article
stringlengths 0
745k
| abstract
stringlengths 0
92.7k
|
|---|---|
# 1#2#3_#3 # 1#2#3_#3 # 1#1 # 1#2 # 1#2m_#1^(#2 ) # 1#2n_#1^(#2 ) # 1 without doubt , the rogers
ramanujan identities @xmath8 with @xmath9 for @xmath10 and @xmath11 , are among the most beautiful and intriguing results of classical mathematics . since their independent discovery by rogers @xcite and ramanujan @xcite many different methods of proof have been developed .
a particularly fruitful approach was initiated by schur @xcite .
the key idea is that identities of the rogers ramanujan type are in fact limiting cases of polynomial identities .
for example , the polynomial identities @xmath12 which hold for arbitrary @xmath13 , yield the rogers ramanujan identities ( [ rogersram ] ) in the limit @xmath14 when @xmath15 is restricted to @xmath16 .
here @xmath17 denotes the largest integer not exceeding @xmath18 , and the gaussian- or @xmath15 polynomial is defined by @xcite @xmath19 using the elementary formulae @xmath20 and @xmath21 it is easy to verify that both the left- and right - hand - sides of equation ( [ finiterr ] ) satisfy the recurrence . given appropriate initial conditions , this recurrence has a unique solution hence establishing the polynomial identity and its limiting form ( [ rogersram ] ) .
identities of the rogers ramanujan type occur in various branches of mathematics and physics .
first , their connection with the theory of ( affine ) lie algebras @xcite and with partition theory @xcite has led to many generalizations of ( [ rogersram ] ) . to illustrate these connections ,
it is for example easily established that for @xmath22 the right - hand - side of ( [ rogersram ] ) can be rewritten in a more algebraic fashion as is used in the context of both classical and affine lie algebras , we will throughout this paper write @xmath23 to mean its classical ( counter)part . ]
@xmath24 with @xmath25 the root lattice , @xmath26 the weyl group and @xmath27 the weyl vector of the classical lie algebra a@xmath28 , @xmath29 the dedekind eta function and @xmath30 similarly , if @xmath31 denotes the number of partitions of @xmath32 with each successive rank in the interval @xmath33 $ ] , then by sieving methods the generating function of @xmath34 can be seen to again yield the right - hand - side of ( [ rogersram ] ) provided we choose @xmath35 and @xmath36 @xcite .
second , rogers
ramanujan identities also appear in various areas of physics .
most notable is perhaps the fact that ( [ virchar ] ) can be identified as the normalized rocha - caridi form @xcite for the identity character @xmath37 of the virasoro algebra .
indeed , each pair of positive integers @xmath38 , @xmath39 with @xmath38 and @xmath39 coprime , labels a minimal conformal field theory @xcite of _ central charge _ @xmath40 given by ( [ cc ] ) .
another branch of physics where rogers ramanujan type identities have occurred is in the theory of solvable lattice models @xcite . among other works , in refs .
@xcite andrews , baxter and forrester ( abf ) encountered generalized rogers
ramanujan identities in their corner transfer matrix ( ctm ) calculation of one - point functions of an infinite series of restricted solid - on - solid ( rsos ) models .
in addition , in refs .
@xcite kedem _ et al .
_ conjectured many identities motivated by a bethe ansatz study of the row transfer matrix spectrum of the 3-state potts model .
interestingly though , it is in fact the combination of the approaches of ref .
@xcite and ref .
@xcite to solvable models that leads to polynomial identities of the type ( [ finiterr ] ) . in computing one - point functions of solvable rsos models using ctms along the lines of ref .
@xcite one is naturally led to the computation of so - called one - dimensional configuration sums .
these configuration sums take forms very similar to the right - hand - side of ( [ finiterr ] ) . on the other hand , in performing bethe ansatz , and more particularly thermodynamic bethe ansatz ( tba ) computations , one
is led to expressions of a similar nature to the left - hand - side of ( [ finiterr ] ) . starting with the abf models and pursuing the lines sketched above , melzer conjectured @xcite an infinite family of polynomial identities similar to those in ( [ finiterr ] ) . in the infinite limit
these identities again lead to rogers ramanujan type identities , but now for virasoro characters @xmath41 of unitary minimal models , i.e. , for characters with the rocha - caridi right - hand - side form ( in the sense of ( [ rogersram ] ) ) @xmath42j}- q^{p(p+1)j^2+[(p+1)r+ps]j+rs } \right\ } \label{xrs}\ ] ] labelled by the conformal weights @xmath43 ^ 2 - 1}{4p(p+1 ) } \quad \qquad 1\leq r \leq p-1 ,
\quad 1\leq s \leq p. \label{weights}\ ] ] the corresponding left - hand - side forms of these characters were conjectured earlier in the work of kedem _ et al .
_ @xcite and their _ finitization _ in @xcite again provided a method of proof . for @xmath44 and 4 the proof was carried out in ref .
@xcite and berkovich @xcite subsequently generalized this to all @xmath45 for @xmath46 . in this paper
we generalize melzer s approach to finding polynomial identities from solvable lattice models . by considering the ctm as well as tba calculations of various higher rank generalizations of the abf models ,
we are led to conjectures for polynomial identities labelled by the lie algebras of a , d and e type . in the infinite limit
our @xmath6=a - d - e polynomial identities lead to rogers ramanujan type expressions for the branching functions associated with the gko coset pair @xcite @xmath47 the rest of this paper is organized as follows . in the next section
we define polynomial expressions @xmath48 following from the tba calculations of bazhanov and reshetikhin @xcite .
since the polynomials are defined as restricted sums over the solutions of @xmath6 type constraint equations , we adopt the terminology of ref .
@xcite and call these polynomials _ fermionic_. in section [ sec3 ] , we define analogous polynomial expressions , originating from the ctm calculations of refs .
@xcite for the same solvable lattice models as considered by bazhanov and reshetikhin . again following the terminology of ref .
@xcite , we call these polynomials _ bosonic _ , and denote them by @xmath49 .
our conjectures can then be formulated as @xmath50 in section [ sec4 ] we study the @xmath14 behaviour of our polynomial identities , thereby deriving a - d - e type rogers ramanujan identities . in the subsequent two sections we provide some indications for the correctness of our conjectures . in section [ sec5 ]
we establish the expected level - rank duality for @xmath6=a@xmath2 and in section [ sec6 ] we show that in the @xmath7 limit the a - d - e rogers
ramanujan identities lead to the correct dilogarithm identities .
finally , in section [ sec7 ] , we summarize our results and discuss possible generalizations to non simply - laced lie algebras and to more general cosets than those listed under ( [ gko ] ) .
[ sec2 ] let us now turn to the definition of the fermionic a - d - e polynomials as follow from the tba calculations of ref .
@xcite .
we denote the cartan matrix of the simply - laced lie algebra @xmath1=a@xmath51,d,@xmath51e by @xmath52 and the corresponding incidence matrix by @xmath53 , choosing the labelling of the nodes of the a - d - e dynkin diagrams as shown in figure 1 .
we furthermore let @xmath54 and @xmath55 be the number of ` particles ' and ` anti - particles ' of type @xmath56 and colour @xmath57 , respectively , satisfying the following constraint system : @xmath58 . \label{tbaconstr}\ ] ] here the particles are labelled from @xmath59 and the colours from @xmath60 .
the variable @xmath38 in the first kronecker delta is 1 except for @xmath6=e@xmath61 when we have @xmath62 .
we note that the above equation is a parameter independent version of the constraint equations for densities of strings and holes in the tba calculations of refs .
@xcite , and given a set @xmath63 it determines a companion set @xmath64 .
of course , only for special sets @xmath65 does it follow that @xmath66 again consists of only non - negative integers .
we now define the fermionic polynomials @xmath48 as the following sum over the solutions to the constraint system ( [ tbaconstr ] ) : @xmath67 here the prime signifies the additional constraints @xmath68 or , equivalently , @xmath69 with @xmath70 for @xmath6=a@xmath2 , @xmath71 and @xmath72 for @xmath6=d@xmath2 , @xmath73 or , equivalently , 5 for e@xmath74 and @xmath73 or , equivalently , 7 for e@xmath75 . since all entries of @xmath76 are integers there is no additional constraint for @xmath6=e@xmath77 .
we remark that for @xmath6=a@xmath28 the fermionic polynomials coincide with a special case of those defined in refs .
@xcite .
[ sec3 ] as mentioned in the introduction , the bosonic polynomials all arise naturally in corner transfer matrix calculations of order parameters of solvable lattice models of rsos type @xcite .
one of the key steps in such a calculation is the evaluation of the _ one - dimensional configuration sums _ @xmath78 the function @xmath79 herein is determined by the boltzmann weights of the relevant solvable model , and each pair @xmath80 is assumed to be admissible in the sense explained below . except for the case @xmath6=e@xmath74 , the bosonic a - d - e polynomials turn out to coincide with one particular configuration sum . for e@xmath74
it can be defined as a linear combination of two such sums .
in this case we have to consider the family of a@xmath81 rsos models as introduced @xcite and solved @xcite by jimbo , miwa and okado ( jmo ) , generalizing the rsos models of abf @xcite which are recovered for @xmath82 . before we present jmo s result for the one - dimensional configuration sums @xmath83 we need some notation . for the affine
lie algebra a@xmath81 we let @xmath84 denote the fundamental weights , @xmath85 the weyl vector and @xmath86 the null root .
we have the usual inner product on @xmath87 by @xmath88 for a general element @xmath89 its level is defined as @xmath90 and hence , according to @xmath91 , all fundamental weights have level 1 .
a weight @xmath57 belongs to the set @xmath92 of _ level-@xmath93 dominant integral weights _
if @xmath94 i.e. , @xmath95 .
we finally introduce the vectors @xmath96 , @xmath97 @xmath98 where we have set @xmath99 .
clearly @xmath100 . an ordered pair of weights @xmath101
both in @xmath102 is said to be _
admissible _ and denoted @xmath103 , if @xmath104 for some @xmath105 .
we now come to the definition of the one - dimensional configuration sums @xmath106 of the a@xmath81 rsos models at level @xmath93 .
they are given by ( [ 1dcs ] ) , with @xmath107 where the sum is over all sequences @xmath108 of admissible level-@xmath93 dominant integral weights .
the function @xmath79 in ( [ 1dcs ] ) takes the values @xmath109 using standard recurrence methods @xcite , jmo found @xcite the following expression for @xmath110 : @xmath111 with @xmath112 the affine weyl group and , for @xmath113 , @xmath114 } \mult{l}{\lambda}{q } = q^{|\lambda-\lambda_{j-1}|^2/2 + ( l - j+1)(l - j+1+n)/(2n ) } \mult{l}{\lambda}{q } .\ ] ] the @xmath15-multinomial coefficient in the definition of @xmath115 reads @xmath116 though equation ( [ xan ] ) defines bosonic polynomials for any @xmath117 and @xmath56 , we only consider the special case @xmath118 and @xmath119 and define is obtained via admissible sequences of @xmath120 steps on @xmath92 it may be better to employ the usual young tableau notation assigning a tableau of signature @xmath121 to each element of @xmath92 . in the convention of ref .
@xcite we then would have @xmath122 , @xmath123 and @xmath124 . ]
@xmath125 our conjecture is that @xmath126 for all @xmath13 , @xmath127 . for @xmath82
this was previously conjectured and partially proven in ref .
a full proof for @xmath82 can be found in ref .
@xcite . to test the conjecture we have generated both fermionic and bosonic polynomials @xmath128 and @xmath129 for all @xmath130 and for extensive ranges of @xmath120 using _ mathematica
_ @xcite . for @xmath6=d@xmath2
we turn to the hierarchy of d@xmath81 rsos models , again introduced by jmo @xcite .
the result for the one - dimensional configuration sums @xcite is almost completely analogous to that for a@xmath81 as described in the previous subsection , and hence we only point out the relevant changes .
first of all , the inner product in the weight space changes to @xmath131 with @xmath132 , @xmath133 and @xmath134 .
we again introduce vectors @xmath135 , @xmath136 , defined by ( [ evec ] ) for @xmath137 and @xmath138 .
two exceptions to ( [ evec ] ) are @xmath139 and @xmath140 , and we note that @xmath141 for all @xmath142 . a pair @xmath101 , @xmath143 is admissible if @xmath144 , where @xmath92 again denotes the level-@xmath93 dominant integral weights .
the values taken by the function @xmath79 read @xcite @xmath145 together with the exceptions @xmath146 and @xmath147 .
again by application of recurrence methods it was shown in ref .
@xcite that @xmath83 can be expressed as the sum over the affine weyl group as defined in ( [ xan ] ) , with @xmath115 therein given by @xmath148 } \mult{l}{\eta}{q},\ ] ] with @xmath149 for @xmath150 and @xmath151 for @xmath137 , and with the same definition of the multinomial coefficient as in ( [ qmulti ] ) . as in the previous case , we only consider the bosonic polynomials obtained by specializing @xmath118 and @xmath119 in ( [ xan ] ) , and set @xmath152 we then conjecture that @xmath153 for all even @xmath13 .
this conjecture has again been tested extensively for all @xmath130 using _ mathematica _
@xcite .
we now come to the exceptional simply - laced lie algebras : e@xmath74 , e@xmath75 and e@xmath77 .
unfortunately , in this case we do not have an algebraic formulation of the bosonic polynomials , and as an immediate consequence we only have conjectures for @xmath154 .
in all three cases to follow we use the following definition of the @xmath15-multinomial @xmath155 define the polynomials + & & [ be6 ] + it is to be noted that 5 is the integer value of the weight @xmath156 and hence that the above form corresponds to a finitization of the _ extended _ identity character ( with respect to the virasoro algebra ) . the bosonic expression ( [ be6 ] ) can be identified with the normalized sum of the two configurations sums ( @xmath157 and @xmath158 in the notation of ref .
@xcite ) of the dilute a@xmath74 lattice model @xcite .
our conjecture is that @xmath159 , for all @xmath13 .
we note that for @xmath154 the restriction ( [ constronm ] ) is implied by the constraint system ( [ tbaconstr ] ) and as a result we can drop the prime in the sum ( [ adepoly ] ) .
we have checked the correctness of the e@xmath74 polynomial identity for all @xmath160 by direct expansion .
define the polynomials b^ _ 7_q(l ) = _ j , k=-^ \ { q^20j^2+j+k(k+10j)- q^20j^2 + 9j+1+k(k+10j+2 ) } .
this bosonic expression is related to one of the configuration sums ( @xmath157 in the notation of @xcite ) of the dilute a@xmath161 model @xcite .
our conjecture is that @xmath162 , for all @xmath13 .
tests have again confirmed the polynomial identity for all @xmath163 .
define the polynomials @xmath164 comparison with the one - dimensional configuration sums for the dilute a@xmath165 model in regime 2 as computed in ref .
@xcite shows that @xmath166 .
the assertion is now that @xmath167 , for all @xmath13 .
the e@xmath77 polynomial identity has in fact been proven @xcite , and is directly related to the e@xmath77 structure of the critical ising model in a magnetic field @xcite .
[ sec4 ] in this section we present the rogers ramanujan type identities that follow by taking the @xmath14 limit in the various polynomial identities listed in the previous section . to put the results in the context of coset conformal field theories based on the gko construction of equation ( [ gko ] )
, we first give a brief reminder on theta functions and _ branching rules _ , see e.g. , refs .
@xcite . for @xmath168
the classical @xmath6 theta function of characteristic @xmath169 and degree @xmath170 is defined as denotes a simply - laced lie algebra . ]
@xmath171 with @xmath25 the root lattice of @xmath6 .
then , according to the weyl - kac formula @xcite , the character of highest weight representation @xmath172 of a @xmath173 kac - moody algebra is given by @xmath174 with , @xmath175 the weyl group and @xmath176 the dual coxeter number of @xmath6 .
we now come to the definition of the _ branching functions _ @xmath177 for dominant integral weights @xmath178 , @xmath179 and @xmath180 , @xmath181 , via the decomposition or branching rule @xmath182 the branching functions can be expressed in terms of the @xmath6 theta functions as @xmath183^{\,\rg } } \sum_{w \in \overline{w } } \mbox{sgn}(w ) \ ; \theta_{- ( p+1)(\overline{a}+\overline{\rho } ) + p w(\overline{b}+\overline{\rho}),p(p+1)}(0,\tau),\ ] ] with @xmath184 . from the definition ( [ gtheta ] ) of the theta functions
this can be rewritten in a form generalizing equation ( [ xrs ] ) for the unitary minimal virasoro characters @xcite @xmath185^{\,\rg } } \sum_{\alpha \in q } \ , \sum_{w\in \overline{w } } \mbox{sgn}(w ) \ ; \ ; q^{\displaystyle \case{1}{2 } \ , p ( p+1 ) \left|
\alpha -\frac{(p+1)\overline{r } - p \ , w(\overline{s})}{p(p+1)}\right|^2 } , \label{bf}\ ] ] with @xmath186 , @xmath187 and @xmath188 .
the lowest order term in this expression occurs for @xmath189=id , @xmath190 , and thus @xmath191 generalizing ( [ cc ] ) and ( [ weights ] ) to @xmath192 ^ 2-\frac{g \ : \mbox{dim } { \cal g}}{12 } } { 2p(p+1)}\ ; . \label{delg}\end{aligned}\ ] ] we now consider the @xmath14 limit of the polynomial identities as conjectured in sections [ sec2 ] and [ sec3 ] . eliminating the @xmath193 from the fermionic polynomials ( [ adepoly ] ) using ( [ tbaconstr ] ) , and then letting @xmath194 by @xmath195 , yields _ lf_q^g(l ) f_q^g= . .^q^
_ a , b=1^ _ j , k=1 ^ -1 ( ) ^-1_a , b c^_j , k _
a=1^ _ j=2 ^ -1 .
[ finf ] here we have introduced the symbol @xmath196 to mean @xmath197 with @xmath198 included for later convenience .
some caution has to be taken in interpreting the primed sum in ( [ finf ] ) .
no longer is the sum over all solutions to the constraint system ( [ tbaconstr ] ) , but simply over all @xmath199 .
the prime still denotes the condition ( [ constronm ] ) where it is to be noted that we can no longer drop the prime for @xmath6=e@xmath74 .
the @xmath14 limit of the bosonic polynomials @xmath129 defined in ( [ ban ] ) was considered in ref .
the result reads @xmath200 where the branching function is that of the previous subsection based on @xmath6=a@xmath2 .
in taking the infinite limit of the d@xmath2 polynomials ( [ bdn ] ) we get exactly the same result as in ( [ binf ] ) with a@xmath2 replaced by d@xmath2 @xcite and interpreting the branching function as that of @xmath6=d@xmath2 . for the exceptional cases we only have @xmath154 and a simplification occurs as the exceptional branching functions at level 2 reproduce the ordinary virasoro characters @xmath41 in the bosonic representation of equation ( [ xrs ] ) . to illustrate this consider for example @xmath6=e@xmath77 .
then @xmath201 and @xmath202 and we compute from ( [ ccg ] ) @xmath203 and from ( [ delg ] ) @xmath204 here the weights on the left - hand - side are to be understood as those of ( [ delg ] ) based on @xmath6=e@xmath77 and the weights on the right - hand - side as those of equation ( [ weights ] ) or , equivalently , as those of ( [ delg ] ) with @xmath6=a@xmath28 . hence writing the three infinite forms of the bosonic polynomials in terms of virasoro characters yields @xcite @xmath205 with the above definitions ( [ finf])-([delisdel ] ) the @xmath6 rogers ramanujan type identities can be written as @xmath206 for @xmath154 and arbitrary @xmath6 this has been conjectured by kedem _
. @xcite . for @xmath6=a@xmath28 and arbitrary @xmath93 this
was again conjectured by kedem _
_ , this time in ref .
also the general form of ( [ grr ] ) can be inferred from refs .
@xcite , but details such as the precise form of the restrictions ( [ constronm ] ) on the sum in ( [ finf ] ) were not conjectured in full generality . [ sec5 ] an important consideration in the confirmation of the conjectured polynomial identities is their behaviour under the transformation @xmath207 . in particular , for the @xmath6=a@xmath2 case we show that the identity @xmath208 displays the correct invariance under the simultaneous transformations @xmath209 provided we choose @xmath210 .
let us first consider how the general fermionic sum ( [ adepoly ] ) transforms under inversion of @xmath15 . to do
so we rewrite ( [ tbaconstr ] ) in a form expressing the number of anti - particles in terms of the number of particles , and vice versa @xmath212 now , upon using the inversion @xmath213 we find that the exponent of @xmath15 in the expression for @xmath214 reads @xmath215 substituting ( [ nasm ] ) to eliminate @xmath193 this becomes @xmath216 \left[\m{k}{b}-l\ : \delta_{b , p } \left(c^{\a}\right)^{-1}_{k,1 } \right].\end{aligned}\ ] ] then , using ( [ masn ] ) , we arrive at @xmath217 we now carry out a transformation of variables . replacing @xmath218 and @xmath219 followed by @xmath220 and @xmath221 yields @xmath222 of course
, now the sum is over all solutions of @xmath223 , \label{tbadual}\ ] ] where the prime denotes the additional condition @xmath224 or , equivalently , @xmath225 in the following we denote the fermionic sum ( [ adedual ] ) without the irrelevant factor in front of the summation symbol by @xmath226 .
we note the apparent similarity of @xmath227 and @xmath228 under interchange of @xmath6 and a@xmath229 in either one of the two expressions .
we now proceed to consider the inversion properties of the polynomial expressions in the case @xmath6=a@xmath2 only . to explicitly exhibit the dependence of the fermionic polynomials on a@xmath2 as well as on a@xmath229 , we write @xmath230 . then , since @xmath149 and @xmath231 a@xmath2 , we clearly have that @xmath232 , given we choose @xmath120 such that @xmath233 .
this duality of the fermionic polynomials under the transformation ( [ levelrank ] ) is an example of so - called _ level - rank _
duality @xcite . of course ,
for our conjecture @xmath234 to be correct , also the bosonic polynomials @xmath129 in ( [ ban ] ) must remain invariant under transformation ( [ levelrank ] ) . as pointed out in ref .
@xcite this is indeed the case .
apart from the overall factor @xmath235 we again have level - rank duality provided @xmath233 .
it is in fact precisely this level - rank duality for @xmath6=a@xmath2 that led us to the conjecture ( [ adepoly ] ) .
that is , from the already known and proven @xcite polynomial identity for a@xmath28 at level @xmath236 one can immediately infer a polynomial identity for a@xmath229 at level 2 .
to then write down polynomial identities for general rank , level and @xmath6 is a matter of straightforward generalization .
[ sec6 ] it was first shown by richmond and szekeres @xcite that in studying rogers
ramanujan identities in the limit @xmath237 , one obtains identities involving dilogarithms .
we here show that all a - d - e identities of the previous section indeed yield the expected dilogarithm identities .
in particular , we will show that for @xmath7 the rogers
ramanujan identities imply the identities @xmath238 corresponding to the coset conformal field theories ( [ gko ] ) with central charges @xmath40 as given by ( [ ccg ] ) . here
@xmath239 is defined as a sum over the rogers dilogarithm function @xcite @xmath240 \d{\zeta } = -\int_0^z \frac{\log(1-\zeta)}{\zeta}\ : \d{\zeta } + \frac{1}{2 } \log z \ , \log ( 1-z ) \label{dilog}\ ] ] as follows : @xmath241 the numbers @xmath242 in the above sum are the solutions to the tba equations @xcite @xmath243 where by definition @xmath244 . before proceeding to derive ( [ adedilog ] ) from the rogers ramanujan identities ( [ grr ] ) , let us mention that the above dilogarithm identities were first conjectured in ref .
@xcite in the computation of central charges of a - d - e tba systems . for the occurrence of these same identities in related work on tba , see e.g. , refs .
a proof of ( [ adedilog ] ) for @xmath6=a@xmath2 has been given by kirillov in ref .
@xcite .
to establish ( [ adedilog ] ) , we follow the working of nahm _ et al . _ @xcite ( see also refs .
@xcite ) in evaluating @xmath245 , @xmath7 using steepest descent . writing @xmath246
we have @xmath247 treating the @xmath248 as continuous variables , we now approximate the integration kernel by @xmath249 differentiating with respect to @xmath250 to find the saddle point results in the tba equations ( [ tbaxi ] ) , with @xmath251 defined by @xmath252 using the definition ( [ dilog ] ) of the dilogarithm function @xmath120 , the simple relation @xmath253 and the pentagonal relation @xcite @xmath254 yields @xmath255 \nonumber \\ & & \quad + \frac{1}{2 } \sum_{a=1}^{\rg } \sum_{j=1}^{\ell-1 } \left\ { \log q \sum_{b=1}^{\rg } \sum_{k=1}^{\ell-1 } \left(\cg\right)^{-1}_{a , b } c^{\a}_{j , k } \m{j}{a } \ , \m{k}{b } \right .
\label{beforecancel } \\ & & \quad \left .
\vphantom{\left(\cg\right)^{-1}_{a , b } } + \left(p_j^{(a ) } + \m{j}{a } \right ) \log \left ( 1-q^{p_j^{(a ) } + \m{j}{a } } \right ) - p_j^{(a ) } \log \left(1-q^{p_j^{(a ) } } \right ) -\m{j}{a } \log \left(1-q^{\m{j}{a } } \right ) \right\}. \nonumber\end{aligned}\ ] ] the variables @xmath256 in this expression are given by @xmath257 and , by ( [ defp ] ) and ( [ defxi ] ) , they satisfy @xmath258 clearly , the @xmath256 , @xmath259 satisfy the same tba equations as the @xmath260 , @xmath59 , but with @xmath93 replaced by @xmath236 .
the following elementary manipulations serve to show that the last two lines in ( [ beforecancel ] ) cancel as a consequence of the tba equations ( [ tbaxi ] ) : + & & = _ a , b=1^ _ j ,
k=1 ^ -1 ( ) ^-1_a , b c^_j , k ( ) + & & = _ a=1^
_ j=1 ^ -1 ( 1-_j^(a ) ) - _ a , b=1^ _ j , k=1 ^ -1 ( ) ^-1_a , b c^_j , k ( _ k^(b ) ) + & & = _ a=1^
_ j=1 ^ -1 \ { p_j^(a ) ( 1-q^p_j^(a ) ) + ( 1-q^ ) - ( p_j^(a ) + ) ( 1-q^p_j^(a ) + ) .
- ( p_j^(a ) + _ b=1^ _ k=1 ^ -1 ( ) ^-1_a , b c^_j , k ) ( _ j^(a ) ) } + & & = _ a=1^ _
j=1 ^ -1 \ { p_j^(a ) ( 1-q^p_j^(a ) ) + ( 1-q^ ) -(p_j^(a ) + ) ( 1-q^p_j^(a ) + ) } . as a result of this
we have @xmath261 \nonumber \\ & = & -m \log q - \frac{\pi^2}{6 \log q } \left [ s^{\ , \cal g}(\ell-1 ) + s^{\ , \cal g}(1 ) - s^{\ , \cal g}(\ell ) \right ] .
\label{simple}\end{aligned}\ ] ] finally we have to fix the value of @xmath15 at the saddle point . from @xmath262
this is found to be @xmath263.\ ] ] returning to the expression ( [ am ] ) hence yields the following result for the asymptotics of @xmath264 @xmath265 m/6}\right ) .
\label{sss}\ ] ] to actually obtain the identity ( [ adedilog ] ) we have to show that in addition to ( [ sss ] ) we also have @xmath266 with @xmath40 given by ( [ ccg ] ) .
however , from the automorphic properties of the @xmath6 theta functions this can indeed be established @xcite .
that is , by carrying out the modular transformation @xmath267 which relates the @xmath237 to the @xmath268 limit of the branching functions ( [ bf ] ) then using for @xmath269 that @xmath270 , the central charge follows as given in ( [ ccg ] ) .
in fact , it is precisely the prefactor @xmath271 in the expansion ( [ bfexp ] ) of the branching functions that ensures proper modular invariance of the branching rule ( [ brule ] ) . before ending this section
let us make some concluding remarks .
we have presented the dilogarithm identities ( [ adedilog ] ) as special sums over the solution to the tba equations ( [ tbaxi ] ) avoiding the problem of explicitly solving these equations . in ref .
