonumber \\ & \quad\cdot\left[\left(n_{\mathbf{q}}+1\right)e^{-i\omega_{\mathbf{q}}\tau}-\left(n_{\mathbf{q}}\right)e^{\text{+}i\omega_{\mathbf{q}}\tau}\right]\cdot\left[\left(n_{\mathbf{\mathbf{q}}'}+1\right)e^{-i\omega_{\mathbf{q}'}\tau}-\left(n_{\mathbf{q}'}\right)e^{\text{+}i\omega_{\mathbf{q}'}\tau}\right]\langle\hat{k}_{i , i}^{10}\left(\tau\right)\hat{k}_{j , j}^{01}\left(0\right)\rangle_{t}.\label{eq : correlation_g-1}\end{aligned}\ ] ] with @xmath401e^{+i\omega_{\mathbf{q}}\tau}+\left(n_{\mathbf{q}}+1\right)\cos\left[q_{x}(i - j)d\right]e^{-i\omega_{\mathbf{q}}\tau}\right]\right\ } .\label{eq : kikj}\end{aligned}\ ] ] the first term in correlation function ( [ eq : correlation_g-1 ] ) is a single - phonon process , which is similar to fermi s golden rule except a renormalization factor @xmath253 .
the second term describes higher order process involving two phonons , which are absent in fermi s golden rule .
the corresponding relaxation rate @xmath238 is calculated by eq .
( [ eq : decay_rate ] ) . at zero temperature
, this corresponds to spontaneous relaxation process : @xmath402e^{\text{-}i\omega_{\mathbf{q}}\tau}\cdot\langle\hat{k}_{i , i}^{10}\left(\tau\right)\hat{k}_{j , j}^{01}\left(0\right)\rangle_{t}\nonumber \\ & + 2\cdot{\bf \text{re}}\int_{0}^{\infty}d\tau e^{i\varepsilon_{{\rm p}}^{\delta}\tau/\hbar}\cdot\sum_{\mathbf{q}\mathbf{\mathbf{q}'}}\hbar^{2}\omega_{\mathbf{q}}\omega_{\mathbf{\mathbf{q}'}}\left(m_{\mathbf{q}}^{01}\right)\left(m_{\mathbf{\mathbf{q}'}}^{01}\right)\left(\lambda_{\mathbf{q}}^{0}-\lambda_{\mathbf{q}}^{1}\right)\left(\lambda_{\mathbf{q}'}^{0}-\lambda_{\mathbf{q}'}^{1}\right)\nonumber \\ & \quad\cdot\sin\left[q_{x}(i - j)d\right]\sin\left[q'_{x}(i - j)d\right]e^{-i\omega_{\mathbf{q}}\tau}e^{-i\omega_{\mathbf{q}'}\tau}\cdot\langle\hat{k}_{i , i}^{10}\left(\tau\right)\hat{k}_{j , j}^{01}\left(0\right)\rangle_{t}.\label{eq : decay_rate_1}\end{aligned}\ ] ] the second part , which involves higher order correlations with two phonons being emitted , is neglected in the total decay rate in eq .
( [ eq : gamma_polaron ] ) .
199ifxundefined [ 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 10.1007/bf02650179 [ * * , ( ) ] link:\doibase 10.1126/science.273.5278.1073 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.81.3108 [ * * , ( ) ] link:\doibase 10.1038/nphys2275 [ * * , ( ) ] link:\doibase 10.1103/revmodphys.80.885 [ * * , ( ) ] link:\doibase 10.1038/nphys2259 [ * * , ( ) ] link:\doibase 10.1038/nphys1102 [ * * , ( ) ] link:\doibase 10.1038/nature08482 [ * * , ( ) ] link:\doibase 10.1038/nature09378 [ * * , ( ) ] link:\doibase 10.1126/science.1201351 [ * * , ( ) ] link:\doibase 10.1038/nature08609 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.109.205303 [ * * , ( ) ] link:\doibase 10.1103/physreva.84.051401 [ * * , ( ) ] link:\doibase 10.1088/1367 - 2630/14/10/105013 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.109.223001 [ * * , ( ) ] link:\doibase 10.1088/1367 - 2630/14/3/033019
[ * * , ( ) ] link:\doibase 10.1103/physrevlett.93.120404 [ * * , ( ) ] link:\doibase 10.1088/0953 - 8984/5/44/017 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.105.266605 [ * * , ( ) ] link:\doibase 10.1016/0003 - 4916(59)90003-x [ * * , ( ) ] http://www.springer.com/gp/book/9780306463389[__ ] ( , ) link:\doibase 10.1103/physrevb.69.024301 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.109.236402 [ * * , ( ) ] link:\doibase 10.1103/physrevb.88.060301 [ * * , ( ) ] http://link.springer.com/article/10.1007/bf01578247 [ * * , ( ) ] link:\doibase 10.1103/revmodphys.59.1 [ * * , ( ) ] link:\doibase 10.1103/physrevb.87.144201 [ * * , ( ) ] link:\doibase 10.1103/physrevb.90.075110 [ * * , ( ) ] link:\doibase 10.1103/physrevb.37.9423 [ * * , ( ) ] link:\doibase 10.1103/physrevb.44.421 [ * * , ( ) ] link:\doibase 10.1109/stsm.1994.836136 [ * * , ( ) ] link:\doibase 10.1038/35087518 [ * * , ( ) ] link:\doibase 10.1088/0034 - 4885/72/6/066501 [ * * , ( ) ] link:\doibase 10.1007/978 - 3 - 642 - 01896 - 1 [ _ _ ] , , vol .
( ) link:\doibase 10.1007/s10948 - 012 - 1523 - 3 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.109.250501 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.102.223201 [ * * , ( ) ] link:\doibase 10.1038/nature08865 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.105.133202 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.107.243201 [ * * , ( ) ] link:\doibase 10.1038/nphys2373 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.110.160402 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.108.190502 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.96.180402 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.96.180403 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.102.030408 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.106.115305 [ * * , ( ) ] link:\doibase 10.1088/1367 - 2630/12/7/073044 [ * * , ( ) ] link:\doibase 10.1103/physreva.86.043614 [ * * , ( ) ] link:\doibase 10.1103/physrevb.77.125101 [ * * , ( ) ] link:\doibase 10.1103/physreva.80.033607 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.102.230402 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.103.170402 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.109.140402 [ * * , ( ) ] http://iopscience.iop.org/article/10.1088/0034-4885/77/3/034401/meta;jsessionid=2bf6bdecb08739a51259ecaa373b360e.c1 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.115.135302 [ * * , ( ) ] link:\doibase 10.1103/physreva.92.053605 [ * * , ( ) ] link:\doibase 10.1103/physreva.71.033605 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.96.210401 [ * * , ( ) ] link:\doibase 10.1103/physreva.76.011605 [ * * , ( ) ] link:\doibase 10.1088/1367 - 2630/10/3/033015 [ * * , ( ) ] link:\doibase 10.1103/physrevb.80.184504 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.105.045303 [ * * , ( ) ] link:\doibase 10.1103/physreva.82.063614 [ * * , ( ) ] link:\doibase 10.1038/nphys2561 [ * * , ( ) ] link:\doibase 10.1103/physreva.88.013613 [ * * , ( ) ] link:\doibase 10.1103/physreva.88.053601 [ * * , ( ) ] link:\doibase 10.1103/physreva.88.053632 [ * * , ( ) ] link:\doibase 10.1103/physreva.89.033615 [ * * , ( ) ] http://arxiv.org/abs/1410.8833 [ ( ) ] link:\doibase 10.1103/physreva.90.013618 [ * * , ( ) ] link:\doibase 10.1103/physreva.89.053617 [ * * , ( ) ] link:\doibase 10.1103/physreva.90.063610 [ * * , ( ) ] link:\doibase 10.1038/srep12124 [ * * , ( ) ] http://arxiv.org/abs/1410.5691 [ ( ) ] link:\doibase 10.1103/physrevlett.115.160401 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.115.125302 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.114.203001 [ * * , ( ) ] http://arxiv.org/abs/1508.02646 [ * * , ( ) ] link:\doibase 10.1103/physreva.92.033612 [ * * , ( ) ] http://arxiv.org/abs/1510.04934 [ ( ) ] link:\doibase 10.1103/physrevlett.100.050402 [ * * , ( ) ] link:\doibase 10.1088/1367 - 2630/11/5/055045 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.110.223002 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.109.235309 [ * * , ( ) ] link:\doibase 10.1126/science.1220314 [ *
* , ( ) ] link:\doibase 10.1103/physrevlett.111.080501 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.107.135303 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.109.235301 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.111.070401 [ * * , ( ) ] link:\doibase 10.1103/physreva.90.013602 [ * * , ( ) ] link:\doibase 10.1016/j.physrep.2006.03.001 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.81.938 [ * * , ( ) ] link:\doibase 10.1103/physreva.74.013616 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.101.170401 [ * * , ( ) ] link:\doibase 10.1103/physreva.69.022306 [ * * , ( ) ] link:\doibase 10.1063/1.449469 [ * * , ( ) ] link:\doibase 10.1103/physrevb.61.4592 [ * * , ( ) ] link:\doibase 10.1103/physrevb.69.144302 [ * * , ( ) ] link:\doibase 10.1103/physrevb.73.085120 [ * * , ( ) ] link:\doibase 10.1063/1.439425 [ * * , ( ) ] link:\doibase 10.1063/1.449372 [ * * , ( ) ] link:\doibase 10.1063/1.449373 [ * * , ( ) ] link:\doibase 10.1021/jp1093492 [ * * , ( ) ] link:\doibase 10.1103/physreva.88.032119 [ * * , ( ) ] link:\doibase 10.1103/physreva.89.053610 [ * * , ( ) ] link:\doibase 10.1088/1367 - 2630/13/10/103029 [ * * , ( ) ] link:\doibase 10.1103/physreva.84.052727 [ * * , ( ) ] link:\doibase 10.1103/physreva.88.063630 [ * * , ( ) ] link:\doibase 10.1103/physreva.90.033627 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.113.260403 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.97.220403 [ * * , ( ) ] link:\doibase 10.1088/1367 - 2630/9/2/044 [ * * , ( ) ] link:\doibase 10.1103/physrevb.46.301 [ * * , ( ) ] link:\doibase 10.1103/physreva.73.043608 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.101.050402 [ * * , ( ) ] link:\doibase 10.1209/0295 - 5075/82/30004 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.102.025301 [ * * , ( ) ] http://store.elsevier.com/product.jsp?lid=0{&}iid=5{&}sid=0{&}isbn=9780123850317[__ ] ( ) p. | in this work we study the formation and dynamics of polarons in a system with a few impurities in a lattice immersed in a bose - einstein condensate ( bec ) .
this system has been experimentally realized using ultracold atoms and optical lattices . here
, we consider a two - band model for the impurity atoms , along with a bogoliubov approximation for the bec , with phonons coupled to impurities via both intraband and interband transitions .
we decouple this frhlich - type term by an extended two - band lang - firsov polaron transformation using a variational method .
the new effective hamiltonian with two ( polaron ) bands differs from the original hamiltonian by modified coherent transport , polaron energy shifts , and induced long - range interaction .
a lindblad master - equation approach is used to take into account residual incoherent coupling between polaron and bath .
this polaronic treatment yields a renormalized inter - band relaxation rate compared to fermi s golden rule . for a strongly coupled two - band frhlich hamiltonian
, the polaron is tightly dressed in each band and can not tunnel between them , leading to an _ inter - band self - trapping _ effect . |
two young compact star clusters have been observed within a few tens of parsecs of the galactic center : the arches cluster ( object 17 , nagata et al .
1995 ) and the quintuplet cluster ( afgl2004 , nagata et al . 1990
; okuda et al . 1990 ) , for which excellent observational data are available . in terms of structural parameters size , mass , density profile and ages , these systems may represent the galactic counterparts to ngc2070 ( r136 ) , the central star cluster in the 30doradus region in the large magellanic could ( massey & hunter 1998 ) . the arches and quintuplet clusters lie behind thick layers of absorbing material , hinting that many more such systems may exist .
recently , dutra & bica ( 2000 ) have reported from the 2mass survey a total of 58 star cluster candidates within @xmath7pc ( in projection ) of the galactic center .
a number of important questions make these clusters worthy of detailed study , among them : ( 1 ) how are such clusters related to globular clusters ? ( 2 ) how do they contribute to the total star formation rate in the galaxy ? ( 3 ) are their mass functions in reality intrinsically flat , as is suggested by observations ? ( 4 ) how far are these clusters from the galactic center ? ( 5 ) how many are hidden , still waiting to be discovered ? in this letter
we summarize the results of a series of simulations of young compact star clusters in the vicinity of the galactic center , and address specifically the last item on this list .
we found that the lifetimes of our model clusters depend very sensitively on their distances from the galactic center .
this is mainly due to the larger size of tidally limited clusters lying farther from the galactic center , resulting in longer relaxation times and therefore longer lifetimes .
the majority of our models are visible only for the first part of their lifetimes , and are likely to be indistinguishable from the stellar background at later times .
we find that the true number of young compact star clusters within 200 pc of the galactic center is at least 10 but could easily exceed 50 . a more comprehensive paper , exploring all of these questions in more detail and presenting the results of an extensive parameter - space study , is in preparation ( portegies zwart et al .
, 2000b ) .
we study the evolution of our model clusters by following the equations of motion of all stars by direct integration , at the same time taking into account the internal evolution of both stars and binary systems .
the `` starlab '' software environment within which this work was performed is described in detail by portegies zwart et al .
( 2000a ; see also http://www.sns.ias.edu/@xmath8starlab ) . the special - purpose grape-4 system ( makino et al .
1997 ) was used to accelerate the computation of gravitational forces between stars
. observed parameters for the arches and quintuplet clusters are listed in table[tab : observed ] .
these clusters have masses of @xmath9 and are extremely compact , with half - mass radii @xmath10pc ( figer , mclean & morris 1999 ) . the projected distance from the arches to the galactic center
is about 34pc ; the quintuplet cluster lies somewhat farther out , at @xmath850pc .
[ cols= " < , < , > , > , > , > , > , > " , ] [ tab : n12kinit ] the mass of the galaxy within the clusters orbit at a distance ( @xmath11pc ) is taken to be ( mezger et al .
1999 ) @xmath12 .
\label{eq : mgal}\ ] ] this mass distribution determines the strength and geometry of the local galactic tidal field ( for details see portegies zwart et al .
the evolution of the cluster is followed using the starlab kira integrator and the binary evolution program ( portegies zwart et al .
for each selected distance to the galactic center we carried out three calculations .
one series of runs was carried out with identical initial realizations of the system ( stellar masses , positions and velocities , with a total mass of 7432 ) .
the same initial model can be used at several galactocentric distances because the shape of the zero - velocity surface does not depend sensitively on distance to the galactic center .
table[tab : n12kinit ] gives results of these calculations .
for each galactocentric distance we also performed two additional calculations ( for a total of 9 runs ) with different initial realizations of the systems .
these calculations were performed to study the uncertainties in cluster lifetimes , and to ascertain the reproducilility of our results .
the calculations with different initial realizations produced roughly 10% spreads in core collapse times ( @xmath13 ) and cluster lifetimes ( @xmath14 ) . for reasons of economy ,
stars were removed from all calculations when they exceeded a distance of 3 from the cluster center .
figure[fig : spz_w147n12k_tm ] shows the evolution of cluster mass and number of stars for the models listed in table [ tab : n12kinit ] which began with identical initial conditions but different galactocentric distances .
perhaps not surprisingly , clusters located farther from the galactic center live considerably longer than those closer in .
the longer lifetime of the more distant cluster is mainly a consequence of its longer relaxation time .
scaling the relaxation time at the tidal radius of model r34w4 to a distance of 150pc results in a lifetime of @xmath15myr ( @xmath16 ) , which is slightly smaller than the @xmath17myr lifetime of the models r150w4 . mass loss from stellar evolution , which is more prominent in model r150w4
because the cluster survives longer , seems to be a minor factor in driving the evolution .
the number of stars in each model ( dashed lines , see figure1 ) decreases more rapidly than the total mass ( solid lines )
. thus the mean mass of the stars within the cluster increases gradually with time .
although clusters like the arches and quintuplet systems are very compact , it may still be hard to see them near the galactic center because the projected foreground and background stellar density is so high .
integration of the local stellar density ( obtained by differentiating eq.[eq : mgal ] ) along the line of sight then gives the surface density .
portegies zwart et al.(2000b ) perform this calculation numerically and arrive at a surface density at 34 pc of about 3000pc@xmath18 .
while the contrast in the surface density of the cluster relative to that of the background is perhaps an oversimplified measure of the cluster s observability , the results of this simple comparison are quite instructive , and a more comprehensive consideration of luminosity density leads to essentially similar overall conclusions .
figure[fig : spz_w17r34_rho_phm ] shows the evolution of the surface density within the projected half mass radius for models r34w4
( dots ) , r90w4 ( dashes ) and r150w4 ( solid ) .
the two points with error bars indicate the projected half - mass densities for the arches ( left ) and quintuplet ( right ) clusters .
the horizontal dotted line gives the background surface density at a projected distance of 34pc from the galactic center .
the projected densities of the two observed clusters are between 3 ( for the quintuplet ) and 50 ( for arches ) times higher than the background ; clusters with densities below a few times the projected background stellar density may well remain unnoticed .
a simplified model for the evolution of these star clusters may be constructed as follows .
the initial relaxation time at the tidal radius may be calculated using eq.[eq : mgal ] and spitzer s ( 1987 ) expression as @xmath19 the constant is obtained by substitution of the appropriate units .
the mass of the cluster decreases almost linearly in time ( see figure[fig : spz_w147n12k_tm ] ) as @xmath20 here @xmath21 and @xmath22 is the age at which the cluster dissolves in the tidal field of the galaxy .
the projected surface density within the half - mass radius of the cluster is @xmath23 here , @xmath24 is the mass contained within the projected half - mass radius , and @xmath25 depends on the density profile , but is always smaller than unity . for a king model with @xmath26 , we find @xmath27 .
substitution of eqs.[eq : l1 ] and [ eq : mgal ] into eq.[eq : sigma ] gives @xmath28,\ ] ] the projected surface density @xmath29 decreases with time because the cluster mass decreases and the half - mass radius of the cluster increases .
substitution of eq.[eq : mass ] into eq.[eq : sigma ] gives : @xmath30 the half - mass radius increases by about a factor of two during the first half mass relaxation time and remains roughly constant thereafter . in our simple model
we implement this by allowing @xmath31 to increase by a factor two in the first half mass relaxation time and to remain constant at later times .
the fact that the half - mass radius remains roughly constant at late times , and does not decrease as @xmath32 as would be expected for a tidally limited system , is a consequence of our use of the total mass , rather than the mass within the jacobi surface ( see figure[fig : spz_w147n12k_tm ] ) , in determining both and @xmath29 .
the resulting surface density evolution agrees very well with our calculations .
figure[fig : rgct_dens ] shows the evolution of the projected density as a function of distance from the galactic center for our simple model . clusters close to the galactic center are compact enough to be easily visible ( dark shades ) for a large fraction of their lifetimes , but they dissolve quickly .
clusters farther from the galactic center live much longer , but their surface densities are generally low , lying well below the background for most of their lifetimes .
the two well - observed clusters both lie in the lower left corner ( high density region ) of figure[fig : rgct_dens ] .
the arches and quintuplet clusters inhabit only a small portion of the available parameter space . however , the region they populate is most favorable for finding clusters , because the projected surface densities of such clusters are high .
we estimate the number of clusters like the arches and the quintuplet based on the fact that both clusters lie at the lower left corner ( youngest and most compact ) in figure[fig : rgct_dens ] .
we assume that all clusters with ages less than @xmath33myr and within 50pc of the galactic center have been found by observers , and all outside this region are as yet undetected .
we further assume that clusters populate the triangular area in figure[fig : rgct_dens ] more or less uniformly probably not a bad assumption , since the galactic mass is roughly proportional to radius ( see equation[eq : mgal ] ) . with these assumptions ,
the number of yet - to - be - found clusters is ( @xmath3460myr@xmath35200pc)/(5myr@xmath3550pc)=24 times more than the number of known clusters .
thus , we expect that the total number of clusters in this region would be around 50 . a conservative lower limit to the number of hidden clusters
may be obtained using the same technique , but adopting the observed projected density of the quintuplet system as the limiting contrast at which such clusters can be discovered .
the projected density of the quintuplet exceeds the background density by about a factor three .
figure[fig : rgct_dens ] then suggests that about 20% of the available surface area harbors visible cluster . a lower limit to the total number of clusters within 200pc of the galactic center
would then be around 10 .
these estimates are in excellent agreement with the results of dutra & bica ( 2000 ) mentioned earlier , although the above reasoning suggests that even these 58 candidate clusters may represent only a small fraction of the number actually present .
finally , we note that a population of 100 clusters with masses of @xmath36 each and a maximum lifetime of @xmath37myr implies a star formation rate of 0.01 per year , enough to build up the entire bulge of @xmath37 within the 10gyr age of the galaxy .
spz is grateful to the institute for advanced study , drexel university and tokyo university for their hospitality and the use of their grape-4 hardware .
this work was supported by nasa through hubble fellowship grant hf-01112.01 - 98a awarded by the space telescope science institute , which is operated by the association of universities for research in astronomy , by the research for the future program of japan society for the promotion of science ( jsps - rftp97p01102 ) and by nasa atp grants nag5 - 6964 and nag5 - 9264 .
part of this letter was written while spz , sm and ph were visiting the american museum of natural history .
they acknowledge the hospitality of their astrophysics department and visualization group . | we study the evolution and observability of young compact star clusters within @xmath0pc of the galactic center .
calculations are performed using direct integration on the grape-4 , including the effects of both stellar and binary evolution and the external influence of the galaxy .
the results of these detailed calculations are used to calibrate a simplified model applicable over a wider range of cluster initial conditions .
we find that clusters within 200pc of the galactic center dissolve within @xmath1 myr .
however , their projected densities drop below the background density in the direction of the galactic center within @xmath2myr , effectively making these clusters undetectable after that time .
clusters farther from the galactic center but at the same projected distance are more strongly affected by this selection effect , and may go undetected for their entire lifetimes .
based on these findings , we conclude that the region within 200 pc of the galactic center could easily harbor some 50 clusters with properties similar to those of the arches or the quintuplet systems .
# 1 # 1*[#1 steve ] * # 1*[#1 steve ] * # 1*[#1 simon ] * # 1*[#1 simon ] * # 1*[#1 piet ] * # 1*[#1 piet ] * # 1*[#1 jun ] * # 1*[#1 jun ] * @xmath3 massachusetts institute of technology , cambridge , ma 02139 , usa , _ hubble fellow _
+ @xmath4 department of astronomy , university of tokyo , 7 - 3 - 1 hongo , bunkyo - ku , tokyo 113 - 0033 , japan + @xmath5 dept . of physics , drexel university , philadelphia , pa 19104 , usa + @xmath6 institute for advanced study , princeton , nj 08540 , usa + subject headings : gravitation methods : n - body simulations stellar dynamics stars : evolution globular clusters : individual ( arches , quintuplet ) |
the theoretical description of relaxation and decoherence processes in open quantum systems often leads to a non - markovian dynamics which is determined by pronounced memory effects @xcite .
strong system - environment couplings @xcite , correlations and entanglement in the initial state @xcite , interactions with environments at low temperatures and with spin baths @xcite , finite reservoirs @xcite , and transport processes in nano - structures @xcite can lead to long memory times and to a failure of the markovian approximation .
a systematic approach to non - markovian dynamics is provided by the projection operator techniques @xcite which are extensively used in nonequilibrium thermodynamics and statistical mechanics @xcite .
these techniques are based on the introduction of a certain projection superoperator @xmath0 which acts on the states of the total system .
the superoperator @xmath0 is the mathematical expression for the idea of the elimination of degrees of freedom from the complete description of the states of the total system : if @xmath1 is the full density matrix of the composite system , the projection @xmath2 represents a certain approximation of @xmath1 which leads to a simplified effective description of the dynamics through a reduced set of variables .
the projection @xmath2 is therefore referred to as the relevant part of the density matrix . with the help of the projection operator techniques
one derives closed dynamic equations for the relevant part of the density matrix .
the equation of motion for @xmath2 can either be the nakajima - zwanzig equation @xcite , an integrodifferential equation with a retarded memory kernel , or else the time - convolutionless master equation , which is a time - local differential equation of first order involving a time - dependent generator @xcite . in most cases these equations
are used as starting point for the derivation of effective master equations through a systematic perturbation expansion with respect to the strength of the system - environment coupling . in the standard approach to the dynamics of open systems
one chooses a projection superoperator which is defined by the expression @xmath3 , where @xmath4 represents the reduced density matrix of the open system , @xmath5 denoting the trace over the environmental hilbert space , and @xmath6 is some fixed environmental state .
a superoperator of this form projects the total state @xmath1 onto a tensor product state , i. e. , onto a state without any statistical correlations between system and environment .
we emphasize that this ansatz does _ not _ imply ( as is sometimes claimed ) that one completely ignores all system - environment correlations .
it only presupposes that all correlations which are present in the initial state or are generated during the time - evolution can be treated as perturbations within the framework of the projection operator techniques .
the projection onto a tensor product state is widely used in studies of open quantum systems .
it is often applicable in the case of weak system - environment couplings .
usually , the perturbation expansion is restricted to the second order ( known as born approximation ) , from which one derives , with the help of certain further assumptions , a markovian quantum master equations in lindblad form @xcite . in this paper
the quantum dynamics of an open system is said to be non - markovian if the time - evolution of its reduced density matrix can not be described ( to the desired degree of accuracy ) by means of a closed master equation with a time - independent generator in lindblad form . a possible approach to large deviations from markovian behavior consists in carrying out the perturbation expansion to higher orders in the system - environment coupling (
several examples are discussed in ref .
however , this approach is often limited by the increasing complexity of the resulting equations of motion . moreover
, the perturbation expansion may not converge uniformly in time , such that higher orders only improve the quality of the approximation of the short - time behavior , but completely fail in the long - time limit @xcite .
there is however a further promising strategy : to treat highly non - markovian processes in a more efficient way one can replace the tensor product state used in the standard born approximation by a certain correlated system - environment state .
this approach has been proposed by esposito and gaspard @xcite and by budini @xcite to derive effective master equations within second order perturbation theory that describe strong non - markovian effects .
it has been demonstrated in ref .
@xcite that this idea can be formulated in terms of a positive projection superoperator @xmath0 which projects any state onto a correlated system - environment state , i. e. , onto a state that contains statistical correlations between certain system and environment states .
this formulation allows an immediate application of the projection operator techniques to correlated system - environment states , and to carry out the perturbation expansion to higher orders in a systematic way .
an example is discussed in ref .
@xcite , where the master equations of second and of fourth order corresponding to a correlated projection superoperator have been constructed .
the application of a correlated projection superoperator implies that the relevant part @xmath2 can no longer be expressed in terms of the reduced density matrix alone .
hence , employing a correlated projection superoperator one enlarges the set of relevant variables to capture those statistical correlations that are responsible for strong non - markovian effects . in the present paper
we discuss this idea of using correlated projection superoperators in the analysis of non - markovian dynamics . on the basis of certain general physical conditions we derive in sec .
[ sec - cps ] a representation theorem for a class of correlated projection superoperators that are appropriate for the application of the projection operator techniques .
a central problem of the theory of non - markovian processes is the formulation of appropriate master equations that preserve the normalization and the positivity of the density matrix ( see , e. g. , the discussion in refs .
@xcite ) . in sec .
[ dynamics ] we develop the general structure of such master equations which results form the application of a correlated projection superoperator . given a superoperator that projects onto a separable quantum state one can construct an embedding of the underlying dynamics into a lindblad dynamics on a suitably extended state space . employing this embedding
we derive a general class of physically acceptable master equations which represents a generalization of the lindblad theory to the regime of highly non - markovian quantum dynamics .
section [ conclu ] contains some conclusions .
the hilbert spaces of the open system @xmath7 and of its environment @xmath8 are denoted by @xmath9 and @xmath10 , respectively .
the state space of the composite system is given by the tensor product @xmath11 .
states of the composite system are represented by density matrices @xmath1 on @xmath12 satisfying @xmath13 and @xmath14 , where @xmath15 is the trace taken over the total state space .
the reduced density matrix @xmath16 of subsystem @xmath7 is given by the partial trace taken over the hilbert space @xmath10 , i. e. @xmath4 .
correspondingly , the partial trace over @xmath9 will be denoted by @xmath17 .
a superoperator @xmath0 is a linear map @xmath18 which takes any operator @xmath19 on the total state space @xmath12 to an operator @xmath20 on @xmath12 .
we consider here superoperators with the following properties .
the map @xmath0 is a projection superoperator : @xmath21 it is this formal property that allows the application of the projection operator techniques . for an efficient performance of these techniques
the projection @xmath2 should represent a suitable approximation of @xmath1 .
a natural minimal requirement is therefore that for any physical state @xmath1 the projection @xmath2 is again a physical state , i. e. , a positive operator with unit trace .
this means that @xmath0 is a positive and trace preserving map , namely @xmath13 implies @xmath22 , and @xmath23 .
we consider projection superoperators of the following general form : @xmath24 where @xmath25 denotes the unit map acting on operators on @xmath9 , and @xmath26 is a linear map that takes operators on @xmath10 to operators on @xmath10 .
a projection superoperator of this form leaves the system @xmath7 unchanged and acts nontrivially only on the variables of the environment @xmath8 . as a consequence of the positivity of @xmath0 and of condition ( [ itimeslambda ] )
the map @xmath26 must be @xmath27-positive , where @xmath27 is the dimension of @xmath9 ( see , e. g. , ref .
@xcite ) . in the following
we use the stronger condition that @xmath26 is completely positive , because completely positive maps allow for a simple mathematical characterization ( see sec .
[ sec - representation ] ) .
we discuss the implications of these conditions . from eqs .
( [ projection ] ) and ( [ itimeslambda ] ) we get that @xmath26 itself must be a projection , namely @xmath28 .
moreover , since @xmath0 is trace - preserving , the map @xmath26 must also be trace - preserving . hence , we find that @xmath26 represents a completely positive and trace - preserving map ( cpt map , or quantum channel ) which operates on the variables of the environment and has the property of a projection . a further physically reasonable consequence of the positivity of @xmath26 and of eq .
( [ itimeslambda ] ) is that @xmath0 maps separable ( classically correlated ) states to separable states , which means that the projection does not create entanglement between system and environment .
an important goal is , of course , the determination of the reduced density matrix @xmath16 of the open quantum system .
using eq .
( [ itimeslambda ] ) and that @xmath26 is trace - preserving we get the relation @xmath29 this relation connects the density matrix of the reduced system with the projection of a given state @xmath1 of the total system .
it states that , in order to determine @xmath16 , we do not really need the full density matrix @xmath1 , but only its projection @xmath2 .
thus , @xmath2 contains the full information needed to reconstruct the reduced system s state .
we derive a representation theorem for the projection superoperator @xmath0 from the basic conditions formulated in sec .
[ gen - cond ] [ see eq .
( [ projection - genform ] ) below ] . since @xmath26 is supposed to be a cpt map one could use , of course , the kraus - stinespring representation @xcite for completely positive maps .
however , for our purposes another representation is much more appropriate , which will be derived now .
we will use the following fact from linear algebra . consider a linear operator @xmath30 which acts on some hilbert space @xmath31 and has the property @xmath32 .
then there exist linear independent vectors @xmath33 and linear independent vectors @xmath34 such that @xmath35 and @xmath36 for all @xmath37 . conversely , given two linear independent sets @xmath38 and @xmath39 of vectors in @xmath31 with @xmath35 , then eq .
( [ def - l ] ) defines a linear operator with the property @xmath32 .
note that we neither require that the @xmath33 or the @xmath34 are orthogonal , nor that @xmath40 is hermitian .
let us apply this fact to linear maps on the hilbert - schmidt space , i. e. , we take @xmath31 to be the vector space of operators on @xmath10 with the scalar product : @xmath41 then we find that any linear map @xmath26 can be represented in the form @xmath42 with two sets @xmath43 and @xmath44 of linear independent operators on @xmath10 , and that the condition @xmath28 is satisfied if and only if @xmath45 . since @xmath26 preserves the hermiticity of operators , the operators @xmath46 and @xmath47 can be chosen to be hermitian .
hence , we obtain the representation : @xmath48 where @xmath43 and @xmath44 are two sets of linear independent hermitian operators satisfying : @xmath49 the condition that @xmath26 is trace - preserving takes the form : @xmath50 finally , we have to formulate the condition of the complete positivity of the map @xmath26 .
a given map @xmath26 is completely positive if and only if @xmath51 where @xmath52 is a maximally entangled vector in @xmath53 , and @xmath54 is an orthonormal basis for @xmath10 . to evaluate condition ( [ cond - cp ] )
we first note that @xmath55 on using the representation ( [ lambda ] ) one finds @xmath56 where @xmath57 denotes the transpose of the operator @xmath46 with respect to the given basis @xmath54 . inserting eq .
( [ deriv-2 ] ) into eq .
( [ deriv-1 ] ) we obtain @xmath58 we conclude that a necessary and sufficient condition for @xmath26 to be completely positive is given by the inequality @xmath59 employing eqs .
( [ lambda ] ) and ( [ itimeslambda ] ) we obtain the following representation for the projection superoperator @xmath0 , @xmath60 given observables @xmath46 and @xmath47 that satisfy eqs .
( [ bjai ] ) , ( [ trace - preserving ] ) , and ( [ cond - pos ] ) , this equation defines a projection superoperator which fulfills the general conditions formulated in sec .
[ gen - cond ] .
conversely , given a projection which fulfills the conditions of sec .
[ gen - cond ] , there exist observables @xmath46 and @xmath47 satisfying eqs .
( [ bjai ] ) , ( [ trace - preserving ] ) , and ( [ cond - pos ] ) such that eq .
( [ projection - genform ] ) holds .
there are in general many different sets of operators @xmath46 , @xmath47 that represent a given @xmath0 .
if we have a particular set of such operators , then the operators @xmath61 represent the same projection , where @xmath62 and @xmath63 are real , non - singular matrices related by @xmath64 . within the standard approaches one considers a projection superoperator of the form @xmath65 where @xmath6 is any fixed environmental density matrix .
using a projection of this form one assumes that the states of the total system may be approximated by certain tensor products , describing states without statistical dependencies between system and environment .
the projection ( [ p - standard ] ) naturally fits into the general scheme developed above if we take a single @xmath66 and a single @xmath67 .
the conditions ( [ bjai ] ) , ( [ trace - preserving ] ) , and ( [ cond - pos ] ) are then satisfied and eq .
( [ projection - genform ] ) obviously reduces to eq .
( [ p - standard ] ) in the general case , a projection @xmath2 of the form of eq .
( [ projection - genform ] ) does not represent a simple product state .
we therefore call such @xmath0 correlated projection superoperators .
they project onto states that contain statistical correlations between the system @xmath7 and its environment @xmath8 . in the following
we will consider the case that one can find a representation of the projection with positive operators @xmath68 and @xmath69 .
equation ( [ cond - pos ] ) is then trivially satisfied .
without restriction we may assume that the @xmath47 are normalized to unit trace , such that condition ( [ trace - preserving ] ) reduces to the simple form @xmath70 .
under these conditions @xmath0 projects any state @xmath1 onto a state which can be written as a sum of tensor products of positive operators . in the theory of entanglement
( see , e. g. , ref .
@xcite ) such states are called separable or classically correlated @xcite . using a projection superoperator of this form ,
one thus tries to approximate the total system s states by a classically correlated state . the general representation of eq .
( [ projection - genform ] ) includes the case of projection superoperators that project onto inseparable , entangled quantum states .
we will not pursue here this possibility further , and restrict ourselves to positive @xmath46 and @xmath47 in the following .
a straightforward example for a separable projection superoperator is obtained through the following construction @xcite .
we take any orthogonal decomposition of the unit operator @xmath71 on the state space of the environment , i. e. , a collection of ordinary projection operators @xmath72 on @xmath10 which satisfy @xmath73 then we choose : @xmath74 it is easy to verify that with this choice the conditions ( [ bjai ] ) , ( [ trace - preserving ] ) , and ( [ cond - pos ] ) are satisfied .
the explicit form of the projection superoperator is given by @xmath75 given a projection superoperator we define the relevant states as the states in the range of @xmath0 , i. e. , for which the relation @xmath76 holds .
we see that these states are of the form @xmath77 where the @xmath78 may be any positive matrices such that @xmath79 .
hence , the manifold of the relevant states is determined by the operators @xmath47 .
one can transfer this concept from states ( schrdinger picture ) to observables ( heisenberg picture ) . a hermitian operator @xmath80 on the total state space is said to be a relevant observable if the relation @xmath81 holds true for all states @xmath1 .
this means that the expectation value of a relevant observable in any state of the composite system is left unchanged under the application of the projection superoperator @xmath0 . introducing the adjoint map
@xmath82 ( defined with respect to the hilbert - schmidt scalar product for operators on the total state space ) we can reformulate this condition as @xmath83 hence , the relevant observables are those observables which are invariant under the application of the adjoint projection . from the representation ( [ projection - genform ] ) we get : @xmath84 from which we infer that the relevant observables must be of the form @xmath85 where the @xmath86 are arbitrary observables of the subsystem @xmath7 .
the structure of the relevant observables is thus determined by the operators @xmath46 .
the dynamics of the total system is given by @xmath87 where @xmath88 denotes the unitary time - evolution operator . given a projection superoperator of the form of eq .
( [ projection - genform ] ) one introduces the dynamical variables : @xmath89 since the @xmath46 are positive operators , we have @xmath90 , and by use of eq .
( [ consistency ] ) and of the normalization @xmath91 we find that the reduced density matrix is obtained from @xmath92 the normalization condition reads : @xmath93 hence , the state of the reduced system is determined by a certain set of unnormalized density matrices @xmath94 .
we consider initial conditions of the following form , @xmath95 this equation means that the initial state belongs to the manifold of the relevant states . as a consequence
there is no inhomogeniety in the nakajima - zwanzig or the time - convolutionless master equation , although @xmath96 describes a correlated system - environment state . on using eqs .
( [ neumann ] ) , ( [ rho - i ] ) , and ( [ init ] ) we get : @xmath97 to be specific we assume that the index @xmath98 takes on the values @xmath99 .
we define a vector @xmath100 whose components are given by the dynamical variables @xmath78 : @xmath101 equation ( [ rho - i - evol ] ) then defines a dynamical transformation of the form @xmath102 i. e. , we get a one - parameter family of dynamical maps @xmath103 , where @xmath104 is equal to the identity map . for each fixed @xmath105
the map @xmath106 transforms any initial vector @xmath107 with positive components @xmath108 into some vector @xmath109 with positive components @xmath90 .
of course , this transformation also preserves the normalization condition ( [ normierung ] ) .
it is important to emphasize that @xmath106 is _ not _ a quantum dynamical map in the conventional sense of the theory of open systems , simply because it is _ not _ an operation on the space of states of the reduced system , but rather a map on the space of vectors @xmath100 . the transition from @xmath107 to the reduced density matrix @xmath110 is connected with a loss of information on the initial correlations , such that from the mere knowledge of @xmath111 the dynamical behavior can not be reconstructed , in general . in other words ,
the evolution from @xmath111 to @xmath112 for @xmath113 is not a map , i. e. , there is no prescription which assigns to each @xmath111 a unique @xmath112 .
the application of the time - convolutionless projection operator technique leads to a closed dynamic equation for the relevant part of the density matrix : @xmath114 where @xmath115 is a linear generator which depends , in general , explicitly on time . if one uses a projection of the form of eq .
( [ projection - genform ] ) this yields a system of equations of motion for the densities @xmath94 , @xmath116 with an explicitly time - dependent generator @xmath117 for each @xmath98 which depends linearly on the input arguments @xmath118 .
let us suppose that these generators may be approximated by time - independent generators such that we can write @xmath119 the family @xmath106 then represents a semigroup of dynamical transformations .
of course , this semigroup assumption for @xmath106 does not imply that the dynamics of the reduced density matrix yields a semigroup .
a typical situation in which the semigroup assumption is valid is given by a projection superoperator of the form of eq .
( [ proj - new ] ) , where the @xmath72 project onto certain energy bands of the environment , describing a structured reservoir .
the semigroup assumption then presupposes that all intra- and inter - band transitions may be described by means of time - independent rates obtained from fermi s golden rule @xcite .
our aim is to determine the general structure of the generators @xmath120 . to this end
, we demand that eq .
( [ master - tcl ] ) preserves the positivity of all components @xmath78 , i. e. , given positive initial data @xmath108 the corresponding solution satisfies @xmath90 for all times . a convenient way of formulating the dynamical transformation is the following .
we introduce an auxiliary hilbert space @xmath121 and an orthonormal basis @xmath122 for this space .
then we can identify the vector @xmath100 introduced in eq .