@xcite an elegant algebraic formulation of the solution for @xmath6=a@xmath2 and d@xmath2 has however been given , based on the work of ref .
@xcite on the representations of yangians . for @xmath6=a@xmath2
the solution is especially simple @xmath272 and can in fact easily be checked by direct substitution in ( [ tbaxi ] ) . for the exceptional cases no explicit form of the solutions
is known , but some conjectures towards a solution have been made in ref .
@xcite .
another remark to be made is that the identities ( [ adedilog ] ) are in fact consequences of the stronger identities @xcite @xmath273 where we recall that @xmath184 in ( [ ccg ] ) .
[ sec7 ] in this paper we have presented conjectures for polynomial identities of a - d - e type following the ` solvable lattice model ' approach of melzer @xcite ( see also ref .
all polynomials identities can be viewed as finitizations of rogers ramanujan type identities for branching functions associated with the gko pair @xmath274 at levels @xmath236 , 1 and @xmath93 , respectively .
apart from extensive computer tests we have corroborated our conjectures by studying the behaviour under level - rank duality for @xmath6=a@xmath2 and by establishing the expected dilogarithm identities following from the a - d - e rogers ramanujan in the asymptotic limit @xmath7 .
given that we have no proofs of our conjectures a lot of course remains to be done .
first of all it is to be noted that we have only presented fermionic polynomials associated with one particular choice of one - dimensional configuration sum , or , equivalently , we have only given fermionic finitizations of one particular branching function associated with the above mentioned coset pair .
this is in contrast to melzer s original work where fermionic polynomials for all virasoro characters @xmath41 were conjectured .
in fact , the method of proof for the a@xmath28 polynomial identities @xcite relies heavily on the completeness of a set of fermionic polynomials in order to apply recurrence methods similar to those used for obtaining the bosonic polynomials . however , despite some efforts to find fermionic counterparts to all configuration sums @xmath106 we did not succeed to do so in general .
so , for example , for @xmath6=a@xmath2 we only managed to find fermionic forms corresponding to configuration sums of the form @xmath275 with @xmath276 , @xmath277 .
even so , assuming that we would have been able to establish completeness , it still seems that the recurrence method developed in ref .
@xcite is far from ideal to tackle the general a@xmath2 case . in this respect , exploiting higher rank partition theory might be a more promising way to go .
( for recent progress on proofs of polynomial identities for finitized virasoro characters using solely partition theoretic arguments , see ref .
@xcite . ) setting aside the problem of proof , it is quite clear that many of the results of this paper admit further generalization . instead of defining the fermionic polynomials ( [ adepoly ] ) based on the constraint system ( [ tbaconstr ] ) ,
we could have started with the more general equation @xmath278 \label{tbas}\ ] ] as considered in parameter dependent form in ref .
the parameter @xmath279 herein provides the generalization to the cosets @xmath280 and by letting @xmath38 taking any of the values @xmath281 , instead of fixing it as in section [ sec2 ] , we can obtain fermionic finitizations to several and not just 1 branching function associated with ( [ gkos ] ) .
the problem is then of course to also define appropriate bosonic polynomials to match the fermionic expressions . for the case @xmath6=a@xmath28 ,
@xmath282 these should be provided by ctm calculations of ref .
@xcite for the fused abf models . in ref .
@xcite an even further generalization to ( [ tbaconstr ] ) was proposed , extending ( [ tbas ] ) to the case of non - simply laced lie algebras . clearly , in defining the appropriately generalized form of the fermionic expression ( [ adepoly ] )
, this should for @xmath46 correspond to the one - dimensional configuration sums for the b@xmath283 and c@xmath283 rsos models of ref .
in fact , quite possibly the only change that has to be made in ( [ adepoly ] ) is the replacement of @xmath93 by @xmath284 to account for the fact that not all simple roots @xmath285 have equal length . here
@xmath286 is defined by normalizing @xmath287 and setting @xmath288 , @xmath57 being the @xmath57-th simple root .
we hope to address some of the above mentioned problems and generalizations in future publications .
we acknowledge many helpful and stimulating discussions on rogers ramanujan identities with omar foda .
we also thank omar foda , peter forrester and yu - kui zhou for kind interest and encouragements .
this work is supported by the australian research council .
a. berkovich , fermionic counting of rsos - states and virasoro character formulas for the minimal series @xmath289
. exact results .
preprint bonn - he-94 - 04 , hep - th/9403073 .
to appear in _ nucl .
phys . _ * b431 * ( 1994 ) . a sketch of the proof has been given in : s. o. warnaar and p. a. pearce , exceptional structure of the dilute a@xmath165 model : e@xmath77 and e@xmath75 rogers
ramanujan identities .
preprint melbourne university , hep - th/9408136 . to appear in _
j. phys . a : math . | we conjecture polynomial identities which imply rogers ramanujan type identities for branching functions associated with the cosets @xmath0 , with @xmath1=a@xmath2 , d@xmath2 @xmath3 , e@xmath4 @xmath5 . in support of our conjectures
we establish the correct behaviour under level - rank duality for @xmath6=a@xmath2 and show that the a - d - e rogers
ramanujan identities have the expected @xmath7 asymptotics in terms of dilogarithm identities .
possible generalizations to arbitrary cosets are also discussed briefly . -10 mm # 1 |
the mgii @xmath0 doublets are among the absorption features commonly seen in the spectra of distant quasars that are produced by intervening gaseous clouds along the quasar lines of sight .
their rest - frame absorption equivalent width @xmath9 , ranging from @xmath10 to @xmath11 ( e.g. steidel & sargent 1990 ; nestor @xmath12 2005 ; prochter @xmath12 2006a ) , is found to represent the underlying gas kinematics ( e.g. petitjean & bergeron 1990 ; churchill @xmath12 2000 ) .
based on comparisons of the abundance ratios of various associated ions , these mgii absorption transitions are understood to arise primarily in photo - ionized gas of temperature @xmath13 k ( bergeron & stasnska 1986 ; hamann 1997 ) and neutral hydrogen column density @xmath14 ( churchill @xmath12 2000 ; rao @xmath12 2006 ) . in addition , surveys for galaxies near the observed mgii absorbers have often uncovered luminous galaxies at projected distances ( @xmath15 kpc ) and velocity separations ( @xmath16 @xmath17 ) from the absorbers ( bergeron 1986 ; lanzetta & bowen 1990 , 1992 ; steidel @xmath12 1994 ; zibetti @xmath12 2005 ; nestor @xmath12 2007 ; kacprzak @xmath12 2007 ) .
the relatively large associated @xmath18 and the small separation between mgii absorbers and galaxies along common lines of sight indicate that mgii absorbers may offer a sensitive probe of photo - ionized halo gas around galaxies at redshifts from @xmath19 to @xmath20 in the optical spectral window .
understanding the physical origin of these mgii absorbers bears significantly on all efforts to apply their known statistical properties for constraining the baryon content of dark matter halos on different mass scales ( tinker & chen 2008 ) .
various theoretical models have been developed in the past that describe the nature of the absorbers in the context of gas accretion , including ram pressure stripped gas from accreted satellites ( wang 1993 ) , gravitationally bound cold gas in halo substructures ( e.g. , sternberg @xmath12 2002 ) , and condensed cold clouds in a thermally unstable hot halo ( mo & miralda - escude 1996 ; maller & bullock 2004 ; chelouche @xmath12 2007 ) .
insights to the origin of mgii absorbers can be obtained from their clustering amplitude on large scales ( @xmath21 mpc ) .
the clustering of the absorbers is a consequence of dark matter halos in which they are found .
measurements of the galaxy and absorber cross correlation function offer a means of quantifying the mean halo mass of the absorbers .
high - mass halos are expected to be highly clustered , while low mass halos on average have weaker clustering strength .
the cross
correlation function of mgii absorbers and luminous red galaxies ( lrgs ) have been measured at @xmath22 by bouch @xmath12 ( 2006 ) , who found that mgii absorbers of @xmath23 cluster strongly with lrgs and that weaker absorbers with @xmath24 on average arise in dark matter halos that are 10 times more massive than those of stronger absorbers with @xmath25 .
both the observed clustering amplitude and the inverse correlation between the mean halo mass and absorber strength are difficult to interpret .
for example , the frequency distribution function of mgii absorbers show that there are on average 10 times more @xmath26 absorbers than those of @xmath27 ( e.g. nestor @xmath122005 ; prochter @xmath12 2006a ) which , when combined with the clustering measurements , implies that the majority of absorbers reside in massive , highly biased halos . at the same time
, we expect that observed mg@xmath28 ions originate primarily in photo - ionized gas of temperature @xmath13 k and the halo gas in massive dark matter halos becomes too hot for abundant mg@xmath28 to survive . in tinker & chen ( 2008 ;
hereafter tc08 ) , we developed a new technique that adopts the halo occupation framework for understanding the origin of qso absorption - line systems .
specifically , the technique adopts a model density profile for the mg@xmath28 ions in individual dark matter halos .
the `` cold '' baryon content of individual dark matter halos , as probed by the presence of mg@xmath28 , is then constrained through matching the space density and clustering amplitude of dark matter halos with the observed frequency distribution function of mgii absorbers and their clustering amplitude .
our model allows the possibility that a predominant fraction of the gas in massive halos is shock heated to the virial temperature of the halo and becomes too hot to host abundant mgii absorbing clouds . within the hot halo
, we further allow the possibility that some dense , cold clouds may penetrate through as seen in high resolution cosmological simulations ( kravtsov 2003 ) .
the result of this halo occupation analysis is the probability function @xmath29 that characterizes , for each dark matter halo of @xmath30 , the total probability of finding a mgii absorber of equivalent width @xmath31 .
the halo occupation analysis shows that observations , including both the mgii absorber frequency distribution function and their clustering amplitude , demand a rapid transition in the halo gas content at @xmath32 .
below @xmath33 , halos contain predominantly cold gas and therefore contribute significantly to the observed mgii statistics . beyond @xmath33 ,
the cold gas fraction is substantially reduced and presumably the halo gas becomes too hot to maintain a large contribution to mgii absorbers . in order to reproduce the observed overall strong clustering of the absorbers and the inverse correlation between @xmath9 and halo mass @xmath30 , roughly 5% of the gas in halos up to @xmath34 is required to be cold .
it is understood under our model that _
the @xmath9 vs. @xmath30 inverse correlation arises mainly as a result of an elevated clustering amplitude of @xmath35 absorbers , rather than a suppressed clustering strength of @xmath36 absorbers . _
the clustering amplitude of weak absorbers is elevated because of the presence of cold streams in massive hot halos .
the amount of cold gas in clusters is small , therefore these halos can only contribute to weak absorbers .
the initial results of our halo occupation analysis demonstrate that combining known statistics of dark matter halos with a simple model for their gas content can already reproduce the statistical properties known for mgii absorbers .
it provides a simple prescription for populating baryons in dark matter halos that can be compared directly to results from numerical simulations and offer insights for understanding the physics of gas accretion in halos of all mass scales .
however , some uncertainties remain in the halo occupation analysis .
first , a key parameter that constrains the distribution of mg@xmath28 ions in individual dark matter halos is their incidence rate , @xmath37 , per halo .
it specifies the total probability of detecting an absorber in a halo of mass @xmath30 . in the halo occupation analysis of tc08
, we find that dark matter halos of @xmath38 on average have 100% incidence rate of mg@xmath28 ions , @xmath39 , at @xmath40 for mg@xmath28 ions and @xmath41 at larger radii . corresponds to the radius , within which the enclosed mean density is @xmath42 and @xmath43 is the background density .
it is motivated by estimates of the virial radius both from the spherical collapse model and from @xmath44-body simulations that predict @xmath45 . ] in halos of lower masses , @xmath46 declines sharply . while a 100% incidence rate shows that every dark matter halo hosts a uniform gaseous halo of mg@xmath28 ions with size @xmath8 , the low @xmath46 is more difficult to interpret because @xmath46 represents a mean value averaged over all halos of mass @xmath30 .
a low @xmath46 may be a result of only a small fraction of dark matter halos containing extended distributions of mg@xmath28 ions or a result of a small covering factor of mg@xmath28 in all halos .
in addition , we have assumed in our initial analysis that the gaseous extent is related to halo mass according to @xmath47^{\beta}$ ] , and @xmath48 physical kpc and @xmath49 .
this scaling relation follows the theoretical expectation between virial radius and halo mass , but a higher @xmath8 would naturally lower @xmath50 for a fixed frequency distribution function of mgii absorbers .
second , the physical origin of the cold clouds probed by the mgii absorption transitions is ambiguous .
while the mass scale at which halo gas is found to experience a transition from a cold - mode dominated state to a hot - mode dominated state agrees well with the expectation of theoretical models established to characterize the gas accretion history in dark matter halos ( e.g. birnboim & dekel 2003 ; kere @xmath12 2005 ; dekel & birnboim 2006 ; birnboim @xmath12 2007 ) , the observed cold - hot transition can also be interpreted as a declining formation efficiency of cool clouds in a hot halo with increasing halo mass ( e.g. mo & miralda - escude 1996 ; maller & bullock 2004 ) .
both of these scenarios predict that halo gas is primarily cold at @xmath51 .
on the other hand , bouch @xmath12 ( 2006 ) have argued that the inverse correlation between @xmath9 and clustering amplitude is suggestive of the absorbers arising in non - virialized gas flows , such as starburst winds . while dense clumps in starburst driven outflows are expected to contribute to some fraction of the observed mgii absorbers , an important quantity to specify
is the significance of this fraction .
if a large fraction of mgii absorbers originate in cold dense clumps in starburst winds , then one must also account for the more complex star formation physics in efforts to apply known mgii statistics for constraining the gas content of dark matter halos .
we defer to 5.3 for a more detailed discussion of the caveats in interpreting the known properties of mgii absorbers as evidence to support their origin in starburst winds .
direct constraints on the covering fraction of mgii absorbing gas versus halo mass not only help to break the degeneracy between the fraction of dark matter halos containing cold baryons and the extent of cold gas in individual halos , in principle they also serve to discriminate between the starburst wind scenario and gas accretion . under the starburst wind scenario ,
superwind driven outflows are expected to proceed over a finite angular span along the minor axis ( e.g. heckman @xmath12 1990 ; veilleux @xmath12 2005 ) , and result in only partial covering of the halos .
in addition , because at a given epoch only some fraction of galaxies are found at starburst or post - starburst stages , we would expect a fraction of dark matter halos to contain extended mgii absorbing gas due to outflows . to obtain empirical constraints on the extent and covering fraction of mg@xmath28 ions around galaxies of different luminosity and mass ,
we have initiated a survey of mgii absorbers around known galaxies at small projected distances ( @xmath52 kpc ) to a background qso .
accounting the presence or absence of mgii absorbers for an unbiased sample of galaxies at different impact parameters yields a statistical estimate of the gas covering fraction around field galaxies .
a detailed comparison between the absorber strength and galaxy properties further constrains the density profile and extent of the gas .
the primary objectives of our study are ( 1 ) to improve the uncertainties of our halo occupation analysis by obtaining empirical measurements of @xmath8 and @xmath53 , and ( 2 ) to examine whether starburst driven outflows are the predominant mechanism for producing the observed mgii absorbers by obtaining an empirical constraint of the gas covering factor @xmath50 . here
we present the initial results of our study .
this paper is organized as follows . in section 2
, we describe the design of our experiment for constraining the extent and covering fraction of mg@xmath28 ions . in section 3
, we describe the imaging and spectroscopic data available for our experiment . in section 4
, we examine the correlation between mgii absorption strength and galaxy properties , such as the impact parameter and luminosity .
we compare our results with previous studies and discuss the implications of our analysis in section 5 .
we adopt a @xmath54cdm cosmology , @xmath55 and @xmath56 , with a dimensionless hubble constant @xmath57 throughout the paper .
accurate estimates of the extent and covering fraction of mg@xmath28 ions around galaxies require an unbiased survey of mgii absorbers in the vicinity of _ known _ galaxies .
nearly all studies published in the literature to date focus on surveys of galaxies that correspond to known mgii absorbers ( e.g. bergeron 1986 ; lanzetta & bowen 1990 , 1992 ; steidel @xmath12 1994 ; zibetti @xmath12 2007 ; kacprzak @xmath12 2007 ) , with the exception of an on - going study described in tripp & bowen ( 2005 ) . searching for galaxies that give rise to known mgii absorbers bears a large uncertainty due to contamination from galaxies that are correlated with the true absorbing galaxies .
in addition , these surveys based on known presence of mg@xmath28 ions do not yield constraints on the incidence of these ions in galactic halos .
we are conducting a survey of mgii absorbers at the locations of known galaxies close to the line of sight toward a background qso .
we identify galaxies at impact parameters @xmath58 .
the value of @xmath59 is selected to allow identifications of galaxies at impact parameters beyond the size of extended mgii halos , in order to constrain the extent and covering fraction of the gas . at the same time , we require @xmath59 to be small enough in order to minimize the contamination of galaxies that do not give rise to the absorbers but are merely correlated with the absorbing galaxies through clustering . to estimate the contamination rate of correlated galaxies as a function of projected distance , we adopt the distribution of dark matter halos identified in cosmological simulations as a guide .
we use the @xmath60 output of a @xmath61 co - moving mpc simulation described in tinker @xmath12 ( 2007 ) .
first , we identify for each random line of sight dark matter halos that have @xmath62 and are located at @xmath63 from the sightline . from tc08 , we consider these halos likely to host a mgii absorber along the line of sight .
this is also consistent with the results from the halo occupation distribution models of galaxies found in deep2 and sloan digital sky survey from zheng , coil , & zehavi ( 2007 ; hereafter zcz07 ) .
these authors found that the minimum mass of a dark matter halo that on average hosts a galaxy of @xmath64 is @xmath65 .
next , for each potential mgii absorber , we search for neighboring halos that are at @xmath66 @xmath17 and projected distance less than @xmath67 from the absorber and may therefore contaminate the search of the true absorber . finally , we repeat this exercise over 10,000 random lines of sight and calculate the frequency of finding contaminating halos as a function of @xmath67 .
@xmath17 and at projected distances less than @xmath67 from an absorber .
the errorbars represent the 68% scatter that was estimated using a jackknife resampling technique . ] the result of our exercise is presented in figure 1 , where we show the probability of contamination due to correlated galaxies versus impact parameter .
the errorbars represent the 68% scatter of the mean from 16 jackknife subsamples .
figure 1 shows that at @xmath68 physical kpc there is a @xmath69 30% chance that one would mis - identify a correlated galaxy as the true absorber . combining the contamination rate estimates with previous search reports that indicate a mean gaseous extent of @xmath70 physical kpc for luminous galaxies ( bergeron 1986 ; lanzetta & bowen 1990 , 1992 ; steidel @xmath12 1994 ) , we carry out our study based on a sample of galaxies that are within @xmath71 kpc from a background qso .
to study the extent and covering fraction of mg@xmath28 ions in galactic halos , it requires survey data of both galaxies and absorbers along common lines of sight toward background quasars .
we have selected seven qso fields for which galaxies data are already available , and echelle spectra of the background qsos have also been obtained either from our own observations or through the eso science data archive .
we describe available qso spectra and galaxy data separately in the following sections .
echelle spectroscopic observations of the qsos , pks0122@xmath720021 ( @xmath73 ) , he 0226@xmath724110 ( @xmath74 ) , pks0349@xmath721438 ( @xmath75 ) , and pks0454@xmath722203 ( @xmath76 ) were obtained using the mike echelle spectrograph ( bernstein @xmath12 2003 ) on the magellan clay telescope in 2007 october .
the spectrograph contains a blue camera and a red camera , allowing a full wavelength coverage from near - ultraviolet 3350 through near - infrared 9400 . depending on the brightness of the qso ,
the observations were carried out in a sequence of one to three exposures of duration 600 s to 900 s each .
the mean seeing condition over the period of integration was @xmath77 .
we used a @xmath78 slit and @xmath79 binning during readout , yielding a spectral resolution of fwhm @xmath80 @xmath17 at wavelength @xmath81 and @xmath82 @xmath17 at @xmath83 .
the data were processed and reduced using the mike data reduction software developed by burles , prochaska , & bernstein .
wavelengths were calibrated to a thar frame obtained immediately after each exposure and subsequently corrected to vacuum and heliocentric wavelengths .
flux calibration was performed using a sensitivity function derived from observations of the flux standard eg131 . the flux calibrated spectra have a mean @xmath84 per resolution element across the entire spectral range .
echelle spectra of the qsos , pks1354@xmath851933 ( @xmath86 ) , pks1424@xmath721150 ( @xmath87 ) , and pks1622@xmath8523 ( @xmath88 ) were obtained using the ultraviolet and visual echelle spectrograph ( uves ; dodorico @xmath12 2000 ) on the vlt / ut2 telescope and were retrieved from the eso data archive .
observations of pks1354@xmath851933 ( program i d 076.a@xmath720860 ) and pks1424@xmath721150 ( program i d 075.a@xmath720841 ) were obtained using a @xmath78 slit and the dichroics # 1 b346@xmath85r580 and # 2 b437@xmath85r860 in a sequence of two to three exposures of duration 180s to 300 s each .
the data were binned @xmath79 during readout .
observations of pks 1622@xmath8523 ( program i d 069.a@xmath720371 ) were obtained using a @xmath78 slit and the dichroic # 1 b346@xmath85r580 in a sequence of three exposures of duration 4900 s each .
the data were also binned @xmath79 during readout .
the uves spectra were processed and calibrated using the standard data reduction pipeline .
the combined spectra of the three qsos cover a spectral range from 3300 to 6600 , with a pixel scale of 0.035 per pixel and @xmath89 per pixel at @xmath90 .
the high @xmath91 and high spectral resolution mike and uves spectra together allow us to search for mgii absorption features at redshifts @xmath92 to a limiting absorption strength @xmath93
along the lines of sight toward the seven background qsos .
p1incccccrc pks0122@xmath720021 & 01:25:28.8 & @xmath7200:05:56 & 1.0765 & 16.4 & mike & @xmath94 & 33509400 he0226@xmath724110 & 02:28:15.2 & @xmath7240:57:16 & 0.4950 & 15.2 & mike & @xmath95 & 33509400 pks0349@xmath721438 & 03:51:28.5 & @xmath7214:29:09 & 0.6163 & 16.2 & mike & @xmath96 & 33509400 pks0454@xmath722203 & 04:56:08.9 & @xmath7221:59:09 & 0.5335 & 16.1 & mike & @xmath97 & 33509400 pks1354@xmath851933 & 13:57:04.4 & @xmath8519:19:07 & 0.7200 & 16.0 & uves & @xmath95 & 33006800 pks1424@xmath721150 & 14:27:38.1 & @xmath7212:03:50 & 0.8060 & 16.5 & uves & @xmath98 & 33006600 3c336 & 16:24:39.1 & @xmath8523:45:12 & 0.9274 & 17.5 & uves & @xmath99 & 33006600 for the purpose of the study described in 2
, we need to identify a representative sample of galaxies at @xmath52 kpc from the background qsos .
galaxies in six of the seven qso fields listed in table 1 have been observed and analyzed by chen @xmath12 ( 1998 , 2001a ) .
specifically , galaxies in the fields around pks0122@xmath720021 , pks0349@xmath721438 , and pks0454@xmath722203 , pks1354@xmath851933 , and pks1424@xmath721150 were observed as part of the survey program to study the gaseous extent of galaxies based on the presence or absence of corresponding and civ absorption features in the spectra of the background qsos ( lanzetta @xmath12 1995 ; chen @xmath12 1998 , 2001a ; chen @xmath12 2001b ) .
the authors carried out a redshift survey of galaxies brighter than @xmath100 mag at angular distances @xmath101 .
the survey depth enabled findings of galaxies with luminosities greater than @xmath102 and impact parameters less than @xmath103 kpc at the redshifts @xmath104 .
the galaxy sample is therefore ideally suited for our present study .
galaxies in the field of pks1622@xmath852352 were first published in steidel @xmath12 ( 1997 ) as part of a program to determine the redshifts of galaxies found in deep hubble space telescope ( hst ) images that are brighter than @xmath105 .
images of individual galaxies were also analyzed by chen @xmath12 ( 1998 ) .
galaxies in the field of he0226@xmath724110 were observed in an on - going program to uncover galaxies brighter than @xmath106 mag at @xmath107 for understanding the cross - correlation between galaxies and ovi @xmath108 absorbers ( chen @xmath12 2008 in preparation ) .
the two galaxy samples satisfy the selection criterion of an unbiased galaxy survey in qso fields that does not rely on known knowledge of the presence of an mgii absorber .
we therefore include the galaxy data in our sample .
images of the galaxies are presented in figure 2 .
the data were obtained using the wide field and planetary camera 2 ( wfpc2 ) and the f702w filter ( see also chen @xmath122001a ) on board hst , with the exception of the images of galaxies in the field around he0226@xmath724110 that were obtained using imacs and the @xmath109-band filter on the magellan baade telescope .
the sample of galaxies identified at @xmath52 kpc from the lines of sight toward seven background qsos is presented in table 2 , where we list for each galaxy in columns ( 1)(7 ) the field , right ascension and declination offsets from the qso @xmath110 and @xmath111 , redshift @xmath112 , impact parameter @xmath67 , apparent @xmath109-band magnitude , and absolute @xmath113-band magnitude @xmath114 . system . ] measurement uncertainties in @xmath109 and @xmath115 were typically @xmath116 dex .
our initial galaxy sample consists of 23 galaxies at redshifts from @xmath117 to @xmath118 with a median of @xmath119 to @xmath120 with a median of @xmath22 , covering a comoving distance of @xmath121 gpc .
therefore these galaxies are not part of a coherent structure .
there is no known galaxy clusters along the sightline toward 3c336 and there is no physical mechanism that suggests that slight overdensities on scales of 0.1 - 1 gpc would affect gas properties within individual halos . ] .
we have excluded galaxies that are within @xmath122 @xmath17 from the background qsos , where most of the gas is expected to be highly ionized due to the qso proximity effect ( e.g. bajtlik @xmath12 1988 ) .
the rest - frame @xmath113-band magnitudes of the galaxies range from @xmath123 to @xmath124 with a median of @xmath125 , corresponding to @xmath126 for @xmath127 that characterizes the blue galaxy population at @xmath128 ( e.g. faber @xmath12 2007 ) . to establish a galaxy mgii absorber pair sample for the subsequent analysis
, we examine the echelle spectra of the background qsos and search for the corresponding mgii absorption doublet at the locations of the galaxies presented in 3.2 .
specifically , we accept absorption lines according to a @xmath129 detection threshold criterion , which is appropriate because the measurements are performed at a small number of known galaxy redshifts .
next , we form galaxy and absorber pairs by requiring a velocity difference in the field around 1354@xmath851933 is @xmath130 @xmath17 from the identified mgii absorber . however ,
including the redshift measurement uncertainty of 300 @xmath17 reported in ellingson @xmath12 ( 1991 ) , we consider this pair a match . ]
@xmath131 @xmath17 .
finally , we measure 3-@xmath132 upper limits to the 2796- absorption equivalent widths of galaxies that are not paired with corresponding absorbers .
the procedure identifies 13 galaxy and absorber pairs and 10 galaxies that do not produce mgii absorption lines to within sensitive upper limits .
impact parameter separations of the galaxy and absorber pairs range from @xmath133 to @xmath134 physical kpc with a median of @xmath135 kpc .
the corresponding mgii absorption strength of each of the 23 galaxies in our sample is listed in columns ( 8)(9 ) of table 2 .
lrrccccccc 0122@xmath720021 & @xmath136 & @xmath137 & @xmath138 & @xmath139 & @xmath140 & @xmath141 & & @xmath142 & @xmath143 + 0226@xmath724110 & @xmath144 & @xmath144 & @xmath145 & @xmath146 & @xmath147 & @xmath148 & & ... & @xmath149 + & @xmath150 & @xmath151 & @xmath152 & @xmath153 & @xmath154 & @xmath155 & & @xmath156 & @xmath157 + 0349@xmath721438 & @xmath158 & @xmath159 & @xmath160 & @xmath161 & @xmath140 & @xmath162 & & @xmath163 & @xmath164 + & @xmath165 & @xmath166 & @xmath167 & @xmath168 & @xmath169 & @xmath162 & & ... & @xmath170 + 0454@xmath722203 & @xmath171 & @xmath172 & @xmath173 & @xmath174 & @xmath175 & @xmath176 & & ... & @xmath170 + & @xmath177 & @xmath178 & @xmath179 & @xmath180 & @xmath181 & @xmath124 & & @xmath182 & @xmath183 + & @xmath184 & @xmath185 & @xmath186 & @xmath187 & @xmath188 & @xmath162 & & ... & @xmath170 + 1354@xmath851933 & @xmath189 & @xmath190 & @xmath191 & @xmath192 & @xmath193 & @xmath194 & & @xmath195 & @xmath196 + & @xmath197 & @xmath198 & @xmath199 & @xmath200 & @xmath201 & @xmath162 & & ... & @xmath170 + 1424@xmath721150 & @xmath202 & @xmath203 & @xmath204 & @xmath205 & @xmath206 & @xmath162 & & @xmath207 & @xmath208 + 3c336 & @xmath209 & @xmath210 & @xmath211 & @xmath212 & @xmath213 & @xmath214 & & @xmath215 & @xmath216 + & @xmath217 & @xmath218 & @xmath219 & @xmath220 & @xmath221 & @xmath222 & & @xmath223 & @xmath224 + & @xmath225 & @xmath190 & @xmath226 & @xmath227 & @xmath228 & @xmath229 & & ... & @xmath230 + & @xmath231 & @xmath232 & @xmath233 & @xmath234 & @xmath235 & @xmath236 & & @xmath237 & @xmath238 + & @xmath239 & @xmath240 & @xmath241 & @xmath242 & @xmath243 & @xmath229 & & ... & @xmath170 + & @xmath244 & @xmath245 & @xmath246 & @xmath247 & @xmath248 & @xmath249 & & @xmath250 & @xmath251 + & @xmath252 & @xmath253 & @xmath254 & @xmath255 & @xmath256 & @xmath257 & & @xmath258 & @xmath259 + & @xmath137 & @xmath260 & @xmath261 & @xmath262 & @xmath263 & @xmath264 & & @xmath265 & @xmath266 + & @xmath267 & @xmath268 & @xmath269 & @xmath270 & @xmath271 & @xmath272 & & ... & @xmath170 + & @xmath273 & @xmath274 & @xmath275 & @xmath276 & @xmath277 & @xmath278 & & @xmath279 & @xmath280 + & @xmath166 & @xmath281 & @xmath275 & @xmath282 & @xmath283 & @xmath148 & & ... & ... + & @xmath284 & @xmath203 & @xmath285 & @xmath286 & @xmath287 & @xmath288 & & ... & @xmath289 + & @xmath290 & @xmath291 & @xmath292 & @xmath293 & @xmath294 & @xmath295 & & ... & @xmath149 +
in this section , we examine the correlation between galaxies and mgii absorbers identified along common lines of sight , using the sample of 13 galaxy mgii absorber pairs and 10 galaxies at @xmath52 kpc that do not give rise to mgii absorption to a sensitive upper limit .
the choice of @xmath71 kpc in selecting galaxies at @xmath296 for our analysis is adopted , in order to minimize the contamination of galaxies that are correlated with the true mgii absorbing galaxies through clustering and to maximize the fraction of physical galaxy and absorber pairs in our sample .
in addition to the simulation results presented in 2 , this criterion is further supported by the stronger clustering signal at small ( @xmath297 kpc ) impact parameters of lrgs ( bouch @xmath12 2004 ) .
this strong signal is a natural consequence of cross - correlating objects inside the same dark matter halos of the lrgs . to examine the existence of a fiducial relationship between the strength of mgii absorption and galaxy impact parameter , we present in figure 3 the distribution of @xmath9 with respect to @xmath67 .
circles represent elliptical or s0 galaxies , triangles represent early - type spiral galaxies , and squares represent late - type spiral galaxies .
galaxy morphological types are adopted from the classifications of chen @xmath12 ( 1998 , 2001a ) that were based on a 2d surface profile analysis of the high resolution hst / wfpc2 images .
we have also classified the galaxy at ( @xmath298 ) and the galaxy at ( @xmath299 ) from the qso he0226@xmath724110 as a late - type and an early - type spiral galaxy based on their respective spectral features .
points with arrows in figure 3 indicate @xmath129 upper limits . versus galaxy impact parameter @xmath67 for a sample of 23 galaxies .
circles represent elliptical or s0 galaxies , triangles represent early - type spiral galaxies , and squares represent late - type spiral galaxies .
points with arrows indicate @xmath129 upper limits . ]
figure 3 shows that there is a moderate anti - correlation between @xmath9 and @xmath67 , i.e. weaker mgii absorption strength is observed for galaxies at larger impact parameter .
we note that the obvious outlier of a mgii absorber with @xmath300 at @xmath301 , currently assigned to a galaxy at @xmath302 kpc , is a known damped absorber of @xmath303 ( steidel @xmath12 1997 ; rao @xmath12 2006 ) .
although the galaxy satisfies our sample selection criterion , it is unlikely that the galaxy at @xmath302 kpc is physically related to the absorber due to the large neutral gas content for two reasons .
first , all radio detected sources in non - cluster environment from deep 21 cm surveys have coincident optical counterparts ( e.g. doyle @xmath12 2005 ) .
second , deep 21 cm observations of the m31 halo reveal no hi clouds of @xmath304 ( e.g. thilker @xmath12 2004 ) .
this is also roughly consistent with the expected contamination rate at this projected separation presented in figure 1 . to assess the significance of this anti - correlation
, we perform a generalized kendall test that accounts for the presence of non - detections and find that the null hypothesis , in which @xmath9 is randomly distributed with respect to @xmath67 , can be ruled out at a @xmath305% confidence level .
however , the scatter of the @xmath9 vs.@xmath67 anti - correlation is clearly large .
specifically , 10 of the 23 galaxies at @xmath52 kpc do not produce a corresponding mgii absorber to a sensitive upper limit , implying a mean covering factor of the mg@xmath28 ions @xmath306 % at @xmath297 kpc . at @xmath307 kpc , we find @xmath308 % , similar to the preliminary result reported in tripp & bowen ( 2005 ) .
we note , however , that the rest - frame @xmath113-band magnitudes of the galaxies range from @xmath123 to @xmath124 with a median of @xmath125 , corresponding to a range in intrinsic @xmath113-band luminosity from @xmath309 to @xmath310 with a median of @xmath126 for @xmath127 at @xmath128 ( e.g. faber @xmath12 2007 ) .
because more massive galaxies are expected to reside in more massive halos and therefore possess larger gaseous halos , we examine whether accounting for galaxy intrinsic luminosity can improve upon the anti - correlation displayed in figure 3 .
following tc08 , we adopt an isothermal profile of finite extent @xmath8 for representing the density of the mg@xmath28 ions around individual galaxies . the expected mgii absorption
equivalent width , @xmath311 as a function of galaxy impact parameter can be calculated following @xmath312 at @xmath313 and @xmath314 otherwise .
the core radius @xmath315 is defined to be @xmath316 and does not affect the expected @xmath317 at large @xmath67 .
the extent of mgii absorbing gas scales with the luminosity of the absorbing galaxy according to @xmath318 following the definition of equation ( 2 ) , @xmath319 characterizes the gaseous extent of @xmath320 galaxies .
the gaseous extent of fainter / more luminous galaxies differs from @xmath319 according the scaling relation . to determine the values of @xmath321 , @xmath319 , and @xmath53 that best describe the data , we perform a maximum - likelihood analysis that includes the upper limits in our galaxy and mgii absorber pair sample .
the likelihood function of this analysis is defined as @xmath322 ^ 2 \right\ } \right)\times \\ \nonumber & & \left ( \prod_{i=1}^m \int_{y_i}^{-\infty } dy ' \exp \left\ { -\frac{1}{2 } \left [ \frac{y ' - \bar{y}(\rho_{i } , l_{b_i})}{\sigma_i } \right]^2 \right\ } \right ) , \ ] ] where @xmath323 , @xmath324 , and @xmath325 is the measurement uncertainty of @xmath326 .
the first product of equation ( 3 ) extends over the @xmath327 measurements and the second product extends over the @xmath328 upper limits .
( this definition of the likelihood function is appropriate if the residuals about the mean relationship are normally distributed . ) because @xmath325 may include significant intrinsic scatter ( which presumably arises due to intrinsic variations between individual galaxies ) as well as measurement error , @xmath325 is taken to be the quadratic sum of the cosmic scatter @xmath329 and the measurement error @xmath330 @xmath331 where the intrinsic scatter is defined by @xmath332 \right\}^2 - \sigma_{m_i}^2 \right).\ ] ] because @xmath329 depends on the maximum - likelihood solution @xmath333 , the maximum - likelihood solution is obtained iteratively with respect to equations ( 1 ) and ( 2 ) .
the maximum - likelihood analysis yields @xmath334 and @xmath335 and @xmath336 for the best - fit @xmath337 .
the errors represent the 1-@xmath132 uncertainties of the best - fit parameters .
in addition , we estimate a cosmic scatter of @xmath338 dex . the best - fit isothermal model is presented in figure 4 ( solid curve ) . versus galaxy impact parameter @xmath67 scaled by galaxy @xmath113-band luminosity .
symbols are the same as those in figure 3 .
the solid curve represents the best - fit model based on an isothermal density profile of the mg@xmath28 ions , while the dotted curve shows the expectation from an nfw density profile .
the errorbar in the upper - right corner represents the intrinsic scatter of the best - fit model , in comparison to the observations , as defined in equation ( 5 ) . ]
the results of our maximum likelihood analysis indicate that the extent of mgii gaseous halo scales with galaxy @xmath113-band luminosity according to @xmath339 and @xmath340 , and that typical @xmath320 galaxies possess an extended mgii halo of radius @xmath341 kpc .
it is worth noting that the best - fit @xmath342 kpc characterizes the gaseous extent of @xmath320 galaxies , beyond which no mg@xmath28 ions are found to the limit of @xmath343 . adopting the best - fit scaling relation , the gaseous extent of @xmath344 galaxies is expected to be @xmath345 kpc .
figure 4 shows the relation of @xmath346 vs.luminosity scaled @xmath67 , in which seven galaxies at @xmath347 kpc do not have a corresponding mgii absorption feature to the limit of @xmath93 ang . these non - detections provide the strongest constraints on @xmath319 . to examine possible bias in @xmath319 as a result of the initial sample definition that consider only galaxies at @xmath348 kpc , we repeat the likelihood analysis adopting @xmath349 kpc and find that the best - fit results remain the same . choosing a smaller @xmath59 reduces the sample substantially and
the parameters become poorly constrained .
furthermore , we note that the results are also insensitive to the choice of model gaseous profile .
we obtain similar results , adopting a navarro - frenk - white ( nfw ) profile commonly used for describing the profiles of dark matter halos ( navarro @xmath12 1997 ) .
the best - fit nfw profile is also presented in figure 4 ( dotted curve ) .
comparisons of observations and different model profiles show that an isothermal model profile of the underlying gaseous halo provides a good fit to the observed @xmath9 vs. @xmath67 relation .
in summary , figure 4 demonstrates that accounting for the scaling relation of gaseous extent with galaxy luminosity , the scatter in the @xmath9 versus @xmath67 relation is substantially reduced .
furthermore , the @xmath9 versus @xmath67 relation scaled by galaxy intrinsic luminosity allows us to estimate a mean covering factor of mg@xmath28 ions average across galaxies of a wide range of luminosity .
we find @xmath7 % within @xmath8 , including the uncertainty in @xmath319 from equation ( 7 ) .
using the sample of 13 galaxy mgii absorber pairs and 10 galaxies at @xmath52 kpc that do not give rise to mgii absorption to a sensitive upper limit , we find that galaxies at larger impact parameters produce on average weaker mgii absorbers .
this anti - correlation is substantially improved , when accounting for the intrinsic luminosity of individual galaxies .
in addition , there exists a distinct boundary at @xmath3 , beyond which no mgii absorbers with @xmath350 are found .
a maximum likelihood analysis shows that the observations are best described by an isothermal density profile of the gas and a scaling relation @xmath351 kpc with a mean covering factor of @xmath7 % at @xmath313 .
the best - fit profile applies to galaxies of @xmath309 to @xmath310 and therefore predominantly cold - mode halos . for higher mass halos ,
we expect from tc08 that the corresponding absorption strength for a given @xmath67 is reduced , namely a smaller @xmath321 in equation ( 1 ) . in this section
, we discuss the implications drawn from the results of our analysis . adopting the best - fit scaling relation and the gaseous profile described in equation ( 1 ) , we derive a corresponding halo size of @xmath352 physical kpc for galaxies of @xmath353 at @xmath6 .
this halo size is consistent with the finding of lanzetta & bowen ( 1990 ) and the expectation inferred from known mgii frequency function ( nestor @xmath12 2005 ; see the descriptions in
5.2 ) , but is larger than what was reported in steidel ( 1995 ) .
steidel ( 1995 ) reported a characteristic size of mgii halo of @xmath354 kpc for @xmath320 ( corresponding to @xmath355 for the cosmology adopted in this paper ) galaxies , but it is not clear at what @xmath9 limit the halo size was estimated . at @xmath356 kpc , we expect @xmath357 from the best - fit model .
the high gas covering factor derived from our analysis is consistent with what is inferred from the sample of kacprzak @xmath12 ( 2008 ) . applying the best - fit luminosity scaling relation of gaseous extent , we find that four of the 28 galaxies within the expected gaseous extent for @xmath358 absorbers in the kacprzak @xmath12sample have a corresponding mgii weaker than the 0.3 threshold .
the inferred covering fraction is @xmath359% . at the same time , this high covering fraction disagrees with the value reported in tripp & bowen ( 2005 ) , the only other on - going survey program designed to constrain gas covering fraction by searching for corresponding mgii absorbers at the locations of known _ field _ galaxies ( 1995 ) that include a dominant fraction of galaxies in dense cluster environment .
both studies reported a substantially lower covering fraction , @xmath360 % for @xmath361 absorbers .
the low covering fraction of strong absorbers is understood , because galaxies clusters reside in massive dark matter halos of @xmath362 and the fraction of `` cold '' gas in these massive halos is expected to be reduced ( see tc08 ) . ] .
the authors obtained follow - up spectra of background qsos that are located within @xmath363 physical kpc of 20 known galaxies at @xmath364 .
they found that 10 of the 20 galaxies do not produce mgii absorbers to within a 2-@xmath132 upper limit of @xmath365 , and concluded that the incidence of mg@xmath28 ions is @xmath366 % at @xmath367 kpc in individual galactic halos .
because more luminous galaxies have larger extended mgii halos according to the scaling relation presented in equations ( 2 ) , ( 6 ) , and ( 7 ) , this discrepancy can be understood if many of the galaxies in tripp & bowen are located at impact parameters that are larger than their expected @xmath8 .
recall that we obtained a similar estimate of @xmath306 % at @xmath368 kpc ( or @xmath308 % at @xmath369 kpc ) , _ prior to _ accounting for the scaling relation between @xmath8 and @xmath370 in individual galaxies ( 4 ) .
in addition , the sample in tripp & bowen includes galaxies of luminosity ranging from @xmath371 to @xmath372 .
those galaxies that are more luminous than @xmath320 are also expected to reside in more massive halos . at @xmath373 , @xmath372 galaxies
are expected to reside in @xmath374 halos ( tinker @xmath12 2007 ) .
if the host dark matter halos are substantially more massive than @xmath375 , then we expect from the halo occupation analysis in tc08 that they do not contribute significantly to strong mgii absorbers . in summary ,
further inspections of the properties of these galaxies and estimates of the gas covering fraction @xmath50 as a function of radius and @xmath9 are crucial for resolving this discrepancy .
the scaling relation and the constraint of @xmath376 together also allow us to test whether the mgii absorbers trace a significant and representative portion of the galaxy population or merely a special class such as those galaxies undergoing a starburst episode .
if extended mgii halos are a generic feature of field galaxies at all epochs , then we can derive the expected number density of mgii absorbers by combining known space density of galaxies ( versus luminosity ) and the product of the cross section and covering factor of the mg@xmath28 ions .
an agreement with observational results from surveys of mgii absorbers would challenge the scenario that attributes a large fraction of these absorbers to starburst systems .
the predicted number density of mgii absorption systems arising in the extended gaseous halos of galaxies may be evaluated according to @xmath377 where @xmath378 is the speed of light , @xmath379 is the galaxy luminosity function , @xmath132 is the gas cross section for producing mgii absorbers of @xmath380 that scales with galaxy @xmath113-band luminosity , and @xmath50 is the halo covering factor . substituting the scaling relationship according to equations ( 2 ) , ( 6 ) , and ( 7 ) , we find @xmath381 ^ 2\,(1 + z)^2}{\sqrt{\omega_{\rm m}\,(1+z)^3 + \omega_{\rm \lambda } } } \\ & & \times\int_0^\infty d\left(\frac{l_b}{l_{b_*}}\right)\,\left(\frac{l_b}{l_{b_*}}\right)^{2\,\beta}\,\phi(l_b , z)\ , , \nonumber\end{aligned}\ ] ] where @xmath382 is the extent of mgii halos that corresponds to @xmath383 . to evaluate equation ( 9 ) , we adopt the luminosity function of the blue galaxy population at @xmath384 from faber @xmath12(2007 ) , which is characterized by @xmath385 , @xmath386 mpc@xmath387 , and a faint - end slope @xmath388 .
including galaxies of @xmath389 and our best - fit @xmath319 and @xmath376 , equation ( 9 ) yields @xmath390 for @xmath391 absorbers , @xmath392 for @xmath393 absorbers , and @xmath394 for @xmath395 absorbers at @xmath60 .
the predictions based on galaxies more luminous than @xmath396 agree well with the observed @xmath397 from nestor @xmath12 ( 2005 ) , lending strong support for the hypothesis that the mgii absorbers trace a significant and representative portion , rather than a sub - class , of the galaxy population .
the model profile of equation ( 1 ) has been proved to be a good representation of the galaxy and absorber pairs at @xmath398 kpc , when the scaling relation of equations ( 2 ) , ( 6 ) , and ( 7 ) is accounted for .
an important test is to examine whether the model remains a good fit at smaller @xmath67 and strong absorbers . identifying galaxy absorber pairs at small impact parameters
is difficult , because the glare of the background qso often prohibits identification of faint galaxies at angular distances @xmath399 that corresponds to @xmath400 physical kpc at @xmath401 .
recent studies of intervening absorption - line systems along the lines of sight toward the optical afterglows of distant @xmath402-ray bursts ( grb ) have uncovered a number of strong mgii absorbers ( e.g.prochter @xmath12 2006b ) . unlike qsos ,
grb afterglows disappear after a while , permitting exhaustive searches of faint galaxies at @xmath399 that give rise to strong mgii absorbers along the lines of sight
. a redshift survey of galaxies along the sightline toward grb060418 ( @xmath403 ) has uncovered the absorbing galaxies for three strong mgii absorbers at @xmath404 ( pollack @xmath122008 , in preparation ) , similar to the redshift range covered by our sample .
the rest - frame @xmath113-band magnitudes of the galaxies range from @xmath405 to @xmath406 ; impact parameters of the galaxies range from @xmath407 kpc to @xmath408 kpc . adopting the best - fit scaling relation ,
we include the three galaxy mgii absorber pairs in the @xmath9 versus @xmath67 plot for comparison ( star points in figure 5 ) .
the good agreement between the best - fit model and the absorbers found along grb sightlines provides further support for our model at small impact parameters . versus @xmath67 correlation scaled by galaxy @xmath113-band luminosity presented in figure 4 , but including pairs from pollack @xmath12 ( 2008 ; star points ) and bouch @xmath12 ( 2007 ; points with horizontal errorbars ) for comparisons .
the solid curves represent the best - fit model based on an isothermal density profile , with @xmath409 ( top curve ) and @xmath316 ( bottom curve ) .
the pairs of pollack @xmath12 are for galaxies and mgii absorbers identified at @xmath410 along the sightline toward grb060418 .
the errorbars in these data points are smaller than the size of the symbols .
the pairs of bouch @xmath12 are galaxies found at the positions of known mgii absorbers based on the presence of h@xmath411 emission lines .
errorbars represent the uncertainties of the inferred @xmath412 from the @xmath412 versus @xmath413 correlation published in tresse @xmath12 ( 2002 ) .
crosses represent the absorbers in the bouch sample that are likely ( @xmath414 % probability ) to be a damped absorber of @xmath415 based on the presence of strong feii and mgi transitions ( rao @xmath122006 ) . ]
[ [ are - strong - w27961- - absorbers - at - zsim-1-produced - primarily - in - starburst - outflows ] ] are strong [ @xmath416 ] absorbers at @xmath417 produced primarily in starburst outflows ?
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ as described in 1
, the large line width of @xmath416 mgii absorbers has recently been interpreted as due to outflow motions in starburst driven winds .
this scenario was initially motivated by the observed inverse correlation between @xmath9 and the clustering amplitude of mgii absorbers at @xmath418 ( bouch @xmath12 2006 ) . in tc08 , we have shown that this inverse correlation can be understood as due to an elevated clustering amplitude of @xmath35 absorbers from contributions of survived cold clouds in massive hot halos , rather than a suppressed clustering strength of @xmath36 absorbers .
in addition , we have shown in
4 , 5.1 , and 5.2 that extended mgii halos are a generic feature of field galaxies over a wide luminosity range .
the predicted number density of the absorbers from adopting the scaling relation , equations ( 2 ) , ( 6 ) , and ( 7 ) , and the galaxy luminosity function agrees well with observations .
specifically , we expect @xmath419 for @xmath391 absorbers , in comparison to the observed @xmath420 for @xmath421 absorbers at @xmath422 from nestor @xmath12 ( 2005 ) . while there is no clear indicator one can apply to unambiguously distinguish between gas inflows and outflows in the distant universe , we discuss several caveats related to the starburst outflow scenario .
if the absorbers are produced in starburst driven outflows , then a large fraction of the absorbing galaxy population might be expected to exhibit disturbed morphology ( mobasher @xmath12 2004 ) .
the mgii absorbing galaxies in our sample exhibit regular disk morphologies in the hst images shown in figure 2 .
in particular , the strong absorber found with @xmath423 at @xmath120 toward 3c336 is associated with an edge - on disk galaxy at @xmath424 kpc .
the three mgii absorbers of @xmath416 found toward grb060418 also exhibit regular disk structures in high - resolution images obtained both in space using hst and on the ground using lgao on the keck telescopes ( pollack @xmath12 2008 , in preparation ) .
kacprzak @xmath12(2007 ) also showed that there is little correlation between @xmath346 and galaxy asymmetry for mgii absorbers of @xmath425 .
while some strong mgii absorbers have been presented to show evidence of arising in expanding superbubble shells ( bond @xmath12 2001 ) , available high - resolution images of other strong mgii absorbers do not support outflow being principally responsible for the observed strong absorbers .
this conclusion is also consistent with the spectral properties of mgii - selected galaxies published by guillemin & bergeron ( 1997 ) on the other hand , bouch @xmath12 ( 2007 ) presented new observations from a targeted search for corresponding h@xmath411 emission in the vicinity of known mgii absorbers .
these authors successfully identified h@xmath411 emission near 14 of 21 mgii absorbers with @xmath426 at @xmath417 .
the integrated raw in our analysis , because the dust content is expected to vary substantially in @xmath417 galaxies ( e.g. tresse @xmath12 2002 ) and is unknown for galaxies in the bouch @xmath12 sample .
bouch @xmath12 ( 2007 ) applied a constant correction to all observed flux in their calculation , resulting in a simple scaling of the observed values .
the scatter in their measurements induced by dust remains the same . for consistency ,
we consider in the following discussion only observed values without extinction correction both for the absorbing galaxies and for comparison results , such as the luminosity function and the @xmath427 correlation , from the literature . ]
luminosities range from @xmath428 erg s@xmath429 to @xmath430 erg s@xmath429 .
the impact parameters range from @xmath431 kpc to @xmath356 kpc . adopting a mean extinction correction @xmath432 mag
, the authors inferred a mean star formation rate of @xmath433 m@xmath434 yr@xmath429 for these galaxies .
the luminosity function at @xmath417 has been studied by a number of groups ( see doherty @xmath12 2006 for a list of references ) .
it is characterized by a schechter function of @xmath435 erg s@xmath429 , @xmath436 , and @xmath388 . the galaxies detected in by bouch @xmath12correspond to @xmath437 at @xmath417 , and therefore have only modest @xmath413 . integrating the luminosity function , we obtain a mean space density of 0.034 @xmath438 mpc@xmath387 for galaxies of brighter than @xmath439 .
this is already 50% of all galaxies more luminous than @xmath440 at @xmath417 , according to the best - fit luminosity function of the blue galaxy population from faber @xmath12 ( 2007 ) .
this exercise shows that emission is commonly seen in @xmath417 galaxies and the presence of modest @xmath413 does not argue for the presence of starburst outflows , unless the outflow material is long - lived . for comparison
, we include the 14 galaxy and absorber pairs from bouch @xmath12 ( 2007 ) in the @xmath9 versus @xmath67 plot in figure 5 , applying the same scaling relation described in equations ( 2 ) , ( 6 ) , and ( 7 ) .
we estimate @xmath412 of these galaxies based on the published @xmath427 correlation in tresse @xmath12 ( 2002 ) .
we adopt the observed @xmath413 and the scaling relation @xmath441 that best describes the data in tresse @xmath12 ( 2002 ) . at small impact parameters , the model predictions depend sensitively on the choice of @xmath315 in equation ( 1 ) .
in addition to the fiducial model , we also include in the figure a model with @xmath409 .
nine of the 14 mgii absorbers are considered likely ( @xmath442 % ) damped absorbers ( dlas ) of @xmath415 based on the presence of strong feii and mgi transitions ( rao @xmath12 2006 ) .
dlas are the high - redshift analog of neutral gas regions that resembles the disks of nearby luminous galaxies ( e.g.wolfe @xmath12 2005 ) .
we therefore expect that the impact parameter distribution of galaxy - dla pairs represents the typical extent of hi disks at high redshifts .
we highlight the potential dlas in the bouch @xmath12 sample as crosses in figure 5 . after applying the best - fit scaling relation for mgii absorbers , we find that the impact parameters of these galaxy and potential dla pairs range from @xmath443 kpc to as high as @xmath444 kpc .
the three largest separation pairs have @xmath445 kpc , where @xmath446 is the luminosity scaled projected distance .
the scaling of impact parameter is justified , because smaller / fainter galaxies have on average smaller disks ( e.g. cayatte @xmath12 1994 ) .
the large impact parameters , together with the likely high @xmath18 , suggest that the galaxies responsible for these absorbers ( presumably at smaller angular distances from the qsos ) may still be missing . for the remaining sample ,
our model is in reasonably good agreement with their data . in summary
, we find a lack of empirical measurements that argue for starburst outflows to be a dominant mechanism for producing the observed strong mgii absorbers at @xmath417 . on the other hand , both the observed number density of mgii absorbers and the @xmath9 versus @xmath67 correlation can be explained by extended mgii halos around typical field galaxies .
the fraction of mgii absorbers originating in starburst outflows is therefore expected to be small at @xmath417 , unless starburst outflows are a common feature in field galaxies .
we caution , however , that the same conclusion can not be automatically applied to absorbers at @xmath447 before a similar analysis is performed .
the sharp decline of @xmath448 found at @xmath449 kpc in figure 4 clearly indicates that mg@xmath28 ions have a finite extent in the extended gaseous halos around galaxies .