( [ vec - rho ] ) with a density matrix on the extended space @xmath123 : @xmath124 this density matrix can be regarded as a block diagonal matrix with blocks @xmath78 along the main diagonal .
moreover , the reduced density matrix @xmath16 is obtained by the partial trace of @xmath100 taken over the auxiliary space .
hence , the auxiliary space represents an additional degree of freedom which expresses the statistical correlations introduced through the projection superoperator @xmath0 .
the dynamical transformation @xmath106 can be viewed as a cpt operation on the extended space that preserves the block diagonal structure . to guarantee the conservation of positivity
we therefore impose the condition that there exists a lindblad generator @xmath125 on the extended space which also preserves the block diagonal structure , i. e. , which has the property @xmath126 if such a lindblad generator exists , the solution of eq .
( [ master - tcl ] ) can be written as @xmath127 this implies the required conservation of positivity of the components @xmath94 for all times @xmath128 . in mathematical terms , eq .
( [ embedding ] ) can be interpreted as an embedding of the non - markovian dynamics into a lindblad dynamics on the extended state space .
we now demonstrate that the simple condition expressed by eq .
( [ l - prop ] ) fixes the structure of the generators @xmath120 to a large extend .
_ theorem . _
a lindblad generator @xmath125 on the extended state space with the property ( [ l - prop ] ) exists if and only if the generators @xmath120 are of the special form @xmath129 + \sum_{j \lambda } \left ( r^{ij}_{\lambda } \rho_j r^{ij\dagger}_{\lambda } - \frac{1}{2 } \left\ { r^{ji\dagger}_{\lambda } r^{ji}_{\lambda } , \rho_i\right\ } \right ) \nonumber\end{aligned}\ ] ] with hermitian operators @xmath130 and arbitrary system operators @xmath131 . _ proof . _
assume that a lindblad generator @xmath125 on the extended state space with the property ( [ l - prop ] ) is given .
as for any lindblad generator we have @xmath132 + \sum_{\lambda } \left ( r_{\lambda } \varrho r^{\dagger}_{\lambda } - \frac{1}{2 } \left\ { r^{\dagger}_{\lambda } r_{\lambda } , \varrho \right\ } \right),\ ] ] were @xmath133 and the @xmath134 are operators on the extended state space .
these operators can always be written as sums over tensor products : @xmath135 substituting these relations into eq .
( [ lindblad - gen ] ) and using the expression ( [ extended - space ] ) for @xmath100 one easily shows that @xmath136 where @xmath137 hence , in order for condition ( [ l - prop ] ) to be satisfied we must have @xmath138 for all @xmath139 , and , in particular , @xmath140 . setting @xmath141 in eq .
( [ d - ik ] ) we get the form ( [ gen - k ] ) for the generator @xmath120 , where @xmath142 is hermitian .
suppose now that @xmath120 is of the form of eq .
( [ gen - k ] ) . then we define operators on the extended state space through @xmath143 where @xmath144 is hermitian .
with the help of these operators we define a lindblad generator by means of @xmath145 + \sum_{ij\lambda } \left ( s^{ij}_{\lambda } \varrho s^{ij\dagger}_{\lambda } - \frac{1}{2 } \left\ { s^{ij\dagger}_{\lambda } s^{ij}_{\lambda } , \varrho \right\ } \right).\ ] ] it is easy to check that this lindblad generator indeed has the required property ( [ l - prop ] ) .
this proves the theorem .
inserting eq .
( [ gen - k ] ) into eq .
( [ master - tcl ] ) we obtain the master equation @xmath146 \nonumber \\ & & + \sum_{j \lambda } \left ( r^{ij}_{\lambda } \rho_j r^{ij\dagger}_{\lambda } - \frac{1}{2 } \left\ { r^{ji\dagger}_{\lambda } r^{ji}_{\lambda } , \rho_i\right\ } \right).\end{aligned}\ ] ] under the condition of the existence of an embedding into a lindblad dynamics on the extended state space [ see eq .
( [ embedding ] ) ] , this is the general form for the equations of motion of the dynamical variables @xmath94 , where the @xmath130 are arbitrary hermitian operators , and the @xmath131 are arbitrary system operators .
several master equations proposed recently in the literature are of the general form of eq .
( [ gen - master ] ) . the simplest special case of this equation
is obtained if we have only a single component @xmath147 such that we can omit the indices @xmath98 and @xmath148 . the master equation ( [ gen - master ] ) then obviously reduces to an ordinary master equation for the reduced density matrix in lindblad form .
thus , eq .
( [ gen - master ] ) can be viewed as a generalization of the gorini - kossakowski - sudarshan - lindblad theorem @xcite .
if we choose @xmath149 the master equation ( [ gen - master ] ) takes the form @xmath150 hence , we get an uncoupled system of equations of motion with a lindblad generator @xmath151 for each component @xmath78 .
although each component @xmath94 follows its own markovian - type dynamics , the dynamics of the reduced density matrix @xmath112 is generally highly non - markovian .
master equations of this uncoupled form have been derived by budini and applied to various models for the dynamics of open systems in structured reservoirs @xcite . in the general case
an equation of the form ( [ gen - master ] ) involves a coupling between the components @xmath78 .
an example of such an equation has been derived in @xcite from a specific microscopic system - environment model .
the model describes a two - state system with ground state @xmath152 , excited state @xmath153 and level distance @xmath154 , which is coupled to an environment @xcite .
the environment consists of a large number of energy levels arranged in two energy bands of width @xmath155 ( see fig .
[ model - fig ] ) .
the lower energy band contains @xmath156 levels , the upper band @xmath157 levels .
the total hamiltonian of the model is taken to be @xmath158 @xmath159 is the free system hamiltonian , where @xmath160 denote the usual raising and lowering operators of the two - state system .
the free hamiltonian of the environment is given by @xmath161 and the system - environment interaction is described by @xmath162 the index @xmath163 labels the levels of the lower energy band and @xmath164 the levels of the upper band .
the overall strength of the interaction is parameterized by the constant @xmath165 .
the coupling constants @xmath166 are independent and identically distributed complex gaussian random variables with zero mean and unit variance . ]
we employ a projection superoperator of the form @xcite @xmath167 where @xmath168 ( @xmath169 ) denotes the projection onto the lower ( upper ) environmental energy band .
this is a projection of the form given in eq .
( [ proj - new ] ) .
it projects onto a state in which the environmental state @xmath170 is correlated with the system state @xmath171 .
the second order of the time - convolutionless projection operator technique leads to the master equation ( written in the interaction picture ) @xmath172 where @xmath173 .
this is a coupled system of first - order differential equations for the two unnormalized density matrices @xmath174 and @xmath175 .
it can be written in the form of eq .
( [ gen - master ] ) by taking @xmath176 , @xmath177 , @xmath178 , and @xmath179 . as has been demonstrated in ref .
@xcite through a comparison with numerical simulations of the full schrdinger equation , this master equation yields an excellent approximation of the reduced system s dynamics . the second term on the right - hand side of eq .
( [ master1 ] ) describes changes of @xmath174 which are due to downward transitions of the two - state system .
these are only possible if the lower band of the environment is populated , i. e. , if the environment is in the state @xmath180 which is correlated with @xmath174 .
for this reason the second term of eq .
( [ master1 ] ) involves the density @xmath174 .
correspondingly , the first term on the right - hand side of eq .
( [ master1 ] ) describes changes of @xmath174 caused by excitations of the two - state system .
such excitations are only possible if the environment is in the state @xmath181 which is correlated with @xmath175 .
therefore , it is the density @xmath175 that enters the first term of eq .
( [ master1 ] ) .
analogous statements hold for eq .
( [ master2 ] ) .
hence , we see that the transitions described by eqs .
( [ master1 ] ) and ( [ master2 ] ) exactly conserve the total number of excitations ( see sec . [ cons ] ) .
let us investigate the behavior of the excited state population @xmath182 which is defined by the expression @xmath183 equations ( [ master1 ] ) and ( [ master2 ] ) lead to @xmath184 this equation exhibits a strong non - markovian character because the initial data appear as inhomogeneous terms on its right - hand side .
these terms express a pronounced memory effect : they imply that the dynamics never forgets its initial condition , i. e. , that the process is governed by an infinite memory time .
a further remarkable feature of eq .
( [ pop ] ) derives from the fact that the initial data on its right - hand side can not be expressed solely in terms of the matrix elements of the reduced density matrix .
this means that at any time @xmath105 , and even in the limit @xmath185 , the process is strongly influenced by the statistical correlations of the initial state .
hence , the influence of the initial correlations never dies out and is present even in the stationary state .
recently , esposito and gaspard @xcite have derived a master equation from a microscopic system - reservoir model within second order perturbation theory , which is also of the form of eq .
( [ gen - master ] ) . in their derivation the index
@xmath98 plays the role of the energy @xmath186 of the reservoir .
the corresponding density matrix @xmath187 describes a system state which is correlated with a certain reservoir state of energy @xmath186 . if the open system represents again a two - state system , the master equation of esposito and gaspard can be written as . here
we keep the discrete notation and employ a slightly different treatment of the lamb shift terms . ]
@xmath188 \\ & & \hspace{-8 mm } + \sum_{\varepsilon ' } \left ( \gamma_1(\varepsilon,\varepsilon ' ) \sigma_+ \rho_{\varepsilon ' } \sigma_- - \frac{\gamma_2(\varepsilon',\varepsilon)}{2 } \left\ { \sigma_+\sigma_- , \rho_{\varepsilon } \right\ } \right . \\ & & \left .
+ \gamma_2(\varepsilon,\varepsilon ' ) \sigma_- \rho_{\varepsilon ' } \sigma_+ - \frac{\gamma_1(\varepsilon',\varepsilon)}{2 } \left\ { \sigma_-\sigma_+ , \rho_{\varepsilon } \right\ } \right).\end{aligned}\ ] ] here , @xmath189 are certain transition rates which are determined by the parameters of the microscopic model , and @xmath190 is the system hamiltonian including an @xmath186-dependent lamb - type energy renormalization .
one easily checks that this master equation can indeed be brought into the general form of eq .
( [ gen - master ] ) by taking @xmath191 and @xmath192 .
a master equation of the general form ( [ gen - master ] ) has been derived recently by budini @xcite .
this author suggests introducing an extra degree of freedom @xmath193 which modulates the interaction between the reduced system @xmath7 and its environment @xmath8 . under the assumptions that @xmath8 may be treated as a markovian reservoir and that the dynamics of the populations decouples from the dynamics of the coherences of @xmath193 , one arrives at the master equation ( [ gen - master ] ) . in a certain sense
the introduction of an additional degree of freedom @xmath193 corresponds to the extended state space @xmath123 used in sec .
[ structure ] to construct the embedding into a lindblad dynamics .
an advantage of the present formulation is that it avoids the use of a microscopic model for the extra degree of freedom and that it clearly shows the basic physical condition underlying the master equation ( [ gen - master ] ) : this condition is the existence of an effective representation of the dynamics through a projection onto separable , classically correlated quantum states . what are the physical conditions under which the use of a standard projection onto a tensor product state is not sufficient to correctly describe the reduced system s dynamics for a given system - environment model ?
while this important question seems to be difficult to answer in full generality and certainly deserves further investigations , important hints can be obtained already by an investigation of the time - convolutionless perturbation expansion @xcite .
if the two - point environmental correlation functions do not decay rapidly in time the second order of the expansion can not , of course , be expected to give an accurate description of the dynamics .
for instance , this situation arises for the spin star model discussed in ref .
@xcite , where the second - order generator of the master equation increases linearly with time such that the born - markov approximation simply does not exist . the example investigated in ref .
@xcite demonstrates that there are even cases in which the standard markov condition is satisfied although the product - state projection completely fails if one truncates the expansion at any finite order . in such cases strong non - markovian dynamics
is induced through the behavior of higher - order correlation functions .
hence , one can judge the quality of a given projection superoperator only by an investigation of the structure of higher orders of the expansion .
the standard projection and the corresponding lindblad equation are not reliable if higher orders lead to contributions that are not bounded in time , signifying the non - uniform convergence of the perturbation expansion @xcite .
for many physical models one knows certain quantities @xmath194 which are exactly ( or at least approximately ) conserved under the time - evolution . a great advantage of the formulation by means of the master equation ( [ gen - master ] )
is that it allows the implementation of the corresponding conservation laws .
for instance , the master equation derived in ref .
@xcite reflects the conservation of the ( uncoupled ) total system energy . to formulate the conservation of @xmath194 one chooses the operators @xmath46 in such a way that @xmath194 is a relevant observable
[ see sec .
[ rel - obs ] ] . according to eq .
( [ o - rel ] ) this means that @xmath194 can be written in the form @xmath195 then we have the exact relation @xmath196 hence , we can express the conservation of @xmath194 on the level of the master equation by means of the conservation law @xmath197 by use of the master equation ( [ gen - master ] ) this yields a relation between the operators @xmath130 and @xmath131 : @xmath198 + \sum_{j \lambda } \left ( r^{ji\dagger}_{\lambda } c^j_s r^{ji}_{\lambda } - \frac{1}{2 } \left\ { r^{ji\dagger}_{\lambda } r^{ji}_{\lambda } , c^i_s \right\ } \right ) = 0.\ ] ] thus , known conserved quantities lead to constraints on the choice of the operators that enter the master equation .
an example is given by the quantity @xmath199 which counts the total number of excitations for the model discussed in sec .
[ examples ] .
this quantity is exactly conserved under the evolution generated by the hamiltonian ( [ h - model ] ) .
writing @xmath200 we see that @xmath194 is indeed a relevant observable for the projection ( [ newprojection ] ) , i. e. , we have @xmath201 .
the corresponding conservation law takes the form @xmath202 this fact can be used to motivate the projection superoperator : with the choice of eq .
( [ newprojection ] ) one ensures that the projection superoperator leaves invariant the conserved quantity and that the effective description respects the corresponding conservation law .
we have discussed the theoretical description of non - markovian quantum dynamics within the framework of the projection operator techniques .
it has been shown that an efficient modelling of strong non - markovian effects is made possible through the use of correlated projection superoperators @xmath0 that take into account statistical correlations between the open system and its environment .
we have formulated and explained some general physical conditions which demand , essentially , that @xmath0 can be expressed in terms of a projective quantum channel that operates on the environmental variables .
on the basis of these conditions a representation theorem for correlated projection operators [ eq . ( [ projection - genform ] ) ] has been derived . employing a correlated projection superoperator instead of the usual projection onto a tensor product state ,
one enlarges the set of dynamical variables from the reduced density matrix @xmath16 to a collection of densities @xmath78 describing system states that are correlated with certain environmental states . by means of an embedding of the non - markovian dynamics into a lindblad dynamics on a suitably extended state space
, we have derived the general structure of a master equation [ eq . ( [ gen - master ] ) ] which governs the dynamics of the @xmath78 and models strong non - markovian effects , while preserving the physical conditions of normalization and positivity .
a particularly important feature of the master equation is that it is able to describe very long and even infinite memory times , as well as large correlations in the initial state .
we have also discussed the role of known conserved quantities .
such quantities may be helpful to find an appropriate projection superoperator by demanding that they be relevant observables for @xmath0 .
once this has been achieved one can express the corresponding conservation laws on the level of the effective description provided by the master equation .
the semigroup assumption used in the derivation of the master equation ( [ gen - master ] ) is not really necessary .
in fact , the present formulation can easily be extended to the case of an explicitly time - dependent lindblad generator on the extended state space .
the resulting master equation is then again of the form of eq .
( [ gen - master ] ) , where the operators @xmath203 and @xmath204 now depend explicitly on time .
the result expressed by eq .
( [ gen - master ] ) could be particularly useful for a phenomenological approach to non - markovian dynamics : for arbitrary @xmath130 and @xmath131 this equation represents a physically acceptable master equation because it preserves the normalization of the reduced density matrix and transforms positive initial components @xmath205 into positive components @xmath206 .
we emphasize that the argument leading to the form ( [ gen - master ] ) of the master equations is non - perturbative .
given a certain projection superoperator @xmath0 defining the densities @xmath78 , the only assumption entering the derivation is the existence of an embedding of the dynamics into a lindblad dynamics on the extended state space .
the projection superoperators used for the description of non - markovian dynamics in sec . [ dynamics ] have a special feature .
namely , they project any given state onto a classically correlated state .
if a non - markovian dynamics can be approximated by use of such a superoperator within low orders of the perturbation expansion , one can conclude that the true states of the total system can be represented effectively through classically correlated states and that genuine quantum correlations ( entanglement ) may be treated as perturbations . on the other hand , the representation theorem of sec .
[ sec - representation ] includes the case of projection superoperators that project onto nonseparable ( entangled ) quantum states . for such superoperators
the arguments that led to the master equation ( [ gen - master ] ) are not applicable .
it remains an important open problem to extend the formulation developed here to the case of nonseparable projection superoperators .
such an extension could enable a systematic investigation of the dynamical significance of genuine quantum correlations in non - markovian processes .
99 h. p. breuer and f. petruccione , _ the theory of open quantum systems _ ( oxford university press , oxford , 2002 )
. f. haake and r. reibold , phys .
a * 32 * , 2462 ( 1985 ) . b. l. hu , j. p.
paz and y. zhang , phys .
d * 45 * , 2843 ( 1992 ) .
h. grabert , p. schramm , and g .- l .
ingold , phys .
168 * , 115 ( 1988 ) .
p. telmachovi and v. buek , phys .
a * 64 * , 062106 ( 2001 ) ; phys .
a * 67 * , 029902(e ) ( 2003 ) .
j. schliemann , a. khaetskii , and d. loss , j. phys .
: condens .
matter * 15 * , r1809 ( 2003 ) .
j. gemmer and m. michel , europhys .
lett . * 73 * , 1 ( 2006 ) .
j. gemmer , m. michel , and g. mahler , _ quantum thermodynamics _ , lecture notes in physics , vol .
657 ( springer , berlin , 2004 ) .
m. michel , g. mahler and j. gemmer , phys .
lett . * 95 * , 180602 ( 2005 ) .
s. nakajima , progr .
. phys . * 20 * , 948 ( 1958 ) .
r. zwanzig , j. chem . phys . * 33 * , 1338 ( 1960 ) .
f. haake , _ statistical treatment of open systems _ , springer tracts in modern physics , vol .
66 ( springer , berlin , 1973 ) .
r. kubo , m. toda , and n. hashitsume , _ statistical physics ii .
nonequilibrium statistical mechanics _ ,
( springer , berlin , 1991 ) . f. shibata , y. takahashi , and n. hashitsume , j. stat .
* 17 * , 171 ( 1977 ) ; s. chaturvedi and f. shibata , z. phys .
b * 35 * , 297 ( 1979 ) ; f. shibata and t. arimitsu , j. phys . soc .
jap . * 49 * , 891 ( 1980 ) ; c. uchiyama and f. shibata , phys .
e * 60 * , 2636 ( 1999 ) ; a. royer , phys .
lett . a * 315 * , 335 ( 2003 ) ; h. p. breuer , phys .
a * 70 * , 012106 ( 2004 ) . v. gorini , a. kossakowski , and e. c. g. sudarshan , j. math
* 17 * , 821 ( 1976 )
. g. lindblad , commun .
. phys . * 48 * , 119 ( 1976 ) .
h. spohn , rev .
phys . * 53 * , 569 ( 1980 ) .
h. p. breuer , d. burgarth , and f. petruccione , phys .
b * 70 * , 045323 ( 2004 ) .
m. esposito and p. gaspard , phys .
e * 68 * , 066112 ( 2003 ) .
m. esposito and p. gaspard , phys .
e * 68 * , 066113 ( 2003 ) .
a. a. budini , phys .
e * 72 * , 056106 ( 2005 ) .
h. p. breuer , j. gemmer , and m. michel , phys .
e * 73 * , 016139 ( 2006 ) . s. m. barnett and s. stenholm , phys .
a * 64 * , 033808 ( 2001 ) .
a. a. budini , phys .
a * 69 * , 042107 ( 2004 ) .
a. shabani and d. a. lidar , phys .
a * 71 * , 020101(r ) ( 2005 ) .
s. maniscalco and f. petruccione , phys .
a * 73 * , 012111 ( 2006 ) .
v. gorini , a. frigerio , m. verri , a. kossakowski , and e. c. g. sudarshan , rep .
* 13 * , 149 ( 1978 ) .
k. kraus , ann .
( n.y . ) * 64 * , 311 ( 1971 ) . w. f. stinespring , proc .
* 6 * , 211 ( 1955 ) .
g. alber , t. beth , m. horodecki , p. horodecki , r. horodecki , m. rtteler , h. weinfurter , r. werner , and a. zeilinger , _ quantum information _ ( springer - verlag , berlin , 2001 ) .
r. f. werner , phys .
a * 40 * , 4277 ( 1989 ) .
a. a. budini , phys .
a * 74 * , 053815 ( 2006 ) . | a systematic approach to the non - markovian quantum dynamics of open systems is given by the projection operator techniques of nonequilibrium statistical mechanics .
combining these methods with concepts from quantum information theory and from the theory of positive maps , we derive a class of correlated projection superoperators that take into account in an efficient way statistical correlations between the open system and its environment .
the result is used to develop a generalization of the lindblad theory to the regime of highly non - markovian quantum processes in structured environments . |
the perturbative series for qcd observables is not convergent , the @xmath1-order coefficients exhibiting @xmath2 growth at large @xmath3 . in the borel plane
there are singularities along the real axis : ultraviolet ( uv ) renormalons along the negative real semi - axis , and infrared ( ir ) renormalons lying on the integration contour , along the positive real semi - axis .
the latter render the borel integral ambiguous , the ambiguity being structurally identical to ambiguities arising in the non - perturbative operator product expansion ( ope ) @xcite .
consequently , all - orders perturbation theory is only well - defined if supplemented by the non - perturbative ope , allowing the ambiguities to cancel . in practice
all - orders calculations are only possible in the `` leading-@xmath4 approximation '' , in which the @xmath1-order perturbative coefficient is recast as an expansion in powers of @xmath5 , the first beta - function coefficient ( in su@xmath6 qcd , with @xmath7 active quark flavours ) . the `` leading-@xmath4 '' term ( proportional to @xmath8 )
is then used to approximate the @xmath1-order coefficient @xcite .
crucially this term can be identified to all - orders from large @xmath7 calculations , involving a restricted set of diagrams in which a chain of @xmath3 fermion bubbles is inserted into a basic `` skeleton '' diagram , in all possible ways . in this way
one can perform an all - orders leading-@xmath4 resummation by explicitly evaluating the borel transform , and defining the borel sum , suitably regulated to control the ir renormalon ambiguities .
such results have been used in the past to try to assess the likely accuracy of fixed - order perturbation theory for quantities such as the @xmath9 ratio at low values of the c.m .
energy @xmath0 ( where the qcd coupling is not so small ) , and for its tau - decay analogue @xmath10 ( where again , at the scale of the @xmath11 mass , the coupling is relatively large ) @xcite .
dis sum rules have also been studied in this context @xcite . in matching to the exact fixed - order perturbative calculations ( nlo and nnlo coefficients are available for @xmath12 , @xmath10 and dis sum rules ) , one faces the problem that the all - orders leading-@xmath4 result is only renormalization scheme ( rs ) invariant if the one - loop form of the coupling is used , whereas the exact result must involve the higher - loop coupling .
the matched leading-@xmath4 resummed result consequently depends on the assumed renormalization scale at which the matching takes place . to avoid this ambiguity
it has been argued in the past that one should remove scale dependence by a resummation of scale logs to all - orders ; employing the effective charge approach @xcite , or the `` complete renormalization group improvement '' ( corgi ) approach @xcite .
these approaches involve scheme - invariants which can be approximated at the leading-@xmath4 level and resummed @xcite .
in this paper we will use the corgi approach to perform leading-@xmath4 resummations for dis sum rules .
the aim will be to assess the reliability of fixed - order nnlo perturbation theory for these quantities , at the rather small values of @xmath0 for which they have been measured @xcite .
in addition we shall used a recently proposed model for power corrections , based on the qcd skeleton expansion @xcite , to obtain fits for the non - perturbative contributions .
the three physical quantities we are interested in are the gls sum rule @xcite , the polarized bjorken sum rule @xcite , and the unpolarized bjorken sum rule @xcite .
these quantities can be written as a parton model result plus perturbative corrections ( denoted by calligraphic letters ) , @xmath13 @xmath14 @xmath15 where @xmath16 @xcite , and in the case of @xmath17 we have neglected contributions from ` light - by - light ' scattering diagrams .
the perturbative corrections take the form of an expansion in powers of the strong coupling @xmath18 , @xmath19 these expressions can be evaluated at fixed - order , and the coefficients @xmath20 and @xmath21 are known up to @xmath22 @xcite ( nnlo ) . @xmath23
the leading-@xmath4 approximation arises from rearranging the expansion as one in powers of @xmath7 .
we can write the @xmath1-order coefficients in the form , @xmath24}{n}_{f}^{n}+{k}_{n}^{[n-1]}{n}_{f}^{n-1}+\ldots+{k}_{n}^{[0]}\;,\ ] ] and similarly for @xmath25 .
the leading @xmath7 coefficient @xmath26}$ ] can be computed exactly to all - orders as noted above .
one can replace @xmath7 by @xmath27 to obtain an expansion in powers of @xmath4 , @xmath28 the leading-@xmath4 term @xmath29 can then be used to approximate @xmath30 @xcite . to see
how well this approximation works we can compare the known exact results for @xmath31 , @xmath32 , and @xmath33 , @xmath34 , with the leading-@xmath4 approximations @xmath35 , @xmath36 and @xmath37 , @xmath38 .
we shall assume the @xmath39 scheme with renormalization scale @xmath40 .
for @xmath41 one has the exact coefficients @xmath42 to be compared with the leading-@xmath4 approximations @xmath43 the leading-@xmath7 coefficients of course agree exactly by construction , but one sees agreement in sign and @xmath44 level agreement in magnitude for the sub - leading coefficients as well .
the corresponding coefficients for @xmath45 are @xmath46 to be compared with the leading-@xmath4 approximations @xmath47 again one sees that sub - leading coefficients are well - reproduced in sign and magnitude .
this can be partially understood since it is possible to derive a fully normalized asymptotic result for the @xmath48}$ ] and @xmath49}$ ] coefficients for _ large _ @xmath3 and _ fixed _ @xmath50 .
such a result for the qcd adler function of vacuum polarization , @xmath51 , was derived in ref .
@xcite from a consideration of the four - fermion operators controlling ultraviolet renormalon singularities .
it can be easily extended to the dis sum rules since the same dominant four - fermion operator is involved @xcite . if one writes the leading-@xmath4 term as @xmath52}{n}_{f}^{n}+{\hat{k}}_{n-1}^{[n-1]}{n}_{f}^{n-1}+\ldots+{\hat{k}}_{n}^{[0 ] } \\ { u}_{n}^{(l)}&=&{\hat{u}}_{n}^{[n]}{n}_{f}^{n}+{\hat{u}}_{n-1}^{[n-1]}{n}_{f}^{n-1}+\ldots+{\hat{u}}_{n}^{[0]},\end{aligned}\ ] ] then one has the asymptotic result for large @xmath3 and fixed @xmath50 @xmath53}&\approx&{\hat{k}}_{n}^{[n - r]}\left(1+o\left(\frac{1}{n}\right)\right)+o\left(\frac{1}{n^2}\right ) \\ { u}_{n}^{[n - r]}&\approx&{\hat{u}}_{n}^{[n - r]}\left(1+o\left(\frac{1}{n}\right)\right)+o\left(\frac{1}{n^2}\right).\end{aligned}\ ] ] the @xmath54 are terms sub - leading in the number of colours , @xmath55 .
the asymptotics agree up to @xmath56 terms in the `` planar approximation '' @xcite , where at each order in @xmath7 only the highest power of @xmath55 is retained .
of course this asymptotic result does not explain the @xmath44 level agreement observed on comparing eqs .
( 8) , ( 9 ) , ( 12 ) and ( 13 ) with eqs .
( 10 ) , ( 11 ) , ( 14 ) and ( 15 ) , for such small values of @xmath57 .
the agreement is also remarkably good for the @xmath58}$ ] and @xmath59}$ ] coefficients which correspond to @xmath60 , or the large-@xmath55 limit .
similar remarks apply for the adler function @xmath51 and some speculative ideas as to how this might arise are given in ref .
whilst this good agreement for given powers in the @xmath7 expansion is interesting , of more relevance is how well the overall @xmath31 , @xmath32 and @xmath33 , @xmath34 coefficients are approximated , since one would not expect the @xmath44 accuracy to survive the addition of the terms .
we give below the exact and leading-@xmath4 coefficients for @xmath61 . for @xmath41 one
has @xmath62 whilst for @xmath45 one has @xmath63 as expected from the above discussion one finds good @xmath64 level agreement for @xmath60 which gradually worsens to factor of @xmath65 level agreement at @xmath66 .
the resummation of leading-@xmath4 terms to all - orders , which we shall denote as @xmath67 , can be accomplished using the borel sum method , @xmath68(z)=\sum_{n=0}^{\infty}\frac{k_{n}^{(l)}}{n!}z^{n}&\qquad\longrightarrow\qquad&{\ensuremath{\mathcal{k } } } _ { \infty}(a)\simeq\int_{0}^{\infty}dz e^{-z / a}b[{{\ensuremath{\mathcal{k } } } } ^{(l)}](z ) .
\label{bt}\end{aligned}\ ] ] the borel transforms of the three above quantities in the leading-@xmath4 approximation can be calculated from the results in refs .
@xcite and @xcite .
one finds , @xmath69(z)= \frac{4/9}{\big{(}1+\frac{z}{z_{1}}\big{)}}-\frac{1/18}{\big{(}1 + \frac{z}{z_{2}}\big{)}}+\frac{8/9}{\big{(}1-\frac{z}{z_{1}}\big{)}}-\frac{5/18}{\big{(}1- \frac{z}{z_{2}}\big{)}},\end{aligned}\ ] ] and , @xmath70(z)=\frac{1/6}{\big{(}1+\frac{z}{z_{2}}\big{)}}+\frac{4/3}{\big{(}1-\frac{z}{z_{1}}\big{)}}-\frac{1/2}{\big{(}1-\frac{z}{z_{2}}\big{)}}.\end{aligned}\ ] ] the borel transforms have simple pole singularities on the positive @xmath71-axis ( ir renormalons ) , and on the negative @xmath71-axis ( uv renormalons ) ; the positions are at @xmath72
. one can easily compute the pv regulated leading-@xmath4 borel sums , @xmath73 and @xmath74 @xcite , @xmath75 } , \label{eq : kinf}\end{aligned}\ ] ] and , @xmath76}. \label{eq : uinf}\end{aligned}\ ] ] ei@xmath77 is the exponential integral function defined ( for @xmath78 , uv renormalons ) as , @xmath79 and for @xmath80 , by taking the pv of the integral ( ir renormalons ) .
for these all - orders resummations to be rs ( scale)-invariant we need to take @xmath81 as the one - loop coupling , @xmath82 the all - orders results of eqs .
( [ eq : kinf ] ) and ( [ eq : uinf ] ) are derived in the so - called @xmath83-scheme ( @xmath39 with renormalization scale @xmath84 ) , and hence @xmath85 in eq .
( [ eq : olc ] ) will refer to that in the @xmath83-scheme defined by , @xmath86 as we noted above there is a problem if we wish to match these leading-@xmath4 resummations with exact higher - order nlo and nnlo calculations for @xmath41 and @xmath45 , which involve the higher - loop coupling .
we shall avoid the matching ambiguity by employing the corgi approach @xcite .
the standard rs - dependent coupling @xmath87 satisfies the rg - equation , @xmath88 here @xmath5 and @xmath89 are universal and rs - invariant , the higher coefficients , @xmath90 , are rs - dependent and may be used , together with dimensional transmutation parameter , @xmath85 , to label the scheme .
the corgi approach consists of resumming to all - orders the rg - predictable terms available to a given fixed - order of calculation .
the standard perturbative expansion is then replaced with the following expression , @xmath91 which removes the renormalization scale dependence of eq .
( [ alo ] ) .
we will use a tilde to denote corgi - ized results .
initially we shall deal with @xmath41 but the results are easily generalized to @xmath45 .
the coefficients @xmath92 are renormalization scheme invariant quantities , each of which can be derived from an @xmath93lo calculation of the coefficients of @xmath41 , and of the beta function equation coefficients , @xmath94 . for @xmath22 and 3
they have the form , @xmath95 the coupling @xmath96 used in eq .
( [ eq : corgiallo ] ) has the form @xcite , @xmath97}\big{)}},\label{eq : aok } \label{eq : a}\end{aligned}\ ] ] and is the solution to the beta function equation in the t hooft scheme @xcite ( a scheme where @xmath98 ) . here
@xmath99 is the lambert @xmath99 function , defined implicitly by @xmath100 .
the `` @xmath101 '' subscript on @xmath99 denotes the branch of the lambert @xmath99 function required for asymptotic freedom , the nomenclature being that of ref .
@xcite . as a result of adopting the corgi approach
, we use a new scale parameter @xmath102 which is rs invariant but dependent on the observable under consideration .
it is related to the @xmath39 scale parameter by , @xmath103 here @xmath104 denotes the nlo perturbative coefficient with scale choice @xmath105 .
the exponential term in the above equation has the effect of resumming a set of rg predictable terms present in the full perturbative expansion , it also reproduces the @xmath106 term present in eq .
( [ alo ] ) but absent from eq .
( [ eq : corgiallo ] ) .
the difference between @xmath85 and @xmath107 is due to different conventions for integrating the beta - function equation , @xmath85 being the standard convention , and @xmath107 being the convention favoured in ref .
@xcite .
equation ( [ eq : corgiallo ] ) can easily be adapted to provide a prediction for the unpolarized bjorken sum rule of eq .
( [ ub ] ) .
we simply use the equivalent coefficients of @xmath45 in eqs .
( [ eq : x2 ] ) and ( [ eq : x3 ] ) and the following coupling , @xmath108}\big{)}},\label{eq : aou}\end{aligned}\ ] ] where , @xmath109 equation ( [ eq : kinf ] ) can be modified so that is resums the leading-@xmath4 components of eq .
( [ eq : corgiallo ] ) and essentially becomes an all - orders corgi result .
we define the coupling @xmath110 through , @xmath111 where @xmath112 is calculated in the v - scheme .
the coupling @xmath96 in eq .
( [ eq : av ] ) is that of eq .
( [ eq : aok ] ) .
the all - orders corgi result can now be obtained from eq .
( [ eq : kinf ] ) using eq .
( [ eq : av ] ) , @xmath113 where @xmath114 , in analogy with eq .
( [ eq : leadingb ] ) , is the leading-@xmath4 part of @xmath92 , for example : @xmath115 we can also improve upon @xmath116 by adding to eq .
( [ eq : kcorinf ] ) the known sub - leading-@xmath4 part of @xmath117 , @xmath118 equation ( [ eq : allo+ ] ) now contains all the information we have about the perturbative coefficients for @xmath41 at all - orders .
analogous expressions also hold for @xmath119 and @xmath120 . in ref .
@xcite , we noted that the one chain result of eq .
( [ eq : kinf ] ) , is finite at the landau pole @xmath121 , and remains so for values of @xmath122 below @xmath85 .
remarkably , this finiteness at @xmath123 holds when we use the t hooft coupling of eq .
( [ eq : aok ] ) .
furthermore , both of these results have the same values at their respective landau poles , i.e. @xmath124 similar relations apply to @xmath45 and also to the adler - d function .
this conclusion is altered slightly for the actual corgi result of eq .
( [ eq : kcorinf ] ) because of the use of @xmath125 . in this case
the result remains finite at @xmath126 , but has a different value to that in eq .
( [ eq : finite ] ) we can also consider nlo and nnlo corgi , fixed - order approximations which can then be compared with all - orders resummations , @xmath127 by comparing these with the all - orders predictions , we can asses the reliability of fixed - order perturbative predictions . before doing
so we consider the accuracy of the leading-@xmath4 approximation for the corgi invariants @xmath128 and @xmath129 , as we did for the perturbative coefficients earlier .
for @xmath130 one finds that @xmath131 notice that the invariants expanded in powers of @xmath4 contain a @xmath132 term @xcite and so it is necessary to multiply by a factor of @xmath4 to obtain the expansion in powers of @xmath7 .
as we found for the perturbative coefficients in eqs .
( 20 - 27 ) there is reasonably good agreement in sign and magnitude of the coefficients with astonishingly good @xmath133 level agreement for the @xmath60 large-@xmath55 coefficient in this case .
the numerical values of the @xmath128 invariants for @xmath134 are @xmath135 we see good agreement for @xmath60 which gradually worsens for increasing @xmath7 .
note that @xmath128 changes sign between @xmath136 and @xmath66 , and so big cancellations are involved in this region . for @xmath129
we find @xmath137 the agreement appears to be significantly worse in this case although the pattern of signs is reproduced .
the numerical values are @xmath138
the data available for @xmath17 and @xmath139 span a range of energy which includes the bottom quark mass threshold .
it is therefore necessary for us to choose a method for evolving the above expressions through this threshold .
we adopt the approach detailed in @xcite . at a particular energy scale
we treat all quarks with masses less than that scale as ` active ' but massless and we ignore quarks with masses greater than that scale . as a consequence , the coefficients @xmath20 and @xmath21 are now @xmath140-dependent and hence they will depend on the c.m .
energy scale , @xmath122 .
we also perform matching of the coupling at @xmath141 . at lo and nlo this
amounts to demanding continuity of the coupling at the threshold but at nnlo and beyond , this continuity is violated .
this matching forces us to adopt different values of the scale parameter in different @xmath140 regions .
this is governed by the following equations , @xmath142 where @xmath143 and @xmath144 can be obtained from the results in ref .
@xcite . here
, @xmath145 is the scale parameter in the region where @xmath140 quarks are active , @xmath146 is the pole mass of the @xmath147 quark and @xmath148 is simply @xmath4 evaluated for @xmath140 quark flavours
. we must be careful how we apply this matching to the different results we have obtained .
equation ( [ eq : matching ] ) is an @xmath149 result and hence we carry out the matching for @xmath150 and then convert to the various other scales we have defined in eqs .
( [ eq : lv ] ) , ( [ eq : lk ] ) and ( [ eq : lu ] ) .
@xmath151 is an nlo result and hence , when applying eq .
( [ eq : matching ] ) to this result , we omit @xmath144 . the results in eqs .
( [ eq : kcorinf ] ) , ( [ eq : allo+ ] ) and ( [ eq : nnlo ] ) are all , at least , nnlo and hence we use the full result from eq .
( [ eq : matching ] ) .
[ f : fig3 ] we wish to compare these results with relevant experimental data .
data is available for the gls sum rule @xcite and the polarized bjorken sum rule @xcite , where the points we plot are those arising from the analysis of @xcite .
we plot the data in figs .
[ fig1 ] and [ fig2 ] along with our predictions for the full observables .
no experiment has so far measured the unpolarized bjorken sum rule , however the possibility exists that it may be extracted from experiments at a future neutrino factory @xcite .
we have taken @xmath152 mev , corresponding to the world average value @xmath153 @xcite . within the considerable error
bars we see that the different versions of corgi all fit the data tolerably well .
the important conclusion is that below @xmath154 @xmath155 the different approaches give drastically different predictions .
this indicates that fixed - order perturbation theory can not be trusted for these lower energies .
even though we have no experimental data points for @xmath156 we have plotted the different versions of corgi in fig .
[ f : fig3 ] .
we see that fixed - order perturbation theory can not be trusted below @xmath157 @xmath155 . a surprising feature evident from figs .
1 - 3 is that even at @xmath158 the fixed - order nlo and nnlo results remain extremely close , leading to the superficial conclusion that fixed - order perturbation theory is working well .
in fact as can be seen from eqs .
( 54,56 ) the corgi invariant @xmath159 for both @xmath160 and @xmath161 changes sign between @xmath136 and @xmath66 and so fortuitously is very small in this region . when @xmath162 and higher invariants are resummed the potential unreliability of fixed - order perturbation theory is revealed .
we can contrast these corgi resummations with a leading-@xmath4 resummation in the @xmath39 scheme . if we choose the scale @xmath163 we have , analogous to eq .
( [ eq : kcorinf ] ) , @xmath164 here @xmath165 denotes the full higher - loop ( three - loop if matching to nnlo ) @xmath39 coupling at scale @xmath163 , and the @xmath166 coefficients are the @xmath39 coefficients with scale @xmath163 .
the coupling @xmath167 is then defined by , @xmath168 we can match this all - orders resummation to the exact nlo and nnlo perturbative coefficients , obtaining , in analogy with eq .