this is reminiscent of the sharp boundary between the civ absorbing and non - absorbing regions around @xmath450 galaxies reported in chen @xmath12(2001b ) .
given the nature of a photo - ionized gas ( bergeron & stasnska 1986 ; hamann 1997 ) , we examine whether the finite extent of mg@xmath28 ions can be understood as due to photo - ionization of the absorbing clouds at large radii . in a two - phase halo model ( e.g. mo & miralda - escud 1996 ) ,
cold clouds that are responsible for producing the absorption features are pressure confined in a hot medium . as the total gas density declines toward larger radii , the density of the cold clouds also decreases and the clouds become optically thin to the ultraviolet photons that can quickly ionize the mg@xmath28 ions ( [email protected] ev ) . with a known background radiation field , the location where the ionization transition occurs may be adopted to constrain the density of cold gas at large radii .
we carry out the photo - ionization analysis , using the cloudy software ( ferland @xmath12 1998 ; version 06.02 ) , and calculate the ionization fractions of mg@xmath28 ions for a grid of models with different total gas density and metallicity .
we adopt the spectral shape of the uv background radiation field of haardt & madau in cloudy for the incident ionizing flux , and scale the radiation intensity to the background uv radiation field inferred from qso proximity effect ( scott @xmath12 2002 ) .
we adopt a plane parallel geometry for the clouds with a thickness of 100 pc , which is motivated by the observed coherent length of mgii absorbers toward lensed qsos ( rauch @xmath12 2002 ) .
the slab of gas is illuminated by the ionization radiation field on both sides .
ions and h@xmath452 atoms versus total gas density @xmath453 for photo - ionized gas .
the values are calculated using the cloudy software ( ferland @xmath12 1998 ; version 06.02 ) for gaseous clouds of plane parallel geometry and 100 pc thickness .
the input ionizing radiation intensity is set to be @xmath454 erg s@xmath429 hz@xmath429 @xmath455 sr@xmath429 suitable for the @xmath456 universe ( scott @xmath12 2002 ) .
we perform the calculations , assuming solar ( dotted curve for mg and dot - dashed curve for h ) and @xmath457 solar metallicity ( solid curve for mg and dashed curve for h ) . ]
figure 6 shows the expected fractions of mg@xmath28 ions ( solid and dotted curves ) as a function of the total gas density @xmath453 .
we include the calculated fraction of h@xmath452 atoms ( dashed and dash - dotted curves ) for comparison .
the solid and dashed curves are for gas of @xmath457 solar metallicity , while the dotted and dash - dotted curves are for gas of solar metallicity .
our photo - ionization calculations indicate that indeed the fraction of mg@xmath28 declines linearly at @xmath458 for a mean radiation intensity of @xmath454 erg s@xmath429 hz@xmath429 @xmath455 sr@xmath429 ( which is appropriate for the observed background radiation field at @xmath456 scott @xmath12 2002 ) .
however , the transition is not as sharp as what is seen for hydrogen atoms . including the estimated ionization fraction from the cloudy models and adopting the density profile parameterized in equation ( 1 ) , we derive the expected correlation between @xmath346 and @xmath67 and find that photo - ionization is insufficient for explaining the presence of the sharp decline in the observed mgii absorption strength .
this extent of mgii gaseous halos ( @xmath459 kpc around typical @xmath320 galaxies ) is remarkably similar to the extent of civ gaseous halo ( @xmath460 kpc around typical @xmath320 galaxies , after correcting for the @xmath54-cosmology ) reported by chen @xmath12(2001b ) . on the basis of 14 galaxy and civ absorber pairs and 36 galaxies that do not produce corresponding civ absorption lines to within sensitive upper limits
, these authors find that the extent of civ - absorbing gas around galaxies scales with galaxy @xmath113-band luminosity as @xmath461 kpc ( corrected for the @xmath54 cosmology ) .
in addition , there exists a sharp boundary between civ absorbing and non - absorbing regions at @xmath462 .
but the scatter in the observed civ absorption strengths at @xmath463 between different galaxies appears to be significantly larger than what is seen in figure 4 , and there is a lack of anti - correlation between the observed cvi absorbing strength and galaxy impact parameter .
given the distinct ionization potentials of the c@xmath464 and mg@xmath28 ions , _ the location where the abrupt transition between absorbing and non - absorbing regions occurs can be understood as a critical radius below which cool clouds can form and stablize in an otherwise hot halo_. this two - phase model to interpret qso absorption line systems was formulated in mo & miralda - escud ( 1996 ) and later re - visited by maller & bullock ( 2004 ) .
taking into account the appropriate photo - ionization condition of the cold clouds , mo & miralda - escud presented the expected mgii and civ absorption strength versus radius for halos of different mass .
their models showed a lack of radial dependence on the civ absorption strength and a tight correlation between mgii @xmath346 and @xmath67 for halos of a wide range of mass .
the differences in the radial dependence of different ions is understood as being due to the photo - ionization condition of cold clouds in the uv background radiation field .
these model expectations agree with observational findings very well .
the agreement between observations and a simple two - phase model is encouraging , although it is not clear how the two - phase model applies to cold - mode halos ( e.g. dekel & birnboim 2006 ) . measurements of the relative abundances between c@xmath464 and mg@xmath28 as a function of radius will provide additional support for the origin of these metal - line absorbers , further constraining the metallicity and density of halo gas around galaxies .
our initial halo occupation analysis ( tc08 ) included two principal assumptions for the extent of mg@xmath28 .
first , we assumed that the extent of the cold gaseous halo is @xmath465 physical kpc for @xmath466 halos , which is roughly @xmath467 of the halo radius @xmath468 at @xmath469 .
recall that the standard definition of a dark matter halo is an object with a mean interior density of 200 times the background density .
second , the gaseous extent scaled with the halo mass as @xmath470 and @xmath471 , following the expectation of the scaling relation between halo mass and the corresponding virial radius .
the results presented in 4 offer a direct test of these assumptions . to compare the empirical constraints of equations ( 6 ) & ( 7 ) with our initial model assumptions , we first derive the corresponding dark matter halo mass @xmath30 and their halo radius @xmath468 for galaxies of known absolute @xmath113-band magnitudes .
the adopted @xmath472 for the analysis in
4 corresponds to a magnitude of @xmath473 in the bandpass of the two - degree field galaxy redshift survey ( 2dfgrs , colless @xmath12 2001 ) . the 2dfgrs probes galaxies at @xmath474 .
excluding galaxies that exist as satellites in a cluster environment % for @xmath320 galaxies ( tinker @xmath12 2007 ; zcz07 ) .
in addition , none of the galaxies in our sample are observed to be part of a larger group or cluster .
here we carry out our calculations , assuming that most of galaxies in our sample are central galaxies in their dark matter halos . ] , the mean halo mass for @xmath472 galaxies is @xmath475 at @xmath474 ( tinker @xmath12 2007 ; van den bosch @xmath12 2007 ) . at @xmath417 , galaxies of this magnitude on average reside in @xmath476 halos ( zcz07 ) .
interpolating in @xmath477 , we find that galaxies of @xmath478 are expected to reside in halos of @xmath479 at @xmath480 and that the halo radius for @xmath472 galaxies is @xmath481 physical kpc
. the best - fit @xmath482 from equation ( 7 ) therefore implies @xmath483 .
next , we derive the expected scaling relation between @xmath8 and halo mass @xmath30
. the halo occupation analysis of 2dfgrs galaxies from tinker @xmath12 ( 2007 ) showed that the relationship between @xmath30 and luminosity for galaxies of @xmath484 at @xmath474 is @xmath485 .
this monotonic relationship is appropriate for galaxies that reside at the _ centers _ of their dark matter halos and are the brightest galaxy in the halo .
a similar scaling relation was obtained from deep2 data by zcz07 for galaxies at @xmath417 . because there is a monotonic relationship between halo mass and galaxy luminosity ,
the scaling relation of @xmath8 with @xmath370 in equation ( 2 ) is equivalent to a scaling relation between @xmath30 and @xmath468 that are related by @xmath486 . adopting @xmath487 from equation ( 6 )
, we derive @xmath488^{0.3}$ ] @xmath489 kpc at @xmath490 .
the calculations above show that observations support the initial assumptions in tc08 that the gaseous radius is a constant fraction of the halo radius for all halos and that the gaseous radius scales with halo mass according to @xmath470 with @xmath491 .
we note that our galaxy sample therefore probes halo mass range from @xmath492 to @xmath493 , and that the best - fit scaling relation is driven by galaxies in the cold - hot transition regime .
next , we compare the observed covering fraction of mgii absorbers with the predicted mean of tc08 . as summarized in 1 , we find in the halo occupation analysis of tc08 that dark matter halos of @xmath38 on average have a covering fraction of unity @xmath39 at @xmath494 for mg@xmath28 and @xmath41 at larger radii .
this halo mass range corresponds to a luminosity range of @xmath495 at @xmath496 , following the mass - to - light scaling relation discussed above .
sixteen galaxies in our sample have @xmath497 , eleven of which are at @xmath498 .
we identify a corresponding mgii absorber for every one of these galaxies , indicating a mean covering fraction of @xmath499 % at @xmath500 .
this empirical result agrees with our model expectation . to assess the significance of detecting a mgii absorber in every galaxy in the sample of eleven , we adopt the binomial distribution function and find that we can constrain the underlying mean covering factor at @xmath501 with greater than 95% confidence level .
three of the seven remaining @xmath502 galaxies in our sample have @xmath503 .
two of them do not have corresponding mgii absorbers identified to a sensitive upper limit , implying @xmath504 % at @xmath500 for fainter galaxies ( presumably lower - mass halos ) .
this is also consistent with the expectation of tc08 , where we found a quickly declining @xmath376 for halos of @xmath505 , although the observations have a much lower statistical significance .
in addition , we note that there exists an intrinsic scatter in the fiducial scaling relation between @xmath30 and @xmath370 .
this intrinsic scatter is expected to smear out the exact edge of gaseous halos , particularly when considering an ensemble of objects with a wide range of luminosity . in summary ,
the initial results from the on - going survey of mgii absorbers in the vicinity of known galaxies close to qso lines of sight have allowed us to study in detail the properties of extended gas around galaxies .
the best - fit scaling relation between the extent of mg@xmath28 and galaxy luminosity and the observed mean gas covering fraction together provide strong empirical evidence to support the halo occupation model of tc08 . to improve the statistical significances of various empirical parameters and to obtain a better understanding of the incidence of cold gas in dark matter halos , such as @xmath50 vs. @xmath30 and @xmath50 vs. @xmath67 ,
a larger sample of galaxy and mgii absorber pairs established for a wide range of galaxy luminosity and morphology is necessary . in particular ,
only three galaxies in our current sample would be considered in a `` cold - mode '' halo based on their luminosity and none shows a corresponding mgii absorber .
deep surveys targeting at fainter dwarfs are necessary to probe these `` cold - mode '' halos . finally , we encourage observations of mgii absorbers along the lines of sight toward close qso pairs for obtaining a direct measure of the incidence of mg@xmath28 in individual halos .
we thank j .-
gauthier , j. omeara , m. rauch and g. becker for assistance on obtaining part of the mike spectra presented in this paper .
we thank n. gnedin and m. rauch for helpful comments on an early version of the paper .
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 .
w.c . acknowledges partial support from nasa long term space astrophysics grant nng06gc36 g and an nsf grant ast-0607510 . , c. c. 1995
, in qso absorption lines , proceedings of the eso workshop held at garching , germany , 21 - 24 november 1994 , edited by georges meylan .
springer - verlag berlin heidelberg new york .
also eso astrophysics symposia , 1995 . , p.139 , ed .
g. meylan , 139 | we study the extent and covering fraction of cool baryons around galaxies of different luminosity and mass , based on a survey of mgii @xmath0 absorption features near known galaxies .
the initial sample consists of 13 galaxy and absorber pairs and 10 galaxies that do not produce mgii absorption lines to within sensitive upper limits .
the redshifts of the galaxy and absorber pairs range from @xmath1 to 0.892 with a median of @xmath2 .
we find that galaxies at larger impact parameters produce on average weaker mgii absorbers .
this anti - correlation is substantially improved when accounting for the intrinsic luminosities of individual galaxies . in addition
, there exists a distinct boundary at @xmath3 , beyond which no mgii absorbers are found .
a maximum likelihood analysis shows that the observations are best described by an isothermal density profile and a scaling relation @xmath4 kpc ( or @xmath5 kpc at @xmath6 ) with a mean covering factor of @xmath7 % .
together with the scaling relation between halo mass and galaxy luminosity inferred from halo occupation studies , this scaling of @xmath8 indicates that gas radius is a fixed fraction of the dark matter halo radius .
we compare our results with previous studies and discuss the implications of our analysis for constraining the baryon content of galactic halos and for discriminating between competing scenarios for understanding the nature of the extended gas . |
the term _ sub - riemannian manifold _ means the triple @xmath2 , where @xmath3 is an @xmath4-dimensional manifold , @xmath5 is a smoothly varying @xmath6-dimensional distribution inside the tangent bundle @xmath7 of the manifold @xmath3 with @xmath8 , and @xmath9 is a riemannian metric defined on @xmath5 , i. e. , a positively definite quadratic form .
recently the study of geometric structures , where the riemannian metric @xmath9 on @xmath5 is substituted by a semi - riemannian metric @xmath10 , that is a non - degenerate indefinite metric , started e. g. , in @xcite .
there is no special attribution so far for such kind of manifolds @xmath11 , thus we propose to call them _ sub - semi - riemannian _ manifolds or shortly _ ssr - manifolds_. in the particular case , when the metric @xmath10 has index 1 , an ssr - manifold receives the name _ sub - lorentzian _ manifold by the analogy to lorentzian manifold . in the present article we study an example of @xmath5-type group furnished with the sub - lorentzian metric .
this is an interesting example not only as an almost unique known example of sub - lorentzian manifold but also because it has a precise physical meaning . in the article
we reveal the connection between sub - lorentzian geometry and physics of relativistic electrodynamics basing on the example of @xmath12-type quaternion group equipped with the lorentzian metric .
we also compare characterising features of sub - riemannian and sub - lorentzian geometries . the notion of @xmath5-type groups
was introduced in @xcite .
it is known that riemannian manifolds have applications in classical mechanics .
sub - riemannian manifolds of step 2 ( heisenberg manifolds ) play important role in quantum mechanics .
sub - riemannian geodesics even localy behave very differently from the ones in riemannian geometry , where the energy minimising motion is described by a unique geodesic .
a sub - atomic particle behaves in a way similar to an electron which moves only along a given set of directions .
there can be infinitely many geodesics with different length joining two points .
in its turn , the sub - lorentzian structure underlise the motion in an electro - magnetic field .
just like space and time emerge in special relativity , the electric and magnetic fields can not be considered separately .
the sub - lorentzian structure absorbes both phenomena , the presence of the electro - magnetic field and the the space - time geometry . that is why the @xmath1-type quaternion group with a sub - lorentzian metric is an interesting example to work with . in the introduction in order to explain the main idea we would like to mention the heisenberg group as the simplest noncommutative example of @xmath13-type ( heisenberg type ) groups and its numerous connections with physics .
the heisenberg group is the manifold @xmath14 with the noncommutative group law @xmath15 left - invariant vector fields @xmath16 , @xmath17 are obtained from the group law and span a 2-dimensional distribution @xmath5 which is called horizontal .
the horizontal distribution can be also defined as the kernel of the contact one - form @xmath18 in @xmath14 .
the differential of @xmath19 is the 2-form @xmath20 that satisfies the maxwell s equation @xmath21 for the magnetic field @xmath22 in @xmath23 .
let us define a riemannian metric @xmath24 on @xmath5 .
then the sub - riemannian manifold @xmath25 is also called the heisenberg group .
it turned out that the geodesic equation for geodesics @xmath26 satisfying the non - holonomic constraints @xmath27 coincides up to a constant with the lorentz equation of motion of the charged particle in the magnetic field @xmath22 .
if we change riemannian metric on @xmath5 on the lorentzian one @xmath28 we come to the notion of sub - lorentzian heisenberg group @xmath29 . in this case
the geodesic equation for non - holonomic geodesics coincides with analogue of the lorentz equation for the motion of a charged particle in the electromagnetic field defined by @xmath22 and the lorentzian metric tensor .
the geodesics , metrics properties , and other related questions on @xmath30 were studied in @xcite .
the lack of the dimension of the horizontal distribution on the heisenberg group does not allow to reveal the peculiarity of the applications in the case of the magnetic and electromagnetic fields .
therefore , we chose the analogue of the heisenberg group admitting a 4-dimensional distribution , that we called quaternionic @xmath13-type group .
this example allows also to show similarities and differences between the sub - riemannian and sub - lorentzian geometries .
the article is organized in the following way . in section [ sec:2 ]
we introduce the @xmath13-type quaternion groups endowed with different metrics : riemannian and lorentzian .
we also present the differential equations for the geodesics in both cases .
section [ sec:3 ] is the collection of the definitions related to the motion of charged particles in electro - magnetic fields .
we give the explanation of the geometrical results from the physical point of view .
section [ sec:4 ] is devoted to the solution of geodesic equations , where we fined the explicit formulas for the horizontal and vertical parts of geodesics .
section [ sec:5 ] is dedicated to study of reachable sets by geodesics of different causal types and estimation of the cardinality of geodesics , connecting two different points .
section [ sec:6 ] shows a brief overview of reachable sets for @xmath30 for the sake of comparison with obtained results for the quaternion @xmath1-type group .
we remind that quaternions form a noncommutative division algebra that extends the system of complex numbers .
it is convenient to define any quaternion @xmath33 in the algebraic form @xmath34 , where @xmath35 .
the scalar @xmath36 is called the real part @xmath37 and a vector @xmath38 recieved the name pure imaginary quaternion and is denoted by @xmath39 .
thus @xmath40 . with this notations
it is easy to introduce the noncommutative multiplication @xmath41 between two quaternions @xmath42 and @xmath43 using the usual inner product @xmath44 and vector product @xmath45 in @xmath23 .
namely , @xmath46 notice that this structure suggests itself the analogy with a lorentzian geometry where @xmath47 consists of a time part @xmath48 and space part @xmath49 .
the conjugate quaternion @xmath50 to @xmath33 is @xmath51 .
it is known that quaternion @xmath34 can also be represented in the @xmath52-matrix form @xmath53 where @xmath54 are the basis of quaternion numbers in the representation given by real @xmath55-matrices .
we introduce an @xmath13-type group whose noncommutative multiplication law makes use of quaternion multiplication rule .
let us take the background manifold @xmath3 as @xmath56 and define the noncommutative law by @xmath57 for @xmath58 and @xmath59 from @xmath60 . here
@xmath61 is the imaginary part of the product @xmath62 defined in of the conjugate quaternion @xmath63 to @xmath64 by another quaternion @xmath65 .
the introduced multiplication law makes @xmath66 into a noncommutative lie group with the unity @xmath67 and the inverse element @xmath68 to @xmath58 .
the group law defines the left translation @xmath69 .
let @xmath70 be a standard basis of the tangent space @xmath71 to @xmath3 at @xmath72 .
the basic left - invariant vector fields can be obtained by the action of the tangent map @xmath73 of @xmath74 to the standard basis as @xmath75 , @xmath76 .
then the vector fields @xmath77 span a @xmath78-dimensional distribution @xmath79 , which we call _ horizontal_. the left - invariant vector fields @xmath80 , @xmath81 form a basis of the complement @xmath82 to @xmath79 in the @xmath83 . at each point
@xmath84 the distribution @xmath85 is a copy of @xmath47 .
the commutation relations are as follows @xmath86=-z_1,\quad [ x_0,x_2]=z_3,\quad [ x_0,x_3]=z_2,\notag\\ [ x_1,x_2]=z_2,\quad [ x_1,x_3]=-z_3,\quad [ x_2,x_3]=-z_1.\notag\end{gathered}\ ] ] therefore , @xmath87 and their commutators span the entire tangent space @xmath83 .
this property of the distribution @xmath79 is called bracket - generating of step 2 . the lie algebra with the basis @xmath88 is nilpotent of step 2
. the horizontal distribution @xmath79 can be defined by making use of one - forms .
namely , the one - forms @xmath89 annihilate the distribution @xmath79 .
here @xmath90 , @xmath91 , and @xmath92 is the transposed vector to @xmath93 .
thus , @xmath79 is the common kernel of forms @xmath94 , @xmath95 .
let us consider the external differential of the linear combination @xmath96 .
we get the two - form that is defined in 4-dimensional space @xmath97 where @xmath98 any vector @xmath99 , @xmath72 , is called horizontal , a vector field @xmath100 tangent to @xmath101 at each point @xmath102 is also called horizontal .
an absolutely continuous curve @xmath103\to m$ ] that has its velocity vector @xmath104 tangent to @xmath105 almost everywhere is called horizontal curve .
let us define the riemannian metric @xmath106 on the distribution @xmath79 in such a way that @xmath107 , where @xmath108 is the kronecker symbol . with this
we get the sub - riemannian manifold @xmath109 with the sub - riemannian structure @xmath110 . geodesics ( or normal extremals ) in the sub - riemannian geometry are defined as a projection of solutions of the hamiltonian equations on the underlying manifold .
let us present the corresponding hamiltonian system .
we denote by @xmath111 , @xmath112 the tangent and cotangent space at @xmath113 respectively and by @xmath114 , @xmath115 the corresponding tangent and cotangent bundle .
thus , if @xmath116 then the restriction of @xmath117 on the subspace @xmath118 of @xmath111 is well defined and making use of the inner product we define a hamiltonian function on @xmath112 by @xmath119 where by @xmath120 we denoted the pairing between vector spaces @xmath111 and @xmath112 .
this definition coincides with the definition of the norm of the linear functional @xmath117 over the vector space @xmath118 .
if we write @xmath121 and @xmath122 then the hamiltonian function can be rewritten in the following form @xmath123 we get the corresponding hamiltonian system @xmath124 here @xmath125 , @xmath126 . after the simplification , we obtain that @xmath127 are constant and @xmath128 @xmath129 the solution of these equations and detailed calculations can be found in @xcite .
let us change the positively definite metric @xmath106 on @xmath79 on the lorentz metric ( that is nondegenerate metric of index @xmath130 ) @xmath131 such that @xmath132 we call the triple @xmath133 the _ sub - lorentzian manifold _ or the _ sub - lorentzian @xmath12-type group _ and the pair @xmath134 is named by the _ sub - lorentzian structure _ on @xmath56 .
we define the casual character on @xmath135 .
fix a point @xmath136 .
a horizontal vector @xmath99 is called _ timelike _ if @xmath137 , _ spacelike _ if @xmath138 or @xmath139 , null if @xmath140 and @xmath141 , nonspacelike if @xmath142 .
a horizontal curve is called timelike if its tangent vector is timelike at each point .
spacelike , null and nonspacelike curves are defined similarly .
the choice of the sub - lorentzian metric [ metric ] implies that the horizontal vector field @xmath143 is timelike and other horizontal vector fields @xmath144 , @xmath145 , are spacelike .
we call @xmath143 the _ time orientation _ on @xmath135
. then a nonspacelike vector @xmath99 is called future directed if @xmath146 , and it is called past directed if @xmath147 . throughout this paper f.d .
stands for `` future directed '' , t. for `` timelike '' , and nspc . for `` nonspacelike '' .
we would like to start the description of @xmath135 with finding geodesics , that is by definition , the projections of a solution of the associated hamiltonian system on @xmath135 .
we construct a hamiltonian system with respect to sub - lorentzian metric .
locally the hamiltonian function associated with the lorentzian metric can be defined in the following way : @xmath148 if we use the coordinates for @xmath149 as in the previous subsection , then the hamiltonian becomes @xmath150 here @xmath151 is the matrix given by and @xmath152 is the matrix of the minkowskii metric tensor.@xmath153 the corresponding hamiltonian system takes the form @xmath154 where the paricipating matrix @xmath155 is a constant matrix of the parameters and @xmath156 here by symbol @xmath157 is denoted the expression @xmath158 .
notice that @xmath159 .
after not intricate calculations two first equations of system roll up to the following linear system of ordinary differential equations @xmath160 @xmath161 that gives the equations for geodesics on @xmath135 . here
the conditions are derived from the second line of the system by substituting @xmath162 from the first line of this system .
the exact formulae for obtained geodesics see in @xcite .
in this section we briefly introduce the notion of an electromagnetic field in order to explain the relation between the motion of the charged particle in an electromagnetic field and a sub - lorentzian geodesic .
consider the minkowski spacetime @xmath3 with the lorentzian metric @xmath152 in it .
let @xmath163 be a charged particle in @xmath3 of a unite charge and a constant mass @xmath164 with a trajectory @xmath165 .
charged particles create an electromagnetic field and also respond on the fields created by other particles . an electromagnetic field in @xmath3
can be described by using two 3-dimensional vectors @xmath166 and @xmath167 that express electric and magnetic components respectively .
electromagnetic fileds in the space free of charge satisfies to four maxwell s equations @xmath168 where @xmath169 stands for the time coordinate in @xmath3 and permittivity and permeability are supposed to be constant and equal to 1 .
if we use the covariant formulation then maxwell s equations can be written in the nice symmetric form @xmath170 where @xmath171 is a 2-form field in 4-dimensional spacetime corresponding to antisymmetric electromagnetic tensor field @xmath172 the operator @xmath9 is the exterior derivative , a coordinate and metric independent differential operator , and @xmath41 is the hodge star operator that is linear transformation from the space of @xmath173 form into the space of two - forms defined by the metric in minkowski space , see for instance @xcite .
while maxwell equations describes how electrically charged particles and objects give rise to electric and magnetic fields , the lorentz force law completes that picture by describing the force acting on a moving charged particle in the presence of electromagnetic fields , see , for instance @xcite .
this effect is described by lorentz equation @xmath174 where @xmath175 is the particle s world velocity , @xmath176 its world momentum , @xmath177 is the proper time of the particle , and @xmath171 is an electromagnetic tenzor field .
let @xmath178 be any admissible basis in @xmath3 ; that is orthonormal , @xmath179 responds for the time coordinate and @xmath180 for space coordinates .
as it was mentioned at each point of @xmath3 the linear transformation @xmath171 can be defined in terms of the classical electric and magnetic @xmath181-vectors @xmath182 and @xmath183 at that point . set @xmath184 the transformation @xmath185 which influences on the charged particle
is often called lorentz force . in order to find eigenspaces of @xmath185 that are invariant subspaces of this linear transformation we come to the characteristic equation . @xmath186 where @xmath187 and @xmath188 .
the algebraic combinations @xmath189 and @xmath190 are the same in all admissible frames and are called lorentz invariants . if both of them are equal to zero ( i.e. , @xmath166 and @xmath167 are perpendicular and have the same magnitudes ) : @xmath191 , then @xmath185 is called _ null _ transformation , otherwise @xmath185 is said to be _
regular_. every regular skew - symmetric with respect to the lorentz metric linear transformation
@xmath192 has a 2-dimensional invariant subspace @xmath82 such that @xmath193 .
there exist a basis , that is called the canonical basis , relative to which the matrix of regular skew - symmetric linear tranformation @xmath185 has the form @xmath194 where @xmath195 and @xmath196 are nonnegative real values such that @xmath197 , @xmath198 . now
the eigenvalues of @xmath185 are easy to calculate since the characteristic equation becomes @xmath199 , i. e. , @xmath200 , which has the following solutions : @xmath201 and @xmath202 .
the linear transformation @xmath203 defined by @xmath204,\ ] ] where @xmath205 , i is the identity transformation @xmath206 for every @xmath207 and @xmath208 is the trace of @xmath209 , is called the _ energy - momentum transformation _ associated with @xmath185 .
observe that @xmath210 is symmetric with respect to the lorentzian inner product and is trace - free , i. e. @xmath211 .
the term @xmath212 is called the _ energy density _ in the given frame of reference for the electromagnetic field @xmath171 with the form .
the 3-vector @xmath213 is called the _ poynting 3-vector _ and describes the energy flux of the field .
finally , the @xmath214 matrix @xmath215 is known as _ maxwell stress tensor _ of the field in the given frame .