( [ eq : allo+ ] ) , @xmath169 where in this @xmath170 is the approximated three - loop coupling , given in @xcite .
prediction for the gls sum rule .
here we use the @xmath39 result of eq .
( [ eq : msleadingb ] ) with renormalization scales of @xmath171 and @xmath172 . ] the resulting prediction is plotted in fig . [ f : fig4 ] for the gls sum rule .
three different matching scales corresponding to @xmath173 were chosen . as can be seen
, the matched resummed perturbative result is hopelessly @xmath174-dependent .
the corgi result @xmath175 corresponds to the choice @xmath176 .
however , we note that no nlo matching is required in the corgi approach , since an infinite set of rg - predictable terms involving @xmath31 have been resummed to all - orders , yielding a @xmath177-independent result .
in addition to perturbative corrections it is expected that there will be non - perturbative , higher - twist ( ht ) corrections to the sum rules . in a recent paper @xcite we have studied the infra - red freezing of euclidean observables , including the sum rules , in the language of the one - chain term of the skeleton expansion .
the leading-@xmath4 resummations we have discussed can be written in the form , @xmath178 here @xmath179 is the characteristic function which is piecewise continuous at @xmath180 .
the two terms respectively reproduce the ir renormalon and uv renormalon contributions in the borel sum , but the skeleton expansion result is also defined for @xmath181 where the standard borel representation breaks down .
remarkably , there is continuity and finiteness at @xmath182 . by considering
the compensation of ambiguities between perturbative and non - perturbative corrections alluded to above we are led to an expression for ht corrections in terms of the characteristic function , @xmath183 here @xmath184 is a single overall non - perturbative constant .
this result is obtained from the v - scheme result of eq .
( [ eq : kinf ] ) and hence the appropriate scale parameter for eq .
( [ eq : kht ] ) is @xmath185 .
a similar expression holds for @xmath186 in terms of @xmath187 .
the characteristic functions @xmath179 and @xmath188 can be determined from their borel transforms , using a result from @xcite .
@xmath179 has the form , @xmath189 and the equivalent expression for @xmath161 is , @xmath190 in figs .
[ f : glsfit ] and [ f : pbjfit ] we take the four expressions in eqs .
( [ eq : kcorinf ] ) , ( [ eq : allo+ ] ) , ( [ eq : nlo ] ) and ( [ eq : nnlo ] ) , supplemented by the non - perturbative term in eq .
( [ eq : kht ] ) , and perform fitting to experimental data for both the gls and the polarized bjorken sum rules separately .
we fix @xmath152 mev , as before , and use @xmath191 fitting to obtain the optimal value of the non - perturbative parameters in each case , @xmath192 and @xmath193 .
the fitted parameters are summarized in table 1 .
we note that the coefficient of a @xmath194 ( twist-4 ) power correction to either eq .
( [ kg ] ) or ( [ kp ] ) corresponds to a value of @xmath195 or @xmath196 , respectively .
the corresponding values in @xmath197 are presented in table 1 . for comparison
the central values resulting from a three - point function qcd sum rules fit @xcite are @xmath198 and @xmath199 , respectively .
we see that the power corrections in table 1 resulting from fitting to the resummed @xmath200 all - orders resummations are significantly smaller , although the errors are large .
connections between power corrections for the three dis sum rules have also been explored in @xcite .
.fitted values of the non - perturbative constants @xmath192 and @xmath193 , together with their respective @xmath191 per degree of freedom . [ cols="^,^,^,^,^,^,^ " , ] . ] .
by comparing leading-@xmath4 resummations based on the corgi approach we attempted to infer the validity of fixed - order perturbation theory for the gls and pbj dis sum rules at small energy scales @xmath201 , where data has previously been used to extract @xmath202 @xcite .
figures 1 and 2 indicate that fixed - order perturbation theory and resummed all - orders predictions start to differ drastically for energies below @xmath203 .
the use of the corgi approach ensures that all rg - predictable scale - dependent logarithms are resummed to all - orders , thus avoiding the need for nlo matching which would otherwise make the conclusions about the validity of fixed - order perturbation theory dependent on the chosen matching scale ( as illustrated in fig . 4 ) .
we performed fits to data for power corrections using the model proposed in ref .
@xcite based on the skeleton expansion characteristic function .
the size of power corrections inferred by fitting the all - orders corgi resummations was much smaller than the central values obtained from three - point qcd sum rule estimates of refs .
@xcite .
we should note that in a recent interesting paper @xcite a renormalon analysis of pbj and gls sum rules was also carried out .
the aim in that case was to perform a sophisticated matching of perturbative and non - perturbative effects in the renormalon subtracted approach .
we would like to thank andrei kataev for useful discussions on the topic of dis sum rules and power corrections .
p.m.b . gratefully acknowledges the receipt of a pparc uk studentship . for a review see : m. beneke , phys .
rep . * 317 * , 1 ( 1999 ) , [ hep - ph/9807443 ] ; m. beneke and v. m. braun , published in ` the boris ioffe festschrift- at the frontier of particle physics / handbook of qcd' , edited by m. shifman ( world scientific , singapore , 2001 ) , [ hep - ph/0010208 ] r. m. corless , g. h. gonnet , d. e. g hare , d. j. jeffrey and d. e. knuth , `` on the lambert @xmath99 function '' , advances in computational mathematics * 5 * ( 1996 ) 329 , available from http://www.apmaths.uwo.ca/@xmath204djeffrey/offprints.html . | for the gls and bjorken dis sum rules we compare fixed - order nnlo perturbative qcd estimates , and all - orders resummed estimates , with the available data , in order to assess the reliability of fixed - order perturbation theory at rather small @xmath0 values .
fits are also performed for non - perturbative power corrections using a recently proposed model . |
the recent development of artificial free - standing structures of nanometer dimensions has led to great interest in their mechanical properties .
a wealth of experimental information is now available for nanowires @xcite and nanotubes@xcite , and a computational literature is developing on the subject@xcite .
many of these works make use of results from the continuum theory of elasticity to analyze the behavior of nanometer structures .
however , the applicability of continuum theories to nanoscale objects , where atomic - level inhomogeneities come to the fore , has yet to be explored in depth .
rigorous understanding of the elastic properties of nanoscale systems is crucial in understanding their mechanical behavior and presents an intriguing theoretical challenge lying at the cross - over between the atomic level and the continuum . in the absence of an appropriate theoretical description at this cross - over ,
critical questions remain to be answered including the extent to which continuum theories can be pushed into the nanoregime , how to provide systematic corrections to continuum theory , what effects do different bonding arrangements have on elastic response , and what signatures in the electronic structure correlate with the mechanical properties of the overall structure ?
recently , there have been a number of theoretical explorations of the impact of nanoscale structure on mechanical properties@xcite .
these studies fall under two broad approaches , either the addition of surface and edge corrections to bulk continuum theories@xcite or the extraction of overall mechanical response from atomic scale interactions@xcite .
the latter approach has the distinct advantage of allowing first principles understanding of how different chemical groups and bonding arrangements contribute to overall elastic response , thus opening the potential for the rational design of nanostructures with specific properties .
in coarse graining from interatomic interactions to mechanical response , some works rely upon the problematic decomposition of the total system energy into a direct sum of atomic energies@xcite , which is always arbitrary and particularly inconvenient for connection with _ ab initio _ electronic structure calculations . the remaining works which attempt to build up overall response from atomic level contributions@xcite fail to account properly for the poisson effect . below
we show that failure to account for this effect leads to surprisingly unphysical results .
this manuscript presents the first theory for the analysis of overall mechanical response in terms of atomic - level observables which suffers from neither difficulty from the preceding paragraph .
this analysis allows , for the first time , quantitative understanding of how continuum theory breaks down on the nanoscale , of how to make appropriate corrections , and of how to predict the effects of different bonding arrangements on overall elastic response .
it is well known that the decomposition of overall elastic response into a sum of atomic level contributions is not unique .
we show here , however , that with the additional constraint of dependence of moduli on local environment our definition of atomic level moduli becomes physically meaningful and essentially unique when coarse - grained over regions of extent comparable to the decay range of the force - constant matrix . for concreteness , in this work
we focus on nanowires .
however , we will also describe briefly how this work can be extended to any system with nanometer dimensions .
the manuscript proceeds as follows .
section [ overview ] briefly overviews the present state of the field .
next , as the traditional concept of young s modulus becomes ill - defined on the nanoscale , we begin by carefully defining continuum elastic constants appropriate for nanowires in section [ sec - nr ] .
we then show how to decompose these constants exactly into atomic - level contributions based on true physical observables ( rather than individual atomic energies ) using a straight - forward application of born and huang s method of long waves @xcite , resulting in a decomposition similar in spirit to those in references [ 17 ] and [ 18 ] ( section [ sec - mlw ] ) .
section [ sec - fail ] demonstrates the surprising , radical breakdown of this approach when applied to nanoresonators .
then , in section [ sec - mlwn ] , we identify the source of the difficulty as the poisson effect and present the first analysis of mechanical response truly applicable to nanoresonators .
the manuscript then goes on to applications .
section [ sec - trans ] verifies the physical meaningfulness of our newly defined quantities by verifying that they predict response to modes of strain for which they were not directly constructed .
we then , in section [ sec - fte ] , use our approach to generate a new , much more accurate , relationship between experimentally accessible observables describing response to flexural and extensional strain in nanomechanical resonators .
finally , section [ sec - cbames ] uses this theory to explore possible links between underlying electronic structure and local elastic response .
as the introduction mentions , the literature pursues two broad categories of approach to the study of mechanical properties on the nanoscale , either surface and edge corrections to continuum theory or extraction of overall response from the underlying atomic interactions . in the former category ,
reference [ 13 ] , through scaling arguments and numerical examples , notes that the young s modulus for nanomechanical resonators scales as a bulk term plus surface and edge corrections .
although providing insight and motivation , this work leaves completely open how one should understand these corrections from first principles .
reference [ 14 ] provides a more rigorous study based on separating nanoscale systems into continuum surface and bulk regions .
this latter approach allows prediction of changes in stiffness properties as one approaches nanometer length scales and has the appeal of generating physically motivated correction terms .
however , it relies on the separation of a nanomechanical resonator into bulk and surface continua as an _ ansatz _ and therefore neither predicts when such a picture suffices to give an accurate description nor prescribes further corrections .
references [ 15 - 18 ] , on the other hand , start from the more general atomic level description and then try to unveil physical properties from the underlying atomic description .
it is important to note that these works do not deal directly with nanoscale systems but rather focus on the effects of nanoscale inhomogeneities in _ bulk _ systems .
references [ 15 ] and [ 16 ] concern the elastic properties of grain boundaries .
these works define atomic - level elastic moduli as the second derivative of the energy associated with each atom with respect to strain and then go on to study the behavior of such moduli near grain boundaries .
the difficulty with this approach is that it requires a breakdown into individual contributions from each atom of the total energy of any system .
such an atomic energy is neither observable nor uniquely defined and therefore can not serve as an appropriate basis for theoretical understanding . although references [ 17 ] and [ 18 ] work from valid physical observables , the components of the force - constant matrix , these works focus on bulk - like or mesoscopic scale systems and fail for nanoscale systems for the reasons which we describe in this work . reference
[ 17 ] investigates nonlocal elastic constants on the mesoscopic scale and links them to the underlying atomic interactions .
it then proceeds to define an elastic constant for each atom and studies the behavior of these quantities near surfaces and grain boundaries .
reference [ 18 ] defines a bond frequency from the force - constant matrix from which it deduces the possibility of bond rupture during crack nucleation .
neither of the above works properly accounts for the poisson effect , which we show in sections [ sec - mlw ] , [ sec - fail ] and [ sec - mlwn ] to play a critical role in the elasticity of nanoscale systems .
moreover , straightforward generalization of these works to include this effect fail for the the same reasons as does the related approach which we describe in section [ sec - mlw ] .
the prime difficulty in the application of continuum theory to objects of nanometer cross - section is the loss of the ratio of the inter - atomic spacing to the cross - sectional dimension as a small parameter .
however , so long as the length of an object and the wavelength of the distortions considered both greatly exceed the inter - atomic spacing and the cross - sectional dimension , the object properly may be viewed as a one - dimensional continuum .
although we focus in this work on nanowires , the generalization of the discussion below to nanoscale systems of other dimensionality such as thin plates or nanoscopic objects is straightforward .
viewed as a linear continuum , the free energy per unit length @xmath0 of a nanowire is @xmath1 where @xmath2 is the linear strain of extension , @xmath3 is the radius of curvature and @xmath4 is the rate of twist of the torsion .
the coupling constant @xmath5 is the _ extensional rigidity _ , @xmath6 is the _ flexural rigidity _ and @xmath7 is the _
torsional rigidity_. unlike traditional bulk continuum concepts , the free - energy function eq .
( [ eqn : free ] ) is observable in principle and thus provides an unambiguous operational definition of the rigidities .
we avoid the use of traditional continuum concepts , such as the young s modulus and the cross - sectional area , because such concepts are neither uniquely nor well - defined for nanoscale systems . the rigidities in eq .
( [ eqn : free ] ) are related to the phonon frequencies through @xmath8 where @xmath9 is the frequency for either the longitudinal , transverse or rotational acoustic modes , respectively , @xmath10 is the linear atomic number density , @xmath11 is the wave vector and @xmath12 is the mass of a single atom .
( this work focuses on single species systems for simplicity . )
finally , @xmath13 is defined unambiguously as the mean rotational moment @xmath14 , where the sum ranges over all atoms in the cell , @xmath15 is the number of atoms per unit cell , the wire is assumed to run along the @xmath16-axis and the origin lies on the center line of the wire .
finally we note that although the rigidities in eq .
( [ eqn : free ] ) are well - defined , certain traditional continuum relations between them do not hold .
specifically , we will show below that the traditional continuum relationship between the extensional rigidity @xmath5 and the flexural rigidity @xmath6 fails on the nanoscale , similar to relations which have been recently used in the analysis of experiments@xcite .
one reason for breakdown of traditional continuum relations on the nanoscale is that the continuum perspective course grains away important fluctuations which occur over distances on the order of the inter - atomic spacing . to overcome this shortcoming
, we propose to coarse grain only on distances over which the underlying interatomic interactions vary , the decay length of the force - constant matrix . the straightforward approach to generate such
a theory is the `` method of long waves '' developed by born and huang@xcite , which is somewhat similar to the approaches which have been used previously to defects in bulk systems@xcite .
this section applies the method of long waves to nanoresonators .
the next section shows how , surprisingly , this and related approaches fail in the study of systems with free surfaces and therefore , in general , can not be used to describe nanoscale systems . in section [ sec - mlwn ] , we describe how to go beyond the straightforward application of the `` method of long waves '' in order to achieve a meaningful description .
we focus initially on longitudinal waves and , as noted in section [ sec - nr ] , choose our origin to lie on the center line of the wire , which we let run along the @xmath16-axis . for nanowires ,
the presence of surfaces breaks periodicity in the transverse directions resulting in a one - dimensional crystal with an extremely large unit cell of length @xmath17 , where @xmath15 and @xmath10 are as in section [ sec - nr ] . in all expressions below , boldfaced quantities
are @xmath18-dimensional and arrowed vector quantities are three - dimensional . finally , sums with greek indices range over atoms in the unit cell . to relate the rigidities to the dynamical matrix , we begin similarly to born and huang and choose to factor the bloch phases ( @xmath19 ) out of the representation of the phonon polarization vector @xmath20 , incorporating them into the definition of the dynamical matrix @xmath21 , so that the acoustic phonon polarization vectors are periodic across the cell boundaries .
this ensures a uniform description of the distribution of elastic energy along the axis of the wire . to generate a scalar equation for the phonon frequency , born and huang project the secular equation for the dynamical matrix , @xmath22 against the zeroth - order polarization vector @xmath23}$ ] . here
, however , to more symmetrically represent the distribution of elastic energy , we project against the full polarization vector @xmath20 . equating the frequency @xmath9 in eq .
( [ eqn : sec ] ) with the longitudinal frequency in eq . ( [ eqn : mlw ] ) gives @xmath24^{[2]}}{{\bf u}^\dagger { \bf u } } = \sum_{s , t=0}^{1 } \frac{(-1)^{s}}{n_c } \ : { \bf u}^{\dagger[s ] } \ : { \bf d}^{[2-s - t ] } \ : { \bf u}^{[t ] } , \label{eqn : omegasq}\ ] ] where we have expanded the numerator of the rayleigh quotient to second - order in powers of @xmath25 and where the @xmath26 sub - block of @xmath27}$ ] , which couples atoms @xmath28 and @xmath29 , is @xmath30}]_{\alpha\beta}=\frac{1}{n !
} \sum_{\vec r } { \phi_{\alpha\beta}(\vec r ) \left({\hat z } \cdot(\vec r + \vec \tau_\beta - \vec \tau_\alpha ) \right)^n}.\ ] ] here , @xmath31 is a lattice vector along the @xmath32 , @xmath33 is the @xmath26 sub - block of the force - constant matrix which couples atoms @xmath28 and @xmath29 located at positions @xmath34 and @xmath35 , respectively , and @xmath36 finally , substituting eq .
( [ eqn : dn ] ) and @xmath37}]_\alpha=\hat z$ ] into eq .
( [ eqn : omegasq ] ) , allows us to express @xmath5 as a sum over atoms ( @xmath28 ) in the unit cell and all atoms ( @xmath38 ) in the system , @xmath39}_\alpha } \cdot \phi_{\alpha\beta}(\vec{r})\cdot { \vec u^{[1]}_\beta } \nonumber\\ & & + { { \delta \vec z}_{\alpha\beta } } \cdot \phi_{\alpha\beta}(\vec{r})\cdot { \vec u^{[1]}_\beta } - { \vec u^{[1]}_\alpha } \cdot \phi_{\alpha\beta}(\vec{r})\cdot { { \delta \vec z}_{\alpha\beta } } \mbox{\large\}}. \label{eqn : atmod_sf}\end{aligned}\ ] ] here , @xmath40 , @xmath41}_{\alpha}$ ] is the first - order polarization vector and we refer to the @xmath42 as the `` atomic moduli '' . the atomic moduli as currently defined in eq . ( [ eqn : atmod_sf ] ) provide a useful microscopic analysis of elastic response in bulk systems which is similar in spirit to the decompositions used previously in the study of bulk material systems @xcite . to see that eq .
( [ eqn : atmod_sf ] ) indeed decomposes the overall elastic response of bulk systems into atomic contributions coarse - grained over distances on the order of the decay - length of the force - constant matrix , we note first that for infinite bulk systems , elastic waves are planar .
this implies that the first - order polarization vector ( @xmath43}_{\alpha}$ ] ) is uniform from primitive cell to primitive cell and thus depends only on the local environment of each atom .
next , we note that although the strain terms ( @xmath44 ) scale linearly with distance between atoms , the terms which contribute to the final result are bounded in range by the inter - atomic interactions ( @xmath45 ) .
thus , the atomic moduli depend only on the local atomic environment over distances which the decay of the force - constant matrix determines .
we now demonstrate through direct calculations that the approach outlined in the previous section gives unphysical results when applied to nanoresonators .
specifically , we study the behavior of @xmath46$]-oriented nanoresonators of silicon , which recent _ ab initio _
studies@xcite predict to undergo a size - dependent structural phase transition between the two structures in figure [ fig : wires ] at a cross - section of @xmath473 nm .
( the interested reader may refer to reference [ 21 ] for explicit details of the microscopic structure of these wires . ) initially , we work with the stillinger - weber inter - atomic potential@xcite , which suffices for the exploration of general nanoelastic phenomena and which allows study of cells with many thousands of atoms . later in the manuscript ( section [ sec - cbames ] ) we use the sawada tight - binding model@xcite with modifications proposed by kohyama@xcite to explore the correlation between our local approach and the underlying electronic structure . for all calculations below , we fully relax the atomic coordinates , the periodicity of the wire and , need be , the electronic structure .
finally , we employ periodic boundary conditions along the @xmath16-direction .
figure [ fig : ecomp]a shows that the atomic moduli @xmath42 predicted for nanowires using the straightforward approach of eq .
( [ eqn : atmod_sf ] ) are unphysical in that they depend upon the macroscopic dimensions of the system and not simply on the local environment of each atom .
in particular , the atomic moduli on the surface grow linearly with the diameter of the wire and the moduli in the center of the wire fail to approach the expected bulk limit , @xmath48 , where @xmath49 is the young s modulus in bulk and @xmath50 is the number density in bulk .
( note that these two effects are interelated , as the moduli must sum to give the macroscopic value in the bulk limit . )
[ 0.375 ] for @xmath51 nanowires of varying diameter : ( a ) straightforward theory ( eq . ( [ eqn : atmod_sf ] ) ) and ( b ) new theory ( eq . ( [ eqn : atmod ] ) ) .
the insets denote the approximate diameters of the wires . the value of the atomic modulus ( eq . ( [ eqn : atmod_sf ] ) or eq .
( [ eqn : atmod ] ) ) is along the ordinate and radial distance of the atom from the center line is along the abscissa.,title="fig : " ] + radial distance [ @xmath52 although one has some freedom in choosing the terms used in the perturbation expansion eqs .
( [ eqn : sec ] ) and ( [ eqn : omegasq ] ) , for example to project eq .
( [ eqn : sec ] ) against @xmath53}}$ ] instead of @xmath20 , all such expansions will lead to similar linear scaling along the surface of the wire and approach an incorrect value at the center of the wire .
thus , straightforward application of the method of long waves fails to result in a local , and hence physically meaningful , description of elastic response in nanoscale systems , as will straightforward variations thereon such as those in references [ 17 ] and [ 18 ] .
to cure the difficulties uncovered in the previous section , we proceed by first identifying the cause of the pathological behavior and then exploiting the freedom in eq .
( [ eqn : atmod_sf ] ) to remove this pathology .
the failure of the straightforward approach arises from the fact that elastic waves in nanoresonators , or any system with free surfaces , are not strictly planar .
in particular , the poisson effect , which the first order polarization vector @xmath54}$ ] contains , causes each atom to displace by an amount in direct proportion to its distance from the center line of the system .
( [ eqn : atmod_sf ] ) then leads directly to linear scaling of the atomic moduli at the surfaces of the wire .
defining an atomic elastic reponse dependent soley on the local enivornment requires separation of extensive elastic effects from intensive nanoscopic effects . to seperate the extenive motion in the first order polarization vector from that of its intensive motion , we define the atomic displacements ( @xmath55 ) as the intensive nanoscopic motions , @xmath56}_{\alpha } - \left ( -\sigma_x \hat x \hat x \cdot - \sigma_y \hat y \hat y \cdot \right ) \vec \tau_\alpha , \label{eqn : decomp}\end{aligned}\ ] ] where @xmath57 are the poisson ratios
there are three logical choices for the poisson ratios : some sort of local atomic definition , an overall average for the wire , or the bulk values .
we choose to use bulk poisson ratios for a number of reasons .
first , a locally varying definition makes it impossible to exploit the continuous rotational and translational symmetries in the dynamical matrix , which we find necessary to employ below in constructing atomic moduli with local behavior .
second , only by employing the bulk ( rather than average ) poisson ratios do we find a definition which approaches the appropriate bulk value in the centers of wires of finite width .
finally , we note that we always have the freedom of working with bulk poisson ratios because any motion along the surface in addition to that resulting from bulk poisson effect will not scale extensively with the diameter of the wire and can therefore be incorporated into the intensive atomic . after making the decomposition in eq .
( [ eqn : decomp ] ) , we next employ the continuous rotational and translational symmetries of the dynamical matrix to eliminate all extensive dependencies in eq .
( [ eqn : atmod_sf ] ) .
appendix [ sec - app ] outlines the procedure for doing this , which then transforms eq .
( [ eqn : atmod_sf ] ) into @xmath58 where @xmath59 represents the _ total _ strain between atoms @xmath28 and @xmath29 , @xmath60 and @xmath61 renormalizes as @xmath62 where @xmath63 , is a diagonal @xmath26 matrix with elements @xmath64 , @xmath65 and @xmath66 , respectively and @xmath67 is the kronecker delta .
this new construction ensures that the modulus of each atom depends only upon its local atomic environment because @xmath68 no longer includes extensive motions and , although @xmath69 still depends on relative atomic distances , the renormalized @xmath70 decays as @xmath71 does .
thus , it is now the range of the force - constant matrix which controls the size of the neighborhood upon which each atomic modulus can depend . therefore , the moduli of atoms in the interior now _ must _ correspond to the expected bulk value , the moduli of the atoms on the surface now _ can not _ depend upon the extent of the system , and
the resulting description is physically meaningful .
figure [ fig : ecomp]b illustrates the success of eq .
( [ eqn : atmod ] ) .
the fact that decomposition of elastic response into atomic level contributions is not unique raises questions as to the physical meaning of such a decomposition .
the new decomposition eq .
( [ eqn : atmod ] ) is the first which remains dependent only upon local environment for systems with free surfaces .
any other definition _ which respects locality _ can only redistribute portions of each atom s modulus among other atoms within a region of extent comparable to the decay of the force - constant matrix .
any sum over such a region of the moduli will always be nearly the same .
therefore , coarse grained over such regions , properly localized atomic moduli become physically meaningful .
section [ sec - trans ] and appendix [ sec - appb ] demonstrate this explicitly by comparing the predictions for flexion from either properly localized moduli or straightforwardly defined moduli , respectively . to further demonstrate that alternate local definitions are equivalent in this coarse - grained sense , we have explored alternate local constructions . in particular , while our present construction takes care to employ the continuous symmetries of the dynamical matrix in such a way so as to respect symmetry among the @xmath72 cartesian coordinates , we have also repeated the construction while treating the @xmath73 coordinates symmetrically but not the @xmath16 coordinate and have found nearly identical results for all of the applications below .
we close this section with a brief description of how the above approach extends to any system of nanometer dimensions . to be considered small in this context
, a dimension must be much smaller than the wavelength of the distortions considered .
this manuscript focuses on nanowires , systems with two small dimensions . for a system with one small dimension , for instance a plate with nanometer thickness , the above approach develops in the same way with the one minor change that the definition of @xmath68 ( eq .
( [ eqn : decomp ] ) involves the poisson effect only in the one small dimension . for an object with three small dimensions , for instance adiabatic loading of a nano - object
, the nature of the poisson effect depends upon the mode of loading , and the system should treated as either of the above cases accordingly .
finally , objects of no small dimension , and hence no poisson effect , can be described straightforwardly as bulk - like using eq .
( [ eqn : atmod_sf ] ) .
to establish that our new atomic moduli are not mere convenient mathematical constructions but are physically meaningful , we now consider their transferability to phenomena not considered in their original construction . in particular , we consider flexion , where the elastic distortion is no longer homogeneous throughout the cross - section .
if our atomic moduli are indeed a measure of the local elastic response , then under flexion the free energy per unit length will take the form @xmath74 where @xmath75 is a measure of the longitudinal strain which atom @xmath28 experiences .
for this form to be sensible , the diameter @xmath76 of the wire must not become comparable to the range of the force - constant matrix so that the atoms which contribute to each @xmath42 all experience similar strains @xmath75 . within continuum theory , uniform flexion with radius of curvature
@xmath3 corresponds to a longitudinal strain which varies linearly across the wire , @xmath77 .
( this holds to better than to two parts in @xmath78 for all wires in our study . )
this would then predict a flexural rigidity of @xmath79 figure [ fig : eandf]a shows the fractional error ( @xmath80 ) @xmath81 in predicting the flexural rigidity from eq .
( [ eqn : atmof ] ) , where @xmath6 is determined directly through numerical calculations .
the figure shows that these errors are indeed quite small . note that use of the straightforward definition in eq .
( [ eqn : atmod_sf ] ) with its unusually scaling surface moduli leads to invalid predictions for flexion .
( appendix [ sec - appb ] shows this directly through scaling arguments . ) as a result , to have _ predictive _ power , definitions of atomic moduli must properly account for the poisson effect . to demonstrate that our new approach has greater predictive power than traditional continuum approaches , the figure also shows the fractional error ( @xmath82 ) in predicting the flexural rigidity when using the traditional continuum relation @xmath83 where @xmath84 defines the mean bending moment , which we define unambiguously as @xmath85 the linear behavior of the fractional error @xmath82 as a function of @xmath86 indicates that continuum theory does not properly account for surface effects , a result of the fact that flexion places a larger emphasis on the surface than does extension .
the dramatic improvement from the use of eq .
( [ eqn : atmof ] ) arises because the atomic moduli place proper emphasis on the surface and on the interior , as they properly treat each atomic environment locally .
the atomic moduli therefore properly account for elastic fluctuations along the cross - section of the wire which are on scales too small for traditional continuum theories to capture .
finally , we note that the only appreciable error within the new framework occurs for the smallest wires ( d@xmath87 nm ) . at this point , the cross - sectional dimension becomes comparable to the range of the force - constant matrix , and eq .
( [ eqn : free_at ] ) represents an improper use of the physical concept of atomic moduli . as described above ( second from last paragraph in section [ sec - mlwn ] ) , only sums of atomic moduli over regions of extent comparable to the range of the force - constant matrix carry physical meaning . any sum sensitive to variations over shorter scales , as is eq .
( [ eqn : free_at ] ) when limited to wires narrower than the range of the force - constant matrix , can not be depended upon to lead to meaningful results .
this underscores the fact that properly construed atomic moduli are not truly atomic - level quantities but a concept coarse grained over the range of the force constant matrix .
as we have seen , however , this coarse - graining is on scales significantly smaller than those captured by traditional continuum theory .
we now apply the above defined atomic moduli to derive a new relation for the extensional modulus in terms of the flexural modulus and other experimentally accessible observables , a relation often needed in experimental analyses @xcite .
we then show that the new relation is much more accurate than the standard continuum - theory based relation currently employed in experimental analyses .
finally , we employ our concept of atomic moduli to provide quantitative insight into the improvement of our new relation over the traditional continuum relation . the basis for the following analysis is the fact that eq .
( [ eqn : atmof ] ) gives a very good estimate of the true flexural modulus , as figure [ fig : eandf]a confirms .
using the exact relation for @xmath5 , eqs .
( [ eqn : thermofree ] ) and ( [ eqn : atmod ] ) , and eq . ( [ eqn : atmof ] ) , one derives the following leading - order prediction @xmath88 for the extensional modulus , @xmath89 with a predicted error @xmath90 of @xmath91 + \left[\langle e_s \rangle_{x^{2 } } - \langle e_s \rangle\right ] \right\}. \label{eqn : error}\end{aligned}\ ] ] here , @xmath92 is the number of `` surface '' atoms , defined as those for which @xmath42 differs significantly from the limiting bulk value @xmath93 , @xmath94 is the inertia weighted average surface moduli @xmath95 with the sums ( @xmath96 ) ranging over `` surface '' atoms , @xmath97 is the average surface moduli @xmath98 and @xmath99 where @xmath100 implies sums over all atoms in the unit cell .
this result holds for any division of the atoms into `` surface '' and `` bulk '' to the extent that each `` bulk '' atom has atomic modulus @xmath93 .
figure [ fig : eandf]b shows the relative error @xmath101 , @xmath102 between the extensional modulus @xmath5 determined directly from numerical calculation and as determined from our new relation , eq .
( [ eqn : atefromf ] ) .
( note that @xmath103 exactly when the moduli prediction @xmath104 holds . ) as the relevant point of comparison , the figure also shows the relative error @xmath105 , @xmath106 associated with the traditional continuum result @xmath107 hence , eq .
( [ eqn : atefromf ] ) is much more accurate than the standard continuum result eq .
( [ eqn : cnte ] ) . to understand the improvement of the new relation , eqs .
( [ eqn : thermofree ] ) and ( [ eqn : atmof ] ) may also be combined to yield a prediction for the fractional error in the traditional continuum analysis , @xmath108 + \left[\langle e_s \rangle_{x^{2 } } - \langle e_s \rangle\right ] \right\ } , \label{eqn : errorct}\ ] ] which takes _
precisely _ the same form as eq .
( [ eqn : error ] ) except for the change in the prefactor in the first term from @xmath109 to @xmath110 . for continuous wires of homogeneous circular or regular polygonal cross - section ,
we have exactly @xmath111 .
thus , generally we expect @xmath109 to be close to zero and @xmath112 to be close to unity .
we now note that the term in the first set of square brackets ( @xmath113 $ ] ) in both eq .
( [ eqn : error ] ) and eq .
( [ eqn : errorct ] ) is an average difference between surface and bulk atoms and , therefore , is generally much larger than the term in the second set of square brackets @xmath114 $ ] , which is the difference between two differently weighted averages over the surface atoms .
thus the larger term nearly vanishes in our new relation , eq .
( [ eqn : error ] ) , but not in the traditional continuum relation , eq .
( [ eqn : errorct ] ) . from this analysis
, we see that the reason why the traditional continuum relation has larger errors is that it does not properly differentiate between the local surface and bulk environments .
the atomic moduli also lead to a quick , intuitive argument to understand the improvement of eq .
( [ eqn : atefromf ] ) over eq .
( [ eqn : cnte ] ) . from our results
we know that the surface moduli can be quite different than those of the bulk .
it is also known that flexion places larger emphasis on the surface than does extension .
if the average surface modulus is less / more than that of the bulk then the first term in eq .
( [ eqn : atefromf ] ) ( @xmath115 ) will underestimate / overestimate the extensional rigidity , while the second term will overestimate / underestimate it .
therefore , errors will tend to cancel in the average of the two .
we now explore the local atomic moduli as a link between local elastic properties and the underlying electronic structure . to do this , we have calculated the atomic moduli , eq .
( [ eqn : atmod ] ) , for both the @xmath51 and @xmath116 structures , using the sawada @xcite tight - binding model with modifications proposed by kohyama@xcite .
we have studied various wires , all of which give similar results . for brevity ,
we here only report on wires with cross - sectional diameter @xmath117 nm . the supercell of our calculation is four bulk cubic lattice constants long in the periodic direction , and hence sampling the brillouin zone at the @xmath118 point in the electronic structure calculations is more than sufficient .
the left panels of figures [ fig : tbc2x2 ] and [ fig : tb2x1 ] show the resulting atomic moduli , with values color coded so that yellow corresponds to the bulk value ( 17.1 ev / atom ) .
the figure shows that deviations from this value concentrate near the surface in patterns characteristic of the structure of the wire .
moduli near the surface fluctuate widely , ranging from 6 - 27 ev / atom for the @xmath51 structure and from 4 - 26 ev / atom for the @xmath119 structure . to allow comparison with the underlying electronic structure , the right panels in the figures display the valence charge densities from the tight - binding calculation projected ( integrated along the wire axis ) onto the cross - section of the wire .
( to compute the electron density from the tight - binding coefficients , we employed orbitals from a density functional calculation of the silicon atom . )
the figures display the electron densities using a color map similar to that employed for the atomic moduli .
intriguingly , there is an apparent correlation between the values of the atomic moduli and the underlying electron density . in particular , large / small atomic moduli correlate with regions of large / small electron density in figure [ fig : tbc2x2 ] , indicating that charge distribution along the surface of these wires greatly affects the local elastic properties and thereby the overall elastic response , particularly to flexion which emphasizes surface effects .
figure [ fig : tb2x1 ] exhibits a similar correlation , but not as pronounced . note that , in this figure , coincidence of red atomic moduli just under the surfaces of the first layer of atoms correlates with red charge densities in the same location .
unlike in the classical potential case , where properly defined surface moduli are systematically lower or equal to the bulk value due to decrease in the number of bonds ( figure [ fig : ecomp]b ) , we find that in a quantum model , surface moduli may even greatly exceed the bulk value due to changes in local charge density which can enhance the mechanical strength of bonds ( figures [ fig : tbc2x2 ] and [ fig : tb2x1 ] . )
this contrast underscores both the importance of considering contributions of the electronic structure to mechanical response and the need for a definition of atomic moduli which can be computed from physical observables obtainable from electronic structure calculations .
the ultimate use of atomic level moduli is to understand mechanical response .
table [ table : tab1 ] compares the errors from both continuum theory and the use of atomic - level moduli in predicting mechanical response for the two wires under consideration in this section .
the quantities compared exactly parallel those of the previous section .
the first two columns of the table consider prediction of flexural response from continuum theory and our atomic moduli , respectively , and the second two columns consider prediction of extensional response from flexural response using either the traditional continuum relation or our new relation , respectively .
the table shows that the new relations , eqs .
( [ eqn : atmof ] ) and ( [ eqn : atefromf ] ) are again very accurate .
the table also shows that these predictions are superior to the corresponding continuum results , eqs .
( [ eqn : cntf ] ) and ( [ eqn : cnte ] ) , respectively . .comparison of errors between the traditional continuum theory and the atomic moduli description when predicting flexural response through eqs ( [ eqn : cntf ] ) and ( [ eqn : atmof ] ) ( first and second columns , respectively ) and when inferring extensional response from flexural response through eqs .
( [ eqn : cnte ] ) and ( [ eqn : atefromf ] ) ( third and fourth columns , respectively ) . [ cols="^,^,^,^,^,^",options="header " , ] interestingly , the continuum predictions for the @xmath51 wire are fairly reliable .
the atomic moduli provide an avenue for understanding this as well .
fluctuations in the moduli in the @xmath51 wires are localized and hence average out over regions of extent comparable to the decay of the force - constant matrix . moreover , in this particular case they tend to average to values close to that expected of the bulk . without meaningful fluctuations on the length scales of the decay of the force - constant matrix , we expect continuum theory to perform well for this wire . in contrast , the @xmath120 wire exhibits much more systematic variations in the moduli .
the outermost surface atoms have a consistent and significantly reduced modulus , and there is also a clear significant and systematic variation in the moduli throughout the cross - section of the wire .
because this second wire does exhibit meaningful fluctuation over distances comparable to the decay range of the force constant matrix , we expect traditional continuum relations to give particularly poor results , underscoring the importance of the local atomic - level moduli description .
this manuscript presents the first definition of atomic - level elastic moduli for nanoscale systems which are defined in terms of physical observables , correctly sum to give the exact overall elastic response and depend only on the local environment of each atom .
although these moduli are not necessarily uniquely defined , their sum over regions of extent comparable to the range of the force constant matrix is physically meaningful and may be used to make accurate predictions of mechanical response .
the moduli resulting from our formulation transfer to different modes of strain and correctly account for elastic fluctuations on the nanoscale .
they also lead to a quantitative understanding of when traditional continuum relations breakdown and how to properly correct them properly .
specifically , we demonstrated a more accurate method for relating extensional and flexural properties .
these moduli provide a clear and natural method for distinguishing _ mechanically _ between `` surface '' atoms and `` bulk '' atoms and give insight into the correlation between the local mechanical response and the underlying electronic structure .
finally , these moduli allow the identification of which atomic arrangements lead to more pliant or stiffer response opening the possibility of their use as a tool to aid in the rational design of nanostructures with specific mechanical properties .
segall acknowledges support of his department of physics fellowship and the department s continuing support .
s. ismail - beigi acknowledges support of the mit mrsec program of the national science foundation under award number dmr 9400334 .
computational support was provided by the mit xolas prototype sun cluster and the cornell center for materials research shared experimental facilities , supported through the nsf mrsec program dmr9632275 .
this appendix outlines the use of rotational and translation symmetries of the force constant matrix to reformulate the ill - defined atomic moduli in eq .
( [ eqn : atmod_sf ] ) into the well - defined form in eq .
( [ eqn : atmod ] ) . from continuous translational symmetry ,
all force - constant matrices obey @xmath121 for any vector @xmath122 that is the constant vector for all atoms in the unit cell .
moreover , the @xmath123 rotational symmetries imply @xmath124 where the @xmath125 correspond to one of the @xmath72 cartesian coordinates , @xmath126 is the @xmath127 component of @xmath128 and @xmath129 corresponds to the @xmath130 component of the position vector @xmath131 . the ill - defined atomic moduli , eqs .
( [ eqn : thermofree])-([eqn : atmod_sf ] ) , contain divergent terms which are in one of the following two forms : @xmath132 or @xmath133 all of these terms scale linearly along the surface of the wire and give rise to the linear scaling of the surface moduli with system size evident in figure [ fig : ecomp]a . from eq .
( [ eqn : tran ] ) , one can set eq .
( [ eqn : firstterm ] ) equal to @xmath134 using both eqs .
( [ eqn : tran ] ) and ( [ eqn : rot ] ) , one can set eq .
( [ eqn : secondterm ] ) equal to @xmath135\sigma_t .
\label{equ : secondnew}\end{aligned}\ ] ] the above terms now only depend on relative distances over a range controlled by the decay of the force - constant matrix and therefore no longer scale linearly with the size of the system . combining the above transformations with the separation of the extensive motion from the intensive motion in the first - order polarization vector , eq .
( [ eqn : decomp ] ) , then results in the well defined form eq .
( [ eqn : atmod ] ) .
we now demonstrate that although the moduli defined in eq .