thus , the content of the matrix of @xmath210 determines the energy content of the field @xmath171 in the corresponding basis .
the noncommutative multiplication law of quaternion @xmath13-type group @xmath216 defines the non - integrable distribution @xmath217 or in the covariant language the nonholonomic constraints @xmath218 .
the curvature of the distribution gives rise to the skew symmetric transformation @xmath22 in 4-dimensional space .
independently whether this space has euclidean structure or it is the minkowskii space the antisymmetric 2-form @xmath22 defines the electromagnetic field , since it trivially satisfies the maxwell equations .
we emphasize that the geometry of the nonholonomic manifold @xmath216 is related to the geometry of 4-dimensional space where a constant electromagnetic field acts . given a positively definite metric @xmath106 ( riemannian metric ) on the nonholonomic distribution @xmath79
we obtain a sub - riemannian manifold @xmath31 .
the hamiltonian function @xmath219 in this case is reduced to the lorentz equation in the euclidean space @xmath220 given by .
equation describes a motion of charged particle of unit charge in magnetic field @xmath22 in 4-dimentional euclidean space .
replacement of a positively definite metric @xmath106 on a nondegenerate indefinite metric , the lorentzian metric @xmath131 , leads to the relativistic lorentz equation in the minkowskii space .
it has more connections to physics since it is related to the motion of the charged particle in electromagnetic field in the minkowskii space that is closely connected to the general relativity .
our aim is to find geodesics that are projections on @xmath221 of the solutions to the hamiltonian system for @xmath222 defined on @xmath223 .
the hamiltonian system is reduced to the equations and .
equations are ordinary differential equations in 4-dimensional space with @xmath185 that is skew - symmetric with respect to the lorentzian metric @xmath152 .
this makes us to endow the 4-dimensional space with the lorentzian metric , producing the minkowskii space @xmath224 .
therefore , at each point of @xmath225 we can elect a model of electromagnetic field which corresponds to a linear skew - symmetric transformation @xmath226 that assigns to the world velocity @xmath227 of a charged particle passing through that point the change in world momentum @xmath228 that the particle should expect due to the field . since we can assume that the charge and the mass of the particle equals
1 we get @xmath229 and the lorentz force law becomes @xmath230 that is exactly equations .
if we set @xmath231 , @xmath232 , @xmath233 , then we conclude that equation describes the motion of a particle of unit charge in the constant electromagnetic field with @xmath234 . in this case
one of the lorentz invariant is zero : @xmath235 , and the other @xmath190 is equal to @xmath236 , where @xmath237 .
the matrix @xmath209 given by has zero trace .
therefore , the energy momentum transformation is given up to a constant multiplied by symmetric matrix @xmath209 .
the energy density is equal to @xmath238 .
the poynting 3-vector in our case is zero vector .
the maxwell stress tensor is given by @xmath239 let us find a canonical basis for the matrix @xmath185 .
first we find a plane in @xmath14 in which both of the vectors @xmath166 and @xmath167 are lying , then we rotate it in such a way that it will coincide with a plane @xmath240 in @xmath14 .
that is , let us choose a right - handed orthonormal basis @xmath241 of the space @xmath242 in which @xmath243 .
there are infinitely many planes containing both vectors @xmath166 and @xmath244 .
fix one of these planes : for example , the one passing through the axis @xmath245 : @xmath246 with the help of the rotation @xmath247 we turn it on the angle @xmath248 between the plane and @xmath240 around the axis @xmath249 so that the third coordinate of @xmath166 equals to zero .
consider now a lorentz transformation @xmath250 of a basis @xmath251 , where @xmath252 is given by .
it yields a new admissible coordinate system @xmath253 in which @xmath254 and @xmath255 and matrix @xmath256 can be defined in the following way : @xmath257 where @xmath258 , @xmath259 , and @xmath260 .
since @xmath261 is regular skew - symmetric transformation , then it has a 2-dimensional invariant subspace @xmath175 such that @xmath262
. then @xmath263 is also a 2-dimensional invariant subspace for @xmath185 and there exist real numbers @xmath264 and @xmath265 such that @xcite @xmath266 take any future directed unit timelike vector @xmath267 in @xmath175 , for example @xmath268 , where @xmath269 are eigenvectors of matrix @xmath185 ( see @xcite ) .
then spacelike unit vector @xmath270 in @xmath175 and the real value @xmath264 can be found from the conditions @xmath271 and @xmath272 .
we get @xmath273 and @xmath274 .
now , let @xmath275 be an arbitrary unit spacelike vector in @xmath263 , for example , select @xmath276 .
then construct @xmath277 satisfying @xmath278 and @xmath279 .
we obtain @xmath280 and @xmath281 .
thus , @xmath282 is an orthnormal basis for @xmath225 , which is called _ canonical _ and @xmath283 electric and magnetic fields corresponding to this transformation are @xmath284 and @xmath285 , so that an observer in this frame will measure them in the same direction ( of @xmath286-axis ) and of magnitude @xmath287 .
matrix @xmath288 is a block - type matrix , where @xmath289-block in the left upper corner coincides with a matrix in sub - lorentzian heisenberg case and the right lower @xmath289-block coincides with a matrix in usual sub - riemannian heisenberg case ( see @xcite ) .
let us denote by @xmath290 the matrix which columns are orthnormal basis vectors @xmath291 .
then in new basis @xmath292 and the vector @xmath293 is of the form @xmath294 , where @xmath295 is a vector in old coordinates .
therefore , @xmath296 .
it is clear that real eigenvalues of @xmath185 are @xmath297 from the canonical form .
therefore , eigenspaces are @xmath298 and @xmath299 respectively
. directions @xmath300 are called _ principal null directions _ of @xmath185 .
the lorentz equation in canonical coordinates takes the form @xmath301 it splits into 2 independent systems @xmath302 we wish to solve this system under the initial conditions @xmath303 , @xmath304 .
the solution for @xmath305 is @xmath306 integrating these expressions , we get the solution @xmath307 , that we write in the matrix form as @xmath308 with @xmath309 now , let us introduce following notation : @xmath310 , @xmath311 .
[ hyperbola ] the projection of the geodesic onto @xmath312-plane is a brunch of the hyperbola with the canonical equation @xmath313 since @xmath314 @xmath315 we calculate that @xmath316 .
this expression can be rewritten as stated in [ circle ] the projection of the geodesic onto @xmath317-plane is a circle with the center at @xmath318 of the radius @xmath319 . since @xmath320 @xmath321
we get @xmath322 .
the expression leads to @xmath323 the horizontality conditions in canonical basis have the form @xmath324 where the matrices @xmath325 @xmath326 @xmath327 are skew - symmetric with respect to usual euclidean metric .
more explicitely , @xmath328 , where @xmath329 are coordinates in the canonical basis , @xmath330 are auxiliary expressions @xmath331 , @xmath332 is given by and @xmath333 notice that @xmath334 is an orthogonal transformation in @xmath14 while the matrix @xmath290 represents the orthogonal transformation @xmath220 .
it is more convenient for us to work with the expressions @xmath335 , @xmath336 , @xmath337 .
then @xmath338 , @xmath339 , @xmath340 can be obtained by the orthogonal transformation @xmath334 of @xmath335 , @xmath336 , @xmath337 .
taking into account that @xmath341 and , we calculate @xmath342 @xmath343 , \end{split}\ ] ] @xmath344.\end{split}\ ] ] let us use the following notation for the constants @xmath345 thus , we integrate @xmath346 @xmath347 observe that @xmath348 then the direct calculations yield @xmath349
we wish to describe the set of points in @xmath32 that can be reached from the origin @xmath350 by a geodesic : timelike , lightlike or spacelike .
we can fix the starting point at the origin @xmath351 , since the solutions of the hamiltonian equations are invariant under the left translation .
we start from the simple lemma , that is related to the case of null transformation @xmath185 .
[ lem : theta0 ] if @xmath352 , then the system with initial data @xmath353 , @xmath354 , @xmath355 has the solution @xmath356 , @xmath357 , @xmath358 , @xmath359 .
the projections to @xmath58-space are straight lines that are timelike if @xmath360 , lightlike if @xmath361 , and spacelike if @xmath362 .
the condition @xmath352 immediately implies that @xmath363 and @xmath364 are constant and @xmath356 .
then @xmath365 since the matrix @xmath366 is skew symmetric . the description of the reachable set by causal curves is very complicated in general , therefore , we present here some particular cases .
we mostly reduce our considerations to the sets that can be reached by geodesics , since the geodesics can not change its causal character .
we need to solve the equations @xmath369 with the boundary conditions @xmath370 we find the relation between the initial velocity and the value @xmath371 .
substituting @xmath372 in @xmath373 , we obtain @xmath374.\ ] ] we also have @xmath375 , where @xmath376 from .
putting @xmath372 , we calculate @xmath377.\end{gathered}\ ] ] the expression can be written in the form @xmath378 let us describe all possible initial velocities which lead to vanishing of the norm @xmath379 .
case 1 . @xmath352 . in this case
lemma [ lem : theta0 ] implies @xmath380 and we conclude that if @xmath367 is such that @xmath368 and @xmath381 then there is a unique geodesic , that is lightlike straight line , connecting @xmath367 with @xmath67 .
@xmath382 , @xmath383 .
therefore , @xmath384 . in this case @xmath385 and the possible geodesic is lightlike that remains lightlike for all @xmath386 $ ] .
let us write the equations for these geodesics taking into account the condition on the initial velocity @xmath387 , @xmath388 .
@xmath389 @xmath390 this is not obvious parametrization of straight lines @xmath391 , @xmath392 . in both cases @xmath393 , @xmath95 that implies @xmath394 .
we conclude that the origin can be joined with a point @xmath395 , where @xmath396 by lightlike geodesic that is straight line .
thus , case 2 is a particular situation of the case 1 .
notice that the equations of the horizontal part and of geodesics in @xmath32 coincides up to reparametrization with the equations of horizontal part for geodesics in @xmath30 provided the initial velocity @xmath387 ( see @xcite ) .
the results of cases 1 and 2 may be united in the following statement let @xmath397 be a point such that @xmath398 , @xmath399 , @xmath95 and @xmath400 .
then there is a unique lightlike geodesic joining the origin and @xmath185 which is a straight line .
@xmath401 , @xmath383 .
we write the expression in the equivalent form @xmath402 and consider the following subcases .
3.1 ) let @xmath384 then @xmath403 and in this case all geodesics are spacelike . then implies that @xmath404 , @xmath405 .
3.1.1 ) let us suppose that @xmath406 and @xmath407 , @xmath408 are arbitrary .
the main result is expressed in the following [ theorem2 ] given a point @xmath409 there are uncountably many spacelike geodesics connecting the origin with @xmath185 .
the geodesics are given by equations @xmath410 @xmath411 the geodesics have the lengths @xmath412 , @xmath405 .
we start from the equations for @xmath64-coordinates of geodesics .
since @xmath413 we get @xmath414 by . the projection into @xmath317-plane is @xmath415 which by lemma can be rewritten as the canonical equation of the circle on the @xmath317-plane with the center at @xmath416 of the radius @xmath417
: @xmath418 the number @xmath405 reflects the number of turns along the circle for any fixed initial velocity @xmath419 and the radius of the circle . to exclude @xmath419 from the equations we use and find @xmath420 set @xmath372 in the expression for @xmath421 and
get @xmath422 .
since @xmath423 with @xmath334 given by we get @xmath424 it implies that @xmath425 and @xmath426 .
fixing @xmath405 , we fix the speed @xmath419 of a geodesic , but we still have the choice in the directions of @xmath427 that are parametrized by the unit circle .
it gives uncountably many geodesics .
we get the equations for @xmath428-coordinates from @xmath423 taking into account that the values of @xmath127 related to the values of @xmath429 by .
since the geodesics are spacelike the length @xmath430 of a geodesic @xmath431 can be calculated from the formula @xmath432 notice that in lorentzian heisenberg group @xmath433 there is no geodesic of any causal type joining @xmath434 and @xmath435 with @xmath436 ( see @xcite ) .
3.1.2 ) we suppose now that @xmath437 .
we present first the auxiliary calculations and then formulate the main statement .
we have @xmath438 then the equations of geodesics take the form @xmath439 the relations lead to @xmath440,\nonumber\\ \tilde z_3(t)=\frac{\dot x_0(0)}{(2\pi n)^2 } \left[\dot x_2(0)\big(\big(\cos(2\pi n t)+1\big)\big(e^{2\pi
n t}-1\big)-\sin(2\pi n t)\big(e^{2\pi n t}+1)\big)\big)\right.\nonumber \\
\qquad\;\;\;\quad\quad\left.-\dot x_3(0)\big(\big(\cos(2\pi n t)-1\big)\big(e^{2\pi n t}+1\big)-\sin(2\pi n t)\big(e^{2\pi n t}-1)\big)\big ) \right],\nonumber\end{gathered}\ ] ] where @xmath441 .
setting @xmath372 in we get @xmath442 the equations imply @xmath443 given a point @xmath397 , @xmath444 , @xmath445 , @xmath446 , there are uncountably many spacelike geodesics connecting the origin @xmath351 with @xmath185 .
the equations of geodesics are given by expressions @xmath447 and @xmath448 , where @xmath449,\nonumber \\
\tilde z_3(t)=\frac{1}{2(e^{2\pi n } -1)}\left[\tilde z_3(1)\big((\cos(2\pi n t)+1)(e^{2\pi n t}-1)-\sin(2\pi n t)(e^{2\pi n t}+1)\big)\right.\nonumber \\ \left.\qquad\qquad\qquad\qquad\qquad -\tilde z_2(1)\big((\cos(2\pi n t)-1)(e^{2\pi
n t}+1)+\sin(2\pi n t)(e^{2\pi n t}-1)\big)\right],\nonumber \end{gathered}\ ] ] @xmath450 , and @xmath441 . fix @xmath451 and choose any @xmath452 such that @xmath453 .
we have uncountably many triples that are parameterized by the 3-@xmath9 sphere .
this choice defines the orthogonal transformation @xmath334 given by .
given @xmath429 we find the auxiliarry parameters @xmath454 by @xmath450 .
setting the values of @xmath455 , @xmath419 and @xmath456 from and into the general solutions and and applying the orthogonal transformation @xmath334 to @xmath457 we finish the proof .
we observe that projections of the geodesics onto @xmath312-plane are straight lines @xmath458 .
the projections onto @xmath317-plane are circles from the lemma [ circle ] .
3.1.3 ) for the case @xmath459 in the same way as above we obtain the following given a point @xmath397 , @xmath444 , @xmath460 , @xmath446 , there are uncountably many spacelike geodesics connecting the origin @xmath351 with @xmath185 .
the equations of geodesics are given by expressions @xmath461 and @xmath448 , where @xmath462,\\ \tilde z_3(t)=\frac{1}{2(e^{-2\pi n}+1)}\left[\tilde z_3(1)\big((\cos(2\pi n t)+1)(e^{-2\pi n t}+1)+\sin(2\pi n t)(e^{-2\pi n t}-1)\big)\right.\\ \left.\qquad\qquad\qquad\qquad\qquad -\tilde z_2(1)\big((1-\cos(2\pi n t))(e^{-2\pi n t}-1)+\sin(2\pi n t)(e^{-2\pi n t}+1)\big)\right].\end{gathered}\ ] ] @xmath450 , and @xmath441 . 3.2 ) consider the case @xmath463 , i. e. , the initial velocity @xmath464 is spacelike .
it suggests that the ratio @xmath465 is negative , which contradicts to .
hence , in this case there are no spacelike geodesics with the property .
3.3 ) now suppose @xmath466 .
it implies that the ratio @xmath465 is positive and here we have 3 more subcases : 3.3.1 ) @xmath467 . in this case
the initial velocity appears to be timelike .
there are no points where this can happen because the function @xmath468 is wave decaying with exponential speed on the right halfplane and bounded with 0 from below and with 1 from above .
thus , there are no timelike geodesics with the considered timelike initial velocity . 3.3.2 ) @xmath469 .
then @xmath470 , which is possible if and only if @xmath352 .
this case corresponds to case 1 ) , when the initial velocity is lightlike .
3.3.3 ) @xmath471 .
let us investigate the quantity of hamiltonian geodesics in this case . using and the fact that @xmath472 we calculate that @xmath473 and notice that the ratio @xmath474 depends only on @xmath287 and @xmath475 the function @xmath476 is nonnegative , has countably many points @xmath477 where @xmath478 vanishes and on each interval @xmath479 , @xmath480 the function @xmath481 has a unique critical point @xmath482 . on each of these intervals the function
@xmath481 strictly increases from @xmath483 to @xmath484 , and then , strictly decreases from @xmath484 to @xmath483 .
let us denote the numerator @xmath485 by @xmath486 and study this function .
its derivative @xmath487 turns into @xmath483 if @xmath488 @xmath489 the latter equation has infinitely many solutions @xmath477 , @xmath480 , where @xmath490 and all the other roots @xmath491 and @xmath492 , @xmath493 . since @xmath494 and @xmath495 , @xmath496
, we conclude that function @xmath497 has minimums at @xmath477 , maximums at points @xmath498 , and it is no - negative for @xmath499 , see figure [ fig : f1 ] .
, width=170,height=170 ] the denominator @xmath500 is an increasing function , passing through @xmath483 only at @xmath352 , since the derivative @xmath501 is always positive and equals to @xmath483 only at @xmath352 .
hence , @xmath10 is positive for @xmath502 .
since @xmath503 for @xmath504 and since @xmath505 , @xmath493 the function @xmath506 equals to @xmath483 only at points @xmath477 , @xmath493 and is positive at all the other points @xmath507 and @xmath504 .
the point @xmath490 needs additional consideration .
using taylor decomposition of function @xmath481 near @xmath508 we get that @xmath509 , which goes to @xmath483 if @xmath510 .
therefore , @xmath511 . let us consider the derivative @xmath512 .
we already know that @xmath477 are the solutions of @xmath513 , since @xmath514 , therefore , they are local extremal points of function @xmath481 . since @xmath515 is nonnegative for @xmath502 and @xmath516 we conclude that the points @xmath477 are minimums of function @xmath481 .
moreover , since @xmath481 is a smooth function on @xmath517 , on each interval @xmath479 , @xmath480 , it reaches its maximums , which we denote by @xmath482 .
let us make one more observation . since @xmath10 is a monotonically increasing function , on each interval @xmath518 $ ] , @xmath493 it has its local minimums at points @xmath519 . then , taking into account that @xmath497 has local maximums at points @xmath498 , @xmath496 , we can estimate the function @xmath481 on each of these intervals from above by the value @xmath520 , @xmath493 .
from here we obtain that @xmath521 with @xmath522 .
therefore , the function @xmath515 decays to @xmath483 when @xmath287 goes to infinity .
note , that we can estimate @xmath523 from above by @xmath524 .
the graph of function @xmath481 is on the figure [ fig : f2 ] . ,
width=226,height=226 ] this lemma allows us to prove the following theorem .
[ general ] let @xmath397 be a point such that @xmath368 , @xmath525 and @xmath523 is a global maximum of @xmath481 .
suppose that @xmath526 and that @xmath287 is a solution of the equation @xmath527 then there exist more than one spacelike geodesics joining the origin with @xmath185 .
the equations of geodesics are given by @xmath528 and @xmath529 if @xmath530 then there are no geodesics of any causal type joining @xmath483 and @xmath185 . given the values @xmath367 of the end point we find @xmath287 as a solution to the equation . for any of these solutions
we get for @xmath372 from the expression @xmath531 @xmath532 from here we calculate @xmath533 @xmath534 we change in general equations the values of @xmath535 and @xmath536 by and get . next ,
for each solution to we find an orthogonal transformation of @xmath428-space that fixes @xmath338-coordinate and therefore leave the expression @xmath537 invariant . therefore using ,
we substitute there the necessary combinations by and and get .
we observe that according to lemma [ circle ] projections of geodesics on @xmath317-plane are circles with center at @xmath538 and radius @xmath319 .
the projections on @xmath312-plane are suitable brunches of hyperbolas passing through @xmath67 , see lemma [ hyperbola ] .
the parameters of the circles and hyperbolas can be rewritten in terms of the final point of a geodesic by making use of .
notice that if @xmath539 , the quantity of solutions to is growing and the values of solutions tends to @xmath519 . in this case shows that the part of the velocity @xmath540 tends to @xmath483 and we come into particular situation of the theorem [ theorem2 ] .
theorem [ general ] gives an answer about existence of spacelike geodesics and allows us estimate their cardinality .
unfortunately the estimation is not complete since we use only the orthogonal transformations in @xmath428-space leaving invariant @xmath338-coordinate .
the difficulty is pure technical , since it is complicate to find the relation between the coordinates of finite point and the values of @xmath287 . in the case of orthogonal transformations in @xmath428-space leaving invariant @xmath338-coordinate the relation
is expressed by equation .
given a point @xmath397 on the surface @xmath368 , @xmath541 , there are no timelike geodesics joining @xmath483 to @xmath185 .
let us use the hyperspherical coordinates @xmath544 : @xmath545 then the horizontality conditions become @xmath546 then , @xmath547 a smooth curve @xmath548 is horizontal with constant @xmath428-coordinates if and only if @xmath548 is a straight line in a 4-dimensional affine subspace : @xmath549 with @xmath550 and @xmath551 .
let @xmath548 be a horizontal curve with constant vertical components .
then the equation @xmath552 implies @xmath553 > from the first equation we conclude that @xmath554 is constant . in the case
when @xmath555 and @xmath556 , @xmath557 , we see that @xmath558 and @xmath559 are constants @xmath560 and @xmath561 . therefore , horizontal components of the curve @xmath548 are of the form @xmath562 if @xmath563 , @xmath564 , then @xmath565 and if @xmath566 , @xmath567 , then @xmath568 conversely , let us assume that @xmath569 , where @xmath570 are some constants .
set @xmath571 .
observe , that @xmath572 for any vector @xmath573 .
then , @xmath574 this implies that @xmath428-coordinates are constants .
in this section we would like to compare the results obtained for @xmath1-type quaternion group with lorentzian metric and 3-dimensional lorentzian heisenberg group @xmath30 .
we remind that @xmath30 is a triple @xmath575 , where @xmath14 is equipped with noncommutative multiplication law @xmath576 the subbundle @xmath12 is a span of two left invariant vector fields @xmath577 , @xmath578 , for which @xmath579=z=\frac{\partial}{\partial z}$ ] , and @xmath10 is a lorentzian metric on @xmath12 defined by @xmath580 let @xmath581 be a point such that @xmath582 @xmath583 , @xmath584 , @xmath585 . then there is a unique future directed (
past directed ) geodesic , joining @xmath586 with the point @xmath185 .
let @xmath587 be a solution of the equation @xmath588 then the equations of timelike future directed geodesic @xmath103\to \mathbb{h}^1_l$ ] are @xmath589 let us define the following sets @xmath590 then there exists a unique geodesic connecting the point @xmath586 with a point @xmath290 that belongs to one of the sets @xmath591 , @xmath592 or @xmath593 . particularly , if @xmath594 , then the geodesic is timelike , if @xmath595 , then the geodesic is spaselike , and if @xmath596 , then the geodesic is lightlike .
the connectivity on @xmath30 is dependent on the solutions @xmath287 of the equation @xmath597 where the function @xmath598 is strictly decreasing on the interval @xmath599 from @xmath600 to @xmath130 .
it means that if the point @xmath581 is such that @xmath601 , i. e. if @xmath602 , then there are no geodesics of any causal type joining the origin with @xmath185 . in particular , there are no hamiltonian geodesics joining the origin with the points of the surfaces @xmath603 and @xmath604 .
we observe that in @xcite it is shown that it is possible to find nonhamiltonian lightlike geodesics joining @xmath483 with the points of these surfaces .
thus , in contrast with @xmath135 , the heisenberg lorentzian group has the property of uniqueness of geodesics starting from the origin with given tangent vector .
it happens due to lower dimension of the spacelikepart of @xmath30 .
grochowski m. _ normal forms of germs of contact sub - lorentzian structures on @xmath606 .
differentiability of the sub - lorentzian distance function .
_ j. dynam .
control systems * 9 * ( 2003 ) , no .
4 , 531547 .
kobayashi s. ; nomizu k. _ foundations of differential geometry .
_ reprint of the 1963 original .
wiley classics library . a wiley - interscience publication .
john wiley & sons , inc .