( [ eqn : atmod_sf ] ) indeed sum to give the correct overall extensional response , their unphysical scaling properties lead to invalid physical predictions when used in other contexts and , therefore , that they are ill - defined . in particular , we use simple scaling arguments to prove that these ill - defined moduli give a nonnegligible error in predicting the flexural rigidity in the continuum limit . consider a circular wire whose radius @xmath3 is sufficiently large such that continuum theory applies . from figure
[ fig : ecomp]a we see that the surface moduli defined by eq .
( [ eqn : atmod_sf ] ) scale linearly with @xmath3 .
because of this , and in order to arrive at an analytic result for the error in predicting the flexural rigidity , we can assume , to a good approximation , that all surface moduli @xmath136 are proportional to their radial distance @xmath137 from the center of the wire , @xmath138 here , @xmath139 is the same for all surface atoms .
next , we define @xmath140 as the inner radius such that all atoms at position @xmath141 are in the bulk and all atoms with positions @xmath142 are on the surface .
finally , we define @xmath143 .
because of the facts that the surface moduli scale linearly with the system and that the sum of all atomic moduli must equal the extensional rigidity , we conclude that the moduli in the bulk region of the wire , derived from eq .
( [ eqn : atmod_sf ] ) , @xmath144 can not equal the average value of the bulk material @xmath93 .
( figure [ fig : ecomp]a also evidences this behavior . )
the two above facts imply that the following equality must hold , @xmath145 or to leading order in @xmath146 , @xmath147 in the continuum limit , the traditional continuum relation eq .
( [ eqn : cntf ] ) holds , and therefore the fractional error in predicting the flexural rigidity from the ill - defined atomic moduli eq .
( [ eqn : atmod_sf ] ) is equal to @xmath148 or to leading order in @xmath146 , @xmath149 using eq .
( [ eqn : ext ] ) to solve for @xmath150 , eq .
( [ eqn : flx ] ) becomes @xmath151 therefore , in the continuum limit , the ill - defined moduli do not approach the correct result and thus give a prediction which is even worse than that of traditional continuum theory . | we initiate the development of a theory of the elasticity of nanoscale objects based upon new physical concepts which remain properly defined on the nanoscale .
this theory provides a powerful way of understanding nanoscale elasticity in terms of local group contributions and gives insight into the breakdown of standard continuum relations .
we also give two applications . in the first ,
we show how to use the theory to derive a new relation between the bending and stretching properties of nanomechanical resonators and to prove that it is much more accurate than the continuum - based relations currently employed in present experimental analyses . in the second
, we use the new approach to link features of the underlining electronic structure to the elastic response of a silicon nanoresonator . |
the study of chemical reaction networks is an essential component of the emerging fields of systems and synthetic biology @xcite .
traditionally chemical reaction networks were modeled in the deterministic setting , where the dynamics is represented by a set of ordinary differential equations ( odes ) or partial differential equations ( pdes ) . in the study of intracellular chemical reactions , some chemical species are present in low copy numbers .
since the behavior of individual molecules is best described by a stochastic process , in the low molecular copy number regime , the copy numbers of the molecular species itself is better modeled by a stochastic process than by odes @xcite . only in the limit of large molecular copy numbers ,
one expects the deterministic models to be accurate @xcite .
while our work in this paper is focused on biochemical reaction networks as primary examples , we emphasize that the mathematical framework of reaction networks can also be used to describe a wide range of other phenomena in fields such as epidemiology @xcite and ecology @xcite .
suppose @xmath0 is a parameter ( like ambient temperature , cell - volume , atp concentration etc . ) that influences the rate of firing of reactions .
let @xmath1 be the @xmath0-dependent markov process representing the reaction dynamics , and suppose that for some real - valued function @xmath2 and observation time @xmath3 , our output of interest is @xmath4 .
this output is a random variable and we are interested in determining the sensitivity of its expectation @xmath5 w.r.t .
infinitesimal changes in the parameter @xmath0 .
we define this sensitivity value , denoted by @xmath6 , as the partial derivative @xmath7 determining these parametric - sensitivity values are useful in many applications , such as , understanding network design and its robustness properties @xcite , identifying critical reaction components , inferring model parameters @xcite and fine - tuning a system s behavior @xcite . generally the sensitivities of the form can not be directly evaluated , but instead , they need to be estimated with monte carlo simulations of the dynamics @xmath1 .
many methods have been developed for this task @xcite , but they all rely on exact simulations of @xmath1 that can be performed using schemes such as gillespie s _ stochastic simulation algorithm _ ( ssa ) @xcite .
this severely constrains the computational feasibility of these sensitivity estimation methods because these exact simulations become highly impractical if the rate of occurrence of reactions is high @xcite , which is typically the case .
the main difficulty is that that exact simulation schemes keep track of each reaction event which is very time - consuming . to avoid this problem ,
tau - leaping methods have been developed that proceed by combining many reaction - firings over small time intervals @xcite in such a way that insignificant stochastic effects are discarded in favor of computational speed .
tau - leap methods haven been shown to produce good approximations of the reaction dynamics , at a small fraction of the computational cost of exact simulations @xcite .
their accuracy has also been investigated theoretically in many papers @xcite .
our goal in this paper is to develop a method that leverages the accuracy and computational efficiency of tau - leap methods for the purpose of estimating sensitivity values of the form .
our method , called _ tau bernoulli path algorithm _ ( @xmath8bpa ) , works with any underlying tau - leap simulation scheme and it is based on a novel integral representation of parameter sensitivity @xmath6 that we derive in this paper .
we provide computational examples that show that using @xmath8bpa we can often _ trade - off _ a small amount of bias for large savings in the overall computational costs for sensitivity estimation .
we prove that the bias incurred by @xmath8bpa depends on the step - size in the same way as the bias of the tau - leap scheme chosen for simulations .
moreover if we substitute the tau - leap simulations in @xmath8bpa with the exact ssa generated simulations , then we recover a new unbiased method for sensitivity estimation that is similar to the method in @xcite . for the sake of comparison
, we also propose the _ tau - leap versions _ of certain commonly used finite - difference estimators ( see @xcite ) that approximate the infinitesimal derivative in by a finite - difference ( see ) .
such estimators are computationally faster than @xmath8bpa but they suffer from two sources of bias ( finite - differencing and tau - leap approximations ) unlike @xmath8bpa which only incurs bias from the latter source .
our examples indicate that the biases from these two sources often pile on top of each other , making these finite - difference estimators much more inaccurate in comparison to @xmath8bpa .
this paper is organized as follows . in section [ sec : prelim ] we formally describe the stochastic model for reaction dynamics and the sensitivity estimation problem .
we also discuss the existing sensitivity estimation methods , introduce the tau - leap simulation schemes and explain the rationale for using such simulations in sensitivity estimation .
the main results of this paper including the description of our novel tau - leap sensitivity estimation method @xmath8bpa is contained in section [ sec : mainres ] . in section [ sec : num_examples ] we provide computational examples to compare our method with other methods and finally in section [ sec : conc ] we conclude and provide directions for future research .
consider a reaction network with @xmath9 species and @xmath10 reactions .
we describe its kinetics by a continuous time markov process whose state at any time is a vector in the non - negative integer orthant @xmath11 comprising of the molecular counts of all the @xmath9 species .
the state evolves due to transitions caused by the firing of reactions .
we suppose that when the state is @xmath12 , the rate of firing of the @xmath13-th reaction is given by the _ propensity _ function @xmath14 and the corresponding state - displacement is denoted by the stoichiometric vector @xmath15 .
there are several ways to represent the markov process @xmath16 that describes the reaction kinetics under these assumptions .
we can specify the generator ( see chapter 4 in @xcite ) of this process by the operator @xmath17 where @xmath18 is any bounded real - valued function on @xmath11 .
alternatively we can express the markov process directly by its random time - change representation ( see chapter 7 in @xcite ) @xmath19 where @xmath20 is a family of independent unit rate poisson processes .
since the process @xmath16 is markovian , it can be equivalently specified by writing the _ kolmogorov forward equation _ for the evolution of its probability distribution @xmath21 at each state @xmath12 : @xmath22 this set of _ coupled _ ordinary differential equations ( odes ) is termed as the _ chemical master equation _ ( cme ) in the biological literature @xcite .
as the number of odes in this set is typically infinite , the cme is nearly impossible to solve directly , except in very restrictive cases .
a common strategy is to estimate its solution using _ pathwise _ simulations of the process @xmath23 using monte carlo schemes such as gillespie s ssa @xcite , the _ next reaction method _ @xcite , the _ modified next reaction method _ @xcite , and so on .
while these schemes are easy to implement , they become computationally infeasible for even moderately large nertworks , because they account for each and every reaction event . to resolve this issue ,
tau - leaping methods have been developed which will be described in greater detail in section [ tau_leap_methods ] .
we now assume that each propensity function @xmath24 depends on a real - valued system parameter @xmath0 . to emphasize this dependence we write the rate of firing of the @xmath13-th reaction at state @xmath12 as @xmath25 instead of @xmath14 .
let @xmath26 be the markov process representing the reaction dynamics with these parameter - dependent propensity functions . as stated in section [ sec : intro ] , for a function @xmath27 and an observation time @xmath28 , our goal is to determine the sensitivity value @xmath29 defined by .
this value can not be computed directly for most examples of interest and so we need to find ways of estimating it using simulations of the process @xmath26 .
such simulation - based sensitivity estimation methods work by specifying the construction of a random variable @xmath30 whose expected value is close " to the true sensitivity value @xmath6 , i.e. @xmath31 once such a construction is available , a large number ( say @xmath32 ) of independent realizations @xmath33 of this random variable @xmath30 are obtained and the sensitivity is estimated by computing their empirical mean @xmath34 as @xmath35 this estimator @xmath34 is a random variable with mean and variance @xmath36 respectively . for a large sample size @xmath32 ,
the distribution of @xmath37 is approximately gaussian with mean @xmath38 and variance @xmath39 , due to the central limit theorem .
the _ standard deviation _
@xmath40 , which is the square root of the variance , measures the _ statistical spread _ of the estimator @xmath41 , that is inversely proportional to its _
statistical precision_. to safeguard against the possibility of different sample realizations producing very different estimates , the sample size @xmath32 must be large enough to ensure that @xmath40 is small relative to @xmath38 . if this holds , then @xmath34 is a reliable estimator for the true sensitivity value @xmath6 because it is very likely to assume a value close to its mean @xmath42 which in turn is close to @xmath6 ( see ) . in practice
both @xmath38 and @xmath40 are unknown , but we can estimate them as @xmath43 and @xmath44 the performance of any sensitivity estimation method ( say @xmath45 ) depends on the following three key metrics that are based on the properties of random variable @xmath30 : 1 . the _ bias
_ @xmath46 , which is the error incurred by the approximation .
the _ variance _ @xmath47 of random variable @xmath30 .
3 . the _ computational cost _
@xmath48 of generating one sample of @xmath30 the bias @xmath49 can be positive or negative , and its absolute value @xmath50 can be seen as the upper - bound on the statistical accuracy that can be achieved with method @xmath45 by increasing the sample size @xmath32 @xcite . as mentioned before , the square - root of the variance @xmath51 measures the statistical precision of the method @xmath45 which determines the number of samples @xmath32 that is needed to produce a reliable estimate .
the computational cost @xmath48 is the cpu time required for constructing one realization of @xmath30 , and hence the total cost of the estimation procedure with @xmath32 samples is @xmath52 .
the goal of a good estimation method is to simultaneously minimize the three quantities @xmath50 , @xmath51 and @xmath48 .
this creates various conflicts and trade - offs among the existing sensitivity estimation methods as we now discuss .
a sensitivity estimation method @xmath45 is called _ biased _ if @xmath53 .
the most commonly used biased methods are the _ finite - difference schemes _ which approximate the infinitesimal derivative in the definition of parameter sensitivity ( see ) by a finite - difference of the form @xmath54 for a small perturbation @xmath18 .
the processes @xmath55 and @xmath56 represent the markovian reaction dynamics with values of the sensitive parameter set to @xmath0 and @xmath57 respectively .
these two processes can be simulated independently @xcite but it is generally better to couple them in order to reduce the variance of the associated estimator .
we briefly review two such coupling strategies called _ common reaction paths _ ( crp )
@xcite and _ coupled finite differences _
( cfd ) @xcite . replacing each @xmath58 by @xmath59 in we obtain the random time - change representation of the @xmath0-dependent process
@xmath55 as @xmath60 where @xmath61 is the initial state and @xmath62 is a family of independent unit rate poisson processes . here each @xmath63 captures the intermittency in the firing times of reaction @xmath13 .
one can regard @xmath63 as specifying the _ reaction path _ for the @xmath13-th reaction . in crp , the process @xmath64 is coupled to process @xmath55 by assigning it the same reaction path for each reaction . in other words , the random time - change representation @xmath56 is given by @xmath65 where the @xmath63-s are the same as in .
in cfd , the processes @xmath55 and @xmath66 are also coupled by their random time - change representations , in such a way that they share many common reactions .
this is achieved using the split - coupling defined as @xmath67 ds\right ) \zeta_k \\
\label{cfd2 } \
\textnormal { and } \qquad x_{\theta + h } ( t ) & = x_0 + \sum_{k = 1}^k y_k \left ( \int_{0}^t \lambda_k ( x_\theta(s ) , \theta ) \wedge \lambda_k ( x_{\theta + h}(s ) , \theta + h ) ds\right ) \zeta_k \notag \\ & + \sum_{k = 1}^k y^{(2)}_k \left ( \int_{0}^t \left [ \lambda_k ( x_{\theta+h}(s ) , \theta+h ) - \lambda_k ( x_\theta(s ) , \theta ) \wedge \lambda_k ( x_{\theta + h}(s ) , \theta + h ) \right ] ds\right ) \zeta_k , \end{aligned}\ ] ] where @xmath68 and @xmath69 is again family of independent unit rate poisson processes .
the finite - difference approximation for the true sensitivity value can be expressed as the expectation @xmath70 of the following random variable @xmath71 the three metrics ( bias , variance and computational cost ) based on this random variable define the performance of crp and cfd . since both these methods estimate the same quantity @xmath72 , they have the same bias ( i.e. @xmath73 ) . however in many cases it is found that the cfd coupling is _
tighter _ than the crp coupling , resulting in a lower variance of @xmath74 ( i.e. @xmath75 ) ( see @xcite ) . for each realization of @xmath74 ,
both crp and cfd require simulation of a coupled trajectory @xmath76 in the time interval @xmath77 $ ] .
the computational costs of such a simulation is roughly @xmath78 , where @xmath79 is the cost of _ exactly _ simulating the process @xmath55 using gillespie s ssa @xcite or a similar method .
is usually smaller for cfd in comparison to crp ( i.e. @xmath80 ) , because the cfd coupling is such that if @xmath81 for some @xmath82 , then this equality will hold for the remaining time - interval @xmath83 $ ] , allowing us to directly set @xmath84 without completing the simulation in the interval @xmath83 $ ] . ]
finite - difference schemes introduce a bias in the estimate whose size is proportional to the perturbation value @xmath18 ( i.e. @xmath85 ) , but the constant of proportionality can be quite large in many cases , leading to significant errors even for small values of @xmath18 @xcite .
unfortunately we can not circumvent this problem by choosing a very small @xmath18 because the variance is proportional to @xmath86 ( i.e. @xmath87 ) .
therefore if a very small @xmath18 is selected , the variance will be enormous and the sample - size @xmath32 required to produce a statistically precise estimate will be very large , imposing a heavy computational burden on the estimation procedure @xcite .
this trade - off between bias and variance is the main drawback of finite - difference schemes and there does not exist a strategy for selecting @xmath18 that optimally balances these two quantities . note that unlike bias and variance , the computational cost of generating a sample ( i.e. @xmath88 or @xmath89 ) does not change significantly with @xmath18 , thereby ensuring that as @xmath18 varies , the total computational burden varies linearly with the required number of samples @xmath32 .
apart from finite - difference schemes , there exists another biased method @xcite that approximates the sensitivity value as @xmath90 which is obtained by replacing the output @xmath91 with its _ average _ value in the time interval @xmath92 $ ] .
such an averaging is performed to _ regularize _ the sensitivity value , so that it can be estimated using _ pathwise - differentiation_. here the parameter @xmath93 is the half - width of the _ regularizing window_. it determines the bias and affects the variance in the same way as parameter @xmath18 for finite - difference schemes , which implies that this method also has similar trade - offs as the finite - difference schemes . a sensitivity estimation method @xmath45
is called _ unbiased _ if @xmath94 .
the main advantage of unbiased methods is that the estimation can in principle be made as accurate as possible by increasing the sample size @xmath32 .
the first unbiased method for sensitivity estimation is called the _ likelihood ratio _ ( lr ) method ( or the _ girsanov transformation _ method ) @xcite , which works by estimating the @xmath0-derivative of the probability distribution of @xmath55 . the lr method is easy to implement and the computation cost of generating each sample is roughly @xmath79 the cost of _ exact _ simulation of the process @xmath55 .
the main issue with lr is that generally the variance of its associated random variable @xmath30 is very large ( i.e. @xmath95 ) , as has been observed numerically @xcite and also investigated theoretically in a recent paper @xcite . due to this large variance , lr method becomes infeasible for most examples of interest , as the number of samples needed to obtain a statistically precise estimate is very high . with this in mind
, we now discuss a couple of new unbiased methods that are based on an explicit formula for parameter sensitivity that we now present .
let @xmath96 be the markov process representing reaction dynamics and let @xmath97 denote its successive jump times starting with @xmath98 . for any state @xmath99 and time
@xmath100 define @xmath101 which is essentially the image of @xmath2 under the markovian semigroup associated with process @xmath55 at time @xmath102 .
let @xmath103 and for any @xmath104 define @xmath105 theorem 2.3 in @xcite shows that @xmath6 exactly satisfies with the random variable @xmath106 defined by @xmath107 note that @xmath108 contains a contribution at each jump time @xmath109 ( before time @xmath3 ) of process @xmath55 and this contribution is given by the function @xmath110 which does not have an analytical formula for most examples . hence in order to use this result for sensitivity estimation
, we need to estimate all these contributions using _ additional _ paths of the process @xmath55 .
this is accomplished in @xcite using a scheme called _ auxiliary path algorithm_(apa ) that estimates these quantities in parallel using a small number @xmath111 of additional paths in such a way that the overall estimate remains unbiased . in comparison to the other unbiased method lr
, the computational cost of generating each sample for apa is much higher ( i.e. @xmath112 ) but at the same time its variance is much lower ( i.e. @xmath113 ) . in @xcite
many examples are provided where apa outperforms lr , because the benefit of lower variance more than compensates for the disadvantage of higher computational cost per sample .
the main drawback of apa is that it tries to estimate _ all _ the unknown quantities of the form @xmath114 that appear in , which is very difficult because their number is proportional to the number of jumps of the process @xmath55 in the time interval @xmath77 $ ] , which can be very large even for small networks . in @xcite a new method is proposed that overcomes this problem by _ randomly selecting _ a small number of these quantities and estimating them using additional paths , without introducing a bias in the estimate .
this method is called _ poisson path algorithm _ ( ppa ) because the selection mechanism is based on the values of certain poisson random variables that are generated on the run " .
due to this extra randomness , the sample variance for ppa is generally greater than apa ( i.e. @xmath115 ) but the computational cost for realizing each sample is much lower ( i.e. @xmath116 ) .
moreover in comparison to apa , ppa is far easier to implement and has lower memory requirements , making it an attractive unbiased method for sensitivity estimation . in @xcite
it is shown using many examples that for a given level of statistical accuracy , ppa can be more efficient than lr and also the finite - difference schemes cfd and crp .
the computational cost of generating each sample in ppa is roughly @xmath117 , where @xmath111 is a small number that upper - bounds the expected number of unknown quantities that will be estimated using additional paths .
for both apa and ppa , the parameter @xmath111 serves as a _ trade - off _ factor between the computational cost and the variance - as @xmath111 increases , the cost also increases but the variance decreases .
however note that both these methods remain unbiased for any choice of @xmath111 .
the foregoing trade - off relationships for the existing sensitivity estimation methods are summarized in table [ tab1:tradeoff ] .
.trade - off relationships among the bias @xmath118 , variance @xmath119 and the computational cost @xmath120 for existing sensitivity estimation methods . here
@xmath18 is the perturbation size for finite - difference schemes @xcite , @xmath93 is the regularization half - width for the pathwise differentiation method @xcite and @xmath111 quantifies the number of auxiliary paths for apa @xcite and ppa @xcite .
the cost of _ exactly _ simulating the underlying process is @xmath79 . [
cols="^,^,^,^,^",options="header " , ]
estimation of parameter sensitivities for stochastic reaction networks in an important and difficult problem . the main source of difficulty is that all the estimation methods rely on exact simulations of the reaction dynamics performed using gillespie s ssa @xcite or its variants @xcite .
it is well - known that these simulation algorithms are computationally very demanding as they track each and every reaction event which can be very cumbersome .
this issue represents the main bottleneck in the use of sensitivity analysis for systems modeled as stochastic reaction networks .
the aim of this paper is to develop a method , called _ tau bernoulli path algorithm _ ( @xmath8bpa ) , that feasibly deals with this issue by requiring only approximate tau - leap simulations of the reaction dynamics , and still providing provably accurate estimates for the sensitivity values .
this method is based on an explicit integral representation for parameter sensitivity that was derived from the formula given in @xcite .
furthermore , by replacing the tau - leap simulation scheme in @xmath8bpa with an exact simulation scheme like ssa , we obtain a new unbiased method ( called ebpa ) for sensitivity estimation , that can serve as the natural limit of @xmath8bpa when the step - size @xmath8 gets smaller and smaller . using computational examples we compare @xmath8bpa with our proposed tau - leap versions of the finite - difference schemes @xcite that are commonly employed for sensitivity estimation .
we find that even though @xmath8bpa is typically _ slower _ than these tau - leap finite - difference schemes , it is usually more accurate .
this is primarily because the two sources of bias ( finite - difference and tau - leap approximations ) often pile on top of each other , thereby compromising the accuracy of an estimate .
this makes @xmath8bpa an appealing method for sensitivity analysis of stochastic reaction networks , where the exact dynamical simulations are computationally infeasible and tau - leap approximations become necessary .
as we argue in section [ using_tau_leap_methods ] , tau - leap simulations provide a natural way to _ trade - off _ estimator bias with gains in computational speed .
therefore it would be of fundamental importance to extend the ideas in this paper and try to _ maximize _ the computational gains from tau - leap simulations while sacrificing the _ minimum _ amount of accuracy . in this context
, we now mention two possible directions for future research .
the method we proposed here , @xmath8bpa , can work with any underlying tau - leap simulation scheme , but for simplicity we examined it with the most basic tau - leap scheme i.e. an explicit euler method with a constant ( deterministic ) step - size and poissonian reaction firings @xcite .
as this tau - leap scheme has several drawbacks ( see @xcite ) , it is very likely that @xmath8bpa can yield much better results if a more sophisticated tau - leap scheme is employed , possibly with random step - sizes @xcite , or with binomial leaps @xcite or using implicit step - size selection @xcite .
we shall explore these issues in a future paper .
another promising direction would be to devise a _
multilevel _ version of @xmath8bpa , where estimators are constructed for a range to @xmath8-values , and are suitably coupled to simultaneously reduce the overall estimator s bias and variance @xcite .
in this section we provide the proofs of the main results in this paper , namely , theorems [ thm : main ] and [ thm - tau - conv ] .
+ let @xmath121 be the filtration generated by the process @xmath122 and let @xmath109 be its @xmath123-th jump time for @xmath124 .
we define @xmath98 for convenience .
since the process @xmath122 is constant between consecutive jump times we can write @xmath125 where @xmath126 and the last equality holds due to linearity of the expectation operator and the fact that @xmath127 if @xmath128 .
given @xmath129 and @xmath130 , the distribution of the random variable @xmath131 has the _ cumulative density function _ given by @xmath132 this shows that for any continuous function @xmath133 we have @xmath134 however using integration by parts we get @xmath135_{s = 0}^{s = t - u } + \int_{0}^{t - u } e^ { -\lambda_0(y,\theta)s } g(s)ds \\ & = - e^{-\lambda_0(y,\theta)(t - u ) } \int_{0}^{t -u } g(s)ds+ \int_{0}^{t - u } e^ { -\lambda_0(y,\theta)s } g(s)ds .\end{aligned}\ ] ] substituting this expression in equation we get @xmath136 taking @xmath137 for all @xmath138 gives us @xmath139 therefore @xmath140 and equation implies that @xmath141 we mentioned in section [ subsec : unbiasedmethods ] that the sensitivity value @xmath142 can be expressed as the expectation of the random variable @xmath108 defined by .
using this fact along with we obtain @xmath143 since @xmath110 is given by we have @xmath144 relation with @xmath145 shows that given @xmath146 and @xmath147 , the following holds @xmath148 therefore @xmath149 since @xmath150 for all @xmath151 . substituting this expression in implies that @xmath152 this completes the proof of this result .
for each @xmath153 define @xmath154 by @xmath155 without loss of generality , we can assume that there exists a @xmath156 such that @xmath157 then due to lemma [ lem - phi - ass123 ] we obtain @xmath158 } |h_k(x , t ) - g_k(x , t)| & \leq \partial \lambda_k(x ) c c_1(p , t,\alpha_1 ) \left((1 + \|x\|^{\xi(p ) } ) + ( 1+\|x+\zeta_k\|^{\xi(p ) } ) \right ) \tau _ { \textnormal{max } } ^\gamma \\ & \leq c^2 c_1(p , t,\alpha_1 ) ( 1 + \|x\|^{p } ) \left((1 + \|x\|^{\xi(p ) } ) + ( 1+\|x+\zeta_k\|^{\xi(p ) } ) \right ) \tau _ { \textnormal{max } } ^\gamma\\ & \leq c_0(p ) c^2 c_1(p , t,\alpha_1 ) \left(1 + \|x\|^{(p+\xi(p))}\right ) \tau _ { \textnormal{max } } ^\gamma , \end{aligned}\ ] ] where @xmath159 is a constant that depends only on @xmath160 as well as @xmath161 . lemma [ lem - phi - ass123 ] also shows that @xmath162 } |h_k(x , t)| & \leq \partial \lambda_k(x ) c c_3(p , t,\alpha_1 ) \left ( ( 1 + \|x\|^p ) + ( 1+\|x+\zeta_k\|^p ) \right)\\ & \leq c_1(p ) c^2 c_3(p , t,\alpha_1 ) ( 1 + \|x\|^{2p } ) \end{aligned}\ ] ] and @xmath163 } |g_k(x ,
t)| \leq c_1(p ) c^2 c_2(p , t ) ( 1 + \|x\|^{2p}),\ ] ] where @xmath164 is again a constant that depends only on @xmath160 and @xmath161 . from and lemma [ lem - phi - ass123 ] it follows that @xmath165 } |\e(h_k(z_{\alpha_0,\beta_0}(x_0,t),t ) ) - \e(h_k(x(t),t))| \leq c_1(p ) c^2 c_3(p , t,\alpha_1 ) c_1(2p , t,\alpha_0 ) \left(1 + \|x_0\|^{\xi(2p)}\right ) \tau _ { \textnormal{max } } ^\gamma.\ ] ] moreover from , we get @xmath166 and hence using assumption 2 , we obtain @xmath167 }
\e(|h_k(x(t),t)-g_k(x(t),t)| ) \leq c_0(p ) c^2 c_1(p , t,\alpha_1 ) c_2(p+\xi(p),t ) \left(1 + \|x_0\|^{p + \xi(p)}\right ) \tau _ { \textnormal{max } } ^\gamma.\ ] ] d. anderson and t. kurtz .
continuous time markov chain models for chemical reaction networks . in h. koeppl , g. setti , m. di bernardo , and d. densmore , editors , _ design and analysis of biomolecular circuits_. springer - verlag , 2011 . c. r. bruno a. walther , joslin l. moore .
the concepts of bias , precision and accuracy , and their use in testing the performance of species richness estimators , with a literature review of estimator performance .
, 28(6):815829 , 2005 .
s. n. ethier and t. g. kurtz . .
wiley series in probability and mathematical statistics : probability and mathematical statistics .
john wiley & sons inc .
, new york , 1986 .
characterization and convergence . | we consider the important problem of estimating parameter sensitivities for stochastic models of reaction networks that describe the dynamics as a continuous - time markov process over a discrete lattice .
these sensitivity values are useful for understanding network properties , validating their design and identifying the pivotal model parameters .
many methods for sensitivity estimation have been developed , but their computational feasibility suffers from the critical bottleneck of requiring time - consuming monte carlo simulations of the exact reaction dynamics . to circumvent this problem one needs to devise methods that speed up the computations while suffering acceptable and quantifiable loss of accuracy .
we develop such a method by first deriving a novel integral representation of parameter sensitivity and then demonstrating that this integral may be approximated by any convergent tau - leap method .
our method is easy to implement , works with any tau - leap simulation scheme and its accuracy is proved to be similar to that of the underlying tau - leap scheme .
we demonstrate the efficiency of our methods through numerical examples .
we also compare our method with the tau - leap versions of certain finite - difference schemes that are commonly used for sensitivity estimations .
+ [ section ] [ theorem]lemma [ theorem]condition [ theorem]proposition [ theorem]remark [ theorem]definition [ theorem]hypothesis [ theorem]corollary [ theorem]example [ theorem]description [ theorem]assumption
| * keywords : * parameter sensitivity ; reaction networks ; markov process ; tau - leap simulations + * mathematical subject classification ( 2010 ) : * 60j22 ; 60j27 ; 60h35 ; 65c05 . |
biological evolution presents many problems concerning interacting multi - entity systems far from equilibrium that are well suited for methods from nonequilibrium statistical physics @xcite . among these are questions concerning the dynamics of the emergence and extinction of species on macroevolutionary timescales @xcite .
traditionally it has been common to treat ecological and evolutionary processes on very different timescales . however , it has recently been realized that evolution often can take place on short timescales , comparable to those of ecological processes @xcite . a well - known example of very rapid evolution is provided by the cichlid fishes of east africa @xcite .
several models have therefore been proposed that , while spanning disparate scales of temporal and taxonomic resolution , consider the complex problem of coevolution of species in a fitness landscape that constantly changes with the composition of the community .
early contributions were simulations of parapatric and sympatric speciation @xcite and the coupled @xmath3 model with population dynamics @xcite .
more recent work includes the webworld model @xcite , the tangled - nature model @xcite and simplified versions of the latter @xcite , as well as network models @xcite .
recently , large individual - based simulations have also been performed of parapatric and sympatric speciation @xcite and of adaptive radiation @xcite .
many of the models discussed above are deliberately quite simple , aiming to elucidate universal features that are largely independent of the finer details of the ecological interactions and the evolutionary mechanisms .
such features may include lifetime distributions for species and communities , as well as other aspects of extinction statistics , statistical properties of fluctuations in diversity and population sizes , and the structure and dynamics of food webs that develop and change with time .
in the present paper we continue our study of a simplified version of the tangled - nature model . in the early studies of this individual - based model of coevolution @xcite ,
the interspecies interactions , which are described by an interaction matrix @xmath4 @xcite , were random and could produce any combination of pair interactions : favorable - favorable , deleterious - deleterious , or favorable - deleterious . under those conditions
the model was found to evolve through a sequence of quasi - stable communities , in which all species interact with mutually favorable interactions , i.e. , mutualistic or symbiotic communities .
for that reason we shall hereafter refer to that version of the model as _ the mutualistic model_. here we instead concentrate on a version that specifically describes the evolution of predator - prey communities .
this restriction is enforced by means of an antisymmetric interaction matrix , so that an interaction that is favorable for one member of a pair is deleterious for the other , and _
vice versa_. many aspects of the dynamics of this predator - prey model are similar to the mutualistic model , such as approximate @xmath0 noise in species diversity and population sizes , and power - law distributions of the durations of communities and the lifetimes of species . however , some of the power - law exponents are different , and , most importantly , the predator - prey model produces communities that take the form of simple food webs .
it also shares with the mutualistic model the property that mean population sizes and stability properties of fixed - point communities can be calculated exactly in the absence of mutations .
comparisons of some aspects of the predator - prey model ( with a much smaller number of potential species than used here ) to those of the mutualistic model have been presented in refs .
the focus of the present paper is a much more detailed discussion of the dynamics and the structure of the resulting food webs ( including comparison with real food webs ) for the predator - prey model with a large number of potential species . for this purpose
we use both exact linear - stability analysis and large - scale kinetic monte carlo simulations .
in particular we wish to study the fluctuations in the statistically stationary state that develops for long times .
a motivation is the hope that understanding of these stationary - state fluctuations can provide information about the system s sensitivity to external perturbations in a way analogous to a fluctuation - dissipation relation @xcite .
we therefore carry out very long simulations .
the rest of this paper is organized as follows .
the model is presented in sec .
[ sec : mod ] .
exact linear - stability analysis is performed in sec .
[ sec : lsa ] , including fixed - point population sizes in sec .
[ sec : fpc ] and stability considerations in sec .
[ sec : stab ] .
numerical results are presented in sec .
[ sec : nr ] , including species abundance distributions and time series of diversities and population sizes ( sec .
[ sec : ts ] ) , power spectral densities ( sec .
[ sec : psd ] ) , species lifetimes ( sec .
[ sec : life ] ) , durations of evolutionarily quiet and active periods ( sec . [ sec : q ] ) , and community structure and stability with comparison with real food webs ( sec .
[ sec : web ] ) .
our conclusions are summarized in sec .
[ sec : concl ] , and the method used to calculate the interaction matrices for systems with a large number of potential species is explained in appendix [ sec : aa ] .
the model considered here is a version of the macroevolution model introduced by rikvold and zia @xcite as a simplification of the tangled - nature model of jensen and coworkers @xcite . in this version ,
the interspecies interactions are constrained to represent a pure predator - prey system . as in ref .
@xcite , selection is provided by the reproduction rates in an individual - based , simplified multispecies population - dynamics model with nonoverlapping generations .
this interacting birth / death process is augmented to enable evolution of new species by a mutation mechanism .
the mutations act on a haploid , binary genome " of length @xmath5 , as introduced by eigen for molecular evolution @xcite .
this bit string defines the species , which are identified by the integer label @xmath6 $ ] .
typically , only a few of these @xmath7 potential species are resident in the community at any one time . during reproduction
, an offspring individual may undergo a mutation that flips a randomly chosen gene ( @xmath8 or @xmath9 ) with a small probability , @xmath10 .
the mutation thus corresponds to diffusional moves from corner to corner along the edges of an @xmath5-dimensional hypercube @xcite .
a mutated individual is assumed to belong to a different species than its parent , with different properties .
genotype and phenotype are thus in one - to - one correspondence in this model .
this is clearly a highly idealized picture , and it is introduced to maximize the pool of different species available within the computational resources .
this picture is justified by a large - scale computational study of the mutualistic version of the model studied in ref .
@xcite , in which species that differ by as many as @xmath11 bits have correlated properties @xcite .
remarkably , this study reveals that the more realistic , correlated model has long - time dynamical properties very similar to the uncorrelated model .
the reproduction probability @xmath12 for an individual of species @xmath13 in generation @xmath14 depends on the individual s ability to utilize the amount @xmath15 of available external resources , and on its interactions with the population sizes @xmath16 of all the species present in the community at that time .
the dependence of @xmath17 on the set of @xmath18 is determined by an _
interaction matrix _ @xmath4 @xcite with offdiagonal elements @xmath19 that are continuously and symmetrically distributed on the interval @xmath20 $ ] in a way defined specifically in the next paragraph .
the elements of @xmath4 are chosen randomly at the beginning of each simulation run and are subsequently kept constant throughout the run ( quenched randomness ) .
( for a discussion of how the matrix elements are created for @xmath21 , in which case the @xmath22 matrix does not fit into the memory of a standard workstation , see appendix [ sec : aa ] .
this method leads to a distribution that is triangular on @xmath20 $ ] . )
in contrast to our previously studied model @xcite , in which the interaction matrix has no particular structure , the predator - prey dynamics is enforced by the requirement that the off - diagonal part of @xmath4 must be antisymmetric .
thus , if @xmath23 and @xmath24 , then species @xmath13 is the predator and @xmath25 the prey , and _ vice versa_. in order to keep the connectance of the resulting communities consistent with food webs observed in nature @xcite the @xmath26 pairs are chosen nonzero with probability @xmath27 .
the nonzero elements in the upper triangle of @xmath4 are chosen independently from the triangular distribution on @xmath20 $ ] , described in appendix [ sec : aa ] .
self - competition is included in the model by choosing the diagonal elements of @xmath28 randomly and uniformly from @xmath29 . the reproduction probability for species @xmath13 , @xmath12 , depends on @xmath15 and the set @xmath30 through the nonlinear form , @xmath31 } \ ; , \label{eq : pi}\ ] ] where @xmath32 here @xmath33 is the cost " of reproduction for species @xmath13 ( always positive ) , and @xmath34 ( positive for primary producers or autotrophs , and zero for consumers or heterotrophs ) is the ability of individuals of species @xmath13 to utilize the external resource @xmath15 .
the latter is renewed at the same level every generation and does _ not _ have independent dynamics .
the total population size is @xmath35 .
[ in contrast , the total number of _ species _ present in generation @xmath14 ( the species richness ) will be defined as @xmath36 . ]
the population - limiting reproduction costs @xmath33 are chosen randomly and uniformly from the interval @xmath37 $ ] . only a proportion @xmath38 of the @xmath7
potential species are producers that can directly utilize the resource .
( for the numerical data shown here , we use @xmath39 . )
thus , with probability @xmath40 the resource coupling @xmath41 , representing consumers , while with probability @xmath38 the @xmath34 are independently and uniformly distributed on @xmath37 $ ] , representing producers of varying efficiency .
in addition to the constraints on @xmath4 mentioned above , we require that producers always are the prey of consumers .
thus , the case @xmath42 and @xmath43 with @xmath44 is forbidden and is changed during setup of the matrix by reversing the signs of the interactions for the pair in question . for large positive @xmath45 , ( small birth cost , strong coupling to the external resources , and more prey than predators ) , the individual almost certainly reproduces , giving rise to @xmath46 offspring . in the opposite limit of large negative @xmath45 , ( large birth cost , weak or no coupling to the external resources , and/or more predators than prey ) , it almost certainly dies without offspring .
the nonlinear dependence of @xmath17 on @xmath45 thus limits the growth rate of the population size , even under extremely favorable conditions .
it also sets a practical negative limit on @xmath45 , below which conditions are so unfavorable that reproduction is virtually impossible .
( a more general version of eq .
( [ eq : delta ] ) , in which population growth is directly limited by a verhulst factor " @xcite or environmental carrying capacity " @xcite as is necessary in models that allow mutualistic interactions , is discussed in ref . @xcite . )
we note that energy dissipation " in this model is achieved through the birth costs @xmath33 and the self - competition terms @xmath47 .
equivalent effects could have been produced by making the positive @xmath48 smaller than the corresponding negative ones by an ecological efficiency " factor between zero and unity , as in the webworld model @xcite .
the normalization of @xmath45 with @xmath49 implies global competition .
this is not very realistic , but it enables us to find exact expressions for the stationary values of the average population sizes in the mutation - free limit .
( see sec .
[ sec : fpc ] . ) the model can thus be used as a benchmark for more realistic ones in future research .
an analytic approximation describing the development in time of the mean population sizes ( averaged over independent realizations ) , @xmath50 , can be written as a set of coupled difference equations , @xmath51 \nonumber\\ & & + ( \mu / l)f\sum_{k(i)}\langle n_{k(i)}(t ) \rangle
p_{k(i)}(r,\{\langle n_j(t ) \rangle \ } ) \ ; , \label{eq : mf}\end{aligned}\ ] ] where @xmath52 is the set of species that can be generated from species @xmath13 by a single mutation ( nearest neighbors " of @xmath13 in genotype space ) .
an advantage of the model studied here is that its fixed - point communities in the mutation - free limit can be found exactly within a mean - field approximation based on eq .
( [ eq : mf ] ) @xcite . to obtain a stationary solution for a community of @xmath53 species
, we must require @xmath54 for all @xmath53 species .
equations ( [ eq : pi ] ) and ( [ eq : delta ] ) then give rise to @xmath53 linear relations , which can be written on the matrix form @xmath55 where @xmath56 , @xmath57 , @xmath58 , and @xmath59 are the column vectors of @xmath60 , @xmath34 , and @xmath61 , respectively ( in all cases including only those @xmath53 species that have nonzero @xmath61 ) , and @xmath62 is the corresponding @xmath63 submatrix of @xmath4 .
( for simplicity , we drop the @xmath64 notation for the average population sizes , and the asterisk superscripts denote fixed - point solutions . )
the solution for @xmath59 is @xmath65 \ ; , \label{eq : ssn}\ ] ] where @xmath66 is the inverse of @xmath67 . to find each @xmath61 ,
we must first obtain @xmath68 , where @xmath69 is an @xmath53-dimensional row vector composed entirely of ones .
multiplying eq .