, new york , 1996 . | we study the existence and cardinality of normal geodesics of different causal types on @xmath0-type quaternion group equipped with the sub - lorentzian metric .
we present explicit formulas for geodesics and describe reachable sets by geodesics of different causal character .
we compare results with the sub - riemannian quaternion group and with the sub - lorentzian heisenberg group , showing that there are similarities and distinctions .
we show that the geodesics on @xmath1-type quaternion groups with the sub - lorentzian metric satisfy the equations describing the motion of a relativistic particle in a constant homogeneous electromagnetic field . |
recently the t2k collaboration reported @xcite evidence at @xmath1 for a non - zero value of the reactor angle @xmath0 in the neutrino mixing matrix .
quantitatively , it was found that @xmath2 at 90% c.l.for @xmath3 , @xmath4 and @xmath5 , @xmath6 and @xmath7 being the atmospheric neutrino mass squared difference and mixing angle , and @xmath8 being the dirac cp violating phase in the neutrino mixing matrix ( see , e.g. @xcite ) . under the same conditions the best fit value of @xmath0 obtained from the data analysis is relatively large : @xmath9 .
subsequently the minos collaboration also reported evidence for a relatively large value of @xmath0 , although with a smaller statistical significance @xcite .
a global analysis of the neutrino oscillation data , including the data from the t2k and minos experiments , performed in @xcite showed that actually @xmath10 at @xmath11 .
the authors of @xcite find : @xmath12 using the `` old '' ( `` new '' ) fluxes of reactor @xmath13 in the analysis .
moreover , it was found in the same global analysis that @xmath14 ( and @xmath15 ) is clearly favored by the data over @xmath16 ( and @xmath17 ) .
the t2k and minos results will be tested in the upcoming reactor neutrino experiments double chooz @xcite , daya bay @xcite and reno @xcite . if confirmed , they will have far reaching implications for the program of future research in neutrino physics .
a relatively large value of @xmath0 opens up the possibilities , in particular , i ) for searching for cp violation effects in neutrino oscillations experiments with high intensity accelerator neutrino beams ( like t2k , no@xmath18a , etc . ) ; ii ) for determining the sign of @xmath6 , and thus the type of neutrino mass spectrum , which can be normally or invertedly ordered ( see , e.g. @xcite ) , in the long baseline neutrino oscillation experiments at accelerators ( no@xmath18a , etc . ) , in the experiments studying the oscillations of atmospheric neutrinos ( see , e.g. @xcite ) , as well as in experiments with reactor antineutrinos @xcite .
a value of @xmath19 is a necessary condition for a successful `` flavoured '' leptogenesis when the cp violation required for the generation of the matter - antimatter asymmetry of the universe is provided entirely by the dirac cp violating phase in the neutrino mixing matrix @xcite . in the present article
we investigate the possibilities to obtain the relatively large value of @xmath0 , suggested by the t2k , minos and the current global neutrino oscillation data , in the case when the diagonalization of the neutrino mass matrix gives a negligible contribution to @xmath0 .
parametrically the latter means a contribution smaller than @xmath20 , where @xmath21 is the cabibbo angle expansion parameter .
the diagonalization of the neutrino mass matrix leads to a negligible @xmath0 , for example , in models with @xmath22 flavour symmetry @xcite and in most models accounting for bimaximal ( bm ) @xcite or tri - bimaximal ( tbm ) @xcite neutrino mixing . for our purposes
, this is equivalent to setting the neutrino mass matrix contribution to @xmath0 to zero , which we will do in our further analysis .
the @xmath0 mixing angle can then be due to the contribution to the neutrino mixing coming from the diagonalization of the charged lepton mass matrix ( see , e.g. @xcite and the references quoted therein , see @xcite as well as , e.g. @xcite . ] ) .
in particular , we will consider the possibility of a `` large '' @xmath0 originating from the charged lepton mass matrix in the context of su(5 ) supersymmetric unification .
the motivation for such a choice are strong and well known . in our case
, this choice will also allow to get a handle on the expected contribution to @xmath0 and on the structure of the charged lepton yukawa coupling matrix .
this is because the su(5 ) symmetry allows to relate the latter and the down quark one , on which further experimental information is available . in minimal
su(5 ) models , the down - quark and transposed charged lepton mass matrices are equal . on the other hand , depending on the ( renormalizable or non - renormalizable )
su(5 ) operator(s ) giving rise to a given yukawa coupling matrix entry , the relation may involve clebsch - gordan ( cg ) coefficients .
such coefficients need to be invoked in order to fix the su(5 ) predictions for the @xmath23 and @xmath24 mass ratios ( unless specific values of @xmath25 , peculiar patterns of susy breaking parameters , or highly asymmetric mass textures are employed @xcite ) , as shown by the authors of @xcite after having identified the possible gut scale mass ratios in a bottom - up approach .
all the possible cg coefficients in su(5 ) and the pati salam group entering such relations arising from dimension five and some dimension six operators are listed in @xcite .
such coefficients not only can give rise to new mass ratios , but can also affect the prediction for @xmath0 @xcite . in this paper , we discuss possible combinations of cg coefficients , which can give a `` large '' @xmath0 even if there is no contribution to @xmath0 from the neutrino mass matrix at all .
the paper is organised as follows . in section 2
we describe the general set up , the determination of the standard neutrino parameters under the assumptions we make , the numerical procedure we used to determine the allowed cg coefficients and the predictions for @xmath0 .
the results are illustrated in section 3 . in section 4
we consider the special case in which the neutrino contribution to the pmns matrix is of the bm or tbm form , in which interesting implications for the leptonic ( dirac ) cp - violation phase can be drawn . in section 5
we summarize .
we start with some definitions and a summary of the assumptions we make regarding the structure of the mixing matrices .
thereby we follow closely the discussion in @xcite .
the pmns lepton mixing matrix @xmath26 is given by @xmath27 where @xmath28 and @xmath29 are the charged lepton and light neutrino mass matrices , respectively , and @xmath30 are diagonal and positive .
the pmns matrix can be parameterized in the standard way as @xmath31 , where @xmath32 is a diagonal phase matrix containing the two physical majorana cp violation phases @xcite . barring correlations among the entries of @xmath28
, the hierarchy of the charged lepton masses translates into the possibility of diagonalizing @xmath28 perturbatively through subsequent @xmath33 unitary rotations in the following order : @xmath34 , where @xmath35 is a unitary rotation in the @xmath36 block .
the 13 rotation is negligible for our purposes and will be set to zero , thus @xmath37 .
the assumption of zero neutrino mass matrix contribution to @xmath0 can be rephrased as @xmath38 in the limit of @xmath39 , or equivalently as @xmath40 up to phase matrices .
note that since @xmath41 does not contribute to @xmath0 , the assumption can be further rephrased as @xmath42 here @xmath43 , where @xmath44 is a cp - violating phase @xcite .
the angles @xmath45 , @xmath46 , @xmath47 and the phase @xmath44 in the expression ( [ eq : assumption ] ) for @xmath26 can be arranged to lie in the intervals @xmath48 $ ] and @xmath49 $ ] , respectively . by commuting with the @xmath50 one , the @xmath51-rotation has induced a non vanishing @xmath0 given by @xmath52 note that we have traded here @xmath53 with @xmath7 in the standard parameterization .
note also that the corrections to @xmath50 and @xmath54 induced by @xmath45 are given by [ eq : corrections ] @xmath55 in the case of 3-@xmath18 mixing under discussion , the magnitude of the cp violation effects in neutrino oscillations is determined @xcite by the rephasing invariant associated with the dirac cp violating phase @xmath8 : @xmath56 the rephasing invariant @xmath57 , as is well known , is a directly observable quantity .
it is analogous to the rephasing invariant associated with the dirac phase in the cabibbo - kobayashi - maskawa quark mixing matrix , introduced in @xcite . in the standard parametrisation of the pmns matrix we find
@xmath58 using eqs .
( [ jcp ] ) and ( [ eq : assumption ] ) we get @xcite @xmath59 thus , to leading order in @xmath60 we have @xmath61 . by comparing the real part of @xmath62 in the two parameterisations
we conclude that , at the same order , @xmath63 we would like now to study the possibility to generate a @xmath45 large enough to induce a @xmath0 in the range indicated by recent experiments in the context of an su(5 ) grand unified theory ( gut ) .
the unification assumption is powerful because it allows to relate the charged lepton and down quark yukawa matrices @xmath66 and @xmath67 .
if all the yukawa entries were generated by renormalizable operators and the mssm higgs fields were embedded in @xmath68 and @xmath69 representations only , we would have @xmath70 , leading to wrong predictions for the fermion mass ratios . in the general case one
has instead @xmath71 .
the coefficients @xmath72 depend on the operators from which the yukawa entries arise .
such values can be constrained to belong to a finite set of rational numbers at the price of assuming that each yukawa entry comes at least dominantly from a single renormalizable or non - renormalizable su(5 ) operator . in this case , the possible values of the @xmath72 coefficients are listed in table [ tab : su5relations ] , see also @xcite . .
summary of possible su(5 ) predictions for the coefficients @xmath72 .
numbers are taken from @xcite , where also the corresponding operators are listed .
[ tab : su5relations ] [ cols="^,^",options="header " , ] as a function of @xmath73 in the cases of tri - bimaximal ( left panel ) and bimaximal mixing ( right panel ) arising from the diagonalisation of the neutrino mass matrix , eqs .
( [ th12bm1 ] ) and ( [ th12tbm1 ] ) . the green , yellow , orange regions correspond to the 1 , 2 , 3@xmath74 allowed ranges of @xmath75 . in the bimaximal mixing case only the negative values of @xmath76 are compatible with the data on @xmath77 .
the vertical straight , dashed , dotted lines denote the 1 , 2 , 3@xmath74 allowed ranges of @xmath73 .
the values of @xmath75 and @xmath73 are taken from @xcite ( see text for details ) .
[ fig : cp1],title="fig : " ] as a function of @xmath73 in the cases of tri - bimaximal ( left panel ) and bimaximal mixing ( right panel ) arising from the diagonalisation of the neutrino mass matrix , eqs .
( [ th12bm1 ] ) and ( [ th12tbm1 ] ) .
the green , yellow , orange regions correspond to the 1 , 2 , 3@xmath74 allowed ranges of @xmath75 . in the bimaximal mixing case only the negative values of @xmath76 are compatible with the data on @xmath77 .
the vertical straight , dashed , dotted lines denote the 1 , 2 , 3@xmath74 allowed ranges of @xmath73 .
the values of @xmath75 and @xmath73 are taken from @xcite ( see text for details ) .
[ fig : cp1],title="fig : " ] the values of @xmath75 , @xmath78 and @xmath79 , obtained in the global data analysis in @xcite are shown for convenience in table [ tab : experimental_pmns_angles ] . in fig .
[ fig : cp1 ] we present graphically the constraints on @xmath80 implied by the data on @xmath75 and @xmath73 in both cases of bm and tbm @xmath81 , using the relations ( [ th12bm1 ] , [ th12tbm1 ] ) , which are exact in @xmath45 .
a few comments are in order .
it follows from our analysis that in the case of bimaximal @xmath81 , the current @xmath82 ( @xmath83 ) experimentally allowed range for @xmath75 requires that @xmath84 , the minimal value being very close to the best fit value found in @xcite . if future data on @xmath0 will show ( taking into account all relevant uncertainties ) that , e.g. @xmath85 , the simple case under discussion of @xmath81 having a bm mixing form and @xmath86 will be ruled out .
further , using the @xmath82 ( @xmath83 ) allowed ranges of both @xmath75 and @xmath73 we find that @xmath76 is constrained to lie in the interval : @xmath87 .
thus , @xmath88 , or @xmath89 , and therefore maximal cp violation in neutrino oscillations , is ruled out in the scheme we are considering .
we get very different constraints on @xmath76 in the case of the tbm form of @xmath81 .
if we use the @xmath82 allowed ranges of @xmath75 and @xmath73 in the analysis , all possible values of @xmath76 are allowed .
the @xmath83 intervals of allowed values of @xmath75 and @xmath73 require that @xmath90 .
thus , maximal cp violation , @xmath89 , is allowed . as a function of @xmath73 in the cases of tri - bimaximal ( left ) and bi - maximal ( right ) mixing from the neutrino sector .
the green , yellow , orange regions correspond to the 1 , 2 , 3@xmath74 allowed ranges of @xmath75 .
the vertical straight , dashed , dotted lines show the 1 , 2 , 3@xmath74 allowed ranges of @xmath73 .
the 1 , 2 , 3@xmath74 allowed ranges of @xmath75 and @xmath73 are the same as in fig .
[ fig : cp1 ] .
[ fig : cp2],title="fig : " ] as a function of @xmath73 in the cases of tri - bimaximal ( left ) and bi - maximal ( right ) mixing from the neutrino sector . the green , yellow , orange regions correspond to the 1 , 2 , 3@xmath74 allowed ranges of @xmath75 .
the vertical straight , dashed , dotted lines show the 1 , 2 , 3@xmath74 allowed ranges of @xmath73 .
the 1 , 2 , 3@xmath74 allowed ranges of @xmath75 and @xmath73 are the same as in fig .
[ fig : cp1 ] .
[ fig : cp2],title="fig : " ] using the constraints on @xmath8 derived above , we have obtained also predictions for the magnitude of the rephasing invariant @xmath91 . the latter has the following form to leading order in @xmath92 in the cases of bm and tbm @xmath81 : @xmath93 the results of this analysis are presented in fig .
[ fig : cp2 ] . as could be expected on the basis of the results obtained for @xmath76 ,
the existing data on @xmath75 and @xmath73 imply that in the case of bm form of @xmath81 , @xmath91 should lie in the interval @xmath94 , using the @xmath82 ( @xmath83 ) ranges , the value @xmath95 being allowed .
we get similar results for the maximal possible value of @xmath96 if @xmath81 exhibits tbm mixing form . however , if we use the @xmath83 allowed ranges of @xmath75 and @xmath73 , we find that @xmath96 has to be non - zero and not smaller than 0.014 : @xmath97 .
values of @xmath98 are potentially accessible by the accelerator experiments t2k , no@xmath18a , etc . planning to search for cp violation effects in neutrino oscillations ( see , e.g. @xcite ) .
in the present article we have considered the possibility that a sizeable value of the neutrino mixing angle @xmath0 , compatible with the recent indications from the t2k , minos and the global neutrino oscillation data , arises primarily from the contribution of the charged lepton sector @xmath99 to the lepton mixing , @xmath100 , where @xmath26 is the pmns matrix and @xmath81 arises from the diagonalisation of the neutrino mass matrix .
this scenario should necessarily be taking place in the context of a wide variety of models predicting a relatively small contribution to @xmath0 from the neutrino sector , say @xmath101 , where @xmath102 is the cabibbo expansion parameter @xmath103 .
the analysis has been performed in the context of su(5 ) supersymmetric models , which are independently well motivated and allow the charged lepton contribution to @xmath0 to be related to the mass and mixing observables in the quark sector .
the naive relation @xmath104 between the charged lepton and down quark mass matrices is not compatible with the indication of a relatively large @xmath0 .
this is not surprising , as the quoted relation is not compatible with the well known values of the charged fermion mass ratios either and it has to be corrected , possibly by clebsch - gordan ( cg ) factors .
this is particularly true in the case ( relevant for us ) of the charged fermions belonging to the first two families , whose small masses can arise from non - renormalizable operators that can well involve su(5 ) breaking sources . under motivated hypotheses for the values of such cg coefficients , we found all the combinations that are experimentally viable .
the main results are summarized in tables [ tab : cg_gen_case_ren ] , [ tab : cg_gen_case_adj ] , [ tab : cg_gen_case_zero ] , and [ tab : cg_symm_case ] . in order to describe the present data
, one needs either a non - vanishing 11 entry in the down quark and charged lepton mass matrices , or cg coefficients from non - renormalizable or from su(5 ) breaking operators in representations larger than the adjoint .
we also studied the specific case in which the contribution from the neutrino sector to the lepton mixing , @xmath81 in eq . ( [ eq : u ] ) , is , up to the phase matrices @xmath105 and @xmath32 , see , eq .
, of the bimaximal ( bm ) or tri - bimaximal ( tbm ) form , while that from the charged lepton sector is assumed to be of the form @xmath86 . under these assumptions ,
@xmath106 , @xmath44 being a cp violating phase , and @xmath32 contains the two majorana cp violating phases . in this case
, it is possible to draw important conclusions on the leptonic ( dirac ) cp - violating phase @xmath107 and on the magnitude of the cp violation effects in neutrino oscillations .
the latter , as is well known , is determined by the rephasing invariant @xmath108 .
our results are summarized in figs .
[ fig : cp1 ] , [ fig : cp2 ] .
we find that in the case of bimaximal @xmath81 , the current @xmath82 ( @xmath83 ) experimentally allowed range for @xmath75 requires that @xmath84 .
if future data on @xmath0 will show ( taking into account all relevant uncertainties ) that @xmath73 has a smaller value , e.g. @xmath85 , the simple case of @xmath109 having a bm mixing form and @xmath86 will be ruled out .
further , using the @xmath82 ( @xmath83 ) allowed ranges of both @xmath75 and @xmath73 we find that @xmath76 is constrained to lie in the interval : @xmath87 .
thus , @xmath88 , or @xmath110 , and therefore maximal cp violation in neutrino oscillations , is essentially ruled out in the scheme we have considered . if it will be confirmed experimentally that @xmath111 , that would imply ( in the standard parameterization of the pmns matrix , in which @xmath112 ) that @xmath113 .
correspondingly , @xmath91 should lie in the @xmath82 ( @xmath83 ) interval @xmath114 , the value @xmath95 being allowed . however , if it will be experimentally established that @xmath115 , that would rule out the case of @xmath86 and @xmath81 having a bm mixing form .
in the case of the tbm form of @xmath81 and @xmath86 , all possible values of @xmath76 are allowed if one uses the @xmath82 allowed ranges of @xmath75 and @xmath73 in the analysis .
the @xmath83 intervals of allowed values of @xmath75 and @xmath73 require that @xmath90 , or @xmath116 .
thus , maximal cp violation , @xmath89 , in both cases is allowed .
we find using @xmath83 allowed ranges of @xmath75 and @xmath73 , that @xmath96 has to be non - zero and not smaller than 0.014 : @xmath97 .
this interval falls in the range of the potential sensitivity of the t2k and the future neutrino neutrino oscillation experiments ( no@xmath18a , etc . ) , designed to search for cp violation effects in neutrino oscillations .
this work was supported in part by the infn program on `` astroparticle physics '' , by the italian miur program on `` neutrinos , dark matter and dark energy in the era of lhc '' and by the world premier international research center initiative ( wpi initiative ) , mext , japan ( s.t.p . ) , by the eu marie curie itn `` unilhc '' ( pitn - ga-2009 - 23792 ) , and by the erc advanced grant no .
267985 `` damesyfla '' . during the finalising stage of the work on the present article , ref .
@xcite appeared , where the authors follow a similar approach in generating a relatively large @xmath0 .
our work goes beyond the analysis performed in @xcite in a number of aspects . in @xcite
only the case @xmath117 is discussed , while we consider the more general case of @xmath118 .
the authors of @xcite use only leading order terms ( without error estimates ) in the expressions for the observables in terms of the parameters of the down quark and charged lepton mass matrices ( i.e. , in eqs .
( 12a ) - ( 12e ) ) and they do not utilise the rather precisely known value of the ratio @xmath119 ( see table [ tab : inputs ] ) .
this rules out some of the possibilities considered in @xcite to be viable . finally , the authors of @xcite draw only qualitatively conclusions about the value dirac cp violating phase @xmath8 in the case of ( tri-)bimaximal mixing arising from the neutrino sector .
we perform a quantitative analysis and determine the allowed ranges ( at 1,2 and 3 @xmath74 ) of values of @xmath76 and of the rephasing invariant @xmath108 in the two cases ( figs .
[ fig : cp1 ] and [ fig : cp2 ] ) .
j. bernabu , s. palomares - ruiz and s.t .
petcov , nucl .
b * 669 * ( 2003 ) 255 ; s.t .
petcov and t. schwetz , _ ibid .
_ b * 740 * ( 2006 ) 1 ; r. gandhi _ et al .
d * 76 * ( 2007 ) 073012 .
petcov and m. piai , phys .
b * 533 * ( 2002 ) 94 ; s. choubey , s.t .
petcov and m. piai , phys .
d * 68 * ( 2003 ) 113006 ; p. ghoshal and s.t .
petcov , jhep * 1103 * ( 2011 ) 058 , and references quoted therein .
f. vissani , hep - ph/9708483 ; v. d. barger , s. pakvasa , t. j. weiler and k. whisnant , phys .
b * 437 * , 107 ( 1998 ) ; a. j. baltz , a. s. goldhaber and m. goldhaber , phys .
lett . * 81 * , 5730 ( 1998 ) ; h. georgi and s. l. glashow , phys .
d * 61 * , 097301 ( 2000 ) ; i. stancu and d. v. ahluwalia , phys .
b * 460 * , 431 ( 1999 ) .
p. f. harrison , d. h. perkins and w. g. scott , phys .
b * 530 * , 167 ( 2002 ) ; phys .
b * 535 * , 163 ( 2002 ) ; z. z. xing , phys .
b * 533 * , 85 ( 2002 ) ; x. g. he and a. zee , phys .
b * 560 * , 87 ( 2003 ) ; see also l. wolfenstein , phys .
d * 18 * , 958 ( 1978 ) .
frampton , s.t .
petcov and w. rodejohann , nucl .
b * 687 * ( 2004 ) 31 ; s.t .
petcov and w. rodejohann , phys . rev . * d71 * ( 2005 ) 073002 .
a. romanino , phys .
d * 70 * ( 2004 ) 013003 [ arxiv : hep - ph/0402258 ] . k.a .
hochmuth , s.t . petcov and w. rodejohann ,
lett b * 654 * ( 2007 ) 177 .
[ arxiv:0706.2975 [ hep - ph ] ] y. h. ahn , h. y. cheng and s. oh , phys . rev .
d * 83 * , 076012 ( 2011 ) [ arxiv:1102.0879 [ hep - ph ] ] ; j. a. escobar , [ arxiv:1102.1649 [ hep - ph ] ] ; s .- f .
ge , d. a. dicus , w. w. repko , phys .
. * b702 * ( 2011 ) 220 - 223 .
[ arxiv:1104.0602 [ hep - ph ] ] ; d. meloni , [ arxiv:1107.0221 [ hep - ph ] ] .
t. blazek , r. dermisek , s. raby , phys .
* d65 * ( 2002 ) 115004 .
[ hep - ph/0201081 ] ; t. blazek , r. dermisek , s. raby , phys .
* 88 * ( 2002 ) 111804 .
[ hep - ph/0107097 ] ; w. altmannshofer , d. guadagnoli , s. raby , d. m. straub , phys .
* b668 * ( 2008 ) 385 - 391 .
[ arxiv:0801.4363 [ hep - ph ] ] ; h. baer , s. kraml , a. lessa , s. sekmen , jhep * 1002 * ( 2010 ) 055 .
[ arxiv:0911.4739 [ hep - ph ] ] ; i. gogoladze , r. khalid , q. shafi , phys .
* d79 * ( 2009 ) 115004 .
[ arxiv:0903.5204 [ hep - ph ] ] ; i. gogoladze , s. raza , q. shafi , [ arxiv:1104.3566 [ hep - ph ] ] ; s. dar , i. gogoladze , q. shafi , c. s. un , [ arxiv:1105.5122 [ hep - ph ] ] .
bilenky , j. hosek and s.t .
petcov , phys .
* b94 * ( 1980 ) 495 . j. schechter and j.w.f .
valle , phys .
d * 22 * ( 1980 ) 2227 ; m. doi _ et al .
b * 102 * ( 1981 ) 323 . p. i. krastev and s. t. petcov , phys .
b * 205 * , 84 ( 1988 ) . c. jarlskog , z. phys .
c * 29 * 491 ( 1985 ) ; phys . rev .
lett . * 55 * , 1039 ( 1985 ) .
l. j. hall , r. rattazzi and u. sarid , phys .
d * 50 * ( 1994 ) 7048 [ arxiv : hep - ph/9306309 ] ; m. s. carena , m. olechowski , s. pokorski and c. e. m. wagner , nucl .
b * 426 * ( 1994 ) 269 [ arxiv : hep - ph/9402253 ] ; r. hempfling , phys .
d * 49 * ( 1994 ) 6168 ; t. blazek , s. raby and s. pokorski , phys .
d * 52 * ( 1995 ) 4151 [ arxiv : hep - ph/9504364 ] .
h. georgi , c. jarlskog , phys .
* b86 * ( 1979 ) 297 - 300 . g. altarelli , f. feruglio and i. masina , nucl .
b * 689 * , 157 ( 2004 ) ; s. f. king , jhep * 0508 * , 105 ( 2005 ) ; i. masina , phys .
b * 633 * , 134 ( 2006 ) ; s. antusch and s. f. king , phys . lett .
b * 631 * , 42 ( 2005 ) ; s. antusch , p. huber , s. f. king and t. schwetz , jhep * 0704 * , 060 ( 2007 ) . | the recent results from t2k and minos experiments point towards a relatively large value of the reactor angle @xmath0 in the lepton sector . in this paper we show how a large @xmath0 can arise from the charged lepton sector alone in the context of an su(5 ) gut .
in such a scenario ( tri-)bimaximal mixing in the neutrino sector is still a viable possibility .
we also analyse the general implications of the considered scenario for the searches of cp violation in neutrino oscillations . |
it was realized about 25 years ago that the sun , our nearest star , is a variable star .
millions of acoustic normal modes ( p - modes ) of the sun are seen to be excited with a typical surface velocity amplitude of only a few cm s@xmath0 , whereas other pulsating stars have a few modes excited to large amplitudes . considering this dramatic difference between the pulsation property of the sun and other variable stars , it should not be surprising that the solar oscillations are excited by a mechanism that is different from the overstability mechanism believed to be responsible for the pulsation of other stars ( overstability can arise for instance when the radiative flux is converted to mechanical energy of pulsation due to an increase of opacity with temperature ) .
a number of early papers in the field proposed that the solar p - modes are excited by some overstability mechanism ( ulrich 1970 , leibacher and stein 1971 , wolf 1972 , ando and osaki 1975 ) .
however , the margin of instability for solar p - modes is found to be small and different ways of handling radiative transfer and/or the interaction of convection with oscillation seems to change the sign of stability e.g. goldreich and keeley 1977a , antia et al .
1982 & 1988 , christensen - dalsgaard and frandsen 1983 , balmforth & gough 1990 , balmforth 1992 ( the last two papers used a sophisticated version of the mixing length theory of gough , 1977 ) ; see cox et al .
1991 for a more complete list of references .
if we assume that the solar p - modes are overstable then their amplitudes grow exponentially until some nonlinear mechanism limits their growth . by considering all possible 3-mode nonlinear couplings amongst overstable and stable p - modes in the sun , which is the most efficient process for saturating the amplitudes of overstable modes , kumar and goldreich ( 1989 ) and kumar , goldreich and kerswell ( 1991 ) showed that the amplitudes of overstable modes saturate at a value that is several orders of magnitude larger than the observed value
this suggests that solar p - modes are linearly stable .
in this article we will assume that solar p - modes are stable , and describe how they can be excited by acoustic waves generated by turbulent convection .
the basic idea is that the broad band acoustic noise generated by the turbulent flow in the convection zone is selectively amplified at frequencies corresponding to the normal mode frequencies of the sun .
the process of wave generation by homogeneous turbulence was first studied systematically and in some detail by lighthill ( 1952 ) .
stein ( 1967 ) and kulsrud ( 1955 ) applied it to the heating of the solar chromosphere / corona by acoustic and mhd waves respectively .
goldreich and keeley ( 1977b ) carried out a careful calculation of the stochastic excitation of solar normal modes by turbulent convection ( for an excellent general review of wave generation due to turbulent fluid please see crighton 1975 ) .
we describe the stochastic excitation process for the simple case of a homogeneous sphere below and later discuss its generalization to the sun ( 2 ) . in
3 we describe various observations that any theory for excitation of solar p - modes must be able to explain and discuss how well the stochastic excitation theory performs when confronted with these observations .
the estimate for the surface velocity amplitude of low degree g - modes , assuming that they are stochastically excited , is given in 4 .
let us consider a homogeneous gas sphere with a surface that reflects acoustic waves .
some fraction of the fluid inside this sphere is assumed to be in the state of turbulence which acts as a source of sound waves .
following lighthill ( 1952 ) we write the perturbed mass and momentum equations as @xmath1 and @xmath2 where @xmath3 and @xmath4 are unperturbed mean sound speed and density of the medium , @xmath5 is fluid displacement , and @xmath6 these equations can be combined to yield the following inhomogeneous wave equation @xmath7 expanding @xmath8 in the terms of normal modes of the system @xmath9 where @xmath10 is displacement eigenfunction of mode @xmath11 which is normalized to unit energy
@xmath12 and substituting this expansion into equation ( [ 3 ] ) we find the following equation for the mode amplitude @xmath13 @xmath14 turbulent flow is crudely described as consisting of critically damped eddies .
the velocity @xmath15 of an eddy of size @xmath16 is related to the largest or the energy bearing eddy ( size @xmath17 and velocity @xmath18 ) by the kolmogorov scaling _ i.e. _ @xmath19 moreover , following lighthill ( 1952 ) , we take @xmath20 .
since the displacement eigenfunction , for low @xmath21 modes , near the surface of the sphere is in the radial direction , therefore equation ( [ 2 ] ) reduces to @xmath22 where @xmath23 is the radial displacement eigenfunction of mode @xmath11 .
the mean energy input rate in mode @xmath11 can be obtained from the above equation and is given by @xmath24 ^ 2 , \eqno(\new)\ ] ] where @xmath25 and @xmath26 are the velocity and size of the eddies which have characteristic time , @xmath27 , approximately equal to the mode period .
this equation is valid not only for the homogeneous gas sphere considered here but also for more general systems including the excitation of solar p - modes by the reynolds stress as discussed below . of course
, we must use the eigenfunction @xmath28 and turbulent velocity appropriate for the system being considered .
it can be easily shown that the solution of the homogeneous wave equation ( eq . [ 4 ] with right side set equal to zero ) , in the limit of large @xmath29 ( mode order ) is @xmath30 where @xmath31 is spherical bessel function , and @xmath32 is a constant factor independent of mode frequency for properly normalized mode eigenfunction ( condition expressed by eq .
[ 6 ] ) . substituting this into equation ( [ 2 ] )
we find @xmath33 let us assume that the turbulent velocity field in the sphere is concentrated in a thin layer of thickness @xmath17 located near the surface of the sphere .
we shall take the size of the largest eddies to be @xmath17 and their rms speed to be @xmath18 .
the p - modes of period greater than @xmath34 are predominantly excited by the largest size eddies and the resultant energy input rate in these modes is proportional to @xmath35 , as can be seen immediately from the above equation
. modes of higher frequency ( @xmath36 ) couple best to inertial range eddies which have characteristic time of order the wave period . making use of the kolmogorov scaling ( eq . [ 8 ] ) and equation ( ) we see that @xmath37 scales as @xmath38 .
thus the energy input rate into p - modes of this homogeneous system shows a break at frequency @xmath39 where the power law index changes by 7.5 .
the generalization of above equations to describe the excitation of solar modes is not difficult .
equation ( 2 ) is replaced by the linearized momentum equation valid for a stratified medium i.e. @xmath40 and the linerized equation of state is @xmath41 where @xmath42 here @xmath43 denotes the background entropy gradient , and @xmath44 is the entropy fluctuation associated with turbulent convection . equation ( ) is the eulerian version of the statement that the lagrangian entropy perturbation is due entirely to turbulent convection . in other words , we approximate the waves as adiabatic ( these equations are adopted directly from goldreich et al .