( [ eq : ssn ] ) from the left by @xmath69 , we obtain @xmath70 where the coefficients @xmath71 have been written with @xmath72 in the denominators in order to remain finite even for near - singular @xmath4 @xcite .
they can be viewed as an effective interaction strength and an effective coupling to the external resource , respectively .
the solution of eq .
( [ eq : quad ] ) is @xmath73 to find each @xmath61 separately , we now only need to insert this solution for @xmath74 in eq .
( [ eq : ssn ] ) . only those @xmath59 that have
all positive elements can represent a _ feasible _ community @xcite .
if @xmath75 or is otherwise singular , the set of equations ( [ eq : ss2 ] ) is inconsistent for @xmath76 , unless @xmath60 and @xmath34 both are independent of @xmath13 ( this case is equivalent to @xmath77 ) .
the only possible stationary community then consists of one single species , the one with the largest value of @xmath78 .
this is a trivial example of competitive exclusion @xcite .
if @xmath78 has the same value for all @xmath53 values of @xmath13 , we have an example of a neutral model @xcite . in ref .
@xcite it was shown that eq .
( [ eq : quad ] ) for fixed @xmath79 and @xmath80 can be seen as a maximization condition for a
community fitness " function , @xmath81 this result would not be particularly remarkable if @xmath79 and @xmath80 were externally fixed parameters .
however , extinctions and mutations provide a mechanism for both parameters to change as old species go extinct and new species emerge . in our numerical simulations
we find that their values evolve toward and then fluctuate around values that maximize @xmath82 , limited only by the internal constraints on @xmath4 and @xmath83 .
in particular , this means that @xmath80 approaches closely to 0 from the negative side @xcite . in the simulations presented in this paper ( see sec . [
sec : nr ] ) averages over the final communities of twelve independent simulation runs yield @xmath84 and @xmath85 , corresponding to @xmath86 . in comparison , for fourteen random , feasible communities obtained as described in sec .
[ sec : weba ] below , the corresponding parameters are @xmath87 and @xmath88 , corresponding to @xmath89 . a detailed discussion of the statistical properties of these and other quantities characteristic of the simulated communities are given in sec . [ sec : webb ] .
the internal stability of an @xmath53-species fixed - point community is obtained from the matrix of partial derivatives , @xmath90 where @xmath91 is the kronecker delta function and @xmath92 are elements of the _ community matrix _
straightforward differentiation yields @xcite @xmath94 \ ; , \label{eq : lam}\ ] ] where @xmath95 is the element of the column vector @xmath96 , corresponding to species @xmath13 . in order for deviations from the fixed point to decay
monotonically in magnitude , the magnitudes of the eigenvalues of the matrix of partial derivatives in eq .
( [ eq : stab ] ) , @xmath97 , where @xmath98 is the @xmath53-dimensional unit matrix , must be less than unity .
the value of the fecundity used in this work , @xmath99 , was chosen to satisfy this requirement for @xmath77 .
since new species are created by mutations , we must also study the stability of the fixed - point community toward invaders . "
consider a mutant invader @xmath100 .
then its multiplication rate , in the limit that @xmath101 for all @xmath53 species @xmath25 in the resident community , is given by @xmath102 } \;. \label{eq : inv}\ ] ] the lyapunov exponent , @xmath103 $ ] , is the _ invasion fitness _ of the mutant with respect to the resident community @xcite .
it will be studied numerically in sec . [
sec : weba ] .
we performed twelve independent , long simulation runs of @xmath104 generations of the model with the following parameters : genome length @xmath105 ( 1048576 potential species ) , external resource @xmath106 , and mutation rate @xmath107 , with connectance parameter @xmath108 and a proportion @xmath39 of the potential species as producers .
these parameters were chosen to represent the realistic situation that the number of species resident in the community at any time is much smaller than the number of potential species ( i.e. , that @xmath109 ) , and also that @xmath110 so that at least one species has a substantial population size . in this parameter range the model is not very sensitive to the exact parameter values @xcite .
we also note that the chosen connectance is above the percolation limit on the @xmath105 dimensional cube of potential genotypes @xcite , so that a finite fraction of the genotypes can be connected by mutations along paths of nonzero interactions .
the very long simulation times were chosen because our main interest is in the statistically stationary dynamics of macroevolution over timescales much longer than the ecological ones of a few generations .
each run therefore starts with a warm - up period " of about one million generations before the @xmath111-generation data - taking period .
details of the simulation algorithm were given in ref .
@xcite .
we collected time series of a number of quantities including several measures of diversity or species richness , as well as population sizes of producer and consumer species .
time series of diversities and population sizes for one representative run are shown in fig .
[ fig : timser ] . ) with parameters @xmath112 , @xmath113 , and @xmath114 , interpolating between log - series and log - normal forms . * ( b ) * histograms of the full - community species richness , the species richness of the core communities , and the shannon - wiener diversity for the sampled communities .
the latter is seen to be an excellent approximation for the species richness of the core communities .
, title="fig : " ] ) with parameters @xmath112 , @xmath113 , and @xmath114 , interpolating between log - series and log - normal forms . *
( b ) * histograms of the full - community species richness , the species richness of the core communities , and the shannon - wiener diversity for the sampled communities .
the latter is seen to be an excellent approximation for the species richness of the core communities . ,
title="fig : " ] , displaying quantities that measure the evolutionary activity : the magnitude of the logarithmic derivative of the shannon - wiener diversity index , @xmath115 ( averaged over 16 generations ) * ( a ) * , and the size of extinctions per generation * ( b)*. the horizontal lines in each part indicate cutoff levels used to define quiet and active periods as discussed in sec .
[ sec : q ] .
, title="fig : " ] , displaying quantities that measure the evolutionary activity : the magnitude of the logarithmic derivative of the shannon - wiener diversity index , @xmath115 ( averaged over 16 generations ) * ( a ) * , and the size of extinctions per generation * ( b)*. the horizontal lines in each part indicate cutoff levels used to define quiet and active periods as discussed in sec .
[ sec : q ] .
, title="fig : " ] in order to filter out noise from the small - population species that are mostly unsuccessful mutations , we use the diversity measure known in ecology as the exponential shannon - wiener index @xcite .
it is defined as the exponential function of the information - theoretical entropy of the population distributions , @xmath116 $ ] , where @xmath117 with @xmath118 for the case of the curves labeled all species " in fig .
[ fig : timser ] . for the producers or consumers , the sums and normalization constant include only the appropriate species .
the utility of the shannon - wiener index is illustrated by the data presented in fig .
[ fig : sad ] . in fig .
[ fig : sad](a ) we show two versions of the species abundance distribution ( sad ) for the model .
this is one of the tools most widely used in ecology to describe the distribution of the number of species over population sizes @xcite .
the sads shown as full lines in the figure represent full communities , from which we have only removed any consumer species that are not connected to the external resource through an unbroken chain of nonzero interactions .
these we term
full connected communities `` ( ' ' full communities " for short ) .
( the removal of disconnected consumer species actually has a numerically insignificant effect on the sads . ) the sads for the full communities are dominated by a large number of species with very small populations , which to a large extent represent unsuccessful mutants .
this was explicitly shown by extracting core communities " in the following way .
communities were sampled every 256 generations , and species with population sizes below eight were excluded .
it was then checked whether each included species also existed with at least this minimum population 256 generations ago , and if this was not the case , the species was removed from the community as unstable .
the fixed - point populations for the community of remaining species were then calculated according to eq .
( [ eq : ssn ] ) , species with negative fixed - point populations ( unfeasible species ) were removed , and the fixed - point calculation was repeated until all species remaining in the community had positive populations .
the sad was then calculated for each of these feasible core communities and averaged over all communities in the twelve independent simulation runs .
this procedure removes most of the low - population species , as shown by the dashed curves in fig .
[ fig : sad](a ) .
the core - community sad appears to be intermediate between fisher s log - series distribution @xcite and preston s log - normal distribution @xcite .
it can be semiquantitatively approximated by fitting the constants @xmath119 , @xmath10 , and @xmath120 in the function @xcite @xmath121 which interpolates between these two limiting forms .
the fit is shown by the dash - dotted curve in the figure .
additional evidence for the agreement between the shannon - wiener diversity index and the species richness of the core communities is shown in fig . [ fig : sad](b ) , where histograms of the two are in excellent agreement and both show a narrow peak near fifteen species ( mean values of 15.7 and 15.3 species , respectively ) , while the raw species richness yields a wide distribution with a mean of 49.3 species .
all three distributions are very well fit by gaussians and are much more symmetric than diversity distributions arising from rossberg et al.s speciation model of food webs @xcite .
time series of additional quantities that indicate the level of evolutionary activity are shown in fig .
[ fig : timser2 ] for the same simulation run as in fig .
[ fig : timser ] .
these are the magnitude of the logarithmic derivative of the shannon - wiener diversity index ( fig .
[ fig : timser2](a ) ) , and the size of extinctions per generation ( fig .
[ fig : timser2](b ) ) .
the latter is defined as the sum of the maximum population sizes reached by the species that go extinct in each generation .
the properties of the fluctuations in the time series were analyzed with several methods , including power spectral densities ( psd ) , lifetimes of individual species , and the durations of quiet and active periods during the evolution .
the results of each are reported in the following subsections .
the error bars are standard errors , estimated from the variations between the individual runs .
the dashed straight line with slope @xmath122 in * ( a ) * is a weighted fit to the psd for the overall diversity over the whole frequency range .
the dashed lines in * ( b ) * and * ( c ) * are guides to the eye with the same slope . ,
title="fig : " ] generations .
the error bars are standard errors , estimated from the variations between the individual runs .
the dashed straight line with slope @xmath122 in * ( a ) * is a weighted fit to the psd for the overall diversity over the whole frequency range .
the dashed lines in * ( b ) * and * ( c ) * are guides to the eye with the same slope . ,
title="fig : " ] + generations .
the error bars are standard errors , estimated from the variations between the individual runs .
the dashed straight line with slope @xmath122 in * ( a ) * is a weighted fit to the psd for the overall diversity over the whole frequency range . the dashed lines in * ( b ) * and * ( c ) * are guides to the eye with the same slope .
, title="fig : " ] psds of the diversity and population - size fluctuations , averaged over the twelve independent simulation runs , are shown in fig .
[ fig : psd ] , both for the total population and for the producers and consumers separately .
the spectra indicate @xmath0 like noise over more than five decades in time
. a weighted fit to the psd for the overall diversity in fig .
[ fig : psd](a ) yields a power law @xmath123 with @xmath124 .
this power is also seen to fit reasonably well , both with the data for the diversities of producers and consumers over the whole frequency range in fig .
[ fig : psd](a ) , and with the psds of all three population measures at low frequencies in fig .
[ fig : psd](b ) , as well as for the extinction measures at low frequencies in fig .
[ fig : psd](c ) .
this suggests that the long - time fluctuations in the diversity , as well as in the population sizes and the extinction measures , obey the same power law on long timescales . on short timescales
the psds for the population sizes have a more complicated structure , possibly indicating overdamped oscillations on a scale of a few hundred generations .
the extinction measures show a wide region of white noise for high frequencies , due to the frequent extinction of unsuccessful mutants .
however , the behaviors for low frequencies appear consistent with the diversities and the population sizes . power law .
data averaged over 12 independent simulation runs . ] the statistics of the lifetimes of individual species are characteristic of the evolution process .
the species lifetime is defined as the time from a particular species enters the community , till it goes extinct ( i.e. , its first return time to zero population size ) .
histograms showing the distributions of species lifetimes , for all species as well as for producers and consumers separately , are shown in fig .
[ fig : speclife ] .
although there are some undulations in these curves , they remain close to a power law @xmath125 with exponent @xmath126 over more than six decades in time , which is the maximum we could expect with the length of our simulations .
the @xmath127 dependence of the lifetime distributions is quite universal .
it is found , e.g. , in our previous studies of the mutualistic model @xcite , and it is in general characteristic of stochastic branching processes @xcite .
lifetime distributions for marine genera that are compatible with a power - law exponent in the range @xmath128 to @xmath129 have been obtained from the fossil record @xcite .
however , the possible power - law behavior in the fossil record is only observed over about one decade in time between 10 and 100 million years and other fitting functions , such as exponential or log - normal , are also possible .
nevertheless , it is reasonable to conclude that the numerical results obtained from complex , interacting evolution models that extend over a large range of time scales support interpretations of the fossil lifetime evidence in terms of nontrivial power laws . .
the data were averaged over 16 generations for each run and then over 12 independent runs .
the dashed curve is a gaussian fit .
the time series of the magnitude of this quantity for one particular simulation run is shown in fig .
[ fig : timser2](a ) . * ( b ) * histograms for the durations of active and quiet periods for various values of the cutoff @xmath130 for @xmath115 , averaged over 12 independent runs .
the dashed , straight line with slope @xmath131 is a weighted fit to the histogram for the quiet periods for @xmath132 between 10 and @xmath133 generations .
the steep histogram curve with only three data points between 10 and 100 generations corresponds to the active periods , which are always very short .
, title="fig : " ] .
the data were averaged over 16 generations for each run and then over 12 independent runs .
the dashed curve is a gaussian fit .
the time series of the magnitude of this quantity for one particular simulation run is shown in fig .
[ fig : timser2](a ) . * ( b ) * histograms for the durations of active and quiet periods for various values of the cutoff @xmath130 for @xmath115 , averaged over 12 independent runs .
the dashed , straight line with slope @xmath131 is a weighted fit to the histogram for the quiet periods for @xmath132 between 10 and @xmath133 generations .
the steep histogram curve with only three data points between 10 and 100 generations corresponds to the active periods , which are always very short .
, title="fig : " ] ( b ) .
data averaged over 12 independent simulation runs .
the straight , dashed line with slope @xmath131 is a guide to the eye , based on the fit to the qss duration distribution based on @xmath115 , shown in fig .
[ fig : dsdt](b ) . ] from the time series shown in figs .
[ fig : timser ] and [ fig : timser2 ] one sees that periods of moderate fluctuations are punctuated by periods of high activity .
the communities corresponding to the lower fluctuation intensities are known as quasi - steady states ( qss ) .
( in the literature on the tangled - nature model @xcite the qss are referred to as quasi evolutionarily steady strategies , or q - ess . )
the cores of these communities correspond to the fixed - point communities of the mutation - free system @xcite , as discussed in sec .
[ sec : ts ] .
one measure of the degree of activity is the time derivative of the entropy or , equivalently , the logarithmic derivative of the total shannon - wiener diversity .
its magnitude is shown as a time series in fig .
[ fig : timser2](a ) , and a histogram is shown in fig . [
fig : dsdt](a ) . while the central part of the distribution is well approximated by a gaussian ,
the heavier wings appear to be exponential or even power - law .
quiet and active periods were defined as contiguous periods during which @xmath115 stayed below or above a cutoff @xmath130 , respectively .
octave - binned histograms for the probability distributions of the durations of quiet and active periods are shown in fig .
[ fig : dsdt](b ) for various cutoffs . for the quiet periods a power law
is seen with exponent near @xmath1 ( a weighted fit between 10 and @xmath133 generations gives @xmath134 with @xmath135 ) and a long - time cutoff that increases with increasing @xmath130 .
the active periods for all values of @xmath130 are very brief in comparison .
( their histogram for @xmath136 is the steep curve with only three data points between 10 and 100 generations in fig .
[ fig : dsdt](b ) . ) as a consequence , the system spends most of its time in qss communities a situation consistent on the community level with eldredge and gould s concept of punctuated equilibria @xcite .
durations of quiet periods could also be obtained from the time series of extinction sizes in fig .
[ fig : timser2](b ) . due to the white noise at short timescales , which was also apparent in the psds in fig .
[ fig : psd](c ) , the power - law behavior is limited to a window of longer times between about 1000 generations and the strongly cutoff - dependent long - time decay . as a result
, this quantity does not provide as clear a quiet - period distribution as the entropy derivative .
we therefore did not perform any independent fit to obtain a power - law exponent for the qss durations measured this way .
see fig .
[ fig : quietext ] .
the decay with time is qualitatively consistent with that observed in fig .
[ fig : dsdt](b ) for the qss duration distributions based on @xmath115 .
it is quite remarkable that the exponent for the qss durations is significantly different from the one for the species lifetimes , which is close to 2 .
this is particularly so because the two exponents appear to be approximately the same ( both near 2 ) for the mutualistic version of the model @xcite .
we believe that the explanation lies in the structure of the qss communities generated by the present evolution process , which take the form of simple food webs .
these are studied in sec .
[ sec : web ] below . and [ fig : timser2 ] .
the horizontal lines correspond to the species label , and the grayscale ( color online ) to the population size .
black : @xmath137 .
gray ( red online ) : @xmath138 $ ] . *
( a ) * : producers . * ( b ) * : consumers .
, title="fig : " ] and [ fig : timser2 ] .
the horizontal lines correspond to the species label , and the grayscale ( color online ) to the population size .
black : @xmath137 .
gray ( red online ) : @xmath138 $ ] . *
( a ) * : producers . * ( b ) * : consumers .
, title="fig : " ] the evolution process in the present model generates dynamic communities , in which species emerge , exist for a shorter or longer time , and eventually go extinct .
the emergence and extinction of a major species are quite fast processes on the evolutionary timescale , and so the vast majority of randomly selected communities are qss communities .
this is confirmed by the short durations of evolutionarily active periods , shown by the corresponding histogram in fig .
[ fig : dsdt](b ) .
diagrams of the population sizes of major producer and consumer species as functions of time in a particular simulation run are shown in figs .
[ fig : species](a ) and ( b ) , respectively . in these figures a horizontal line represents a species .
the beginning of the line represents the emergence of the species , and the end represents its extinction .
the population size is represented by the color .
we see that some species persist for tens of millions of generations , while others are so short - lived as to hardly be visible on the scale of these figures .
this is consistent with the power - law behavior of the species - lifetime distribution ( see fig . [
fig : speclife ] ) .
we also see that producer species appear to emerge and go extinct relatively independently of each other , while there is a significant correlation between the producer and consumer species .
this correlation indicates that extinction of a producer species is likely to trigger a ( limited ) cascade of consumer extinctions .
conversely , a new producer species is likely to quickly acquire a group of consumer species .
the structure of the plots in fig . [ fig : species ] contrasts with that of similar plots for the mutualistic model of ref .
@xcite ( see fig . 2 of that paper ) , in which species tend to emerge as well as go extinct together .
we believe this is the reason for the difference between the exponents for the species lifetimes and the durations of qss communities in this model : the overall community is relatively resilient toward losing or gaining a single species @xcite . as a community
it is more long - lived than the individual species , leading to the smaller value for the exponent of the distribution for the qss durations ( compare figs .
[ fig : speclife ] and [ fig : dsdt](b ) ) .
( a similar relationship has been noted between the lifetimes of orders and their constituent genera in the fossil record @xcite . ) , [ fig : timser2 ] , and [ fig : species ] at times near @xmath139 generations * ( a ) * , near @xmath140 generations * ( b ) * , and at the end of the simulation near @xmath141 generations * ( c)*. the core communities were identified as described in sec .
[ sec : ts ] .
the thickness and head size of the arrows correspond to the magnitude of @xmath48 , and the area of the circles to the stationary population size , as calculated analytically from eq .
( [ eq : ssn ] ) .
light gray ( red online ) lines connect nearest neighbors in genotype space .
labels in red mark species that persist from the first to the last snapshot , * ( a ) * to * ( c ) * , while blue labels mark species that persist from * ( a ) * to * ( b ) * , and and green labels mark species that persist from * ( b ) * to * ( c)*. black labels indicate species that are unique to the particular community . , title="fig : " ] , [ fig : timser2 ] , and [ fig : species ] at times near @xmath139 generations * ( a ) * , near @xmath140 generations * ( b ) * , and at the end of the simulation near @xmath141 generations * ( c)*. the core communities were identified as described in sec .
[ sec : ts ] .
the thickness and head size of the arrows correspond to the magnitude of @xmath48 , and the area of the circles to the stationary population size , as calculated analytically from eq .
( [ eq : ssn ] ) .
light gray ( red online ) lines connect nearest neighbors in genotype space .
labels in red mark species that persist from the first to the last snapshot , * ( a ) * to * ( c ) * , while blue labels mark species that persist from * ( a ) * to * ( b ) * , and and green labels mark species that persist from * ( b ) * to * ( c)*. black labels indicate species that are unique to the particular community . , title="fig : " ] + , [ fig : timser2 ] , and [ fig : species ] at times near @xmath139 generations * ( a ) * , near @xmath140 generations * ( b ) * , and at the end of the simulation near @xmath141 generations * ( c)*. the core communities were identified as described in sec .
[ sec : ts ] .
the thickness and head size of the arrows correspond to the magnitude of @xmath48 , and the area of the circles to the stationary population size , as calculated analytically from eq .
( [ eq : ssn ] ) .
light gray ( red online ) lines connect nearest neighbors in genotype space .
labels in red mark species that persist from the first to the last snapshot , * ( a ) * to * ( c ) * , while blue labels mark species that persist from * ( a ) * to * ( b ) * , and and green labels mark species that persist from * ( b ) * to * ( c)*. black labels indicate species that are unique to the particular community . , title="fig : " ] + three representative qss core communities from the same simulation run shown in figs . [
fig : timser ] , [ fig : timser2 ] , and [ fig : species ] are shown in fig .
[ fig : webs ] .
these are consecutive core communities near @xmath139 , @xmath140 , and @xmath142 generations , respectively , and they are all stable by the eigenvalue criterion discussed in sec .
[ sec : stab ] .
the communities take the form of simple food webs with two trophic levels above the resource node .
communities with three trophic levels are also occasionally observed .
the different branches of the webs are relatively independent of each other , and many consumer species have more than one prey .
( see further quantitative discussion in sec .
[ sec : webb ] . )
both features contribute to the resilience against mass extinctions discussed above .
( exponential of the invasion fitness , eq . ( [ eq : inv ] ) ) for low - density invaders .
each curve is averaged over 14 different core communities .
black : nearest neighbors in genotype space against stable qss communities .
dark gray ( red online ) : all species not in the community against stable qss communities .
light gray ( orange online ) : all species not in the community against random feasible communities .
inset : the part of the distributions for potentially successful invaders , @xmath143 , with the @xmath144-axis on logarithmic scale .
the tails appear nearly exponential . ]
the relative stability of evolved qss core communities with respect to invasion by mutants is illustrated in fig .
[ fig : invader ] .
the results are averaged over 14 qss core communities the three shown in fig .
[ fig : webs ] plus the final communities of the eleven other simulation runs .
this figure shows the multiplication ratio of a small population of invaders ( the exponential function of the invasion fitness ) , given by eq .
( [ eq : inv ] ) .
only about 2.3% of species outside the communities have multiplication ratios larger than unity , and most of these lie between 1.0 and 1.1 .
this percentage does not seem to depend significantly on the hamming distance of the invader from the community .
we note that the species that are removed from the community during extraction of the core community are among these low invasion fitness species .
this is a further indication that the difference between core and full communities is mostly made up by unsuccessful mutants .
we also tested if the evolved core communities are more resilient toward invasion than randomly constructed feasible communities .
such communities are more difficult to construct in this model , than in the mutualistic model @xcite .
to have the same level of statistics as for the qss communities , we produced 14 such communities in the following way .
we started a run with a random sample of 200 species , each with @xmath145 , and evolved the community for 1024 generations _ without _ mutations .
we then tested the remaining community for feasibility and removed species with a negative stationary population according to eq .
( [ eq : ssn ] ) .
the resulting feasible communities had much smaller diversities than the evolved qss communities an average of only 3.6 species per community .
these communities had a significantly larger percentage of potential invaders about 4% , with a maximum multiplication ratio near 1.7 ( see inset in fig .
[ fig : invader ] ) . while there thus is a clear difference between the stability against invaders of qss communities and random feasible communities , the difference is not as large as it is for the mutualistic model of ref .
( see fig . 3 of that paper .
) ( left ) and producer coupling to the external resource , @xmath34 ( right ) , for members of qss core communities ( shaded areas ) and full communities ( lines only ) .
both parameters are selected away from the uniform distributions of the full species pool ( represented by the horizontal dashed line ) . ]
next we turn to a detailed statistical description of the qss core and full connected communities identified in sec .
[ sec : ts ] .
since communities were extracted every 256 generations from twelve independent runs of @xmath111 generations each , data were averaged over @xmath146 communities of each type . however ,
due to the long - time correlations in the evolution process , many of the communities from the same run are identical or similar .
this sampling method thus ensures that the statistics for both communities and individual species are weighted according to their longevities .
first we consider the properties of individual species , and next we turn to the collective properties of the corresponding food webs and a comparison with data for real food webs .
, for qss core and full communities . * ( b ) * the effective interaction strength @xmath80 ( left ) and resource coupling @xmath79 ( right ) for qss core communities , defined in eq .
( [ eq : quad2 ] ) .
, title="fig : " ] , for qss core and full communities . *
( b ) * the effective interaction strength @xmath80 ( left ) and resource coupling @xmath79 ( right ) for qss core communities , defined in eq .
( [ eq : quad2 ] ) .
, title="fig : " ] the individual species are characterized by the birth cost @xmath33 , the self interaction @xmath47 , and for producers by the resource coupling @xmath34 . in the total pool of potential species these
are uniformly distributed on @xmath37 $ ] , @xmath29 , and @xmath37 $ ] , respectively .
not surprisingly , the most prevalent species in the core communities turn out to be the most individually fit " ones in the sense that they have low birth cost and , for producers , relatively strong coupling to the external resource .
the resulting probability densities for @xmath33 and @xmath34 are shown in fig .
[ fig : b - eta ] .
the selection for low birth cost is very strong for members of the core communities , and less so when the full communities are considered .
this is further indication that the species that are ignored when extracting the core communities have higher @xmath33 , and thus overall are less individually fit , than the core species . on the other hand ,
the moderate selection for strong coupling to the external resource ( large @xmath34 ) is approximately equal for producers in both types of communities .
in contrast , the self interactions appear to be evolutionarily neutral , and their probability density remains approximately uniform for the members of both core and full communities ( not shown ) .
the first community - level quantities we consider are the nonzero interspecies interaction strengths , @xmath147 .
histograms for these , based on the same sampling and averaging as previous results , are shown in fig .
[ fig : mij](a ) .
the distribution for the core communities is significantly skewed toward strong interactions , compared to the triangular distribution characteristic of the total species pool , which is shown as a dashed line .
the bias is much weaker when all members of the full communities are considered .
( see appendix [ sec : aa ] for a discussion of the unbiased distribution . )
( main part ) and the directed connectance , @xmath148 ( inset ) , for the model communities . ] the biased distribution of interaction strengths combines with the strongly skewed distributions of @xmath33 and @xmath34 to produce values of the effective interaction strength @xmath80 and effective resource coupling @xmath79 for the core communities ( defined in eq .
( [ eq : quad2 ] ) ) that are in excellent agreement with the expectations expressed in sec .
[ sec : fpc ] . as shown in fig .
[ fig : mij](b ) , @xmath80 is narrowly distributed closely below 0 ( @xmath149 ) , while @xmath79 has a broader distribution over positive values with mean @xmath150 .
these averages are in excellent agreement with those obtained from the final communities of the twelve runs , obtained in sec .
[ sec : ts ] .
next we consider several quantities from network theory that can be compared with data from real food webs .
these include the directed connectance , @xmath148 , where @xmath5 is the number of links ( i.e. , the number of nonzero @xmath48 , @xmath151 pairs ) and @xmath152 is the diversity ( here defined as the species richness ) , the linkage density , @xmath153 , the probability distributions of a species number of prey species ( its generality or indegree ) , number of predator species ( its vulnerability or outdegree ) , and their sum ( its total degree ) @xcite . for comparison with the simulated food webs generated by our evolutionary model , we use data for seventeen empirical webs , including both aquatic and terrestrial communities .
these data sets were kindly provided to us by j. a. dunne .
tabular overviews of the parameters characterizing these communities are available in the literature @xcite .
for all the empirical webs we defined the diversity as the species richness @xmath152 in terms of _ trophic species_. these are obtained by lumping together all taxa that share all their predators and prey @xcite . like
the simulated communities analyzed , the empirical webs contain no disconnected species or subwebs @xcite .
the included empirical food webs are : coachella valley @xcite , el verde rainforest @xcite , scotch broom @xcite , st . martin island @xcite , and u.k .
grassland @xcite ( terrestrial ) ; bridge brook lake @xcite , little rock lake @xcite , and skipwith pond @xcite ( lake or pond ) ; canton creek @xcite and stony stream @xcite ( stream ) ; chesapeake bay @xcite , st .
mark s estuary @xcite , ythan estuary @xcite , and ythan estuary with parasites @xcite , ( estuaries ) ; benguela @xcite , caribbean reef
small @xcite , and north - eastern u.s .
continental shelf @xcite ( marine ) .
as discussed above , the food webs produced by the evolutionary process in this model are relatively small , with average diversity @xmath154 for the qss core communities and approximately 50 for the full communities . as seen in fig .
[ fig : conn ] , the directed connectance @xmath119 is somewhat smaller ( @xmath155 for core and @xmath156 for full communities ) than the input connectance , @xmath27 . however , this is partly because it is calculated in the conventional way as @xmath157 @xcite , rather than as @xmath158 .
the average linkage density is also quite small ( @xmath159 for core and @xmath160 for full communities ) compared to most of the real food webs that have been documented . for comparison
, the seventeen documented food webs have @xmath161 with a range from 25 to 155 , @xmath162 with a range from 0.03 to 0.32 , and @xmath163 with a range from 1.6 to 17.8 .
however , williams and martinez niche model for food - web structure @xcite leads to a scaling hypothesis for the degree distributions ( exact for the niche - model prey distribution and approximate for the predator distribution ) @xcite : @xmath164 where @xmath165 can be either the generality , the vulnerability , or the total degree ( with different scaling functions @xmath166 for each ) . as a consequence , cumulative degree distributions
can also be rescaled by simply dividing the argument by @xmath167 . using this niche - model result as a general scaling hypothesis @xcite
, we can compare the degree distributions for the communities generated by our model with scaled data for the seventeen real food webs . .
the inset on log - linear scale shows that these distributions decay at least exponentially with increasing argument . *
( b ) * same as ( a ) , for the full communities .
average linkage density is @xmath168 . *
( c ) * histograms over individual sampled communities of the correlation coefficient between a species numbers of predators and prey .
black filled with light gray ( yellow online ) : qss core communities .
medium gray ( red online ) : full , connected communities .
dark gray ( blue online ) with drop lines : empirical data for seventeen natural food webs .
, title="fig : " ] .
the inset on log - linear scale shows that these distributions decay at least exponentially with increasing argument . *
( b ) * same as ( a ) , for the full communities .
average linkage density is @xmath168 . *
( c ) * histograms over individual sampled communities of the correlation coefficient between a species numbers of predators and prey .
black filled with light gray ( yellow online ) : qss core communities .
medium gray ( red online ) : full , connected communities .
dark gray ( blue online ) with drop lines : empirical data for seventeen natural food webs .
, title="fig : " ] + .
the inset on log - linear scale shows that these distributions decay at least exponentially with increasing argument . *
( b ) * same as ( a ) , for the full communities .
average linkage density is @xmath168 . *
( c ) * histograms over individual sampled communities of the correlation coefficient between a species numbers of predators and prey .
black filled with light gray ( yellow online ) : qss core communities .
medium gray ( red online ) : full , connected communities .
dark gray ( blue online ) with drop lines : empirical data for seventeen natural food webs .
, title="fig : " ] the degree distributions for the model , individually normalized for each community , and then averaged over all communities , are shown in unscaled form in fig .
[ fig : pk](a ) and ( b ) . as seen from the insets , all three distributions ( prey , predators , and total degree ) decay at least exponentially for large argument .
this property is shared by most real food webs @xcite and models @xcite and indicates that food webs in general are not scale - free networks .
the behavior for @xmath169 is more model and system dependent .
histograms for the correlations between a species generality and vulnerability are shown in fig .
[ fig : pk](c ) , together with data for the seventeen real food webs .
the average correlation coefficients are @xmath170 for core communities and @xmath171 for full communities , more negative than , but not inconsistent with , the real food - web average of @xmath172 .
+ due to the relatively small size of the empirical food webs , it is difficult to extract information from scaled probability densities , which will include empty bins unless a very large bin size is chosen .
this problem is avoided by instead studying the cumulative distribution , @xmath173 $ ] @xcite .
these are shown in fig .
[ fig : cum ] for the seventeen empirical webs , together with averaged results for the empirical webs and the model .
the averaged curves were generated by first scaling the degrees ( predators , prey , and total ) for each web by the value of @xmath167 for that particular web , binning the results in bins of width 0.2 , and normalizing the binned histogram by the species richness @xmath152 of that same web .
the individually scaled , binned , and normalized histograms were then averaged over all webs ( seventeen for the empirical webs and @xmath146 for the model webs ) and finally integrated to produce the average scaled cumulative distributions . for the model , results are shown both for the full and core communities . despite their smaller average diversity and linkage density , the scaled cumulative degree distributions for both types of model communities , which are shown in fig .
[ fig : cum ] , fall well within the range of the empirical data . for scaled degrees below about 1.5 ,
the averaged distributions for the core communities are in quite reasonable agreement with those for the empirical food webs .
the largest deviations are for total degree ( fig .
[ fig : cum](c ) ) at small values of the scaled argument .
this is possibly due to the significantly stronger negative correlations between the numbers of predators and prey ( vulnerability and generality ) for species in the model communities , compared to the empirical ones .
( see fig .
[ fig : pk](c ) . ) for scaled degrees above approximately 1.5 , the averages for the empirical webs are significantly above those for the model webs and appear approximately exponential in the tail ( see insets in fig .
[ fig : cum ] ) .
we note , however , that this behavior in the averages is caused by a small number of webs ; the remaining empirical webs have no species of such highly above - average degrees .
overall , the scaled , cumulative distributions for the present model agree with the empirical data to about the same degree as the niche model @xcite and the speciation model of rossberg et al .
however , it must be admitted that much potentially valuable detail about the food - web structure is lost in calculating the cumulative distributions . a measure of the hierarchical structure of a food web is given by the proportions of basal species ( species with no prey , supported only by the external resource ) , intermediate species ( that have both prey and predators ) and top species ( that have no predators ) .
as seen from the histograms in fig .
[ fig : bit ] , there are wide variations from community to community for each class of species . the high proportion @xmath13 of intermediate species seen for the full communities in fig .
[ fig : bit](b ) ( @xmath174 ) is consistent with most of the real food webs analyzed , including caribbean reef ( @xmath175 ) , u.s .
shelf ( @xmath175 ) , benguela ( @xmath176 ) , scotch broom ( @xmath177 ) , skipwith pond ( @xmath177 ) , coachella valley ( @xmath178 ) , little rock lake ( @xmath179 ) , el verde rainforest ( @xmath180 ) , st .
marks seagrass ( @xmath180 ) , st .
martin island ( @xmath180 ) , and bridge brook lake ( @xmath181 ) , as well as simulations of the web world model @xcite and mean - field analysis of a lotka - volterra predator - prey model with evolution and competition @xcite .
the core communities , on the other hand , have their member species much more evenly distributed among the three classes , as shown in fig .
[ fig : bit](a ) . real food webs with more even distributions are canton creek ( @xmath182 , @xmath183 , @xmath184 ) , stony stream ( @xmath185 , @xmath186 , @xmath187 ) , and chesapeake bay ( @xmath188 , @xmath189 , @xmath190 ) .
thus , at least in this model , and somewhat counterintuitively , on average the intermediate species are both the least stable and the most numerous .
we conclude these comparisons of the structures of food webs evolving in the present model to data for real food webs with some words of caution .
the model was primarily developed to study the long - time temporal fluctuations in diversity and population size , and it was in no way tuned to produce realistic community structures except insofar as they are generally food - web like .
this community structure does , however have profound effects on the dynamics , as it is responsible for the difference between the power - law exponents for the lifetimes of species and communities in this predator - prey model .
one should also note that , while the lack of correlations between the properties of closely related species was shown to have little effect on the long - time dynamics @xcite , we have not considered whether such correlations might affect the detailed community structure .
most likely they would , at least to some degree .
in addition , the growth of real food webs involve a number of mechanisms , besides just evolution and local population dynamics .
most notably , immigration plays an important role in the occurrence of new members of a spatially localized community .
however , due to the large differences between parent species and mutants in the current model , our evolution mechanism may also been seen as covering immigration .
we also note that the simulated communities have significantly smaller diversities , connectances , linkage densities , and population sizes than the real communities , so that our comparisons had to rely heavily on scaling arguments
. it would therefore be desirable in future studies to increase the connectance , number of potential species , and resource size to bring the results closer to the realistic range .
in this paper we have studied in detail an individual - based predator - prey model of biological coevolution , based on the simplified version of the tangled - nature model @xcite that was introduced in ref .
selection is provided by a population - dynamics model in which the reproduction probability of an individual of a particular species depends nonlinearly on the amount of external resources and on the population densities of all other species resident in the community .
new species appear in the community through point mutations in a genome consisting of a string of @xmath5 bits . in the mutation - free limit
, the mean fixed - point population sizes and stability properties of any @xmath53-species community can be obtained exactly by linear stability analysis . while the universal competition effect that enables this analytical treatment is not very realistic
, the exact solutions make the model ideal as a benchmark for more realistic , but also more complicated , models . a preliminary discussion of two more realistic models is found in ref .
@xcite . in the simulations presented here , we used @xmath105 for a total of @xmath191 potential species .
in order to study the statistically stationary properties of the model , we performed long kinetic monte carlo simulations over @xmath104 generations . by studying the stationary fluctuations we hope in the future to gain an understanding of the the system s sensitivity to external perturbations in a way analogous to the fluctuation - dissipation relations of equilibrium statistical mechanics @xcite .
qualitatively , many of the statistical properties of this model are similar to those of the related , mutualistic model studied in refs . @xcite . these include approximate @xmath0 noise in power spectra ( psds ) of diversity and population sizes ( @xmath123 with @xmath192 ) , and power - law distributions for the lifetimes of individual species , as well as of the durations of evolutionarily quiet periods , corresponding to qss communities . however , in contrast to the mutualistic model , the power - law exponents for the species lifetimes and qss durations are different : @xmath125 with @xmath126 for the former ( consistent with a stochastic branching process @xcite ) and @xmath134 with @xmath193 for the latter . in ref .
@xcite it was speculated that the exponent values @xmath194 , @xmath195 , and @xmath196 are consistent with predictions for a zero - dimensional extremal - dynamics model @xcite .
however , this speculation is not consistent with the presumably more accurate estimate for @xmath197 presented here , and it seems advisable to be rather skeptical about any mapping of the current model onto a simple statistical - mechanical extremal - dynamics model .
it is probably more fruitful to consider why @xmath198 and @xmath199 are different for the current predator - prey model , while they coincide for the corresponding mutualistic model . here
we believe the answer lies in the different community structures in the two models . while communities in the mutualistic model are tightly knit with all positive interactions ( see fig .
10 of ref .
@xcite ) , the communities in the present predator - prey model take the form of simple food webs , which are much more resilient toward the loss of a single or a few species . in this sense , this predator - prey model is in much better agreement with real food webs than the mutualistic model @xcite .
our comparisons of the structure of the communities generated by our model with real food webs show both similarities and differences .
a difficulty with such comparisons is the large differences in diversity and linkage density between the simulated and real webs , which necessitated heavy use of scaling arguments in the quantitative comparisons . for comparison of detailed community properties , simulations of models with higher diversity and connectance
are therefore desirable in the future .
nevertheless , our model presents a synthesis of long - term evolutionary dynamics and food - web like communities within an individual - based framework of integrated population dynamics and evolution , that should provide a sound basis for more refined models in the future .
we are grateful to j. a. dunne for providing us with raw data on seventeen empirical food webs .
we also thank two anonymous referees for helpful comments on an earlier version of this paper .
this research was supported in part by national science foundation grant nos .
dmr-0240078 and dmr-0444051 , and by florida state university through the school of computational science , the center for materials research and technology , the national high magnetic field laboratory , and a cofrs summer salary grant .
here we describe an improved version of the method introduced by hall et al .
@xcite to produce pseudorandom matrix elements @xmath48 for values of @xmath5 that are too large for the full @xmath22 matrix @xmath4 to fit into computer memory .
this method permits the matrix elements for a given community to be generated and retained only as needed .
we first present hall et al.s method , point out some problematic features , and then present our modifications .
let @xmath200 be the string of binary digits corresponding to the decimal species label @xmath13 .
this bit string has length @xmath5 , so there are @xmath7 different strings . to generate the matrix element @xmath48 , one first generates a new string of the same length , @xmath201 , where @xmath202 is the logical _ exclusive or _ operator . from this bit string
is generated the corresponding new decimal index @xmath203 .
next one creates two one - dimensional arrays , @xmath204 and @xmath205 , each of @xmath7 random numbers between @xmath1 and @xmath206 .
( for simplicity let the starting index for the arrays be zero . ) since @xmath207 is symmetric in @xmath13 and @xmath25 , asymmetric pseudorandom matrix elements are generated as @xmath208 / 2 \;. \label{eq : mjensen}\ ] ] ( hall et al .
instead use the product of the two random numbers , which gives a pseudorandom number with a slightly different distribution . ) for the two schemes described in the text , each based on a single realization of @xmath4 for @xmath209 , 8 , and 9 . *
( a ) * the scheme described in ref
. @xcite . *
( b ) * the modified scheme introduced here .
, title="fig : " ] for the two schemes described in the text , each based on a single realization of @xmath4 for @xmath209 , 8 , and 9 . *
( a ) * the scheme described in ref .
( b ) * the modified scheme introduced here .
, title="fig : " ] the problem with this method is that it produces strong correlations along the columns of @xmath4 since the second of the two random numbers that produce @xmath48 is the same for all @xmath13 at the same @xmath25 . in fig .
[ fig : corr](a ) we show the resulting correlation function for this scheme as a function of the hamming distance between the pairs of bit strings involved in two different matrix elements . regardless of @xmath5 ,
significant correlations are seen for hamming distances less than five .
a discussion of how to calculate such correlation functions is found in ref .
@xcite . to reduce these correlations between matrix elements involving closely related genotypes , we here modify the scheme as follows .
we extend array @xmath211 to a length of @xmath212 and define @xmath48 as @xmath213 / 2 \;. \label{eq : x}\ ] ] the addition of @xmath203 in the index of @xmath214 ensures that the matrix element depends in an erratic fashion on both @xmath13 and @xmath25 , while the term linear in @xmath25 ensures that @xmath4 is not symmetric .
the resulting correlation function is shown in fig .
[ fig : corr](b ) . the correlations for elements involving closely related genotypes
are strongly suppressed .
the correlations for a hamming distance of @xmath215 , which are present in both schemes , are of little practical significance .
they are caused by the fact that the @xmath202 operation is invariant under simultaneous bit reversal in both its arguments and could be removed by adding a linear function of @xmath13 in the argument of @xmath216 .
the probability density for @xmath48 is triangular as expected from simple analytical arguments , and the whole interval from @xmath1 to @xmath206 is well sampled , even for relatively small @xmath5 .
this is shown in fig .
[ fig : dist ] . | we study an individual - based predator - prey model of biological coevolution , using linear stability analysis and large - scale kinetic monte carlo simulations .
the model exhibits approximate @xmath0 noise in diversity and population - size fluctuations , and it generates a sequence of quasi - steady communities in the form of simple food webs .
these communities are quite resilient toward the loss of one or a few species , which is reflected in different power - law exponents for the durations of communities and the lifetimes of species .
the exponent for the former is near @xmath1 , while the latter is close to @xmath2 .
statistical characteristics of the evolving communities , including degree ( predator and prey ) distributions and proportions of basal , intermediate , and top species , compare reasonably with data for real food webs . |
the cp - conserving minimal supersymmetric standard model ( mssm ) contains three neutral higgs bosons @xcite , the lighter and heavier cp - even scalars @xmath1 and @xmath2 , respectively , and the cp - odd pseudoscalar @xmath3 . in the presence of cp phases ,
the mssm higgs sector is still cp - conserving at born level .
however loop effects , dominantly mediated by third generation squarks , can generate significant cp - violating scalar - pseudoscalar transitions .
thus the neutral cp - odd and cp - even higgs states mix and form the mass eigenstates @xmath4 , @xmath5 , @xmath6 , with no definite cp parities @xcite .
a detailed knowledge of their mixing pattern will be crucial for the understanding of the mssm higgs sector in the presence of cp - violating phases .
the fundamental properties of the higgs bosons have to be investigated in detail to reveal the mechanism of electroweak symmetry breaking .
the general theoretical formalism for higgs mixing with cp phases is well developed @xcite .
the results have been implemented in sophisticated public computer programmes @xcite , that allow for numerical higher order calculations of the higgs masses , widths and couplings .
it was shown that cp - violating phases can lead to a lightest higgs boson with mass of order @xmath7 gev @xcite , which can not be excluded by measurements at lep @xcite .
cp phases can also change the predictions for the higgs decay widths drastically , see e.g. ref .
@xcite for recent calculations for @xmath8 at higher order . in general , angular distributions of higgs decay products will be crucial to probe their cp parities at colliders , for recent studies see , e.g. , ref .
@xcite .
a particularly important feature of the cp - violating higgs sector is a possible resonance enhanced mixing of the states @xmath2 and @xmath3 .
it is well known that states with equal conserved quantum numbers can strongly mix if they are nearly degenerate , i.e. , if their mass difference is of the order of their widths @xcite .
this degeneracy occurs naturally in the higgs decoupling limit of the mssm , where the lightest higgs boson has standard model - like couplings and decouples from the significantly heavier higgs bosons @xcite . in the decoupling limit ,
a resonance enhanced mixing of the states @xmath2 and @xmath3 can occur , which may result in nearly maximal cp - violating effects @xcite .
such effects can only be analyzed thoroughly in processes in which the heavy neutral higgs bosons can interfere as nearby lying , intermediate resonances @xcite .
the relative phase information of the two higgs states would be lost if only their masses , widths and branching ratios are calculated , e.g. , by assuming that they are produced as single , on - shell resonances at colliders . in previous studies of the cp - conserving @xcite and cp - violating higgs sector @xcite
, it was shown that the interferences of the heavy neutral higgs bosons can be ideally tested in @xmath9 collisions .
since the higgs bosons are resonantly produced in the @xmath10-channel , the muon collider is known to be the ideal machine for measuring the neutral higgs masses , widths , and couplings with high precision @xcite . for a systematic classification of cp observables which test the higgs interference , a preparation of initial muon polarizations _ and
_ the analysis of final fermion polarizations will be crucial .
the cp - even and cp - odd contributions of the interfering higgs resonances to observables can be ideally studied if the beam polarizations are properly adjusted @xcite . for example , for final sm fermions @xmath11 , with @xmath12 , polarization observables have been classified according to their cp transformation properties@xcite . for
non - diagonal chargino pair production , the c - odd observables are obtained by antisymmetrizing in the chargino indices of the cross sections , and complete the necessary set of observables @xcite . finally , the well controllable beam energy of the muon collider then allows to study the center - of - mass energy dependence of the observables around the higgs resonances .
this is an advantage over the photon collider , where direct line - shape scans are not possible @xcite . in this work ,
we extend the study of chargino production at the muon collider @xcite to the cp - violating case .
first results for chargino production in the mssm with explicit cp violation in the higgs sector have been reported in ref .
we classify all cp - even and cp - odd observables for chargino production @xmath13 using longitudinally polarized muon beams .
we analyze the longitudinal chargino polarizations by their subsequent leptonic two - body decays @xmath14 with @xmath15 , and the charge conjugated process @xmath16 .
asymmetries in the energy distributions of the decay leptons allow us to classify all cp - even and a cp - odd observables for chargino decay , which probe the chargino polarization .
similar asymmetries can be defined for the energy distributions of @xmath17 bosons , stemming from the decays @xmath18 .
we find that the muon collider provides the ideal testing ground for phenomenological studies of the higgs interferences , although it might only be built in the far future .
our analysis gives a deeper understanding of the higgs mixings , and allows us to test the public codes regarding the relative phase information of the higgs states in the presence of cp - violating phases .
in section [ formalism ] , we give our formalism for chargino production and decay with longitudinally polarized muon beams . in an effective born - improved approach , we include the leading self - energy corrections into the higgs couplings .
we give analytical formulas for the production and decay cross sections and distributions . we show that also the energy distributions of the chargino decay products depend sensitively on the higgs interference . in section [ asymmetriesforpandd ] , we classify the asymmetries of the production cross section and of the energy distributions according to their cp properties . for the production of unequal charginos @xmath19 , we also define a new set of c - odd asymmetries . in section [ numerical results ] , we present a detailed numerical study of the cross sections , chargino branching ratios , and the cp observables .
we analyze their dependence on @xmath20 , on the cp - violating phase @xmath21 of the common trilinear scalar coupling parameter @xmath3 , and on the gaugino and higgsino mass parameters @xmath22 and @xmath23 .
we complete our analysis by comparing the results with the cp observables obtained in neutralino production @xcite .
we summarize and conclude in section [ summary and conclusions ] .
we study cp violation in the higgs sector in pair production of charginos @xmath24 with longitudinally polarized muon beams .
we analyze the production of charginos at center - of - mass energies of the nearly mass degenerate heavy neutral higgs bosons @xmath5 and @xmath6 .
they will be resonantly produced in the s - channel , see the feynman diagrams in fig .
[ fig : reshiggsmix ] .
the significantly lighter higgs boson @xmath4 is also exchanged in the @xmath10-channel , however is suppressed far from its resonance . together with sneutrino exchange @xmath25 in the @xmath26-channel , and @xmath27 and @xmath28 exchange in the @xmath10-channel , they are the non - resonant continuum contributions to chargino production , see the feynman diagrams in fig . [
fig : feynprod ] . to analyze the longitudinal chargino polarization , we consider the subsequent cp - conserving but p - violating leptonic two - body decay of one of the charginos @xmath29 and the charge conjugated process @xmath30 .
we will focus on the case @xmath31 .
however our results can be extended to @xmath32 , and for the chargino decay into a @xmath17 boson @xmath33 for which we give the relevant formulas in appendix [ chardecay ] .
[ fig : charginoprod.res ] cp violation in the mssm higgs sector is induced by scalar - pseudoscalar transitions at loop level .
the mixing is given by the symmetric and complex higgs mass matrix @xcite @xmath34 - \hat\sigma_{hh}(p^2 ) & \ m_{h}^2 - \hat\sigma_{hh}(p^2 ) \ & - \hat\sigma_{ha}(p^2 ) \\[2 mm ] - \hat\sigma_{ha}(p^2 ) & - \hat\sigma_{ha}(p^2 ) & \ m_{a}^2 - \hat\sigma_{aa}(p^2 ) \
\end{array } \right ) \label{eq : weisskopf}\ ] ] at momentum squared @xmath35 in the tree - level basis of the cp eigenstates @xmath36 .
here @xmath37 , with @xmath38 , are the renormalized self energies of the higgs bosons at one loop , supplemented with higher - order contributions , see ref .
we neglect mixings with the goldstone boson @xmath39 and the @xmath27 boson .
their contribution to the mass matrix is a two - loop effect , see discussion in ref .
their effect on transition amplitudes , which enters already at the one - loop level , is expected to be numerically small , see ref .
and @xmath27 turned out to be as large as those from the non - resonant @xmath1@xmath2 and @xmath1@xmath3 mixing contributions , this would not result in a significant change of our numerical results . ] the higgs propagator matrix @xmath40^{-1},\ ] ] has complex poles at @xmath41 , @xmath42 , where @xmath43 and @xmath44 are the mass and width of the higgs boson mass eigenstate @xmath45 , respectively , with convention @xmath46 . at fixed @xmath47
, we may diagonalize the higgs mass matrix @xmath48 , and consequently the propagator @xmath49 , by a complex orthogonal matrix @xmath50 , @xmath51 the diagonal elements of the propagator @xmath52 are given in the breit - wigner form by @xmath53_{kk } & = & \frac{-i}{s- [ { \rm \bf m_d}(s)]_{kk } } \equiv \frac{-i}{s - m^{\prime 2}_{h_{k}}+ im^{\prime}_{h_{k}}\gamma^{\prime}_{h_{k } } } \equiv \delta^\prime(h_{k } ) , \label{prophiggs_d}\end{aligned}\ ] ] which defines the @xmath10-dependent real parameters @xmath54 and @xmath55 . in the limit @xmath56 , these parameters are the on - shell mass and decay width of the higgs boson @xmath57 , @xmath58 and @xmath59 , respectively . if the @xmath10-dependence of the mass matrix @xmath48 is weak around the resonance @xmath60 of a higgs boson @xmath45 , the @xmath61-element of the propagator can be approximated by @xmath62 the approximations @xmath63 and @xmath64 hold for instance , if no thresholds open around the higgs resonance . in our numerical calculations ,
we evaluate the mass matrix @xmath65 and the diagonalization matrix @xmath66 at fixed @xmath67 with the program feynhiggs 2.5.1 @xcite .
we find that the momentum dependence of @xmath65 is weak around the resonances , and thus @xmath68 to a very good approximation .
for the following discussions , we set @xmath69 for simplicity , and drop the prime index of the propagator in our notation .
the amplitude for chargino production in muon - antimuon annihilation via higgs exchange can be written in the general form @xmath70 & = & \gamma_{}^{(\chi ) } c^{-1}\;c \ , \delta(s ) \,c^{-1 } \ ; c \ ; \gamma_{}^{(\mu ) } \\[2 mm ] & = & \gamma_{}^{(\chi ) } c^{-1 } _ { } \delta_{d}(s ) \ , c_{}\ , \gamma_{}^{(\mu ) } = \gamma_{\rm eff}^{(\chi ) } \delta_{d}(s ) \gamma_{\rm eff}^{(\mu ) } , \label{eq : generaltp}\end{aligned}\ ] ] where @xmath71 and @xmath72 are the one - particle irreducible ( 1pi ) higgs vertices to charginos and muons , respectively .
we have used the basis transformation of the higgs propagator @xmath73 at fixed center - of - mass energy @xmath10 , eq .
( [ eq : diagonalization ] ) , which leads to the definition of the effective 1pi vertices @xmath74 and @xmath75 .
the higgs mixing corrections are thus effectively included via the diagonalization matrix @xmath76 into the 1pi vertices . while the diagonalization of the higgs boson propagator matrix in eqs .
( [ eq : generaltphha])-([eq : generaltp ] ) is not strictly necessary , it yields the transition amplitudes in a diagonalized form . ) , as in , e.g. , refs .
@xcite , for similar studies of standard model fermion pair production . ]
in addition of simplifying the technical evaluation , in the new basis the higgs boson states are approximately the pole - mass eigenstates , as discussed in the preceding section .
we thus obtain compact analytic formulas for the squared amplitudes of the higgs exchange contributions , which we explicitely give in appendix [ ab_coefficients ] .
if the higgs bosons are nearly mass degenerate , the radiative corrections to the propagator matrix are strongly enhanced by the higgs mixing , and constitute the dominant contributions to the transition amplitude @xmath77 ( [ eq : generaltp ] ) .
thus we only include these radiative corrections and neglect specific vertex corrections .
the effective higgs couplings to the initial muons , @xmath78 , and the final charginos , @xmath79 , are then obtained by transforming the tree level couplings @xmath80 and @xmath81 , respectively , with the matrix @xmath76 , see eq .
( [ eq : diagonalization ] ) , @xmath82 c_{l , r}^{h_k\chi_i\chi_j } & = & \tilde c_{k\alpha } c_{l , r}^{h_\alpha\chi_i\chi_j},\quad h_\alpha = h , h , a . \label{eq.hchi.eff.c}\end{aligned}\ ] ] the tree level higgs couplings are defined and discussed , e.g. , in detail in refs .
note that the final state chargino couplings transform with @xmath83 .
if the non - physical phases of the higgs bosons are chosen appropriately , the matrix @xmath76 can be made complex orthogonal , which implies @xmath84 @xcite . with the born - improved effective couplings of eqs .
( [ eq.hmu.eff.c ] ) and ( [ eq.hchi.eff.c ] ) , we can write the higgs exchange amplitude for chargino production @xmath85\phantom{.}\nonumber\\[3 mm ] \times \left [ \bar u(p_{\chi^+_j})\left(c_l^{h_k\chi_i\chi_j}p_l + c_r^{h_k\chi_i\chi_j}p_r \right ) v ( p_{\chi^-_i } ) \right ] \label{thiggsnohel}\end{aligned}\ ] ] in its born - improved form . the amplitudes and lagrangians for non - resonant @xmath28 , @xmath27 and @xmath25 exchange are given in appendix [ nonresamplitudes ] , and the lagrangians for the chargino decays , eqs .
( [ decayell ] ) or ( [ decayw ] ) , are given in appendix [ chardecay ] . in order to calculate the squared amplitude for chargino production ( [ production ] ) and the subsequent decay of chargino @xmath86 , eqs .
( [ decayell ] ) or ( [ decayw ] ) , with initial beam polarizations and complete spin correlations , we use the spin density matrix formalism @xcite . following the detailed steps as given in appendix [ density matrix formalism ] , the squared amplitude in this formalism
can be written as @xmath87 with the propagator @xmath88 of the decaying chargino , see eq .
( [ eq : charginopropagator ] ) . here
@xmath89 denotes the unpolarized production of the charginos and @xmath90 the unpolarized decay .
the corresponding polarized terms are @xmath91 and @xmath92 , and their product in eq .
( [ eq : tsquare ] ) describes the chargino spin correlations between production and decay . with our choice of the spin vectors ,
see eqs .
( [ rhop ] ) and ( [ rhod ] ) , @xmath93 is the longitudinal polarization of @xmath86 , @xmath94 is the transverse polarization in the production plane , and @xmath95 is the polarization perpendicular to the production plane .
the terms @xmath90 and @xmath96 for chargino decay are given in appendix [ chardecay ] .
the expansion coefficients of the squared chargino production amplitude @xmath97 ( [ eq : tsquare ] ) subdivide into contributions from the higgs resonances ( @xmath98 ) , the continuum ( @xmath99 ) , and the resonance - continuum interference ( @xmath100 ) , respectively , with @xmath101 the continuum contributions @xmath102
, @xmath103 are those from the non - resonant @xmath28 , @xmath27 and @xmath104 exchange channels , which we give in appendix [ nonresamplitudes ] .
the contributions from @xmath4 exchange are suppressed by @xmath105 .
the dependence of the resonant contributions @xmath106 and @xmath107 on the longitudinal @xmath108 and @xmath109 beam polarizations @xmath110 and @xmath111 , respectively , is given by and @xmath112 to the transverse polarizations of the chargino vanish for scalar higgs bosons exchange in the @xmath10-channel .
note that there is no interference contribution to the coefficients @xmath89 and @xmath113 . ]
@xmath114 this expression is useful for analyzing the cp properties of @xmath106 and @xmath107 .
the coefficients @xmath115 and @xmath116 , given explicitly in appendix [ ab_coefficients ] , are functions of products of the higgs couplings to muons and charginos , eqs .
( [ eq.hmu.eff.c ] ) and ( [ eq.hchi.eff.c ] ) .
the coefficients also include the product of higgs boson propagators , eq .
( [ prophiggs_d ] ) , which strongly depend on the center - of - mass energy , as well as on the cp phases , which enter via the diagonalization matrix @xmath76 , see the transformation in eq .
( [ eq : diagonalization ] ) .
.[coeffs : candp ] charge ( c ) and parity ( p ) properties of the coefficients @xmath117 , @xmath118 , @xmath119 , and @xmath120 , see eqs .
( [ eq.pr ] ) and ( [ eq.sr ] ) , and the corresponding observables for their determination , as defined in section [ asymmetriesforpandd ] .
the four c - odd observables vanish for diagonal chargino production , @xmath121 . [ cols="^,^,^,^,^",options="header " , ] we center our numerical discussion around scenario cp@xmath122 , defined in table [ scenarioa ] .
inspired by the benchmark scenario cpx @xcite for studying enhanced cp - violating higgs - mixing phenomena , we set @xmath123 , @xmath124 , and a non - vanishing phase @xmath125 .
we thus obtain large contributions from the trilinear coupling parameter @xmath126 of the third generation to the higgs sector , both cp - conserving and cp - violating .
the cp - violating effects are maximized if the mixing of the higgs states with different cp parities is resonantly enhanced .
this can happen naturally in the higgs decoupling limit .
resonant higgs mixing occurs when the diagonal elements @xmath127 and @xmath128 of the higgs mass matrix @xmath48 , eq .
( [ eq : weisskopf ] ) , are similar in size , and thus their difference is of the same order as the off - diagonal element @xmath129 .
when @xmath130 $ ] changes sign , we interpret this condition as a level crossing of the cp eigenstates @xmath2 and @xmath3 @xcite .
the difference of the imaginary parts can be small , @xmath131 , if the higgs channels into heavy squarks are closed , as in our scenario cp@xmath122 .
thus we obtain a large @xmath2@xmath3 mixing even for moderate values of the cp - violating scalar - pseudoscalar self - energy transitions @xmath129 .
this also means that , in contrast to the benchmark scenario cpx @xcite , the higgs mixing can be large even for small values of @xmath22 .
we thus choose @xmath132 and @xmath133 of similar size in order to enhance the branching ratios of the higgs bosons into lighter charginos , which are large only for mixed gaugino - higgsino charginos .
we give the masses of the higgs bosons , charginos , sneutrinos , light neutralinos , and the widths of the higgs bosons for scenario cp@xmath122 in table [ scenarioamasses ] , where we also list the branching ratios for chargino @xmath134 , and the decay factors @xmath135 ( [ etatau ] ) and @xmath136 ( [ etaw ] ) . for non - diagonal chargino production
we consider smaller values of @xmath22 and @xmath23 than in scenario cp@xmath122 , in order for this process to be kinematically allowed , while keeping the same remaining parameters . for the scenario cp@xmath122
, we analyze the dependence of the asymmetries and the cross sections on the center - of - mass energy @xmath138 .
the cp - even and cp - odd observables exhibit a characteristic @xmath20 dependence , mainly given by the product of higgs boson propagators @xmath139 , see eq .
( [ deltadelta ] ) .
a muon collider will have a precise beam energy resolution , and thus enables detailed line - shape scans . in fig .
[ fig : sqrts](a ) , we show the cp - odd production asymmetry @xmath140 ( [ eq : aprodcepo ] ) for chargino production @xmath141 as a function of @xmath138 around the heavy higgs resonances @xmath5 and @xmath6 . at the peak value , @xmath142 gev ,
the interference of the two nearly degenerate higgs bosons is maximal , leading to an asymmetry of up to @xmath143 .
the asymmetry measures the difference of the chargino production cross section @xmath144 for equal positive and negative muon beam polarizations @xmath145 , see eq .
( [ eq : aprodcepo ] ) . in fig .
[ fig : sqrts](b ) , we show the beam polarization averaged cross section @xmath146 , see eq .
( [ eq : sigmacepe ] ) , for @xmath147 ( solid ) , @xmath148 ( dotted ) , and @xmath149 ( dash - dotted ) .
we can observe that the splitting of the two resonances is increased in the presence of cp - violating phases in this scenario .
for @xmath147 and @xmath149 the two resonances are clearly visible in the line shape of the polarization averaged cross section @xmath146 , whereas it assumes the form of a single resonance for @xmath148 , where the higgs bosons are extremely degenerate , see fig .
[ fig : sqrts](b ) .
the higgs boson interference in chargino production also leads to cp - odd and cp - even contributions to the average longitudinal chargino polarizations . in order to analyze the @xmath150 polarization
, one can measure the cp - odd asymmetry @xmath151 ( [ eq : adeccepo ] ) , and the cp - even asymmetry @xmath152 ( [ eq : adecceope ] ) , of the energy distributions of the decay particle @xmath153 in the chargino decay @xmath154 or @xmath155 , respectively . for simplicity , we discuss only the decay into an electron sneutrino , @xmath156 , i.e. @xmath157 . with our choice of slepton sector parameters ,
i.e. , approximate rge relations for the masses , see eqs .
( [ mslr])-([msn ] ) , and a vanishing trilinear coupling for the first two slepton generations , the same asymmetries are obtained for the decay into a muon sneutrino , @xmath158 .
the asymmetries for the decay into a tau sneutrino , @xmath159 , or into a @xmath17 boson , @xmath160 , are obtained by taking into account the decay factors @xmath161 ( [ etatau ] ) and @xmath162 ( [ etaw ] ) , respectively , see eqs .
( [ eq : adecceopedep ] ) and ( [ eq : adeccepodep ] ) . for the decay @xmath163
, we show the @xmath138 dependence of the cp- and cp@xmath164-odd asymmetry @xmath165 in fig .
[ fig : sqrts](c ) for @xmath147 .
the cp- and cp@xmath164-even asymmetry @xmath166 is shown in fig .
[ fig : sqrts](d ) , both for @xmath148 and @xmath147 .
the phase @xmath21 tends to increase the mass splitting of the higgs resonances .
their overlap is now reduced , leading in general to a suppression of the cp - even asymmetry @xmath167 , in particular at the mean energy of the resonances @xmath168 , see the solid line in fig .
[ fig : sqrts](d ) . on the contrary ,
the larger higgs splitting increases the cp - odd asymmetries @xmath165 and @xmath140 .
all asymmetries for production and decay vanish asymptotically far from the resonance region .
the continuum contributions from smuon , photon and @xmath27 exchange to the difference of the cross sections and to the average chargino polarization cancel in the numerator , but contribute in the denominator of the corresponding asymmetries , see their definitions in section [ asymmetriesforpandd ] .
( 16,13 ) ( -3,-10.3 ) ( 5.3,-10.3 ) ( 0,7.)(a ) ( 8,7.)(b ) ( -3,-17.3 ) ( 5.3,-17.3 ) ( 0,0.15)(c ) ( 8,0.15)(d ) in the following sections , we analyze the dependence of the production cross section and the asymmetries on @xmath169 and @xmath21 , and finally on @xmath23 and @xmath22 , fixing all remaining parameters to those of scenario cp@xmath122 .
we fix the center - of - mass energy to @xmath168 , where we expect the largest cp - odd asymmetries @xmath140 and @xmath165 , see figs .
[ fig : sqrts](a ) and ( c ) , respectively . for consistency
, we also choose @xmath168 for the discussion of the cp - even decay asymmetry @xmath167 although it is generally suppressed at this value if cp is violated .
we analyze the dependence of the cp asymmetries on the phase @xmath21 of the trilinear coupling @xmath126 , which is the only source of cp violation in our study .
the cp - odd asymmetries , @xmath140 and @xmath165 see fig .
[ fig : phi](a ) , are approximately maximal , if the mixing of the higgs states is resonantly enhanced , as discussed in section [ sec : enhancement ] .
the degeneracy in the neutral higgs bosons , however , is lifted by the @xmath2@xmath3 mixing , as can be observed from fig .
[ fig : phi](c ) .
a splitting of the order of the higgs widths @xmath170 , shown in fig .
[ fig : phi](d ) , leads to large absorptive phases , which are necessary for the presence of cp@xmath171-odd observables .
the increased higgs mass splitting leads , however , also to lower peak cross sections , and thus to a lower polarization averaged cross section @xmath146 , which we show in fig .
[ fig : phi](b ) for common muon beam polarizations @xmath172 .
( 16,13 . )
( -3,-10.3 ) ( 5.4,-10.3 ) ( .0,7.15)(a ) ( 8.,7.15)(b ) ( -3,-17.3 ) ( 5.4,-17.3 ) ( .0,0.15)(c ) ( 8.,0.15)(d ) ( 2.8,11.8)@xmath140 ( 4.9,11.0)@xmath165 ( 1.5 , 9.0)@xmath167 the asymmetries and cross sections for negative @xmath21 can be obtained from symmetry considerations . since the complex trilinear coupling @xmath126 is the only source of cp violation in our analysis , the cp - odd asymmetries @xmath140 and @xmath165 are odd with respect to the transformation @xmath173 , while the cp - even asymmetry @xmath167 , and the averaged cross section @xmath146 , are even . in fig .
[ fig : asymm.aphia ] , we show contour lines of the averaged cross section @xmath146 and the asymmetries in the @xmath21@xmath169 plane .
the largest cp - odd asymmetries @xmath140 and @xmath165 are obtained for @xmath174 tev . for larger values of @xmath169 ,
the lighter stops become kinematically accessible and @xmath5 decays dominantly into @xmath175 pairs .
this leads to a suppression of the chargino production cross section , and also to a suppression of the higgs mixing for small @xmath22 .
we therefore restrict our discussion to @xmath176 tev .
( 16,15.5 ) ( -2.15,-5.9 ) ( 6,-5.9 ) ( 1,8.15)(a ) ( 9,8.15)(b ) ( -2.15,-13.9 ) ( 6,-13.9 ) ( 1,0.15)(c ) ( 9,0.15)(d ) ( 0.7,14.9)averaged cross section @xmath146 in @xmath177 ( 9.8,14.9)asymmetry @xmath140 in % ( 1.7 , 6.9)asymmetry @xmath167 in % ( 9.8 , 6.9)asymmetry @xmath165 in % as we have observed in fig .
[ fig : phi](a ) , the cp - even asymmetry @xmath167 is in general larger in the cp - conserving limit .
this can be also seen in fig .
[ fig : asymm.aphia](c ) , where the maximum of the asymmetry is obtained for @xmath178 , and @xmath179 . however , this is rather coincidental , and is due to the exact degeneracy of the higgs bosons @xmath2 and @xmath3 .
the couplings of the higgs bosons to the charginos strongly depend on the gaugino - higgsino composition of the charginos , which is mainly determined by the values of @xmath22 and @xmath23 .
for chargino production @xmath180 , we show the cp - odd asymmetry @xmath140 ( [ eq : aprodcepo ] ) in the @xmath22@xmath23 plane in fig .
[ fig : prod11](a ) with @xmath172 .
the absolute maximal value of the asymmetry is constrained by the degree of beam polarization . for @xmath172 the maximum would be @xmath181 , as follows from eq .
( [ eq : amax ] ) .
we observe in fig .
[ fig : prod11](a ) that the asymmetry reaches more than @xmath182 near the chargino production threshold , where the coefficient @xmath183 ( [ eq.a1 ] ) receives large spin - flip contributions . for smaller @xmath22 and @xmath23 ,
the interplay between unitarity and cpt symmetry @xcite suppresses @xmath140 , see ref .
@xcite for a discussion on the related process @xmath184 .
in addition , the higgs boson widths are increased , since decay channels into light neutralinos and charginos open .
the larger widths result in a larger overlap of the higgs resonances , which reduces the absorptive phases in our scenario , and consequently suppresses the cp@xmath185-odd asymmetry @xmath140 . in fig .
[ fig : prod11](c ) , we show the beam polarization averaged cross section @xmath146 ( [ eq : sigmacepe ] ) for chargino production @xmath180 , which reaches up to @xmath186 pb .
in addition the cross section shows no @xmath187-wave suppression at threshold , since @xmath5 and @xmath6 are mixed cp eigenstates .
the corresponding statistical significance , see eq .
( [ eq : saprodcepo ] ) , shown in fig .
[ fig : prod11](b ) for an integrated luminosity of @xmath188 , is largest near threshold . in fig .
[ fig : prod11](d ) , we show the branching ratio for the chargino decay @xmath189 , which reaches more than @xmath190 , as the leptonic decay modes into @xmath22 and @xmath191 .
the main competing channels are the decay into the @xmath17 boson , and those into sleptons @xmath192 , @xmath193 , and @xmath194 , see eq .
( [ chardecaymodes ] ) .
the chargino decay @xmath195 is dominating , see the branching ratios in table [ scenarioamasses ] , and is the only two - body decay channel for @xmath196 , see the contour of the kinematical threshold in fig .
[ fig : prod11](d ) .
( 16,15.5 ) ( -2.15,-5.9 ) ( 6,-5.9 ) ( 1,8.15)(a ) ( 9,8.15)(b ) ( -2.15,-13.9 ) ( 6,-13.9 ) ( 1,0.15)(c ) ( 9,0.15)(d ) ( 1.4,14.9 ) asymmetry @xmath140 in % ( 10.3,14.9)significance @xmath197 ( 0.7 , 6.9)averaged cross section @xmath146 in @xmath177 ( 10 . , 6.9)@xmath198 in % ( 16,7.5 ) ( -2.15,-13.9 ) ( 6,-13.9 ) ( 1.0,0.15)(a ) ( 9.0,0.15)(b ) ( 1.7,6.9)asymmetry @xmath165 in % ( 9.8,6.9)asymmetry @xmath167 in % for the chargino decay @xmath199 , we show the cp - odd and cp - even decay asymmetries @xmath165 ( [ eq : adeccepo ] ) , and @xmath167 ( [ eq : adecceope ] ) , in figs . [ fig : dec.asym.e](a ) and ( b ) , respectively .
as discussed before , the cp - even asymmetry @xmath167 is suppressed by cp - violating effects due to the smaller overlap of the resonances at @xmath168 .
therefore we only find large values of @xmath167 for light neutralinos and charginos in the lower left corner of fig .
[ fig : dec.asym.e](b ) , where the larger higgs widths counter the effect of the larger higgs mass difference . on the contrary , in that region the cp - odd asymmetry @xmath165
is reduced due to smaller absorptive phases .
finally , at threshold the longitudinal polarization of the chargino @xmath107 ( [ eq.sr ] ) vanishes , and thus also both decay asymmetries , as follows from eqs .
( [ eq.b0 ] ) and ( [ eq.b1 ] ) .
the statistical significances of the cp - odd and cp - even decay asymmetries @xmath165 , and @xmath167 , see eqs .
( [ eq : sadecceopoe ] ) , ( [ eq : sadecceopeo ] ) , reach at most @xmath200 and @xmath201 respectively , for an integrated luminosity of @xmath188 .
thus their measurement will be challenging .
nonetheless , only these asymmetries allow to disentangle the contributions @xmath202 and @xmath203 to the longitudinal chargino polarization @xmath107 ( [ eq.sr ] ) .
however these asymmetries can also be measured in the decays @xmath204 , and @xmath205 .
note that the corresponding decay asymmetries are smaller due to the decay factors @xmath206 , which are of the order @xmath207 and @xmath208 .
a measurement of the asymmetries in these decay channels is more involved , since the @xmath17 and @xmath191 reconstruction efficiencies have to be taken into account .
the results obtained for diagonal chargino production @xmath180 in the previous section have similarities with the corresponding asymmetries for neutralino production , see ref .
@xcite , and ref .
@xcite for a detailed review .
for instance , we have obtained similar line - shapes of the cp - odd production asymmetry , @xmath140 , eq .
( [ eq : aprodcepo ] ) , as well as of the cp - odd and cp - even asymmetries for the chargino decays @xmath165 , eq .
( [ eq : adeccepo ] ) , and @xmath167 , eq .
( [ eq : adecceope ] ) , respectively , compared with the analogous asymmetries for neutralino production .
the asymmetries also show a very similar dependence on the common complex trilinear coupling parameter @xmath3 , as well as on the parameters @xmath22 and @xmath23 . in both processes ,
the production asymmetry is enhanced at threshold , whereas the decay asymmetries show a distinct @xmath187-wave suppression .
the similarities are due to similar kinematics , i.e. , production of spin - half charginos or neutralinos , as well as dynamics , i.e. , higgs interference in the production of a c - even final state of massive gauginos and higgsinos .
note that for neutralino or diagonal chargino production only c - even production asymmetries can be defined , whereas after including the c - odd decay , only c - odd decay asymmetries can be probed , see also the detailed discussions in section [ asymmetriesforpandd ] . for non - diagonal chargino production , in addition two c - odd production asymmetries , as well as two c - even decay asymmetries , are accessible , which we discuss in the following section . for non - diagonal chargino production , @xmath210 , we study the @xmath22@xmath23 dependence of the cross section , and of the cp - odd and cp - even production asymmetries @xmath211 , eq .
( [ eq : aprodcepo ] ) , and @xmath212 , eq .
( [ eq : aprodcopo ] ) , respectively . the cp - odd asymmetry @xmath211 is largest at threshold , where it reaches up to @xmath213 , see fig .
[ fig : prod12](a ) .
since the beam polarization averaged cross section reaches up to @xmath214 fb in that region , we obtain a significance of @xmath215 , for an integrated luminosity of @xmath188 , see fig .
[ fig : prod12](c ) .
the cp - even production asymmetry exhibits a characteristic @xmath216 symmetry , see fig .
[ fig : prod12](d ) .
the higgs chargino couplings transform as @xmath217 ( [ eq.hchi.eff.c ] ) under @xmath216 , resulting in a sign change of @xmath212 @xcite .
consequently , for @xmath218 the asymmetry vanishes , see the zero contour in fig .
[ fig : prod12](d ) . on the other hand ,
the asymmetry is nearly maximal for @xmath219 or @xmath220 , see the upper left and lower right corners of fig .
[ fig : prod12](d ) .
the maximum absolute value would be @xmath221 , see eq .
( [ eq : amax ] ) with @xmath222 .
we do not show the cp - odd production asymmetry @xmath223 , eq .
( [ eq : aprodcope ] ) , which is only of order @xmath224 near threshold .
( 16,15.5 ) ( -2.15,-5.9 ) ( 6,-5.9 ) ( 1,8.15)(a ) ( 9,8.15)(b ) ( -2.15,-13.9 ) ( 6,-13.9 ) ( 1,0.15)(c ) ( 9,0.15)(d ) ( 0.7,14.9)averaged cross section @xmath225 in @xmath177 ( 9.8,14.9 ) asymmetry @xmath211 in % ( 2.4 , 6.9)significance @xmath226 ( 9.8 , 6.9 ) asymmetry @xmath212 in % for the subsequent chargino decay @xmath227 , we show the cp - even decay asymmetries @xmath228 , eq . ( [ eq : adecceope ] ) , and @xmath229 , eq . ( [ eq : adeccopo ] ) , in fig . [ fig : dec2.asym.e ] .
both asymmetries are large in a wide region of the @xmath22@xmath23 parameter space .
the asymmetry @xmath229 also exhibits the characteristic @xmath216 symmetry , see fig .
[ fig : dec2.asym.e](b ) , as discussed above for the production asymmetry @xmath212 .
the decay asymmetry however has also contributions from the continuum polarization @xmath230 , see eq .
( [ eq : adeccopodep ] ) .
thus the zero contour in fig .
[ fig : dec2.asym.e](b ) gets slightly shifted from @xmath231 .
since the continuum contributions are small compared to those from the resonance , the @xmath216 symmetry of @xmath229 , however , still holds approximately .
the chargino @xmath232 two - body decays are open in the entire @xmath22@xmath23 plane , with branching ratios up to br@xmath233 .
other competing channels are @xmath234 and @xmath235 , with branching ratios of also up to @xmath236 . for these decay channels
the asymmetries @xmath237 and @xmath238 are accessible . here , the appropriate decay factors @xmath239 ( [ etaw ] ) and @xmath240 ( [ etatau ] ) , respectively , have to be taken into account . the cp - odd decay asymmetries @xmath241 , eq . ( [ eq : adeccepo ] ) , and @xmath242 , eq .
( [ eq : adecceope ] ) , do not exceed @xmath243 .
( 16,7.5 ) ( -2.15,-13.9 ) ( 6,-13.9 ) ( 1.0,0.15)(a ) ( 9.0,0.15)(b ) ( 1.7,6.9)asymmetry @xmath228 in % ( 9.8,6.9)asymmetry @xmath229 in %
in the mssm with cp violation in the higgs sector , we have studied chargino production at the muon collider around the resonances of the two heavy neutral higgs bosons . for
nearly degenerate neutral higgs bosons , as in the higgs decoupling limit , the cp - violating higgs mixing can be resonantly enhanced , which allows for large cp - violating effects .
we have defined the complete set of cp observables to study the higgs interference and their radiatively induced higgs mixings .
the muon collider with a precisely tunable beam energy is an ideal tool to study the strong @xmath20 dependence of the cp asymmetries . using longitudinally polarized muon beams , we have classified the set of cp - even and cp - odd asymmetries for chargino production , as well as for the chargino decays into leptons .
appropriate observables can be defined for the decays into @xmath17 bosons .
in contrast to the production asymmetries , the asymmetries of the decay probe the longitudinal chargino polarization . due to spin correlations between production and decay ,
the chargino polarization depends sensitively on the higgs interference .
thus a different and independent set of asymmetries sensitive to the higgs couplings can be defined , that complement the production asymmetries . in a numerical study
, we have parametrized cp violation in the higgs sector by a common phase @xmath21 of the trilinear scalar coupling parameter .
we have analyzed the dependence of the asymmetries and cross sections on the complex parameter @xmath3 , as well as on the ( real ) gaugino parameters @xmath22 and @xmath23 .
large cp - violating asymmetries are obtained if the @xmath2-@xmath3 higgs mixing is resonantly enhanced .
we have discussed the conditions necessary for such a resonant mixing .
it occurs naturally in the higgs decoupling limit , when the higgs mass difference is of the order of their widths .
we have identified the appropriate class of scenarios which are adequate to study chargino production and decays . for @xmath137 production , the largest cp - odd production asymmetry can go up to @xmath182 for a degree of beam polarization of @xmath244 .
the cp - odd decay asymmetry for the subsequent chargino decay reaches up to @xmath236 . by comparing the asymmetries to the results which are known in the literature from non - diagonal neutralino production ,
we have observed striking similarities .
they are due to similar kinematics , i.e. , production of spin - half charginos or neutralinos , as well as to the dynamics , i.e. , higgs interference in the production of a c - even final state of fermions .
we have completed the numerical study with the discussion of non - diagonal chargino production . here , the c - odd and cp - even asymmetries are almost maximal and reach more than @xmath245 for a degree of beam polarization of @xmath244 .
we find that chargino and neutralino pair production at a muon collider is complementary to similar studies for the production of standard model fermions .
while @xmath246 production is favored for small values of @xmath0 , third generation fermion production is largest for large @xmath0 .
the strong effect of the cp phases on the higgs mixing , and thus on their masses , leads to a center - of - mass energy dependence of the observables which can be ideally studied at the muon collider .
we conclude that chargino production at the muon collider is the ideal testing ground to analyze and comprehend the phenomenology of mssm higgs mixing in the presence of cp - violating phases .
in particular , the crucial interference effects of the two nearly degenerated heavy neutral higgs resonances can well be analyzed and understood .
the proposed set of cp - sensitive observables allows for a systematic test of the cp nature of the heavy neutral higgs resonances , and their couplings to charginos .
we thank sven heinemeyer , herbi dreiner and karina williams for helpful comments and discussions .
ok was supported by the sfb transregio 33 : the dark universe .
fp has been partially supported by the european community marie - curie research training network under contract mrtn - ct-2006 - 035505 `` tools and precision calculations for physics discoveries at colliders '' .
fp thanks the grid infrastructure of the euforia project ( fp7 contract 211804 ) .
ok thanks the instituto de fsica de cantabria ( ifca ) for kind hospitality , where part of this work was completed .