1994 ) . combining equations ( 1 ) and ( [ 3])-([1 ] )
we obtain the following inhomogeneous wave equation , which is the generalization of equation ( 4 ) and describes the stochastic excitation of solar oscillations : @xmath45+\bg\bnabla\bcdot\left(\rho\bxi\right ) = - \bnabla\left({\partial p\over\partial s}\tilde{s}\right)-\bnabla\bcdot\left(\rho\bv\bv\right ) .
\eqno(\new)\ ] ] this equation describes wave generation due to reynolds stress as well as entropy fluctuation . as the entropy of a fluid element fluctuates , so does its volume .
the fluctuating volume is a monopole source for acoustic waves . in a stratified medium
the fluctuating buoyancy force adds a dipole source . by transferring momentum among neighboring fluid elements ,
the reynolds stress acts as a quadrupole source .
the anisotropy of a stratified medium blurs the distinction between monopole , dipole , and quadrupole sources .
it allows for destructive interference between the monopole and dipole amplitudes .
although the monopole and dipole amplitudes are individually larger than the quadrupole amplitude , their sum is of comparable size to that of the quadrupole . that this applies to energy bearing eddies follows directly from equations ( ) and the relation between entropy and velocity fluctuation for convective eddies .
the justification for inertial range eddies requires a subtle argument ( cf . goldreich and kumar , 1990 ) . the new energy equation ( which replaces eq .
[ 10 ] ) is given below @xmath46 where @xmath47 is wave compressibility defined by @xmath48 the shape parameter @xmath49 describes the ratio of the horizontal to vertical correlation lengths of turbulent eddies , and @xmath50 is given by @xmath51 with @xmath52 the mixing length and @xmath17 the pressure scale height .
the factor @xmath53 measures the ratio of the excitation by entropy fluctuations to that due to fluctuating reynolds stress .
note that the frequency spectra of waves excited due to entropy fluctuation and that due to reynolds stress are identical except of course for an over all normalization factor .
the compressibility , @xmath47 , for p - modes near the top of the convection zone , where the excitation takes place , is close to 1 , and the value of @xmath54 in this region is about 10 ( see goldreich et al .
therefore , the excitation of p - modes is dominated by entropy fluctuations ( stein & nordlund 1991 , arrived at the same conclusion using their numerical simulations of solar convection ) . on the other hand
f - modes are nearly incompressible ( @xmath55 ) and so they are not excited by entropy fluctuation .
this is perhaps why the power in f - modes is observed to be smaller than a p - mode of similar frequency .
the observed rate of energy input into solar p - modes can now be readily understood .
one of the differences between the homogeneous gas sphere system and the sun is in the shape of the eigenfunction especially near the surface where the wave excitation takes place .
it can be shown that the radial derivative of the normalized radial displacement eigenfunction for p - modes just below the photosphere scales as @xmath56 for @xmath57mhz and as @xmath58 for @xmath59mhz ( see figure 1 ) .
substituting this scaling in equation ( [ 3 ] ) , or equation ( 10 ) , we find that the energy input rate in the p - modes , at a fixed degree , scales as @xmath60 for @xmath57mhz and as @xmath61 for @xmath59mhz which is in good agreement with the observations ( libbrecht & woodard 1991 ) ; please see goldreich et al .
( 1994 ) for a more detailed analysis and comparison with the observed energy input rate .
a valid theory for the excitation of solar p - modes should be able to explain the observed rate of energy input in different modes . in addition , there are four other observational results that the theory must be able to explain and provide a fit to the data .
these observations are : mode linewidth , the deviation of p - mode line profiles from symmetric lorentzian shape , the statistics for the fluctuation of mode energy , and the presence of peaks in the power spectrum above the acoustic cutoff frequency ( @xmath62mhz ) .
the agreement between the observed and the theoretically calculated energy input rate in solar p - modes due to stochastic excitation was described in the last section .
we describe below the other four observations and compare them with the results of the stochastic excitation theory .
a number of groups have measured p - mode linewidth as a function of mode frequency ( cf .
duvall et al .
1988 , libbrecht 1988 , elsworth et al .
it is found that the mode linewidth at a fixed degree increases monotonically with mode frequency . at 2 mhz
the linewidth is about 0.5 @xmath63hz , or mode lifetime is 20 days ( @xmath64 ) and at 4 mhz the linewidth is 10 @xmath63hz .
the observed linewidth for 2mhz@xmath65mhz increases as @xmath66 whereas numerical calculations show that the mode linewidth due to radiative and turbulent damping increases as @xmath67 for frequencies below @xmath684mhz ( cf .
christensen - dalsgaard & frandsen 1983 , balmforth 1992 ; goldreich and kumar 1991 , give a simple analytical derivation of these results ) .
this suggests that the mode damping at frequencies greater than about 2 mhz is due to some process other than the radiative and turbulent viscosity .
a number of alternate mechanisms have been suggested to account for the observed mode linewidth .
these include modulation of convective flux ( christensen - dalsgaard et al . 1989 ) , scattering of low degree p - modes by turbulent convection to high degree modes ( goldreich and murray 1994 ) , scattering of p - modes by magnetic fields ( bogdan et al . 1996 ) .
goldreich and murray ( 1994 ) have carried out a detailed calculation of the scattering process and find that an almost elastic scattering of p - modes by convective eddies is an important contributor to the mode linewidth at frequencies @xmath69mhz , and the computed linewidth has the same frequency dependence as the observed width ( see also murray , 1993 ) .
recent observational results indicate that the linewidth scales as @xmath70 for @xmath71 mhz ( chaplin et al .
1996 , and tomczyk 1996 ) .
perhaps below 2 mhz there are few modes available for p - modes to scatter into , and thus the linewidth falls off more rapidly with decreasing frequency .
according to jefferies ( personal communication ) the observed mode linewidth peaks at a frequency of about 5mhz , followed by a slight decline , and then remains constant at higher frequencies .
this is a puzzling result for which as far as i know no explanation has been offered .
modes excited by their interaction with a gaussian random field ( turbulent convection ) have fluctuating amplitudes that follows the gaussian distribution .
the correlation time for mode amplitude , which is of order the mode lifetime , is typically much larger than the mode period ( see 3.1 ) or the characteristic time of resonant eddies .
this is because a mode interacts with a large number of eddies each of which contribute only a small fraction of the total energy in the mode .
a good analogy is a pendulum placed in contact with a thermal heat bath of molecules .
the mean energy in the pendulum is one third the mean kinetic energy of molecules , however it takes a large number of collisions ( of order the ratio of the pendulum mass to molecular mass ) in order for the amplitude of the pendulum to change .
the statistics of energy fluctuation in a solar p - mode , if stochastically excited , like the energy of the pendulum placed in a heat bath , follows boltzmann distribution ( see kumar et al .
1988 for a rigorous derivation of this result ) .
at least two different groups ( toutain & frohlich 1992 ; and elsworth et al . 1995 ) have looked for the statistics of energy fluctuation in the solar p - modes and find it to be in good agreement with the theoretical expectation for stochastic excitation i.e. boltzmann or exponential distribution ( see fig .
acoustic waves of frequency less than about 5 mhz ( the acoustic cutoff frequency at the temperature minimum ) are reflected at the solar photosphere and thus trapped inside the sun .
the reflectivity however drops off rapidly at higher frequencies ; at 6mhz about 2% of the incident wave energy is reflected at the photosphere whereas at 7mhz the reflectivity drops to less than 0.3% .
a number of observations indicate that high frequency acoustic waves ( waves of frequency greater than about 5mhz ) suffer little reflection at the chromosphere / corona as well ( duvall et al . 1993 , kumar et al .
1994 , jefferies 1996 ) .
if high frequency acoustic waves were significantly reflected at the chromosphere / corona boundary then the frequency spacing between modes of adjacent order would fluctuate with mode frequency ( see figure 3 ) ; this is because of the interference of waves partially trapped between two cavities above and below the temperature minimum .
the observations , however , show , no evidence for such a behavior ( fig .
3 ) and thus provide an upper limit of about 10% to the reflection at the chromosphere / corona boundary ( kumar et al .
1991 ; jefferies , personal communication ) . in the absence of wave reflection at the solar surface
these high frequency acoustic waves are not trapped in the sun , and thus it was expected that the power spectrum above the acoustic cutoff frequency should be featureless i.e. devoid of peaks .
however , the observed spectra contain very regular peaks that are seen upto the nyquist frequency of observations .
one of the best recent data set obtained at the south pole in 1994 shows peaks extending to almost 11 mhz which makes the length of the spectrum above the acoustic cutoff frequency larger than the observed spectrum below the cutoff frequency !
the existence of these high frequency peaks provides one of the strongest evidence that solar acoustic oscillations above 5 mhz are not excited by some overstability mechanism - folding time must be less than about an hour , and thus these waves can at best be amplified by a factor of @xmath72 as they make one passage through the solar interior .
thus we need a mechanism that provides a large seed amplitude , within a factor of @xmath73 of the observed value , and clearly in this case it seems most natural that the same mechanism generates the full observed amplitude . ] , and since power spectrum varies smoothly from below the acoustic cutoff frequency to above the cutoff frequency we infer that the trapped p - modes in the sun are also not overstable ( kumar et al . 1989 ) . the peaks at high frequencies can be understood as arising quite naturally if waves are stochastically excited .
these peaks form because of the constructive interference between waves propagating from the source ( located in the convection zone ) upward to the photosphere and waves traveling downward from the source that is refracted back up due to increasing sound speed and thus end up at the photosphere ( kumar and lu 1991 ) .
therefore the frequencies of peaks above the acoustic cutoff ( @xmath74 mhz for the sun ) depend on the difference between these two paths or in other words on the depth of acoustic sources .
a good fit to the high frequency power spectrum is obtained by placing sources ( assumed to be quadrupole ) approximately 140 km below the photosphere ( kumar 1994 ) ; please see figure 4 .
it should be emphasized that unlike the lower frequency trapped p - modes ( frequency less than about 5 mhz ) the frequencies of peaks at high frequencies is not a property of the equilibrium model of the sun alone but depends in a sensitive way on the location of sources that excite these oscillations . as discussed in kumar ( 1994 ) if the acoustic sources are assumed to be dipolar instead , then no matter where these sources are placed in the solar convection zone they do not provide a fit to the observed power spectrum .
this suggests that the acoustic sources , at least for the high frequency waves , are not dipole but quadrupole , which is consistent with the work of goldreich et al .
( 1994 ) .
high frequency acoustic waves also provide information about the power spectrum of turbulent convection in the sun ( kumar 1994 ) ; we can constrain the spectrum of turbulent convection in the region where acoustic emission is significant . the theoretical power spectrum shown in figure 4
was computed using the kolmogorov power spectrum of turbulence , _ i.e. _ , @xmath75 .
evidently , this provides a good fit to the observed spectrum between 5.5 and 10 mhz . in order to determine the power law index @xmath76 of solar turbulence , @xmath77 , from the high frequency interference peaks
, we relate the fluctuating velocity , @xmath15 , of sub - energy bearing eddies to the velocity @xmath18 of scale - height size eddies as follows : @xmath78 where @xmath16 is the size of the eddy .
this equation is a generalization of equation ( 8) . using this relation , we find that the frequency dependence of the source function is @xmath79 ( the derivation is similar to the one leading to eq . [ 12 ] ) .
therefore , a change in the spectral index of turbulence from 5/3 to 1.4 decreases the dependence of the acoustic power spectrum on frequency by @xmath80 , which results in a poor fit to the observed spectrum .
we find that the observed high frequency power spectra suggest that the power law index for the solar turbulence lies between 1.5 and 1.7 .
we note that the energy input rate for p modes in the frequency range between 3.5 and 5mhz is proportional to @xmath81 , which is understood most naturally if the spectrum of turbulence near the top of the solar convection zone is taken to be kolmogorov ( goldreich et al .
the result described above extends this range to 10mhz .
the last observational result i would like to describe is the asymmetry of low frequency p - mode line profiles .
the spectrum of individual p - modes is fitted very well by a lorentzian profile
. however , duvall et al . (
1993 ) discovered that the line profiles do not have perfect lorentzian shape and in particular the power spectrum falls off more rapidly on one side of the peak than the other i.e. the lines are asymmetrical .
the data from gong and soho clearly show that lineprofiles for low frequency modes are asymmetrical .
duvall et al .
had also proposed in their original paper an explanation for why the lines are asymmetrical which is found to be basically correct by a number of independent investigations ( gabriel 1992 , 1995 ; abrams & kumar 1996 ) .
the lineprofile for a p - mode of frequency 1.9 mhz and @xmath82 , calculated using jc - d solar model , is shown in figure 5 .
the physical explanation for line asymmetry is simplest when sources lie in the region of the sun where acoustic waves can propagate ( this case does not seem to apply to solar oscillations however which are excited by sources that lie in the evanescent region ) . consider a source lying close to the node of a p - mode . as the frequency of acoustic waves
is varied in the neighborhood of this p - mode frequency the position of the node changes with respect to the source position .
thus waves of frequencies lying symmetrically on the opposite side of the p - mode frequency gets excited to different amplitudes making the resultant power spectrum asymmetrical .
it is clear from this rather simple example that not all p - mode line profiles are expected to be equally asymmetrical ( as is observed ) and also that the degree of asymmetry depends on the location of sources .
in fact duvall et al ( 1993 ) had recognized this in their original paper and used this to determine the depth of sources that are exciting p - modes .
the recent paper of abrams and kumar ( 1996 ) uses a realistic solar model due to christensen - dalsgaard to calculate the p - mode power spectrum and finds that in order to reproduce the magnitude of asymmetry observed by duvall et al .
( 1993 ) the sources responsible for exciting low frequency p - modes should lie about 250 km below the photosphere .
this might appear to be in conflict with the result obtained using high frequency solar oscillations described in 3 . however , the result is in agreement with the theory of stochastic excitation which predicts that lower frequency oscillations are excited deeper in the convection zone where the characteristic eddy time is longer .
line asymmetry causes a slight error to the observational determination of p - mode frequencies which are obtained by fitting a lorentzian function to the power spectra .
abrams and kumar ( 1996 ) find that this frequency error is proportional to the product of mode linewidth and a nondimensional measure of lineasymmetry @xmath83 ( see the figure below ) .
the parameter @xmath83 is obtained by decomposing the observed power spectrum in the neighborhood of a peak corresponding to a mode @xmath76 into even and odd functions . since the odd function is zero at the peak ( by definition ) and again far from the peak , its magnitude has a maximum at some intermediate distance from the peak , typically less than one linewidth .
the ratio of the maximum magnitude of the odd function to the maximum magnitude of the even function is a dimensionless measure of the asymmetry which we denote by @xmath84 i.e. @xmath83 is the percentage line asymmetry of mode @xmath76 .
the sign of @xmath83 is taken to be positive or negative according to whether there is more power on the high- or low - frequency side of the peak , respectively .
there is one feature of the observed line asymmetry that is very puzzling and for which there is no theoretical explanation .
duvall et al .
( 1993 ) reported that the sense of asymmetry reverses in the velocity and the intensity power spectra for the p - mode .
this behavior has been confirmed by the most recent gong data .
the difference between the velocity and the intensity power spectra can arise as a result of line formation in the presence of oscillations .
however , it is not clear what process can cause a reversal of the sign of @xmath83 in the two spectra ; the process has to be extremely frequency sensitive so that it can modify the spectrum in an interval of only a few @xmath63hz .
gravity mode oscillations of the sun are primarily confined to its radiative interior and their observation would thus provide a wealth of information about the energy generating region which is poorly probed by the p - modes . in the past 20 years a number of different groups have claimed to detect g - modes in the sun ( e.g. brookes et al . 1976 ; brown et al . 1978 ; delache & scherrer 1983 ; scherrer et al . 1979 ; severny et al . 1976 , thomson et al . 1995 ; for a detailed review of the observations please see the article by pall , 1991 , and references therein ) , but thusfar there is no consensus that g - modes have in fact been observed .
one of the objectives of the instruments aboard soho ( virgo , golf and soi ) is to search for solar g - modes .
so it should be helpful to estimate the expected surface velocity amplitudes of g - modes in the sun , providing observations and data analysis programs a rough target number to shoot for .
the first question we need to address in this respect is whether solar g - modes are self - excited ( overstable ) or not .
this is dealt with in the following paragraph .
a number of people have investigated the linear stability of solar g - modes ( e.g. dilke & gough 1972 ; rosenbluth & bahcall 1973 ; christensen - dalsgaard et al .
1974 ; shibahashi et al . 1975 ; boury et al . 1975 ; saio 1980 ) .
all of these investigations find that g - modes of radial - order ( @xmath29 ) greater than 3 are stable . however , there is no general agreement about the stability of low order modes ( @xmath85 ) .
if overstable , the g - mode amplitude will increase exponentially with time until nonlinear effects become important and saturate their growth .
kumar and goodman ( 1995 ) have recently investigated 3-mode parametric interaction , a very efficient nonlinear process . using their results
we find that the low order overstable g - modes in the sun will attain an energy of at least 10@xmath86 erg before they are limited by nonlinearities .
the velocity at the solar surface corresponding to this energy is @xmath87 cm s@xmath0 , which is an order of magnitude larger than the observational limit of pall ( 1991 ) .
thus even low order g - modes of frequency greater than about 150 @xmath63hz are unlikely to be overstable
. however , g - modes can be stochastically excited .
a number of people have estimated g - mode amplitude assuming that they are linearly stable and stochastically excited .
keeley ( 1980 ) applied his theory , depeloped with goldreich in 1977 , of the excitation of solar modes to estimate the amplitude of the 160 minute oscillation , and found the theoretical amplitude to be much smaller than claimed by the observations ; much more sensitive observational searches since then have not detected this oscillation ( cf .
pall 1991 ) .
gough ( 1985 ) carried out an application of the energy partition result of goldreich and keeely ( 1977 ) to solar g - modes and estimated the surface velocity amplitude of low @xmath29 and @xmath21 g - modes ( @xmath85 , @xmath88 ) to be about 1 - 2 mm s@xmath0 .
kumar et al .
( 1996 ) estimated the g - mode amplitude using the recent theoretical work of goldreich et al .
( 1994 ) on stochastic excitation of waves , which reproduces the observed energy input rate into solar p - modes of all frequencies ( see 2 ) , and taking into account the radiative and viscous turbulent dampings .
they find the surface velocity amplitude of low order g - modes to be about 0.4 mm s@xmath0 ( see figure 7 ) .
recently anderson has carried out numerical simulation of g - mode excitation as a result of turbulent flow associated with penetrative convection .
he finds that the transverse surface velocity amplitudes of g - modes of degree about 6 is @xmath68 0.2 mm s@xmath0 in the case when he assumes that 10@xmath89 modes are excited by this process ( anderson , 1996 ) .
thus several different calculations suggest that the amplitudes for low degree g - modes are of order 0.5 mm s@xmath0 .
the uncertainty in this estimate is at least a factor of a few . if the nature turns out to be cooperative and the actual amplitudes of solar g - modes are a factor of a few larger than these estimates , then instruments aboard soho have a good chance of detecting g - modes and opening up a new window in the study of the solar core .
i am grateful to peter goldreich for collaboration over a number of years and for his insights on the topic discussed in this review .
i thank joergen christensen - dalsgaard for the use of his solar model , and for pointing out several references .
i am indebted to tom duvall , jack harvey and stuart jefferies for sharing with me their excellent data from the south pole .
this work was supported by a nasa grant nagw-3936 .
abrams , d. , & kumar , p. , 1996 , apj , to appear in nov . | the stochastic excitation of solar oscillations due to turbulent convection is reviewed .
a number of different observational results that provide test for solar p - mode excitation theories are described .
i discuss how well the stochastic excitation theory does in explaining these observations .
the location and properties of sources that excite solar p - modes are also described .
finally , i discuss why solar g - modes should be linearly stable , and estimate the surface velocity amplitudes of low degree g - modes assuming that they are stochastically excited by the turbulent convection in the sun . |
let @xmath1 be a closed riemannian @xmath0-manifold , and let @xmath2 be a morse function , i.e. @xmath3 is a smooth function , all of whose critical points are non - degenerate , and for which distinct critical points have distinct images in @xmath4 .
we define the _ area _ of @xmath3 to the maximum area of any level set @xmath5 over all points @xmath6 .
we define the _ morse area _ of @xmath1 to be the infimum of the area of all morse functions @xmath7 . for hyperbolic @xmath0-manifolds ,
the hyperbolic metric is a topological invariant by mostow rigidity , and the critical points of a morse function determine a handle decomposition of the manifold , so one might hope that morse area is related to a topological measure of complexity defined in terms of handle decompositions of the manifold .
we show that morse area is linearly related to a definition of topological complexity we call _ scharlemann - thompson width _ or _ linear width _ , and which we now describe . for a closed ( possibly disconnected ) surface @xmath8
, we define the complexity , or genus , of @xmath8 to be the sums of the genera of each connected component . for a compact ( possibly disconnected ) surface with boundary , we define the genus of @xmath8 to be the genus of the surface obtained by capping off all boundary curves with discs . we shall write @xmath9 for the number of boundary components of @xmath8 .
a _ handlebody _ is a compact @xmath0-manifold with boundary , homeomorphic to the regular neighborhood of a graph in @xmath10 . up to homeomorphism
, a handlebody is determined by the genus @xmath11 of its boundary surface .
every @xmath0-manifold @xmath1 has a _
heegaard splitting _ , which is a decomposition of the manifold into two handlebodies .
this immediately gives a notion of complexity for a @xmath0-manifold , called the _ heegaard genus _
, which is the smallest genus of any heegaard splitting of the @xmath0-manifold .
there is a refinement of this , due to scharlemann and thompson @xcite , which we now describe .
let @xmath8 be a closed surface , which need not be connected . a _ compression body _
@xmath12 is a compact @xmath0-manifold with boundary , constructed by attaching some number of @xmath13-handles to one side @xmath14 of @xmath15 .
we do not require compression bodies to be connected .
we shall refer to @xmath16 as the _ top boundary _
@xmath17 of the compression body , and the other boundary components of @xmath12 as the _ lower boundary _ @xmath18 .
the lower boundary may be disconnected , even if @xmath12 is connected , and any @xmath13-sphere components are capped off with @xmath0-balls .
in particular , if a maximal number of non - parallel @xmath13-handles are attached , then the resulting compression body is a handlebody , so a handlebody is a special case of a compression body .
a _ generalized heegaard splitting _ , which we shall call a _
linear splitting _ , is a decomposition of a closed @xmath0-manifold @xmath1 into a linearly ordered sequence of compression bodies @xmath19 , such that the upper boundary of an odd numbered compression body @xmath20 is equal to the top boundary of the compression body @xmath21 , and the lower boundary of @xmath20 is equal to the lower boundary of the previous compression body @xmath22 . for the even numbered compression bodies @xmath22 , the top boundary is equal to the upper boundary of @xmath23 , and the lower boundary is equal to the lower boundary of @xmath20 . in the case of the first and last compression bodies
@xmath24 and @xmath25 , the lower boundaries are empty .
let @xmath26 be the sequence of surfaces consisting of the upper boundaries of the compression bodies @xmath23 and @xmath22 .
the complexity @xmath27 of the surface @xmath26 is the genus of @xmath26 , i.e. the sum of the genera of each connected component , and the _ width _ of the linear splitting is the maximum value of @xmath27 over all upper boundaries .
the _ scharlemann - thompson width _
, which we shall also refer to as the _ linear width _ , of a @xmath0-manifold @xmath1 is the minimum width over all possible linear splittings .
as a heegaard splitting is a special case of a linear splitting , the heegaard genus of @xmath1 is an upper bound for the linear width of @xmath1 .
there is a refinement of linear width known as _ thin position _ , which we discuss when we use it in section [ section : flow ] . in order to bound morse area in terms of linear width we shall assume the following result announced by pitts and rubinstein @xcite ( see also rubinstein @xcite ) .
@xcite [ conjecture : pr ] let @xmath1 be a riemannian @xmath0-manifold with a strongly irreducible heegaard splitting .
then the heegaard surface is isotopic to a minimal surface , or to the boundary of a regular neighborhood of a non - orientable minimal surface with a small tube attached vertically in the i - bundle structure .
a full proof of this result has not yet appeared in the literature , though recent progress has been made by colding and de lellis @xcite , de lellis and pellandrini @xcite , and ketover @xcite .
we shall show : [ theorem : morse ] there is a constant @xmath28 , such that for any closed hyperbolic @xmath0-manifold , @xmath29 where the right hand bound holds assuming theorem [ conjecture : pr ] .
our methods are effective , and the constant @xmath30 may be estimated using a bound on the margulis constant for @xmath31 , though we omit the details of this calculation , as our methods seem unlikely to give an optimal constant . in section [ section : metric bounds topology ] , we show how to bound linear width in terms of morse area .
a bound on the morse area of @xmath1 gives a morse function @xmath7 , with bounded area level sets , but with no _ a priori _ bound on the topological complexity of the level sets .
we use a voronoi decomposition of @xmath1 to give a polyhedral approximation of the morse function , which we now describe in a simple case .
let @xmath32 be a voronoi decomposition of @xmath1 in which every voronoi cell @xmath33 is a topological ball , and has size bounded above and below , i.e. there is an @xmath34 such that @xmath35 , where @xmath36 is the center of the voronoi cell .
let @xmath37 be the sublevel set of the morse function , i.e. @xmath38)$ ] .
the sublevel sets are a monotonically increasing collection of subsets of @xmath1 , which start off empty , and eventually contain all of @xmath1 , so in particular , for each voronoi cell @xmath33 there is a @xmath39 such that the volume of @xmath40 is exactly half the volume of @xmath41 , and we shall call @xmath39 the _ cell splitter _ for the voronoi cell @xmath33 . furthermore , we may assume that the @xmath39 are distinct for distinct voronoi cells .
this gives a linear order to the voronoi cells , and we wish to show that constructing the manifold by adding the voronoi cells in this order gives a bounded linear width handle decomposition for @xmath1 .
let @xmath42 be the union of the voronoi cells whose cell splitters @xmath39 are at most @xmath43 .
each voronoi cell is a ball , with a bounded number of faces , so adding a voronoi cell corresponds to adding a bounded number of handles .
it remains to show that the boundary of each @xmath42 has genus bounded in terms of the area of the level set @xmath44 .
let @xmath33 and @xmath45 be two adjacent voronoi cells , with @xmath33 contained in @xmath42 and @xmath45 outside @xmath42 , so their common face is a subset of @xmath46 . consider the sequence of balls @xmath47 , as @xmath48 runs along the geodesic from @xmath36 to @xmath49 .
at least half the volume of @xmath41 is contained in @xmath37 , and at most half the volume of @xmath50 is contained in @xmath37 , so there is an @xmath48 such that exactly half the volume of @xmath51 is contained in @xmath37 , and so there is a lower bound on the area of @xmath52 .
therefore , a bound on the area of @xmath53 gives a bound on the number of faces of @xmath46 . as each face has a bounded number of edges , this gives a bound on the genus of @xmath46 , and hence a bound on the linear width of @xmath1 , though this bound depends on @xmath54 . in order to produce a bound which works for any compact hyperbolic manifold @xmath1
, we use the margulis lemma and the thick - thin decomposition for hyperbolic manifolds .
there is constant @xmath55 , called a margulis constant , such that any compact hyperbolic manifold may be decomposed into a thick part @xmath56 , where each point has injectivity radius greater than @xmath55 , and a thin part , where each point has injectivity radius at most @xmath55 , and which is a disjoint union of solid tori .
if we choose @xmath54 sufficiently small , then we may choose a voronoi decomposition of the thick part in which each voronoi cell has size bounded above and below , and run the argument in the previous paragraph to control the genus of @xmath46 inside the thick part .
we do not control the complexity of @xmath46 in the thin part , but as each component of the thin part is a solid torus , we may cap off @xmath57 with surfaces parallel to @xmath58 , while still obtaining bounds on the genus . in order to bound the number of handles corresponding to adding a voronoi cell , we use a result of kobayashi and rieck @xcite which gives bounds on the topological complexity of the intersection of a voronoi cell with the thin part .
the key problem for the upper bound is that the techniques of pitts and rubinstein use sweepouts , so although their minimax construction produces a sweepout of bounded area , we do not know how to directly replace a bounded area sweepout with a bounded area foliation .
however , the upper bound is obtained in recent work of colding and gabai @xcite , using work of colding and minicozzi @xcite on the mean curvature flow , and we describe their results in section [ section : flow ] .
the authors would like to thank dick canary , david futer , david gabai , joel hass , daniel ketover , sadayoshi kojima , yair minksy , yoav rieck and dylan thurston for helpful conversations , and the tokyo institute of technology for its hospitality and support .
the second author was supported by the simons foundation cgm 234477 and psc cuny tradb-45 - 17 .
in this section we show that we can bound the scharlemann - thompson width of a hyperbolic manifold in terms of its morse area . we will approximate level sets by surfaces which are unions of faces of voronoi cells , and we start by describing the properties of the voronoi decompositions that we will use .
we will approximate the level sets of @xmath3 by surfaces consisting of faces of voronoi cells .
we now describe in detail the voronoi cell decompositions we shall use , and their properties . a _ polygon _ in @xmath31 is a compact convex subset of a hyperbolic plane whose boundary consists of a finite number of geodesic segments .
a _ polyhedron _ in @xmath31 is a convex topological @xmath0-ball in @xmath31 whose boundary consists of a finite collection of polygons .
a _ polyhedral cell decomposition _ of @xmath31 is a cell decomposition in which which every @xmath0-cell is a polyhedron , each @xmath13-cell is a polygon , and the edges are all geodesic segments .
we say a cell decomposition of a hyperbolic manifold @xmath1 is _ polyhedral _ if its preimage in the universal cover gives a polyhedral cell decomposition of @xmath31 .
let @xmath59 be a discrete collection of points in @xmath0-dimensional hyperbolic space @xmath31 .
the voronoi cell @xmath33 determined by @xmath60 consists of all points of @xmath1 which are closer to @xmath36 than any other @xmath61 , i.e. @xmath62 we shall call @xmath36 the _ center _ of the voronoi cell @xmath33 , and we shall write @xmath63 for the collection of voronoi cells determined by @xmath64 .
voronoi cells are convex sets in @xmath31 , and hence topological balls .
the set of points equidistant from both @xmath36 and @xmath49 is a totally geodesic hyperbolic plane in @xmath31 .