* appendix *
we use the spin density matrix formalism @xcite for the calculation of the squared amplitudes for chargino production , eq .
( [ production ] ) , and decay channels , eqs .
( [ decayell ] ) and ( [ decayw ] ) . the amplitude for chargino production via resonant higgs exchange , eq .
( [ eq : tsquare ] ) , depends on the helicities @xmath247 of the muons @xmath248 and the helicities @xmath249 of the produced charginos , @xmath250 , @xmath251\phantom{.}\nonumber\\ \times \left [ \bar u(p_{\chi^+_j},\lambda_j)\left(c_l^{h_k\chi_i\chi_j}p_l + c_r^{h_k\chi_i\chi_j}p_r \right ) v ( p_{\chi^-_i},\lambda_i ) \right ] .
\label{thiggs}\end{aligned}\ ] ] we include the longitudinal beam polarizations of the muon - beams , @xmath252 and @xmath253 , with @xmath254 in their density matrices @xmath255 \label{eq : density2 } \rho^{+}_{\lambda_{+ } \lambda_{+}^\prime } & = & \frac{1}{2}\left(\delta_{\lambda_{+ } \lambda_{+}^\prime } + { \mathcal p}_{+}\tau^3_{\lambda_{+ } \lambda_{+}^\prime}\right),\end{aligned}\ ] ] where @xmath256 is the third pauli matrix .
the unnormalized spin density matrix of @xmath257 production and @xmath86 decay are given by , @xmath258 the amplitude squared for production and decay is then @xmath259 with the chargino propagator @xmath260 the spin density matrices , eqs .
( [ rhop ] ) and ( [ rhod ] ) , can be expanded in terms of the pauli matrices @xmath261 @xmath262 where we have defined a set of chargino spin vectors @xmath263 . in the center - of - mass system , they are @xmath264 we have chosen a coordinate frame such that the momentum of the chargino @xmath265 is @xmath266 with @xmath267 and the triangle function @xmath268 with @xmath269 . inserting the density matrices , eqs .
( [ rhop ] ) and ( [ rhod ] ) , into eq .
( [ tsquare ] ) , gives the amplitude squared in the form of eq .
( [ eq : tsquare ] ) .
we have expressed the resonant contributions to the spin - density matrix coefficients as functions of the longitudinal @xmath108 and @xmath109 beam polarizations , eqs .
( [ eq.pr ] ) and ( [ eq.sr ] ) , @xmath270 the coefficients @xmath115 and @xmath116 on the r.h.s .
are @xmath271 with the sum over the contributions from the higgs bosons @xmath57 , @xmath272 with @xmath273 , respectively , and @xcite @xmath274_{(kl ) } , \label{eq.a0 } \\ a_1^{kl}~&= & \ !
\phantom{+}\ !
\frac{s}{2 } |\delta_{(kl)}| \big [ |c_\mu^- | |c^+_\chi | f_{ij } \cos ( { { \delta_\mu^- } } + { { \delta_\chi^+ } } + { { \delta_\delta } } ) \nonumber\\ & & \phantom { \ ! \frac{s}{2}|\delta_{(kl)}| } -|c_\mu^- | |c_\chi^{rl } | m_i m_j \cos ( { { \delta_\mu^- } } + { { \delta_\chi^{rl } } } + { { \delta_\delta } } ) \big]_{(kl ) } , \label{eq.a1 } \\ b_0^{kl}~&=&\ !
-\frac{s}{4 } |\delta_{(kl)}| \big [ |c_\mu^+ | |c^-_\chi | \sqrt{\lambda_{ij } } \cos ( { { \delta_\mu^+ } } + { { \delta_\chi^- } } + { { \delta_\delta } } ) \big]_{(kl ) } , \label{eq.b0 } \\ b_1^{kl}~&=&\
! - \frac{s}{4 } |\delta_{(kl)}| \big [ |c_\mu^- | |c^-_\chi | \sqrt{\lambda_{ij } } \cos ( { { \delta_\mu^- } } + { { \delta_\chi^- } } + { { \delta_\delta } } ) \big]_{(kl)}. \label{eq.b1}\end{aligned}\ ] ] we have defined the products of couplings , suppressing the chargino indices @xmath275 and @xmath276 , @xmath277_{(kl ) } , \quad \alpha=\mu , \chi , \label{eq.clambda.mp } \\ c_{\chi(kl)}^{rl } & = & c_r^{h_k\chi\chi } c_l^{h_l\chi\chi\ast}+ c_l^{h_k\chi\chi } c_r^{h_l\chi\chi\ast } = \big [ \big]_{(kl ) } , \label{eq.cchi.rl}\end{aligned}\ ] ] the product of the higgs boson propagators ( [ prophiggs ] ) , @xmath278_{(kl ) } , \label{deltadelta}\end{aligned}\ ] ] and the kinematical functions @xmath279 , and @xmath280 , see eq .
( [ triangle ] ) .
we neglect interferences of the chirality violating higgs exchange amplitudes with the chirality conserving continuum amplitudes , which are of order @xmath281 .
the contributions from @xmath4 exchange are small far from its resonance .
note that the longitudinal chargino polarization , parametrized by the coefficients @xmath282 and @xmath283 of @xmath107 , see eqs .
( [ eq.b0 ] ) and ( [ eq.b1 ] ) , vanishes at threshold as @xmath284 .
we have defined @xmath282 and @xmath283 for the longitudinal polarization of the positively charged chargino @xmath265 in the reaction @xmath250 .
the coefficients are the same for the other produced chargino @xmath285 , due to angular momentum conservation .
the non - resonant chargino production ( [ production ] ) proceeds via @xmath28 and @xmath27 boson exchange in the @xmath10-channel , and muon - sneutrino @xmath25 exchange in the @xmath26-channel , see the feynman diagrams in fig .
[ fig : feynprod ] .
the mssm interaction lagrangians are @xcite : @xmath286\mu , \label{zelel}\\[2 mm ] { \scr l}_{\gamma \tilde\chi^+ \tilde\chi^- } & = & - e a_{\nu } \bar{\tilde\chi}^+_i \gamma^{\nu } \tilde\chi_j^+ \delta_{ij},\quad e>0 , \\[2 mm ] { \scr l}_{z^0\tilde\chi^+\tilde\chi^- } & = & \frac{g}{\cos\theta_w}z_{\mu}\bar{\tilde\chi}^+_i\gamma^{\mu } [ o_{ij}^{'l } p_l+o_{ij}^{'r } p_r]\tilde\chi_j^+,\\[2 mm ] { \scr l}_{\ell \tilde\nu\tilde\chi^+ } & = & - g v_{i1}^ { * } \bar{\tilde\chi}_i^{+c } p_l \ell \tilde{\nu}^{*}+\mbox{h.c.},\quad \ell = e,\mu,\end{aligned}\ ] ] with @xmath287 and @xmath288 . the couplings are @xcite @xmath289 o_{ij}^{'l}&=&-v_{i1 } v_{j1}^{*}-\frac{1}{2 } v_{i2 } v_{j2}^{*}+ \delta_{ij } \sin^2\theta_w,\\[2 mm ] o_{ij}^{'r}&=&-u_{i1}^ { * } u_{j1}-\frac{1}{2 } u_{i2}^ { * } u_{j2}+ \delta_{ij } \sin^2\theta_w,\end{aligned}\ ] ] with the weak mixing angle @xmath290 , and the weak coupling constant @xmath291 , @xmath292
. the complex unitary @xmath293 matrices @xmath294 and @xmath295 diagonalize the chargino mass matrix @xmath296 , @xmath297 , with @xmath298 .
the helicity amplitudes for @xmath28 , @xmath27 and @xmath25 exchange are @xmath299 \nonumber\\ & & \times \left [ \bar{u}(p_{\chi^+_j } , \lambda_j ) \gamma_{\nu } v ( p_{\chi^-_i } , \lambda_i)\right],\label{char : t1}\\ t^p_{\lambda_i\lambda_j\lambda_{+}\lambda_{-}}(z)&= & - \frac{g^2}{\cos^2\theta_w}\delta(z ) \left[\bar{v}(p_{\mu^+},\lambda_{+ } ) \gamma^{\nu } ( l_\mu p_l+r_\mu p_r ) u(p_{\mu^-},\lambda_{- } ) \right ] \nonumber\\ & & \times \left[\bar{u}(p_{\chi^+_j } , \lambda_j ) \gamma_{\nu } ( o^{'l}_{ji } p_l + o^{'r}_{ji }
p_r ) v(p_{\chi^-_i } , \lambda_i ) \right ] , \label { char : t2}\\ t^p_{\lambda_i\lambda_j\lambda_{+}\lambda_{-}}(\tilde{\nu})&= & -g^2 v_{j1 } v_{i1}^{*}\delta(\tilde{\nu } ) \left[\bar{v}(p_{\mu^+},\lambda_{+ } ) p_r v(p_{\chi^+_j } , \lambda_j ) \right ] \nonumber\\ & & \times \left[\bar{u}(p_{\chi^-_i } , \lambda_i ) p_l u(p_{\mu^-},\lambda_{-})\right],\label{char : t3 } \end{aligned}\ ] ] with the propagators @xmath300 and @xmath301 , @xmath302 .
we neglect the @xmath27-width in the propagator @xmath303 for energies beyond the resonance .
the feynman diagrams are shown in fig .
[ fig : feynprod ] . for @xmath304 collisions
, the amplitudes are given in ref .
@xcite .
the continuum contributions @xmath102 are those from the non - resonant @xmath28 , @xmath27 and @xmath25 exchange channels .
the coefficient @xmath102 is independent of the chargino polarization .
it can be decomposed into contributions from the different continuum channels @xmath305 with @xmath306 , \\ p(z\tilde \nu)&=&\frac{2g^4}{\cos^2\theta_w } l_\mu c_{l}e_b^2 \delta ( z)\delta ( \tilde \nu)^{\ast}{\rm re}\big\ { v^{\ast}_{i1}v_{j1 } \nonumber\\ & & \times [ o^{'l}_{ij}(e_{\chi_j^+ } e_{\chi_i^-}-2e_bq\cos\theta+q^2\cos^2\theta ) + o^{'r}_{ij}m_{\chi_j^+ } m_{\chi_i^- } ] \big\},\\ p(\tilde \nu\tilde \nu)&= & \frac{g^4}{2}c_{l } e_b^2(e_{\chi_j^+ } e_{\chi_i^-}-2e_bq\cos\theta+q^2\cos^2\theta ) , \end{aligned}\ ] ] with the scattering angle @xmath307 .
the longitudinal beam polarizations are included in the weighting factors @xmath308 for equal beam polarizations @xmath309 , we thus have @xmath310 . for @xmath304 collisions , the terms for @xmath102 in the laboratory system
are also given in refs .
@xcite , however they differ by a factor of @xmath311 in our notation ( [ eq : tsquare ] ) .
note that the term @xmath102 is invariant under @xmath312 exchange at tree level if cp is conserved .
the continuum contributions @xmath313 ( [ contributions ] ) to the longitudinal @xmath314 polarization for chargino production @xmath250 decompose into @xmath315 with @xmath316,\\ \sigma_p^3(z z)&=&\frac{2g^4}{\cos^4\theta_w}|\delta ( z)|^2e_b^2\big [ ( l_\mu^2 c_{l } + r_\mu^2 c_{r } ) ( |o^{'r}_{ji}|^2-|o^{'l}_{ji}|^2)q(e_{\chi_i^-}+e_{\chi_j^+}\cos^2\theta ) \nonumber\\ & & + ( l_\mu^2 c_{l } - r_\mu^2 c_{r } ) 2{\rm re}\big\{o^{'l}_{ji}o^{'r\ast}_{ji}\big\ } m_{\chi_j^+ } m_{\chi_i^-}\cos\theta \nonumber\\ & & + ( l_\mu^2 c_{l } - r_\mu^2c_{r } ) ( |o^{'l}_{ji}|^2+|o^{'r}_{ji}|^2 ) ( q^2+e_{\chi_j^+ } e_{\chi_i^-})\cos\theta\big ] , \\
\sigma_p^3(z\tilde \nu)&=&\frac{2g^4}{\cos^2\theta_w } l_\mu c_{l } e_b^2 \delta ( z)\delta ( \tilde \nu)^{\ast } { \rm re}\big\{v^{\ast}_{j1}v_{i1 } [ o^{'r}_{ji}m_{\chi_j^+ } m_{\chi_i^-}\cos\theta \nonumber\\ & & -o^{'l}_{ji}(q e_{\chi_i^-}-(q^2+e_{\chi_j^+ } e_{\chi_i^-})\cos\theta + qe_{\chi_j^+}\cos^2\theta ) ] \big\},\\ \sigma_p^3(\tilde \nu\tilde \nu)&= & -\frac{g^4}{2}c_{l } |v_{j1}|^2|v_{i1}|^2 |\delta ( \tilde \nu)|^2 e_b^2 \nonumber\\ & & \times [ q e_{\chi_i^-}-(q^2+e_{\chi_j^+ } e_{\chi_i^-})\cos\theta + qe_{\chi_j^+}\cos^2\theta ] .
\label{eq : snuetsneut}\end{aligned}\ ] ] for @xmath304 collisions , the terms for @xmath313 for chargino @xmath314 in the laboratory system are also given in ref .
@xcite , however they differ by a factor of @xmath311 in our notation ( [ eq : tsquare ] ) . in order to obtain the corresponding polarization terms for the other chargino @xmath317
, one has to substitute @xmath318 and @xmath319 in eqs .
( [ eq : gammagamma])-([eq : snuetsneut ] ) ; in addition , one has to change the overall sign @xcite .
note that for the charge conjugated process @xmath320 , the averaged chargino @xmath321 polarization of the continuum , see eq .
( [ eq : sigmabar ] ) , changes sign , i.e. , @xmath322 . for equal beam polarizations @xmath309 ,
the terms obey @xmath323 , see eq .
( [ eq : clcr ] ) .
the expansion coefficients of the chargino decay matrix ( [ rhod ] ) for the chargino decay @xmath324 , with @xmath325 , are @xmath326 the coefficient @xmath327 for the charge conjugated process , @xmath328 , is obtained by inverting the sign of eq .
( [ sigmad ] ) . with these definitions
, we can rewrite the factor @xmath329 , that multiplies the longitudinal chargino polarization @xmath330 in eq .
( [ eq : tsquare ] ) , @xmath331 where we have used @xmath332 similar results can be obtained for the chargino decay into a @xmath333 , @xmath334 , and @xmath17 boson , @xmath335 @xcite .
the corresponding interaction lagrangians are @xcite @xmath336 { \scr l}_{w^-\tilde\chi^+{\tilde{\chi}^0 } } & = & g w^-_\mu \bar{\tilde{\chi}}_k^{0 } \gamma^\mu ( o_{kj}^l p_l+o_{kj}^r p_r ) \tilde{\chi}_j^{+ } { \tilde{\nu}}_\ell + \mbox{h.c.},\end{aligned}\ ] ] respectively , with the couplings @xmath337 and @xmath338 .
the @xmath339 unitary matrix @xmath340 diagonalizes the neutralino mass matrix @xmath341 in the basis @xmath342 with @xmath343 @xcite .
the expansion coefficients of the chargino decay matrix ( [ rhod ] ) for @xmath344 are @xmath345 and those for @xmath346 are @xmath347 \nonumber\\ & & -6 g^2 { \rm re}(o^l_{kj } o^{r\ast}_{kj } ) m_{\chi_j^{\pm } } m_{\chi_k^0 } , \label{dw } \\[1 mm ] \sigma^a_{d } & = & -g^2 ( |o^l_{kj}|^2-|o^r_{kj}|^2 ) \ , \frac { ( m_{\chi_j^{\pm}}^2-m_{\chi_k^0}^2- 2 m_w^2)}{m_w^2 } \ , m_{\chi_j^{\pm } } ( s^a_{\chi_j^{\pm}}\cdot p_w ) .
\label{sigmadw } \end{aligned}\ ] ] the coefficients @xmath348 for the charge conjugated processes , @xmath349 and @xmath350 , are obtained by inverting the signs of eqs .
( [ sigmadtau ] ) and ( [ sigmadw ] ) , respectively . for the chargino decay @xmath18
, the energy limits of the @xmath17 boson are @xmath351 , see eq .
( [ kinlimits ] ) , with @xmath352 the decay factor ( [ edist2 ] ) is @xmath353 with @xmath354 for the chargino decay @xmath355 , the decay factor ( [ edist2 ] ) is obtained from eqs .
( [ drtau ] ) and ( [ sigmadtau ] ) @xmath356 the coefficients @xmath357 , and also @xmath358 ( [ etatau ] ) , depend on the chargino couplings , as well as on the chargino and neutralino masses , which could be measured at the international linear collider ( ilc ) with high precision @xcite . in order to reduce the free mssm parameters , we parametrize the slepton masses with their approximate renormalization group equations ( rge ) @xcite @xmath359 with @xmath360 the common scalar mass parameter the gut scale .
we obtain cross sections and distributions by integrating the amplitude squared @xmath361 ( [ eq : tsquare ] ) over the lorentz invariant phase space element @xmath362 @xmath363 we use the narrow width approximation for the propagator of the decaying chargino .
the approximation is justified for @xmath364 , which holds in our case with @xmath365 .
note , however , that the naive @xmath366-expectation of the error can easily receive large off - shell corrections of an order of magnitude and more , in particular at threshold or due to interferences with other resonant or non - resonant processes . for a recent discussion of these issues ,
see @xcite explicit formulas of the phase space for chargino production ( [ production ] ) and decay ( [ decayell ] ) , can be found , e.g. , in ref .
the cross section for chargino production is @xmath367 with the triangle function @xmath368 ( [ triangle ] ) , and @xmath369 given in eq . ( [ eq : sigmabar ] ) . the integrated cross section for chargino production ( [ production ] ) and subsequent leptonic decay @xmath370 ( [ decayell ] ) is given by @xmath371 the integrated cross section for chargino production and subsequent decay @xmath372 ( [ decayw ] ) is @xmath373
we define the statistical significances of the c - even cross section observable @xmath374 ( [ eq : aprodcepo ] ) by @xcite @xmath375 where @xmath376 denotes the integrated luminosity for chargino production , and @xmath377 is the c and p symmetrized chargino production cross section , defined in eq .
( [ eq : sigmacepe ] ) .
there is a factor @xmath311 appearing in eq .
( [ eq : saprodcepo ] ) , since the asymmetries require two sets of equal beam polarizations @xmath378 .
there is a factor @xmath379 , since two independent cross section measurements are available for @xmath380 , but only one for @xmath121 .
the significances for the c - odd asymmetries @xmath381 , eqs .
( [ eq : aprodcope ] ) , and ( [ eq : aprodcopo ] ) , are defined by @xmath382 note that the c - odd production asymmetries @xmath383 vanish trivially for @xmath121 .
the significances for the cp - odd decay asymmetries @xmath384 , eqs . ( [ eq : adecceope ] ) and ( [ eq : adeccepo ] ) , are defined by @xcite @xmath385 and similarly for the cp - even asymmetries @xmath386 , eqs .
( [ eq : adecceope ] ) and ( [ eq : adeccopo ] ) , @xmath387 with @xmath388 or @xmath17 , and @xmath389 the associated sneutrino or neutralino , respectively .
the effective integrated luminosity is @xmath390 , with the detection efficiency @xmath391 of the leptons or @xmath17 bosons in the chargino decay @xmath392 or @xmath393 , respectively .
there is a factor @xmath394 appearing in the significances , eqs .
( [ eq : sadecceopoe ] ) and ( [ eq : sadecceopeo ] ) , since the asymmetries require two sets of equal beam polarizations @xmath378 , as well as two decay modes , @xmath395 and the charge conjugated decay @xmath396 .
for an ideal detector a significance of , e.g. , @xmath397 implies that the asymmetries can be measured at the statistical 68% confidence level . in order to predict the absolute values of confidence levels
, clearly detailed monte carlo analysis including detector and background simulations with particle identification and reconstruction efficiencies would be required , which is however beyond the scope of the present work .
j. f. gunion and h. e. haber , nucl .
b * 272 * , 1 ( 1986 ) [ erratum - ibid .
b * 402 * , 567 ( 1993 ) ] .
j. f. gunion , h. e. haber , g. l. kane and s. dawson , `` _ the higgs hunter s guide _ '' ( addison - wesley publishing company , redwood city , ca , usa , 1990 ) . m. s. carena and h. e. haber , prog . part .
phys . * 50 * , 63 ( 2003 ) [ arxiv : hep - ph/0208209 ] .
a. djouadi , arxiv : hep - ph/0503173 .
e. accomando _ et al .
_ , arxiv : hep - ph/0608079 . h. fraas , o. kittel and f. von der pahlen , _
`` higgs boson interference in chargino and neutralino production at the muon collider '' _ , in ref .
@xcite , p. 169 .
j. f. gunion , b. grzadkowski , h. e. haber and j. kalinowski , phys .
lett . * 79 * ( 1997 ) 982 [ arxiv : hep - ph/9704410 ] ; + b. grzadkowski , j. f. gunion and j. kalinowski , phys . rev .
d * 60 * ( 1999 ) 075011 [ arxiv : hep - ph/9902308 ] .
d. a. demir , phys .
d * 60 * ( 1999 ) 055006 [ arxiv : hep - ph/9901389 ] ; + s. y. choi , m. drees and j. s. lee , phys .
b * 481 * ( 2000 ) 57 [ arxiv : hep - ph/0002287 ] ; + g. l. kane and l. t. wang , phys .
b * 488 * ( 2000 ) 383 [ arxiv : hep - ph/0003198 ] ; + t. ibrahim and p. nath , phys .
d * 63 * , 035009 ( 2001 ) [ arxiv : hep - ph/0008237 ] ; phys .
d * 64 * ( 2001 ) 035009 [ arxiv : hep - ph/0102218 ] ; phys .
d * 66 * , 015005 ( 2002 ) [ arxiv : hep - ph/0204092 ] ; + s. w. ham , s. k. oh , e. j. yoo , c. m. kim and d. son , phys .
d * 68 * ( 2003 ) 055003 [ arxiv : hep - ph/0205244 ] .
a. pilaftsis and c. e. m. wagner , nucl .
b * 553 * ( 1999 ) 3 [ arxiv : hep - ph/9902371 ] ; + m. s. carena , j. r. ellis , a. pilaftsis and c. e. m. wagner , nucl .
b * 586 * ( 2000 ) 92 [ arxiv : hep - ph/0003180 ] ; + m. s. carena , j. r. ellis , a. pilaftsis and c. e. m. wagner , nucl .
b * 625 * , 345 ( 2002 ) [ arxiv : hep - ph/0111245 ] .
s. y. choi , j. kalinowski , y. liao and p. m. zerwas , eur .
j. c * 40 * ( 2005 ) 555 [ arxiv : hep - ph/0407347 ] . s. heinemeyer , eur . phys .
j. c * 22 * ( 2001 ) 521 [ arxiv : hep - ph/0108059 ] .
g. degrassi , s. heinemeyer , w. hollik , p. slavich and g. weiglein , eur .
j. c * 28 * ( 2003 ) 133 [ arxiv : hep - ph/0212020 ] ; + s. heinemeyer , w. hollik and g. weiglein , eur .
j. c * 9 * ( 1999 ) 343 [ arxiv : hep - ph/9812472 ] ; comput .
commun .
* 124 * ( 2000 ) 76 [ arxiv : hep - ph/9812320 ] . j. s. lee , a. pilaftsis , m. s. carena , s. y. choi , m. drees , j. r. ellis and c. e. m. wagner , comput .
commun .
* 156 * , 283 ( 2004 ) [ arxiv : hep - ph/0307377 ] ; + j. s. lee , m. carena , j. ellis , a. pilaftsis and c. e. m. wagner , arxiv:0712.2360 [ hep - ph ] .
m. carena , j. r. ellis , a. pilaftsis and c. e. m. wagner , phys .
b * 495 * ( 2000 ) 155 [ arxiv : hep - ph/0009212 ] .
r. barate _
[ lep working group for higgs boson searches ] , phys .
b * 565 * , 61 ( 2003 ) [ arxiv : hep - ex/0306033 ] ; + s. schael _ et al .
_ [ aleph collaboration ] , eur .
j. c * 47 * , 547 ( 2006 ) [ arxiv : hep - ex/0602042 ] ; + p. bandyopadhyay , a. datta , a. datta and b. mukhopadhyaya , arxiv:0710.3016 [ hep - ph ] ; + a. belyaev , q. h. cao , d. nomura , k. tobe and c. p. yuan , phys .
lett . * 100 * , 061801 ( 2008 ) [ arxiv : hep - ph/0609079 ] .
k. e. williams and g. weiglein , phys .
b * 660 * ( 2008 ) 217 [ arxiv:0710.5320 [ hep - ph ] ] ; arxiv:0710.5331 [ hep - ph ] .
s. berge , w. bernreuther and j. ziethe , phys .
lett . * 100 * , 171605 ( 2008 ) [ arxiv:0801.2297 [ hep - ph ] ] ; + r. m. godbole , d. j. miller and m. m. muhlleitner , jhep * 0712 * , 031 ( 2007 ) [ arxiv:0708.0458 [ hep - ph ] ] ; + p. s.
bhupal dev , a. djouadi , r. m. godbole , m. m. muhlleitner and s. d. rindani , phys .
lett . * 100 * , 051801 ( 2008 ) [ arxiv:0707.2878 [ hep - ph ] ] ; + s. hesselbach , s. moretti , s. munir and p. poulose , eur .
j. c * 54 * , 129 ( 2008 ) [ arxiv:0706.4269 [ hep - ph ] ] .
a. dobado , m. j. herrero and s. penaranda , eur .
j. c * 17 * , 487 ( 2000 ) [ arxiv : hep - ph/0002134 ] ; + j. f. gunion and h. e. haber , phys . rev .
d * 67 * , 075019 ( 2003 ) [ arxiv : hep - ph/0207010 ] ; + h. e. haber and y. nir , phys .
b * 306 * , 327 ( 1993 ) [ arxiv : hep - ph/9302228 ] ; + h. e. haber , arxiv : hep - ph/9505240 .
b. grzadkowski and j. f. gunion , phys .
b * 350 * , 218 ( 1995 ) [ arxiv : hep - ph/9501339 ] . v. d. barger , m. s. berger , j. f. gunion and t. han , phys .
* 286 * ( 1997 ) 1 [ arxiv : hep - ph/9602415 ] ; + r. casalbuoni , a. deandrea , s. de curtis , d. dominici , r. gatto and j. f. gunion , jhep * 9908 * ( 1999 ) 011 [ arxiv : hep - ph/9904268 ] ; + m. s. berger , phys .
lett . * 87 * , 131801 ( 2001 ) [ arxiv : hep - ph/0105128 ] .
s. dittmaier and a. kaiser , phys .
d * 65 * , 113003 ( 2002 ) [ arxiv : hep - ph/0203120 ] .
h. fraas , f. franke , g. moortgat - pick , f. von der pahlen and a. wagner , eur .
j. c * 29 * ( 2003 ) 587 [ arxiv : hep - ph/0303044 ] .
b. grzadkowski , j. f. gunion and j. pliszka , nucl .
b * 583 * , 49 ( 2000 ) [ arxiv : hep - ph/0003091 ] .
e. asakawa , a. sugamoto and i. watanabe , eur .
j. c * 17 * , 279 ( 2000 ) . v. d. barger , t. han and c. g. zhou , phys . lett .
b * 480 * , 140 ( 2000 ) [ arxiv : hep - ph/0002042 ] .
h. fraas , f. von der pahlen and c. sachse , eur .
j. * c37 * ( 2004 ) 495 [ arxiv : hep - ph/0407057 ] .
o. kittel and f. von der pahlen , phys .
d * 72 * ( 2005 ) 095004 [ arxiv : hep - ph/0508267 ] .
s. y. choi and j. s. lee , phys .
d * 61 * ( 2000 ) 111702 [ arxiv : hep - ph/9909315 ] .
s. y. choi , m. drees , b. gaissmaier and j. s. lee , phys . rev .
d * 64 * ( 2001 ) 095009 [ arxiv : hep - ph/0103284 ] .
j. bernabeu , d. binosi and j. papavassiliou , jhep * 0609 * , 023 ( 2006 ) [ arxiv : hep - ph/0604046 ] .
z. hioki , t. konishi and k. ohkuma , jhep * 0707 * , 082 ( 2007 ) [ arxiv:0706.4346 [ hep - ph ] ] .
h. k. dreiner , o. kittel and f. von der pahlen , jhep * 0801 * ( 2008 ) 017 [ arxiv:0711.2253 [ hep - ph ] ] . v. d. barger , m. s. berger , j. f. gunion and t. han , nucl .
suppl . * 51a * ( 1996 ) 13 [ arxiv : hep - ph/9604334 ] .
proceedings of _ prospective study of muon storage rings at cern _ , eds . b. autin , a. blondel , j. ellis , cern yellow report , cern 99 - 02 , ecfa 99 - 197 , april 30 ( 1999 ) ; + v. d. barger , m. berger , j. f. gunion and t. han , _ in the proceedings of aps / dpf / dpb summer study on the future of particle physics ( snowmass 2001 ) , snowmass , colorado , 30 jun - 21 jul 2001 , pp e110 _ [ arxiv : hep - ph/0110340 ; + a. blondel _ et al . _ , `` _ ecfa / cern studies of a european neutrino factory complex _ , '' cern-2004 - 002 . c. blchinger et al . , higgs working group of the ecfa - cern study on neutrino factory & muon storage rings at cern , _ physics opportunities at @xmath9 higgs factories _ cern - th/2002 - 028 , [ arxiv : hep - ph/0202199 ] ; p. osland and a. vereshagin , phys .
d * 76 * ( 2007 ) 036001 [ arxiv:0704.2165 [ hep - ph ] ] .
k. rolbiecki and j. kalinowski , arxiv:0709.2994 [ hep - ph ] ; + p.
osland , j. kalinowski , k. rolbiecki and a. vereshagin , arxiv:0709.3358 [ hep - ph ] ; + k. rolbiecki and j. kalinowski , arxiv:0710.3318 [ hep - ph ] .
j. s. lee , mod .
a * 22 * , 1191 ( 2007 ) [ arxiv:0705.1089 [ hep - ph ] ] ; + r. m. godbole , s. kraml , s. d. rindani and r. k. singh , phys .
d * 74 * , 095006 ( 2006 ) [ erratum - ibid .
d * 74 * , 119901 ( 2006 ) ] [ arxiv : hep - ph/0609113 ] ; + j. r. ellis , j. s. lee and a. pilaftsis , nucl .
b * 718 * , 247 ( 2005 ) [ arxiv : hep - ph/0411379 ] .
g. a. moortgat - pick , h. fraas , a. bartl and w. majerotto , eur .
j. c * 7 * , 113 ( 1999 ) [ arxiv : hep - ph/9804306 ] .
g. moortgat - pick , `` spin effects in chargino / neutralino production and decay '' , in german , doctoral thesis ( 1999 ) , university of wrzburg , germany .
`` tesla technical design report part iii : _ physics at an e+e- linear collider _ , '' arxiv : hep - ph/0106315 .
a. djouadi , j. lykken , k. monig , y. okada , m. j. oreglia and s. yamashita , `` international linear collider reference design report volume 2 : _ physics at the ilc _ , '' arxiv:0709.1893 [ hep - ph ] .
p. g. harris _ et al .
_ , phys .
* 82 * ( 1999 ) 904 .
b. c. regan , e. d. commins , c. j. schmidt and d. demille , phys .
* 88 * ( 2002 ) 071805 .
m. v. romalis , w. c. griffith and e. n. fortson , phys . rev .
* 86 * ( 2001 ) 2505 [ arxiv : hep - ex/0012001 ] . l. j. hall and j. polchinski , phys .
b * 152 * ( 1985 ) 335 .
k. hagiwara _ et al .
_ , phys .
d * 73 * ( 2006 ) 055005 [ arxiv : hep - ph/0512260 ] .
d. berdine , n. kauer and d. rainwater , phys .
lett . * 99 * , 111601 ( 2007 ) [ arxiv : hep - ph/0703058 ] ; + n. kauer , phys .
b * 649 * , 413 ( 2007 ) [ arxiv : hep - ph/0703077 ] . | we study the pair - production of charginos in the cp - violating minimal supersymmetric standard model at center - of - mass energies around the heavy neutral higgs boson resonances .
if these resonances are nearly degenerate , as it can happen in the higgs decoupling limit , radiatively induced scalar - pseudoscalar transitions can be strongly enhanced .
the resulting mixing in the higgs sector leads to large cp - violating effects , and a change of their mass spectrum . for longitudinally polarized muon beams ,
we analyze cp asymmetries which are sensitive to the interference of the two heavy neutral higgs bosons .
we present a detailed numerical analysis of the cross sections , chargino branching ratios , and the cp observables .
we obtain sizable cp asymmetries , which would be accessible in future measurements at a muon collider . especially for intermediate values of the parameter @xmath0 , where the largest branching ratios of higgs bosons into charginos are expected
, this process allows to analyze the higgs sector properties and its interaction to supersymmetric fermions .
bonn - th-2008 - 06 + * cp - violating higgs boson mixing in chargino production at the muon collider * olaf kittel _ physikalisches
institut der universitt bonn , nussallee 12 , d-53115 bonn , germany _ _ instituto de fsica de cantabria ( csic - uc ) , e-39005 santander , spain _ |
first , we consider a system of non - interacting and weakly interacting particles in a deforming cavity . then we analyze a single particle in a harmonic potential , which is a part of a larger system , and subject to a time - dependent external force .
two additional examples of a classical one - dimensional xy - model and a quantum one - dimensional transverse field ising model will be discussed in the supplementary material . _ single particle in a deforming cavity _ :
let us first consider a very simple system - a single particle bouncing elastically in a cavity .
when the cavity is stationary the energy of the particle is conserved . if the cavity is chaotic there are no other conserved quantities so that in the long - time limit the particle relaxes to a uniform position distribution and an isotropic momentum distribution .
we consider a process where the system is repeatedly driven by deforming the cavity . at the end of each cycle
the cavity comes back to its original shape and the system is allowed to relax in the sense described above ( see fig.[jar2 ] ) . in this setup
the number of degrees of freedom @xmath8 is given by @xmath63 where @xmath64 is the dimensionality of the system . in a single collision with the moving wall the particle s kinetic energy can either increase or decrease .
however , it will always increase on average and eventually the particles velocity will become much greater than the velocity of the wall . then the work per cycle automatically becomes small and the conditions for the validity of the fokker - planck equation are satisfied .
if the cavity is deformed while keeping its volume fixed then a very simple behavior emerges . in this case
it has been shown @xcite that the particles velocity distribution becomes exponential irrespective of the container s shape and the deformation protocol , @xmath65 , where @xmath66 is proportional to time ( number of cycles ) and to the second moment of the velocity of the wall .
moreover if the cavity is sufficiently chaotic successive collisions are uncorrelated and the formalism holds even if the waiting time between cycles approaches zero ( see @xcite for details ) . in this case
the fundamental equation for the energy distribution , @xmath67 , assumes the fokker - planck form of eq . with @xmath68 and @xmath69 where @xmath70 contains information about the mass of the particle , the area or volume of the container and the velocity of the moving walls ( see ref .
@xcite for details ) . using @xmath71
it is easy to verify that eq .
is exactly satisfied with the @xmath7 correction included .
therefore , in the large @xmath64-limit , eq . holds , and we find @xmath72 , consistent with the exponents @xmath73 and @xmath74 . with the @xmath7 correction
included we find : @xmath75 this result also follows from the exactly known single particle distribution @xcite .
this result was also derived for a single light impurity moving in a background of heavy atoms ( lorenz gas ) . in this case
a full microscopic description based on a lorentz - boltzmann equation is possible @xcite .
_ weakly - interacting particles _ : extending the above example , consider @xmath76 weakly interacting particles in a deforming cavity so that @xmath77 .
we assume that during each cycle the particles can be treated as non - interacting , while between cycles the system rethermalizes at a fixed total energy , so that the velocity distribution of the particles becomes maxwell - boltzmann rather than exponential .
a calculation similar to that of @xcite shows that @xmath11 and @xmath20 have the same functional form as in the non - interacting case but with different prefactors .
moreover both @xmath11 and @xmath20 become extensive and the constraint , eq . , is satisfied with the @xmath7 correction becoming negligible
. then eq .
( [ widthbeta ] ) gives the asymptotic result @xmath72 , consistent with the exponents @xmath73 and @xmath74 .
this result is identical to the single particle result in the large @xmath8 limit , despite the very different single particle velocity distributions . in either noninteracting or interacting setups
the two functions , @xmath17 , can be experimentally obtained by measuring the first two cumulants of the work distribution in one cycle : @xmath78 and @xmath79 .
alternatively one can measure the average energy and its variance versus time and determine @xmath11 and @xmath20 as the slopes of these two functions respectively . _ single particle in a time - dependent potential _ :
next we consider a classical particle in a harmonic trap , which is part of a larger system , e.g. a set of @xmath8 identical particles , whose details define @xmath10 and hence @xmath25 .
we assume that the coupling to the rest of the system is weak and unimportant within the duration of a cycle , much like in the weakly - interacting particle gas example above .
in contrast with the previous example , this setup illustrates driving a system with a local perturbation ( which can , however , be applied independently to many different particles ) .
the particle s energy @xmath80 between cycles is given by @xmath81 for simplicity we work in one dimension .
for large @xmath8 the probability distribution for @xmath82 before the drive is @xmath83 .
we consider a driving process which consists of an impulse of magnitude @xmath84 with @xmath85 short enough so that the particle s position does not change appreciably during the drive .
this assumption also guaranties that the coupling to the rest of the system is unimportant during a cycle . under this impulse
the velocity changes according to @xmath86 .
@xmath87 are readily calculated and read @xmath88 verifying fluctuation - dissipation relations ( [ real_ab ] ) .
taking @xmath89 and @xmath90 we find @xmath91 using our previous convention , @xmath92 , we see that @xmath93 .
different values of @xmath94 can be obtained using different impulse forces and systems .
for example , for a system with @xmath95 , such as a fermi liquid or a one - dimensional harmonic system , with @xmath96 we obtain @xmath97 leading to the _ gibbs - like _ regime with @xmath98 .
when @xmath99 we are at the critical regime @xmath48 .
finally for @xmath100 we obtain @xmath101 leading to the _ run - away _ regime where @xmath102 .
the main result of our paper is based on a fluctuation - dissipation relation connecting drift and diffusion of the energy in a driven system ( eq .
( [ real_ab ] ) ) .
this relation is very closely connected to the recently discovered fluctuation theorems .
in fact , in the supplementary material we give a rigorous derivation of eq .
( [ real_ab ] ) using the quantum version of fluctuation theorems , which we extend to our setup of repeated cyclic processes . in the original formulation
the jarzynski relation states that if a system starts from a _ gibbs distribution _ the change in energy of the system , @xmath103 , for cyclic processes obeys the equality @xmath104 . here
the angular brackets denote an average over both different realizations of the process and different initial conditions . as emphasized in ref .
@xcite these relations holds very little information about the first few moments of the distribution of @xmath103 , unless @xmath103 is small . only when it is small , a cumulant expansion of the jarzynski relation up to the second order in @xmath103 recovers the fluctuation - dissipation relation ( [ real_ab ] ) without the @xmath7 correction ( see supplementary material for details ) .
when the energy changes are large its moments are governed by details of the physical process which have to be accounted for ( see @xcite ) . in this paper we ( i )
importantly , overcome the restriction of small energy changes by considering a large change which is a results of many small changes ( leading to eq .
( [ widthbeta ] ) ) and ( ii ) show that relation ( [ real_ab ] ) is independent of the exact form of the initial distribution .
we believe that some of the assumptions of the work can be further relaxed .
in particular , it can be shown that relation ( [ real_ab ] ) is valid for generic ( non - cyclic ) quasi - static process where the system is approximately in a steady state at each moment of time . in this case by @xmath27 is related to the non - adiabatic part of the work @xmath103 .
likewise it is plausible that the assumptions of complete relaxation to a steady state between cycles are not necessary , at least for ergodic systems .
physically these assumptions amount to a loss of correlations between different cycles which is inevitable in ergodic systems .
the assumption of unitary dynamics can be also relaxed .
for example , we can allow measurements of the energy during the protocol , which in quantum systems project the system to one of the energy eigenstates .
such measurements do not invalidate the derivation given in appendix [ extramaterial2 ] .
these points will be addressed in a future study .
the predictions of this paper can be experimentally tested in cold atom systems or in driven nuclear spins .
for example , a trapping potential can be modulated to perform non - adiabatic work on the system .
the resulting energy distribution can then be probed using time of flight experiments .
likewise one can measure energy fluctuations due to a @xmath0 noise in systems of neutral atoms near atom chips or in trapped ions in a setup similar to that discussed in ref . @xcite .
if there is no external cooling or dissipation in the system after its initial preparation ( as is often the case ) then the energy fluctuations induced by the noise should also agree with our predictions .
finally , we comment that energy fluctuations can be measured indirectly through averages or fluctuations of other observables , like the magnetization or correlation functions .
in particular , it is easy to check that the expectation value of every observable @xmath105 and its variance , up to @xmath106 corrections , are given by @xmath107 where @xmath108 is the microcanonical average of @xmath105 at fixed energy @xmath6 and @xmath109 is the variance of @xmath105 in the microcanonical ensemble . thus eq . ( [ widthbeta ] )
will have implications to fluctuations of a wide class of observables in mesoscopic thermally isolated driven systems .
the authors would like to thank g. ortiz for the comment related to a cumulant expansion of the jarzynski equality which plays an important role in the proof .
the authors also acknowledge the support of the nsf dmr-0907039 ( a.p . ) , afosr fa9550 - 10 - 1 - 0110 ( l.d . and a.p . ) , sloan foundation ( a.p . ) , y.k .
thanks the boston university visitors program for its hospitality .
we now outline a particularly simple derivation of eqs . and ( [ real_ab ] ) .
a more rigorous derivation based on fluctuation theorems and the unitarity of the evolution is given in the supplementary material .
our process consists of many repeated cycles during which the control parameter @xmath110 is varied in time , returning to its initial value at the end of each cycle .
we assume that between the cycles the system reaches a steady state ( or a diagonal ensemble @xcite in the quantum language ) so that its state is fully characterized by its energy distribution . in ergodic systems this requirement can be satisfied by waiting between cycles a time which is longer than the relaxation time of the system . in non ergodic ( integrable )
systems this can be achieved by having a long fluctuating time between cycles .
this effectively leads to an additional time averaging which is equivalent to the assumption of starting from a diagonal ensemble .
( for more details about relaxation to asymptotic states in integrable systems see ref .
@xcite and refs .
therein ) . to make this discussion more concrete consider , for example , a compression and expansion of the piston in fig .
[ ovens2 ] according to an arbitrary protocol .
the gas is allowed to relax between the cycles ( when the piston is stationary ) at a fixed energy . for a weakly interacting ergodic gas
such a relaxation implies that the momentum distribution of individual particles assumes a maxwell - boltzmann form together with a randomization of the coordinate distribution . for a noninteracting gas in a chaotic cavity the relaxation implies conservation of the individual energies of each particle and a randomization of the coordinates and directions of their motion . and
finally for noninteracting particles in a regular non - chaotic cavity the relaxation implies a randomization of the coordinates within individual periodic trajectories .
therefore , in the beginning of each cycle there are no correlations between positions and velocities of particles within the available phase space .
if we make an additional assumption about ergodicity then the system between the cycles is fully described by the total energy . as we will see later ,
when we discuss specific examples , this assumption is not always necessary . assuming that during each cycle a small amount of work is carried out on the system the energy distribution @xmath15 can be described by the fokker - planck equation ( [ fp_main ] ) @xcite .
the easiest way to derive the fluctuation - dissipation relation ( [ real_ab ] ) is to note that under very general conditions the only attractor of hamiltonian dynamics is a flat probability distribution for the occupation of different microstates ( see ref .
@xcite for the classical case and ref .
@xcite for the quantum case ) , which is the maximum entropy state .
therefore the energy distribution which is proportional to the many - particle density of states @xmath111 should be stationary under the fokker - planck equation , implying that the current @xmath112 is a constant , which vanishes since @xmath113 for @xmath6 below the ground - state energy . finally we use @xmath114 to obtain eq . .
for a rigorous derivation of this result and the range of its applicability see supplemetary material .
the relation ( [ real_ab ] ) allows us to make general statements about the energy distribution .
in particular , the main result of the paper , eq .
( [ widthbeta ] ) , immediately follows from eqs . and to leading order in an expansion in @xmath7 .
to see this we first multiply eq .
( [ fp_main ] ) by @xmath6 and @xmath115 and integrate over all energies . in this way we obtain the differential equations describing the time evolution of @xmath116 and @xmath117 , where angular brackets stand for averaging over @xmath67 : @xmath118
these two equations can be combined into a single one : @xmath119 if the energy distribution @xmath67 is narrow , as in the case of large systems , we can evaluate the averages above using a saddle - point approximation and eq . . then to order @xmath120 we find : @xmath121 integrating this equation immediately yields eq .
( [ widthbeta ] ) .
let us now comment on the regime of validity of eq .
( [ widthbeta ] ) .
the derivation is based on the fokker - plank equation , eq . , the generalized fluctuation - dissipation relation , eq . , and the saddle - point expansion in eq . .
the validity of the fokker - planck equation relies on the assumption that the work distribution is narrow and the average work for cycle is small .
more specifically this equation is derived from a cumulant expansion of the crook s relation up to second order in the work ( see sec .
[ extramaterial2 ] in supplementary material for details ) .
let us here only mention two necessary conditions justifying the fokker - planck equation and the fluctuation - dissipation relation eq .
( [ real_ab ] ) : ( i ) the third ( and higher order ) cumulant of work per cycle are small @xmath122 .
( ii ) the average work per cycle is smaller than the product of temperature and the specific heat @xmath3 : @xmath123 . as explained in the supplementary material ( sec .
[ extramaterial2 ] ) , if this condition is not satisfied there are corrections of order @xmath124 to eq . .
finally the saddle - point approximation in eq . is justified if the energy fluctuations in the system are small .
this is the case in large or mesoscopic extensive systems .
we also note that our derivation implicitly relies on the assumption of ergodicity within the system .
in particular , we are assuming that @xmath67 is a differentiable function of energy .
in integrable systems this is not necessarily the case @xcite .
then the validity of our results should be checked on a case by case basis .
for example , as we showed above , eq . and the corresponding fokker - planck equation describe the dynamics of a single particle in a cavity . in this case , however , the second term in the rhs of eq . is important and modifies the width of the distribution even if we consider an ensemble of many noninteracting ( and therefore non - ergodic ) particles . 99 i. bloch , j. dalibard , w. zwerger , `` many - body physics with ultracold gases '' , rev .
phys . * 80 * , 885 - 964 ( 2008 ) .
r. blatt and d. j. wineland , `` entangled states of trapped atomic ions '' , nature * 453 * , 1008 - 1015 ( 2008 ) .
petta , j. m. taylor , a. c. johnson , a. yacoby , m. d. lukin , c. m. marcus , m. p. hanson , and a. c. gossard , `` dynamic nuclear polarization with single electron spins '' , phys .
lett . * 100 * , 067601 - 067604 ( 2008 ) .
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 , `` coupling superconducting qubits via a cavity bus '' , nature * 449 * , 443 - 447 ( 2007 ) .
a. polkovnikov , k. sengupta , a. silva , and m. vengalattore , arxiv:1007.5331 , rev .
mod . phys .
( accepted ) .
j. dziarmaga , `` dynamics of a quantum phase transition and relaxation to a steady state '' , adv . in phys . *
59 * , 1063 - 1189 ( 2010 ) .
e. g. dalla torre , e. demler , t. giamarchi , e. altman , `` quantum critical states and phase transitions in the presence of non - equilibrium noise '' , nat . phys . * 6 * , 806 - 810 ( 2010 ) .
f. reif , _ fundamentals of statistical and thermal physics _ , ( waveland pr .
2008 ) . c. jarzynski , `` nonequilibrium equality for free energy differences '' , phys .
78 * , 2690 - 2693 ( 1997 ) .
g. e. crooks , `` nonequilibrium measurements of free energy differences for microscopically reversible markovian systems '' , j. stat . phys . * 90 * , 1481 - 1487 ( 1998 ) .
m. campisi , p. hanggi , p. talkner , `` colloquium : quantum fluctuation relations : foundations and applications '' , arxiv:1012.2268 , rev .
( accepted ) .
n. merhav and y. kafri , `` statistical properties of entropy production derived from fluctuation theorems '' , j. stat .
p12022 ( 2010 ) . n. w. ashcroft and n. d. mermin ,
_ solid state physics _
( brooks cole , 1976 ) .
a. silva , `` statistics of the work done on a quantum critical system by quenching a control parameter '' , phys . rev
* 101 * , 120603 - 120606 ( 2008 ) .
l. santos , a. polkovnokov , m. rigol , `` on the entropy of isolated quantum systems after a quench '' , submitted to phys rev lett .
c. jarzynski and w. j. wiatecki , `` a universal asymptotic velocity distribution for independent particles in a time - dependent irregular container '' , nucl .
a * 552 * , 1 - 9 ( 1993 ) ; c. jarzynski , `` energy diffusion in a chaotic adiabatic billiard gas '' , phys . rev .
e * 48 * , 4340 - 4350 ( 1993 ) .
j. blocki , f. brut and and w. j. swiatecki , `` a numerical verification of the prediction of an exponential velocity spectrum for a gas of particles in a time - dependent potential well '' , nucl .
a * 554 * , 107 - 117 ( 1993 ) ; j. blocki , j. skalski and w. j. swiatecki , `` the excitation of an independent - particle gas classical or quantal by a time - dependent potential well '' , nucl .
a * 594 * , 137 - 155 ( 1995 ) .
l. dalessio , p. l. krapivsky , `` light impurity in an equilibrium gas '' , phys .
e * 83 * , 011107 ( 2011 ) ; m. rigol , v. dunjko , and m. olshanii , `` thermalization and its mechanism for generic isolated quantum systems '' , nature * 452 * , 854 - 858 ( 2008 ) . c. jarzynski , `` diffusion equation for energy in ergodic adiabatic ensembles '' , phys .
a * 46 * , 7498 - 7509 ( 1992 ) .
a. polkovnikov , `` microscopic expression for heat in the adiabatic basis '' , phys .
101 * , 220402 - 220405 ( 2008 ) .
here we will sketch the derivation of eq . relying only on the unitarity of the dynamics in the quantum case and the incompressibility of hamiltonian dynamics ( i.e. liouville s theorem ) in classical case .
our proof will be based on the crooks theorem . as it was noted in previously
( see for example @xcite ) , for a closed system obeying classical hamiltonian dynamics , the crooks theorem relies on the incompressibility of trajectories in phase space ( liouville s theorem ) and microscopic time reversibility .
here we extend this proof to isolated quantum hamiltonian systems with a discrete spectrum .
our proof of the quantum crook s relation bears some similarities with that discussed in ref .
@xcite , and is presented here for completeness .
this will emphasize some important properties of the transition matrix , highlight that the crook s relation does not rely on assumptions related to energy measurements , e.g. at intermediate steps , and extend the fluctuation - dissipation relation to non - canonical distributions .
let us assume that a system prepared in a stationary state , described by a diagonal density matrix in the energy basis , undergoes some process described by a unitary operator @xmath125 . according to standard quantum mechanics the density matrix evolves in time according to @xmath126 .
this means that the diagonal elements of the time evolved density matrix in the new energy basis are given by @xmath127 where we used the fact that the initial density matrix is diagonal and introduced the transition probabilities @xmath128 ( see also ref .
the matrix @xmath129 is doubly stochastic meaning that @xmath130 .
while the first equality is simply the conservation of probability the second is a direct consequence of unitarity .
it is easy to see that this equality is violated if there are losses in the system due to e.g. spontaneous emission .
now let us imagine a time - reversed protocol described by the inverse evolution operator @xmath131
. from the definition of the transition probabilities it is clear that @xmath132 where @xmath133 refers to the reverse process .
let us comment that the transition probabilities also satisfy detailed - balance , @xmath134 , in the two following situations : ( i ) if the hamiltonian of the system is time - reversal invariant at each moment of time and the protocol is time symmetric so that @xmath135 , where @xmath2 is the period of the cycle .
( ii ) if the transition probabilities during one cycle are small and can be computed within first order in an adiabatic perturbation theory @xcite , i.e. a perturbation theory in a basis evolving with the hamiltonian ( this theory also includes ordinary perturbation theory as a particular limit of small amplitude perturbations ) .
let us stress that detailed - balance only plays the role in our proof for deriving the subleading @xmath136 correction in the relation ( [ real_ab ] ) .
to proceed we use the energy distribution : @xmath137 and relate transition probabilities between energy levels to the transition probabilities between energy shells : @xmath138 where @xmath139 is the many - body density of states .
the factor @xmath140 ensures conservation of probability : @xmath141 . the master equation ( [ master ] ) is then given by @xmath142 we now multiply both sides of eq . by @xmath143 and sum over @xmath144 to obtain @xmath145 denoting @xmath146 and using the fact that @xmath147 $ ] we can rewrite the equation above as @xmath148 which is known as the crooks relation @xcite . to prove relation ( [ real_ab ] ) we use eq .
, and expand the entropy and the transition probability @xmath149 in @xmath103 : @xmath150 as in the main text we assume a cyclic process .
note that when @xmath151 is held constant and the system size is increased , the second terms on the rhs of each of the above equations scale as @xmath7 . to leading order in @xmath7 , integrating eq .
( [ crooks - cont ] ) over @xmath152 we obtain the jarzynski like relation @xmath153 , where the brackets represent an average over realizations of the process .
note that we did not make any assumption about an initial gibbs distribution .
taking the logarithm of this jarzynski relation and performing a cumulant expansion we find @xmath154 from which eq .
is obtained by using @xmath155 the condition for the validity of this expansion is that the third cumulant of the work is small : @xmath156 when the additional assumption of the detailed - balance holds we can use @xmath157 in eq .
( [ crooks - cont ] ) .
we point again that detailed - balance is valid for arbitrary symmetric protocols as well as for non - symmetric protocols , provided that the transition probabilities can be computed within first order of adiabatic perturbation theory .
since the work per cycle is assumed to be small , it is expected that the transition probabilities are also small and can be computed perturbatively .
so the assumption of @xmath157 between energy shells is likely to be generic . in particular
, one can check that it is asymptotically satisfied at high energies for the piston example discussed in the main text even for asymmetric protocols . in this case , integrating eq .
( [ crooks - cont ] ) using the expansions ( [ expansion1 ] ) we obtain @xmath158\right>\approx 1-\partial_e\langle w\rangle\ ] ] taking the logarithm of both sides and performing the cumulant expansion of the exponent up to the order @xmath159 we find @xmath160 it is easy to check that up to order @xmath7 the equations above imply eq .
( [ real_ab ] ) : @xmath161 as long as @xmath162 is negligible compared to @xmath163 .
noting that @xmath164 this gives us a necessary condition of validity of relation ( [ real_ab ] ) : @xmath165 namely , the work per unit cycle should be small compared to the temperature multiplied by the specific heat .
we note that even though we derived eq .
( [ real_ab ] ) to the order of @xmath7 , it is actually correct to all orders in @xmath7 .
this relation is valid as long as the conditions ( [ cond1 ] ) and ( [ cond2 ] ) are satisfied . finally let us discuss extension of the fluctuation - dissipation relation ( [ real_ab ] ) to arbitrary distributions . in order to do this
we need to weight eq .
( [ real_ab ] ) with an energy distribution @xmath67 and integrate over energies .
then it is easy to check that in the gaussian approximation we find @xmath166 this relation is clearly a generalization of eq .
( [ real_ab ] ) .
in particular , it reduces to eq .
( [ real_ab ] ) for the microcanonical distribution with @xmath167 and it reduces to the conventional result obtained from the cumulant expansion of the jarzynski relation @xmath168 for the canonical distribution @xmath169 .
in this section we will provide two additional examples illustrating validity of our results to a classical and quantum interacting one - dimensional spin chains .
first we consider an xy - model on a one - dimensional lattice of size @xmath8 . on each lattice site
there is a single degree of freedom , which may be viewed as a two - dimensional unit vector .
the interaction energy between neighboring sites @xmath170 is @xmath171 , where @xmath172 is the difference between the angles at sites @xmath173 and @xmath174 .
the total hamiltonian is given by summing over all nearest neighbors the interaction terms @xmath175 .
to drive the system , we assume that the angle at a specific site @xmath176 is changed by a small amount @xmath177 , which is a fluctuating variable with zero mean , and other sites are unaffected . such a protocol , for example , describes an interaction of the system with an external fluctuating local magnetic field .
the problem can be solved exactly in one - dimension . to do this , note that the change in energy of the system depends only on the two differences @xmath178 and @xmath179 before the drive .
the probability distribution of @xmath179 to order @xmath180 is given by @xmath181.\ ] ] using this expression , it is straightforward to calculate the values of the average work and its fluctuations @xmath182 to order @xmath183 : @xmath184 here to relate @xmath6 and @xmath185 we used the expression for the energy @xcite @xmath186 $ ] , where @xmath187 denotes a modified bessel function of the first kind of order @xmath76 . substituting relations ( [ ab_xy ] ) into eq .
( [ widthbeta ] ) and numerically integrating we obtain @xmath188 as a function of @xmath25 .
the results are shown in fig .
[ fig : xy ] .
note that there are two regimes .
in the low energy regime @xmath27 is to lowest order constant , @xmath189 , and @xmath73 , which gives @xmath190 , and @xmath191 , see table [ summary ] .
the next order correction for small @xmath192 can be obtained using eq .
( [ ab_xy ] ) and gives : @xmath193 .
then at high energies , close to infinite temperature , @xmath194 and again @xmath195 .
this is due to the finite phase - space available to this system , which allows the system to reach a stationary distribution with @xmath196 at a finite total energy density ( @xmath194 ) .
finally we consider a quantum transverse - field ising model in one dimension described by the hamiltonian : @xmath199+h ' , \label{ising_model}\ ] ] where @xmath200 is a weak perturbation breaks integrability of the system but does not affect the dynamics during the driving protocol .
for example , this perturbation can be a weak second nearest - neighbor spin - spin interaction .
the assumption of smallness of @xmath200 is only important to make explicit analytic calculations .
for simplicity we will consider the domain of non - negative values of the transverse field @xmath201 .
this system undergoes a quantum phase transition at @xmath202 @xcite . using a jordan - wigner transformation
the hamiltonian assumes a quadratic form in the fermionic operators and can be fully diagonalized performing a bogoliubov rotation in momentum space @xcite .
the final hamiltonian reads : @xmath203 where @xmath204 and @xmath205 are quasi - particle creation and annihilation operators . here in order to simplify analytic expressions we linearized the spectrum by taking @xmath206 and the relevant momenta are much smaller than the ultra - violet cutoff given by the lattice : @xmath207 .
the ground state of this hamiltonian , which is annihilated by all quasi - particle operators @xmath208 , is factorized into momentum sectors .
the excited states can be obtained by applying various combination of operators @xmath209 to the ground state . if the external time - dependent perturbation is spatially uniform then due to momentum conservation only the excited states corresponding to pairs of excited quasi - particles with opposite momenta obtained by applying @xmath210 to the ground state participate in the dynamics .
moreover excitations to different momentum states are independent and the problem effectively splits into a collection of two levels - systems .
let us emphasize that this system does not become classical even in the infinite temperature limit . in disordered systems this feature was explored e.g. in the context of the many - body localization in ref .
@xcite .
we now perturb the system by changing the amplitude of the transverse - field @xmath211 , @xmath212 $ ] from its initial value , @xmath213 , to an intermediate value , @xmath214 and then back to @xmath213 . here
the parameter @xmath215 sets the velocity of the quench and @xmath216 sets its amplitude . as a result of this process the occupation of the energy levels will change .
because different momentum modes are effectively independent from each other we will consider each two - level system separately .
the presence of the weak integrability breaking perturbation ensures a fermi - dirac redistribution of the energy among different modes between different cycles .
it is similar to the effect of a weak interaction between particles leading to a maxwell - boltzmann single - particle distribution in the piston example or an assumption of weak coupling to the rest of the system in the single oscillator example discussed in the main text .
the ergodicity of small weakly nonintegrable one - dimensional systems was recently tested numerically to a very good accuracy for hard - core bosons and fermions , which are closely related to the transverse field ising model ( see e.g. ref .
@xcite ) . under these assumptions
each cycle starts from the fermi - dirac distribution in each momentum mode . to avoid extra complications related to the singularities of the transition probabilities at the critical point
we will additionally assume that dynamics occurs only in one phase , say @xmath217 ( at finite temperatures we are interested in , this assumption can be further relaxed ) . for slow quenches , if the rate @xmath215 is small : @xmath218 , such that the adiabaticity condition @xmath219 is satisfied for all modes , the dynamics can be solved analytically .
in particular , in ref .
@xcite it was shown , that under these conditions the transition probability between the ground and excited states is approximately equal to ( see eqs . (
20 ) and ( 87 ) in ref .
@xcite ) @xmath220 note that there is an additional contribution to eq .
( [ t_k ] ) which is rapidly oscillating at a frequency @xmath221 ( see eq .
( 20 ) in ref .
@xcite ) and averages to zero either because of adding contributions from different momentum modes or because of slight fluctuations of the quench time @xmath222 from cycle to cycle .
actually the addition of the oscillating term to eq .
( [ t_k ] ) will at most double the transition probability .
as we will see below our results are not affected by the precise form of @xmath223 as long as @xmath223 is small , which is controlled by @xmath215 and decays sufficiently fast with momentum @xmath224 so that the work distribution remains sufficiently narrow . with these transition rates ,
the master equation for the occupation probabilities of different momentum states becomes particularly simple : @xmath225 where @xmath226 and @xmath227 refer to the probabilities of having a pair of fermions with momenta @xmath224 and @xmath228 before and after the quench respectively and likewise , @xmath229 and @xmath230 are the probabilities to have no fermions in the @xmath224 and @xmath228 momentum mode .
due to rethermalization of the system between driving protocols we have @xmath231\over 2\cosh[\beta\epsilon_k^{g_1 } ] } , \rho^k_{ex}={\exp[-\beta \epsilon_k^{g_1}]\over 2\cosh[\beta\epsilon_k^{g_1}]}\ ] ] actually these expressions are somewhat modified due to presence of the inert modes with one fermion in either @xmath224 or @xmath228 mode , but this modification does not affect our conclusions since the inert modes do not participate in transitions only effectively reducing the number of active modes by a factor of two in the regime of interest . from the master equation and the probabilities of occupying initial states above
we find the average work and its second moment for individual momentum modes : @xmath232 from this equation we deduce the relation ( [ real_ab ] ) between @xmath233 and @xmath234 : @xmath235 is satisfied provided that @xmath236 , which is the case for sufficiently slow quenches , and @xmath237 which means that the temperature is big compared to the initial energy of the fermions .
the latter condition is always satisfied if the temperature is large compared to the gap in the system and that the relevant excited modes correspond to small momenta , which is the case for sufficiently slow quenches .
if these conditions are fulfilled then eq .
( [ real_ab ] ) is satisfied for the total work and its variance : @xmath238 in fig .
[ isingplot ] we show the calculated values of @xmath239 and @xmath240 for different initial temperatures for the protocol with @xmath241 and @xmath242 ( the time @xmath222 drops out from the answer if @xmath243 ) .
as the temperature increases the work per cycle , @xmath27 , decreases due to fermion anti - bunching and the relation ( [ real_ab ] ) is satisfied to a very good accuracy as soon as temperature becomes much bigger than the gap .
let us make now a number of remarks on validity of our results and the importance of the protocol in quantum systems .
the general analysis and derivation of the fokker - planck equation in the main paper as well as the relation ( [ real_ab ] ) in sec .
[ extramaterial2 ] of this supplementary material ( see eq.([lrq ] ) ) are based on a quantum formalism . therefore as soon as the conditions @xmath245 and @xmath246 are satisfied eq .
( [ real_ab ] ) should work .
and indeed we illustrated this here for a particular example of a transverse field ising model . instead of
this model we could get e.g. fermions with two - body interactions and would come to similar conclusions .
the smallness of work can be controlled by doing sufficiently slow quenches or by quenches of small amplitude similar to the classical case .
there is an important subtlety related to the second condition of the smallness of the third and higher cumulants of work which distinguishes quantum and classical systems .
in particular , in classical situations it is impossible to give a large energy to a system during a cycle unless the external parameter changes suddenly .
e.g. in a deforming cavity example during each collision with a wall the particle can gain at most velocity @xmath247 , where @xmath248 is the velocity of the wall .
this means that in classical situations the work distribution is typically bounded and the smallness of average work usually implies smallness of its cumulants . in quantum systems
the situation is very different .
namely in any protocol it is possible to give the system arbitrarily large energy . for
smooth protocols , where the external parameter changes analytically in time the transition probability to high energy states decreases exponentially like in the conventional landau - zener problem and such transitions do not affect cumulants of work .
however , for non - analytic protocols where e.g. amplitude , velocity ( like in our case ) or acceleration experience a discontinuity the transition probability to high energy states decreases only algebraically with energy like in eq .
( [ t_k ] ) ( see ref .
@xcite for more details ) .
this means that high enough cumulants of work necessarily diverge . of course real experimental protocols are always smooth and these divergencies are cutoff , however the degree of smoothness introduces a new quantum scale into the dynamics
. thus the work distribution can be effectively wide or narrow in the sense of satisfying the condition @xmath246 depending on the ratio of this new energy scale and temperature .
since this discussion is not directly related to this paper we postpone it until future work . here
we only explicitly analyze another protocol where the coupling changes quadratically in time : @xmath249 . in this case in the slow limit @xmath250 instead of eq .
( [ t_k ] ) we get ( see ref .
@xcite for details ) : @xmath251 as we see indeed the transitions to the higher energy states are suppressed even more than for the linear protocol so that more cumulants of work now converge at small energies . as a result
the relation ( [ real_ab ] ) is satisfied even at smaller temperatures , see fig .
[ isingplot1 ] . in fig .
[ isingplot1 ] we show the values of @xmath239 and @xmath240 obtained using eq . with @xmath241 and @xmath252 ( the time @xmath222 drops out from the answer if @xmath253 ) .
with this choice of the parameters the peak value of @xmath223 in the two protocols ( linear , eq . , and quadratic , eq . )
are equal .
the einstein - relation , @xmath254 , is satisfied when the relative error @xmath255 becomes less than @xmath256 . with this definition
the einstein - relation is satisfied at @xmath257 for the linear and @xmath258 for the quadratic protocol . in the regime of validity of the fluctuation - dissipation relation ( [ real_ab ] ) we can expand expression to first order in @xmath185 to obtain @xmath260 for both the linear and quadratic protocols described above .
moreover , if the temperature @xmath2 is in turn much smaller than the cutoff energy scale given by @xmath261 ( which is nothing but the fermi energy in the fermion representation ) then @xmath262 , which is the case for weakly interacting fermions if @xmath263 . from these considerations
it immediately follows that @xmath95 , @xmath264 and @xmath265
. therefore @xmath266 . at high energies , i.e. close to the infinite temperature limit
, @xmath267 and @xmath268 must tend to @xmath269 .
as in the xy - model this is due to the finite size of the hilbert space ( see discussion after eq . )
this example shows that for weakly interacting fermions ( to which the ising spin chain is equivalent ) it is easy to get distributions significantly narrower than canonical without any special fine tuning .
we expect this to be generic result following from the pauli exclusion principle and suppression of transitions at higher temperatures ( see ref .
@xcite for additional discussion ) .
98 p. pradhan , y. kafri , and d. levine , `` nonequilibrium fluctuation theorems in the presence of local heating '' , phys .
rev e * 77 * , 041129 - 041135 ( 2008 ) m. campisi , p. hanggi , p. talkner , `` colloquium : quantum fluctuation relations : foundations and applications '' , arxiv:1012.2268 , rev . mod .
( accepted ) .
a. polkovnikov , `` microscopic expression for heat in the adiabatic basis '' , phys .
* 101 * , 220402 - 220405 ( 2008 ) .
g. e. crooks , `` nonequilibrium measurements of free energy differences for microscopically reversible markovian systems '' , j. stat .
phys . * 90 * , 1481 - 1487 ( 1998 ) .
s. sachdev , _ quantum phase transitions _
( cambridge university press , cambridge , england , 1999 ) . v. oganesyan and d. huse , `` localization of interacting fermions at high temperature '' , phys .
b * 75 * , 155111 ( 2007 )
. l. f. santos and m. rigol , `` localization and the effects of symmetries in the thermalization properties of one - dimensional quantum systems '' , phys .
e * 82 * , 031130 ( 2010 ) .
c. de grandi and a. polkovnikov , `` adiabatic perturbation theory : from landau - zener problem to quenching through a quantum critical point '' , contribution to `` quantum quenching , annealing and computation '' , eds .
a. das , a. chandra and b. k. chakrabarti , lect .
notes in phys . ,
springer , heidelberg ( 2009 ) , arxiv:0910.2236 . | when an isolated system is brought in contact with a heat bath its final energy is random and follows the gibbs distribution a cornerstone of statistical physics .
the system s energy can also be changed by performing non - adiabatic work using a cyclic process .
almost nothing is known about the resulting energy distribution in this setup , which is especially relevant to recent experimental progress in cold atoms , ions traps , superconducting qubits and other systems . here
we show that when the non - adiabatic process comprises of many repeated cyclic processes the resulting energy distribution is universal and different from the gibbs ensemble .
we predict the existence of two qualitatively different regimes with a continuous second order like transition between them .
we illustrate our approach performing explicit calculations for both interacting and non - interacting systems .
understanding equilibrium and non - equilibrium properties of thermally _ isolated _ systems has become a forefront of research due to experimental developments over the past decade , particularly in cold atom systems @xcite , trapped ions @xcite , and nuclear spins @xcite and superconducting qubits @xcite . in these systems the coupling to external dissipative degrees of freedom is strongly suppressed and irrelevant on accessible time scales .
these systems provide a new and very clean playground where one can investigate fundamental questions in statistical and quantum physics .
moreover , they point to new practical applications , in particular in the context of quantum information .
the experimental studies inspired intensive theoretical research on a variety of topics .
these include equilibration in isolated systems initially driven out of equilibrium by a sudden change in a coupling constant ( a quench ) ; defect ( or energy ) generation during slow nearly adiabatic processes in gapless phases or near singularities , such as quantum phase transitions ( for a review see refs .
@xcite ) ; non - equilibrium quantum phase transitions in the presence of @xmath0 noise @xcite ; and many more .
in this work , we consider the energy distribution of a thermally _ isolated _ system following a non - adiabatic process .
consider two setups where the energy of an isolated system is changed .
in the first , the system is brought into contact with a heat bath until equilibration and then disconnected from it - similar to an oven . in the second the energy of the system
is increased due a non - adiabatic change of some external parameter(s ) - much like a microwave .
the two setups are illustrated schematically in fig .
[ ovens2 ] .
our interest is in the energy distribution of each of the systems at the end of the process .
for the first setup the result is well - known and corresponds to the classic heating mechanism which can be found in any book on thermodynamics ( see e.g. ref .
@xcite ) .
if , as usual , the bath is large compared to the system then the energy distribution of the system becomes gaussian , with a canonical width uniquely determined by the fluctuation - dissipation relation : @xmath1 , where @xmath2 is the temperature and @xmath3 is the specific heat .
this relation is valid for both quantum and classical systems and is independent of the details of the interactions between the system and the bath .
now , consider a second setup where the energy of the system is changed due to a non - adiabatic variation of an external parameter ( say , the electro - magnetic field in the case of a microwave or the motion of the piston in fig .
[ ovens2 ] ) .
this type of heating also inevitably leads to an uncertainty in the final energy of the system .
while it is known that the energy changes in the system obey , even beyond linear response , the recently discovered fluctuation theorems ( see for example , @xcite ) very little is known about the resulting energy distribution in this case @xcite . in a large macroscopic system
the energy distribution is expected to be very narrow and the relative energy fluctuations negligible .
but in small or mesoscopic systems , which are of primary experimental interest ( see e.g. ref .
@xcite ) , the fluctuations can be large and important .
fundamental questions are unanswered : which features of the energy distribution are universal and which features depend on details of the system and driving protocol ?
to what extent can the width of the distribution be controlled ?
for example , can one dynamically increase the energy of an isolated system without increasing the uncertainty in the final energy ? can the fluctuation - dissipation relations , which determine the energy width in the oven - like setup , be extended to the microwave - like setup ? in this paper we begin to address such questions .
specifically , we study a thermally isolated system undergoing a repeated cyclic process , whereby some external parameter @xmath4 in the hamiltonian is changed in time and returns to its original value at the end of each cycle , see fig .
[ ovens2 ] .
we show that under generic assumptions of ( i ) small work per cycle and
( ii ) absence of correlations between cycles ( see detailed discussion in methods ) the variance of the energy distribution @xmath5 at energy @xmath6 assumes a particularly simple form to leading order in @xmath7 , where @xmath8 is the number of degrees of freedom in the system .
it depends only on the microcanonical temperature @xmath9 where @xmath10 is the density of states , and on the average energy change in a cycle at energy @xmath6 , @xmath11 : @xmath12 here @xmath13 is the initial energy of the system and @xmath14 is the initial variance .
this equation is the main result of the paper .
the result ( [ widthbeta ] ) follows from integrating a fokker - planck equation which describs the time evolution of the energy distribution @xmath15 ( see methods and supplementary material for details ) : @xmath16 the change of the energy distribution in one cycle of the protocol is obtained by integrating this equation over the duration of the protocol , set for simplicity to be unity . within this choice @xmath17 represent the average work per cycle and its variance respectively : @xmath18 and @xmath19 . here
the angular brackets denote averaging over realizations of the cycle starting from a fixed initial energy . in general @xmath11 and @xmath20
are protocol dependent functions and are _ a priori _ independent from each other . however , since the system is thermally isolated its time evolution is governed by hamilton s equations of motion in the classical case and the schrodinger equation in the quantum case .
this puts strong constraints on the relation between @xmath11 and @xmath20 similar to the einstein fluctuation - dissipation relations between drift coefficient ( mobility ) and diffusion in open systems @xcite .
in particular , we find @xmath21 for interacting systems with many degrees of freedom the second term on the rhs of this equation is a @xmath7 correction which can be neglected
. this term can be important though in mesoscopic or integrable systems .
eq .
( [ real_ab ] ) was previously suggested for classical systems in ref .
@xcite .
the fluctuation - dissipation relation ( [ real_ab ] ) is derived within a small work assumption ( explicitly @xmath22 , where @xmath23 is the third cumulant of the work , and @xmath24 , see methods and supplementary material ) . as we will show below eq .
( [ real_ab ] ) holds for a very wide class of classical and quantum systems starting from noninteracting particles in a time dependent cavity to fully interacting spin systems .
the main result of the paper eq .
( [ widthbeta ] ) is a direct consequence of this relation ( see methods ) .
several interesting consequences follow from eq . :
( i ) when @xmath11 is constant the energy width depends only on @xmath25 , and not on the amplitude of the drive or other details of the driving protocol .
( ii ) when @xmath11 is not constant , depending on the functional form of @xmath11 and @xmath25 , the variance of the distribution can be larger and surprisingly , even smaller than the width of the equilibrium gibbs distribution at the same mean energy .
in fact , @xmath26 can be made arbitrarily small by a proper choice of @xmath11 .
( iii ) when @xmath27 is a function of the energy density @xmath28 ( with a possible extensive energy independent prefactor like the total number of particles ) , we have @xmath29 , scaling as in equilibrium . for a single quench this result
was noticed e.g. in ref .
@xcite . here
we show that it remains valid after many quenches .
( iv ) the dependence of @xmath30 on @xmath6 displays two qualitatively distinct behaviors with increasing @xmath6 , depending on whether the integral in eq .
( [ widthbeta ] ) diverges or converges as @xmath31 . to illustrate the distinct behaviors associated with point ( iv ) above we consider the generic case where @xmath32 , which is the case for phonons , superfluids or other systems with goldstone bosons , fermi liquids , ideal gases and others .
moreover we measure time in units of the number of cycles carried out ( in what follows we will use time and number of cycles interchangeably ) and assume a simple power law behavior for @xmath11 : @xmath33 as will become clear below the two regimes exist even in cases when @xmath11 and @xmath25 are not power laws .
the values of @xmath34 are constrained by simple thermodynamic arguments to @xmath35 : the lower bound is required by positivity of the specific heat and the upper bound assures that the entropy ( @xmath36 ) is an increasing unbounded function of the energy . to prevent the system s energy from diverging in a finite time we require @xmath37 ( as follows from integrating eq . ) . for simplicity
we also assume @xmath38 and compare the width to the equilibrium canonical width @xmath39 . in this case
the system displays a _
transition _ between two behaviors as the functional form of @xmath11 is changed .
this transition is continuous and is characterized by a diverging time - scale needed to reach the asymptotic regime .
specifically , depending on the sign of @xmath40 , eq .
( [ widthbeta ] ) implies : ( i ) when @xmath41 the width is _ gibbs - like _ with @xmath42 , i.e. the ratio @xmath43 asymptotically approaches a constant value that can be either larger or smaller than one .
note that smaller widths correspond to protocols with large and negative @xmath44 , i.e. to protocols where @xmath11 is a strongly decreasing function of energy .
( ii ) a second _ run - away _ regime occurs when @xmath45 . here
the width increases with a higher power of energy than the canonical width : @xmath46 .
the resulting distribution is significantly wider than the canonical one . given the constraint on the value of @xmath44 , this regime can only be reached if @xmath47 , in particular this regime is unreachable for a classical ideal gas .
the transition between the _ gibbs - like _ and _ run - away _ regimes occurs when @xmath48 which implies @xmath49 .
close to this transition when @xmath50 , there is a divergent time scale ( or number of cycles ) required to reach the asymptotic regime .
this time scale can be obtained by combining eq .
and eq . , see table [ summary ] .
the crossover from the _ gibbs - like _ to the _ run - away _ regime is qualitatively similar to a continuous phase transition , with the diverging time scale being analogous to a divergent relaxation time ( critical slowing down ) in the equilibrium case .
we summarize our results , close to the transition , for the above choices of @xmath25 and @xmath11 in table [ summary ] .
the qualitative difference between the two regimes can also be understood in terms of the entropy of the distribution : @xmath51 , where @xmath52 are the microscopic probabilities to occupy different energy levels . converting this sum into an integral over energies and expanding the resulting expression up to @xmath7 corrections it is straightforward to check @xcite that : @xmath53 where @xmath54 is the equilibrium canonical entropy .
it is easy to see that the correction to the equilibrium entropy is always negative except when the width of the energy distribution coincides with the canonical width . in the _ gibbs - like _ regime
this correction is a constant , while in the _ run - away _ regime it has an explicit energy dependence .
note that under the assumptions used to derive our main result , the integral in eq . can be rewritten in terms of @xmath55 the average entropy change per unit of time ( per - cycle ) , where @xmath56 .
using @xmath57 the nature of the transition between the two regimes assumes an interesting physical interpretation . specifically , the integral in eq .
( [ widthbeta ] ) becomes @xmath58 showing that the _ gibbs - like _ and ( _ run - away _ ) regimes correspond to the entropy growing slower ( faster ) than time squared .
this conclusion does not rely on any assumptions about the specific functional form of @xmath25 and @xmath11 .
.a summary of the results for @xmath59 , @xmath60 with @xmath61 , @xmath37 and @xmath62 .
the width specifies the asymptotic value in units of the equilibrium width at the same energy .
the time scale specifies the characteristic `` relaxation '' time needed to reach the asymptotic regime . [ cols="^,^,^,^",options="header " , ] |
Subsets and Splits