@xmath65 of the voronoi decomposition consists of all points which lie in two distinct voronoi cells @xmath33 and @xmath45 , so @xmath65 is contained in a geodesic plane .
an _ edge _
@xmath66 of the voronoi decomposition consists of all points which lie in three distinct voronoi cells @xmath67 and @xmath68 , which is a geodesic segment , and a _ vertex _
@xmath69 is a point lying in four distinct voronoi cells @xmath70 and @xmath71 . by general position
, we may assume that all edges of the voronoi decomposition are contained in exactly three distinct faces , the collection of vertices is a discrete set , and there are no points which lie in more than four distinct voronoi cells .
we shall call such a voronoi decomposition a _
regular _ voronoi decomposition , and it is a polyhedral decomposition of @xmath31 if every cell is compact . as each edge is @xmath0-valent , and each vertex is @xmath72-valent , this implies that the dual cell structure is a simplicial triangulation of @xmath31 , which we shall refer to as the _
dual triangulation_. the dual triangulation may be realised in @xmath31 by choosing the vertices to be the centers @xmath36 of the voronoi cells and the edges to be geodesic segments connecting the vertices , and we shall always assume that we have done this . in this case the triangles and tetrahedra
are geodesic triangles and geodesic tetrahedra in @xmath31 . given a collection of points @xmath59 in a hyperbolic @xmath0-manifold @xmath1 ,
let @xmath73 be the pre - image of @xmath64 in the universal cover of @xmath1 , which is isometric to @xmath31 .
we say a subset of @xmath31 is _ equivariant _ if it is preserved by the covering translations determined by the quotient @xmath1 . as @xmath73 is equivariant , the @xmath74-skeleton of the corresponding voronoi cell decomposition @xmath32 of @xmath31 is also equivariant , for @xmath75 , as are the @xmath74-skeletons of the dual triangulation .
we now show that the interior of each voronoi cell @xmath76 is mapped down homeomorphically by the covering projection .
suppose @xmath77 is a point in the interior of a voronoi cell @xmath76 with center @xmath48 , so @xmath78 , for any other @xmath79 .
let @xmath11 be a covering translation , which is an isometry , so @xmath80
. as covering translations act freely , this implies that @xmath81 lies in the interior of the voronoi cell corresponding to @xmath82 .
therefore @xmath83 has disjoint translates under the group of covering translations , and so is mapped down homeomorphically into @xmath1 , though the covering projection may identify distinct faces of a voronoi cell under projection into @xmath1 . by abuse of notation , we shall refer to the resulting polyhedral decomposition of @xmath1 as the voronoi decomposition @xmath32 of @xmath1 . by general position
, we may assume that @xmath32 is regular .
the dual triangulation also projects down to a triangulation of @xmath1 , which we will also refer to as the dual triangulation , though this triangulation may no longer be simplicial .
we say a collection @xmath59 of points in @xmath1 is _ @xmath84-separated _ if the distance between any pair of points is at least @xmath84 , i.e. @xmath85 , for all @xmath86 .
let @xmath1 be a compact hyperbolic @xmath0-manifold .
we say a voronoi decomposition @xmath32 is @xmath54-regular , if it is regular , and it arises from a maximal collection of @xmath54-separated points .
we shall write @xmath87 for the closed metric ball of radius @xmath88 about @xmath48 in @xmath1 , @xmath89 which need not be a topological ball . as the cells of an @xmath54-regular voronoi decompositions are determined by a maximal collection of @xmath84-separated points in @xmath1 , each voronoi cell
is contained in a metric ball of radius @xmath54 about its center .
furthermore , as the points @xmath36 are distance at least @xmath54 apart , each voronoi cell contains a metric ball of radius @xmath90 about its center , i.e. @xmath91 one useful property of @xmath54-regular voronoi decompositions is that the boundary of any union of voronoi cells is an embedded surface , in fact an embedded normal surface in the dual triangulation , as we now describe .
a _ simple arc _ in the boundary of a tetrahedron is a properly embedded arc in a face of the tetrahedron with endpoints in distinct edges .
a _ triangle _ in a tetrahedron is a properly embedded disc whose boundary is a union of three simple arcs , and a _ quadrilateral _ is a properly embedded disc whose boundary is the union of four simple arcs .
a _ normal surface _ in a triangulated @xmath0-manifold is a surface that intersects each tetrahedron in a union of normal triangles and quadrilaterals .
let @xmath1 be a compact hyperbolic manifold , and let @xmath32 be an @xmath54-regular voronoi decomposition .
let @xmath92 be a union of voronoi cells in @xmath32 , and let @xmath8 be the boundary of @xmath92 . then @xmath8 is an embedded surface in @xmath1 .
the collection of voronoi cells @xmath92 intersects a tetrahedron @xmath93 in the dual triangulation in a regular neighborhood of the vertices of @xmath93
. if a tetrahedron @xmath93 has one or three vertices corresponding to voronoi cells in @xmath92 , then @xmath8 intersects @xmath93 in a single normal triangle .
if @xmath93 has exactly two vertices corresponding to voronoi cells in @xmath92 , then @xmath8 intersects @xmath93 in a single normal quadrilateral .
therefore @xmath8 consists of at most one triangle or quadrilateral in each tetrahedron , and so is an embedded normal surface .
we shall write @xmath94 for the injectivity radius of @xmath1 at @xmath48 , i.e. the radius of the largest embedded ball in @xmath1 centered at @xmath48 .
we shall write @xmath95 for the injectivity radius of @xmath1 , which is defined to be @xmath96 we shall say a voronoi cell @xmath33 with center @xmath36 is a _ deep _ voronoi cell if the injectivity radius at @xmath36 is at least @xmath97 , i.e. @xmath98 , and in particular this implies that the metric ball @xmath99 is a topological ball .
we shall also call centers , faces , edges and vertices of deep voronoi cells deep .
we shall write @xmath100 for the subset of @xmath32 consisting of deep voronoi cells .
we now show that there are bounds , which only depend on @xmath84 , on the number of faces of a deep voronoi cell , and the number of edges and faces of a deep voronoi cell .
[ prop : bound ] let @xmath1 be a compact hyperbolic @xmath0-manifold with an @xmath54-regular voronoi decomposition @xmath32 , and let @xmath100 be the collection of deep voronoi cells . then there is a number @xmath101 , which only depends on @xmath54 , such that each deep voronoi cell @xmath102 has at most @xmath101 faces , edges and vertices . let @xmath103 be a voronoi cell with center @xmath48 , and with faces @xmath104 .
let @xmath36 be the center of the voronoi cell @xmath105 adjacent to the face @xmath106 .
as @xmath103 is interior , @xmath105 also lies in @xmath100 .
if two voronoi cells share a common face , then the distance between their centers is at most @xmath107 .
therefore all of the centers of the voronoi cells corresponding to the faces of @xmath76 are contained in the metric ball @xmath108 .
this implies that the the balls of radius @xmath90 around the @xmath36 are contained in the metric ball @xmath109 .
as the @xmath110 are all disjoint , this implies that the number of faces is at most @xmath111 note that @xmath112 is also an upper bound for the maximum number of edges in any face of a voronoi cell , because every edge of that face is contained in another face in that cell .
so the total number of edges is at most @xmath113 , and by the formula for euler characteristic , the number of vertices is at most @xmath114 .
therefore we may choose @xmath101 to be @xmath115 .
an similar volume bound argument to the one above proves the following : [ prop : bound2 ] let @xmath1 be a compact hyperbolic @xmath0-manifold with an @xmath54-regular voronoi decomposition @xmath32 . then there is a number @xmath116 , which depends only on @xmath54 , such that for any deep voronoi center @xmath36 , the number of voronoi centers contained in @xmath99 is at most @xmath116 .
we may choose a morse function @xmath7 , such that the complexity of @xmath3 is within some small @xmath117 of the infimum , i.e. @xmath118 for all @xmath119 .
we now describe how to use the morse function @xmath3 to give a linear ordering to the voronoi cells in @xmath32 .
let @xmath1 be a compact hyperbolic @xmath0-manifold , and let @xmath120 be a morse function .
given @xmath119 define the _ sublevel set of @xmath1 at @xmath43 _ , which we shall denote @xmath37 , to be the subset of @xmath1 consisting of the union of all level sets @xmath53 with @xmath121 $ ] , i.e. @xmath122 ) .\ ] ] for @xmath43 sufficiently small , @xmath37 is the empty set , and for @xmath43 sufficiently large @xmath37 is equal to all of @xmath1 .
the region @xmath37 varies continuously in @xmath43 , and is monotonically increasing in @xmath43 .
[ definition : splitter ] let @xmath1 be a compact hyperbolic @xmath0-manifold with an @xmath54-regular voronoi decomposition @xmath32 .
let @xmath7 be a morse function .
for each voronoi cell @xmath33 with center @xmath36 there is a unique @xmath123 such that the surface @xmath124 divides the metric ball @xmath110 exactly in half by volume , i.e. @xmath125 .
we call this @xmath39 the _ cell splitter _ of @xmath33 .
[ definition : generic ] we say that a morse function @xmath7 is _ generic _ with respect to a voronoi decomposition @xmath126 if the cell splitters for distinct voronoi cells @xmath33 correspond to distinct points @xmath123 , and no cell splitter is also a critical point for the morse function .
we say a point @xmath119 is _ generic _ if it is not a critical point for the morse function , and is not a cell splitter .
we may assume that @xmath3 is generic by an arbitrarily small perturbation of @xmath3 , and we shall always assume that @xmath3 is generic from now on . let @xmath1 be a compact hyperbolic @xmath0-manifold with an @xmath54-regular voronoi decomposition @xmath32 , and let @xmath127 be a generic morse function .
let @xmath32 be the voronoi decomposition ordered by the order inherited from the cell splitters @xmath39 .
given @xmath128 , let @xmath37 be the sublevel set of @xmath1 at @xmath43 .
we define @xmath42 , the _ polyhedral approximation to @xmath37 _ , to be the union of the voronoi cells @xmath33 with @xmath129 , and call @xmath130 the _ polyhedral surface _ determined by @xmath119 .
the polyhedral surface @xmath131 is a union of faces of the voronoi cells , and so is a normal surface in the dual triangulation .
we shall write @xmath132 for the number of voronoi faces the polyhedral surface @xmath131 contains .
we shall write @xmath133 for the number of faces in the polyhedral surface @xmath134 , which may have boundary . a schematic picture of a polyhedral surface is given below in figure [ pic : pt ] .
( 12,-4 ) node [ right ] @xmath4 ( 12,8 ) ; ( 9 , 0 ) ( 11,0 ) node [ midway , above ] @xmath3 ; ( -2 , 2 ) .. controls ( 2,3 ) .. ( 4 , 2 ) .. controls ( 6 , 1 ) .. ( 8 , 2 ) node [ right ] @xmath53 ; ( 12 , 2 ) circle ( 0.1 cm ) node [ right ] @xmath43 ; ( -1,1.732 ) node [ midway , below ] @xmath131 + + ( 0:2 ) + + ( 60:2 ) + + ( 0:2 ) + + ( -60:2 ) + + ( 0:2 ) ; ( 3,-1.732 ) node @xmath42 ; in this section , we will show the following bound on the complexity of the polyhedral surface in the deep part @xmath100 .
[ cor : bound ] let @xmath1 be a compact hyperbolic @xmath0-manifold , with an @xmath54-regular voronoi decomposition @xmath32 , deep part @xmath100 , and a generic morse function @xmath7 . for @xmath119 , let @xmath131 be the polyhedral surface associated to @xmath43 .
then there is a constant @xmath30 , which only depends on @xmath54 , such that @xmath135 and @xmath136 in particular , this bounds the genus of @xmath134 as constant times the morse width of @xmath1 , where the constant depends only on @xmath54 .
we start by showing that the area the level sets bounds the number of faces of the polyhedral surface in the deep part @xmath100 .
[ bounded_faces ] let @xmath1 be a compact hyperbolic @xmath0-manifold , with an @xmath54-regular voronoi decomposition @xmath32 , deep part @xmath100 , and a generic morse function @xmath7 .
for @xmath119 , let @xmath131 be the polyhedral surface associated to @xmath43 . then there is a constant @xmath30 , which only depends on @xmath54 , such that @xmath137 let @xmath42 be the polyhedral approximation to @xmath37 .
let @xmath12 be a face of @xmath134 , and let @xmath105 and @xmath138 be the two adjacent voronoi cells in @xmath32 . up to relabeling
we may assume that @xmath105 is contained in @xmath42 , and @xmath138 is not .
let @xmath139 be a geodesic connecting @xmath36 to @xmath49 , and consider @xmath140 , for @xmath141 .
as @xmath105 and @xmath138 are deep , the metric balls @xmath142 , @xmath143 and @xmath140 are all topological balls , isometric to the ball @xmath144 in @xmath145 . at least half of the volume of @xmath110 is contained in @xmath42 , and strictly less than half of the volume of @xmath143 is contained in @xmath42 , so there is some @xmath141 such that exactly half the volume of @xmath140 is contained in @xmath42 .
there is a constant @xmath146 , depending only on @xmath84 , such that any surface dividing a ball in hyperbolic into regions of equal volume has area at least @xmath146 .
in fact , we may take @xmath146 to be the area of the equatorial disc , which is @xmath147 ; see for example bachman , cooper and white @xcite . recall that the voronoi decomposition has a dual triangulation in which each edge is a geodesic segment , and we shall write @xmath148 for the geodesic graph in @xmath1 formed by the @xmath149-skeleton of the dual triangulation
. we shall write @xmath150 for the subset of @xmath148 consisting of vertices corresponding to deep voronoi cells , and edges connecting two deep voronoi cells , and we shall refer to this as the _ deep graph_. each geodesic edge between two deep voronoi cells has length strictly less than @xmath151 .
therefore the choice of geodesic is unique for the voronoi cells in @xmath100 , as its length is smaller than the injectivity radius at each deep voronoi cell center @xmath36 . by proposition [ prop : bound ] , the geodesic dual graph @xmath150 has valence at most @xmath101 . [ claim : area ] consider a collection of points @xmath152 such that each point @xmath153 lies in a distinct edge @xmath154 of the deep graph @xmath150 .
then any ball @xmath155 intersects at most @xmath116 other balls @xmath156 , where @xmath116 is the constant from proposition [ prop : bound2 ] .
if two balls @xmath157 and @xmath158 intersect , then the distance between their corresponding edges @xmath154 and @xmath159 is at most @xmath54 , and so there is a pair of vertices , @xmath160 and @xmath161 with @xmath162 . by proposition [ prop : bound2 ]
, there are at most @xmath116 other vertices within distance @xmath163 of a given vertex .
therefore the total number of balls intersecting @xmath157 is at most @xmath116 , which only depends on @xmath54 .
if there are @xmath164 faces in @xmath131 then there are at least @xmath165 disjoint balls @xmath157 , each containing a part of @xmath53 of area at least @xmath146 .
therefore , the total number of faces is at most @xmath166 where the constants only depend on @xmath54 , as required .
we now show that the bound on the number of faces of @xmath131 in the deep part @xmath100 gives a bound on the genus of @xmath134 .
[ bounded_genus ] let @xmath1 be a compact hyperbolic @xmath0-manifold , with an @xmath54-regular voronoi decomposition @xmath32 , deep part @xmath100 , and a generic morse function @xmath7 .
for @xmath119 , let @xmath131 be the polyhedral surface associated to @xmath43 . then there is a constant @xmath101 , which only depends on @xmath54 , such that : @xmath167 @xmath168 where @xmath101 is the constant from proposition [ prop : bound ] . we shall write @xmath8 for @xmath131 to simplify notation .
the first bound follows as each boundary component must contain at least one edge , so the number of boundary components is at most the number of edges in @xmath169 , which is at most @xmath170 by proposition [ prop : bound ] .
we shall write @xmath171 for the surface @xmath172 with all boundary curves capped of with discs
. recall that the genus of a disconnected surface is the sum of the genera of each component , and this in turn is equal to the number of connected components minus half the euler characteristic , i.e. @xmath173 where @xmath174 is the number of connected components of @xmath171 . as capping off with discs does not change the number of connected components , this is at most the number of connected components of @xmath169 , which is at most the number of faces @xmath175 .
furthermore , capping off boundary components with discs may only increase the euler characteristic , so @xmath176 therefore @xmath177 where @xmath178 and @xmath65 are the numbers of vertices , edges and faces of @xmath172 . as each face of a deep voronoi cell
has at most @xmath101 edges , this implies @xmath179 as we may assume that @xmath101 is at least @xmath13 , this gives the second inequality .
proposition [ cor : bound ] now follows immediately from propositions [ bounded_faces ] and [ bounded_genus ] .
we have constructed surfaces with bounded complexity in the deep part .
the complement of the deep part is contained in a union of solid tori by the margulis lemma , and we now explain how to cap off the surfaces in the deep part with surfaces in the solid tori to produce bounded genus surfaces .
we will use the margulis lemma and the _ thick - thin _
decomposition for finite volume hyperbolic @xmath0-manifolds , which we now review . given a number @xmath180 ,
let @xmath181 be the _ thick part _ of @xmath1 , i.e. the union of all points @xmath48 of @xmath1 with @xmath182 .
we shall refer to the closure of the complement of the thick part as the _ thin part _ and denote it by @xmath183 .
the margulis lemma states that there is a constant @xmath184 , such that for any compact hyperbolic @xmath0-manifold , the thin part is a disjoint union of solid tori , and each of these solid tori is a regular metric neighborhood of an embedded closed geodesic of length less than @xmath185 . we shall call a number @xmath185 for which this result holds a _ margulis constant _ for @xmath145 .
if @xmath185 is a margulis constant for @xmath145 , then so is @xmath55 for any @xmath186 , and furthermore , given @xmath55 and @xmath185 there is a number @xmath117 such that @xmath187 . for the remainder for this section we shall fix a pair of numbers @xmath188 such that there are margulis constants @xmath189 , a number @xmath190 such that @xmath191 , and @xmath192 .
we shall call @xmath188 a choice of _
mv_-constants for @xmath145 .
this choice of constants ensures that the deep part @xmath100 is non - empty .
let @xmath188 be a choice of @xmath193-constants , and consider an @xmath54-regular voronoi decomposition of @xmath1 .
the fact that @xmath194 means that we adjust the boundary of @xmath195 by an arbitrarily small isotopy so that it is transverse to the voronoi cells , and we will assume that we have done this for the remainder of this section .
our choice of @xmath54 implies that the thick part @xmath56 is contained in the voronoi cells in the deep part , i.e. @xmath196 , so in particular @xmath197 is contained in the deep part .
furthermore , as @xmath198 , each deep voronoi cell hits at most one component of @xmath199 .
each boundary component of the surface @xmath200 is contained in @xmath199 , so @xmath200 is a properly embedded surface in @xmath56 .
we now bound the number of boundary components of @xmath201 in terms of the number of polyhedral faces in the deep part , @xmath133 .
[ prop : thick bound ] let @xmath188 be @xmath193-constants , and let @xmath1 be a compact hyperbolic @xmath0-manifold with thin part @xmath199 , an @xmath54-regular voronoi decomposition @xmath32 with deep part @xmath100 , and a generic morse function @xmath3 .
let @xmath131 be a polyhedral surface in @xmath1 .
then there is a constant @xmath101 , depending only on @xmath54 , such that @xmath202 and @xmath203 the properly embedded surface @xmath200 is obtained from @xmath204 by cutting @xmath134 along simple closed curves and discarding some connected components .
this does not increase the genus , which gives the first bound , using proposition [ bounded_genus ] .
each face @xmath12 of a voronoi cell is a totally geodesic convex polygon , and a component of @xmath199 lifts to a convex set in the universal cover @xmath145 , so @xmath205 is a convex subset of @xmath12 .
therefore @xmath206 consists either of a simple closed curve , or a collection of properly embedded arcs which have at most two endpoints in each edge of @xmath12 , so there are at most as many arcs as the number of edges of @xmath12 .
therefore , the number of components of @xmath206 has at most ( number of faces of @xmath134 ) plus ( number of edges of @xmath134 ) components , and this gives the second bound , again using proposition [ bounded_genus ] .
we now wish to cap off the properly embedded surfaces @xmath200 with properly embedded surfaces in @xmath199 to form closed surfaces . for each torus @xmath207 in @xmath208 let @xmath209 be the subsurface consisting of @xmath210 .
let @xmath211 , and we shall call the resulting closed surface the _ @xmath93-capped surface _ @xmath212 .
we now bound the genus of the resulting @xmath93-capped surfaces .
[ prop : bounded_genus ] let @xmath188 be @xmath193-constants , and let @xmath1 be a compact hyperbolic @xmath0-manifold with thin part @xmath199 , an @xmath54-regular voronoi decomposition @xmath32 with deep part @xmath100 , and a generic morse function @xmath3 .
let @xmath131 be a polyhedral surface in @xmath1 , and let @xmath212 be the corresponding @xmath93-capped surface .
then there is a constant @xmath30 , depending only on @xmath54 , such that @xmath213 furthermore , for any finite collection of generic points @xmath214 in @xmath4 , the corresponding @xmath93-capped surfaces @xmath215 may be isotoped to be disjoint . by proposition [ bounded_faces ]
, it suffices to bound the genus of the @xmath93-capped surface in terms of the number of polyhedral faces of the surface in the deep part .
we will show @xmath216 where @xmath101 is the constant from proposition [ prop : bound ] , which only depends on @xmath54 .
each surface @xmath209 is a subsurface of a torus , and so consists of a union of planar surfaces , together with at most one surface which is a torus with ( possibly many ) boundary components .
capping off components of @xmath200 with planar surfaces can not increase the genus by more than twice the number of boundary components , and capping off with punctured tori increases the genus by at most the number of boundary components , plus the number of punctured tori . as each voronoi cell hits at most one component of @xmath199 , there are at most @xmath133 components of the @xmath209 surfaces which may be punctured tori .
this implies @xmath217 using the bounds from proposition [ prop : thick bound ] , we obtain @xmath218 as required .
finally , we show that for any finite collection of generic points @xmath219 in @xmath4 , we may isotope the corresponding @xmath93-capped surfaces to be disjoint . to simplify notation , given a generic point @xmath220 , we will write @xmath221 and @xmath222 for the corresponding polyhedral approximation and polyhedral surface determined by @xmath223 . for any two distinct points
@xmath224 in @xmath4 , the polyhedral approximation @xmath221 is a strict subset of @xmath225 , so @xmath222 and @xmath226 are disjoint normal surfaces .
let @xmath93 be a single solid torus component of @xmath199 .
take a small product neighborhood @xmath227 $ ] , and choose the parameterization such that @xmath228 is equal to @xmath229 , and the product neighborhood is contained in @xmath93 .
let @xmath230 be the subsurface of @xmath229 given by @xmath231 .
let @xmath232 be the properly embedded surface in the product @xmath233 $ ] given by placing @xmath209 at depth @xmath234 , together with a product neighborhood of the boundary @xmath235 connecting @xmath209 to the boundary of @xmath236 , i.e. @xmath237 ) .\ ] ] as the submanifolds @xmath238 are strictly nested , the subsurfaces @xmath209 are also strictly nested , i.e. @xmath239 for @xmath240 , and so the resulting surfaces @xmath241 are disjoint .
we now bound the number of handles between a pair of @xmath93-capped surfaces @xmath242 and @xmath243 whose corresponding points in @xmath223 and @xmath244 in @xmath4 bound an interval containing a single cell - splitter . [ bounded_handles ] let @xmath188 be @xmath193-constants , and let @xmath1 be a hyperbolic @xmath0-manifold with an @xmath54-regular voronoi decomposition @xmath32 , and a generic morse function @xmath7 .
let @xmath245 be a pair of points in @xmath4 , which bound an interval containing a single cell splitter @xmath43 .
let @xmath246 and @xmath247 be @xmath93-capped surfaces corresponding to the level sets for @xmath248 and @xmath249 , bounding regions @xmath250 and @xmath251 , with @xmath252 .
then @xmath251 is homeomorphic to a manifold obtained from @xmath250 by adding at most @xmath253 handles , where @xmath101 is the constant from proposition [ prop : bound ] , which depend only on @xmath54 .
we start with the observation that attaching a compression body @xmath92 to a @xmath0-manifold @xmath254 by a subsurface @xmath8 of the upper boundary component of @xmath92 , requires a number of handles which is bounded in terms of the heegaard genus of @xmath92 , and the number of boundary components of the attaching surface .
[ bounded_handles3 ] let @xmath254 be a compact @xmath0-manifold with boundary , and let @xmath255 , where @xmath92 is a compression body of genus @xmath11 , attached to @xmath254 by a homeomorphism along a ( possibly disconnected ) subsurface @xmath8 contained in the upper boundary component of @xmath92 of genus @xmath11 .
then @xmath256 is homeomorphic to a @xmath0-manifold obtained from @xmath254 by the addition of at most @xmath257 @xmath149-and @xmath13-handles , where @xmath9 is the number of boundary components of @xmath8 .
recall that the genus of a disconnected surface with boundary is the sum of the genus of each closed component obtained by capping off all boundary components with discs .
therefore , the genus of @xmath8 is at most the genus of @xmath92 . for a connected surface of genus @xmath11 with @xmath258 boundary components , cutting along a non - separating arc with endpoints in the same boundary component produces a surface of genus @xmath259 with @xmath260 boundary components .
a planar surface with @xmath258 boundary components may be cut in to at most @xmath258 discs by @xmath261 non - separating arcs .
therefore we may choose at most @xmath262 arcs which cut the surface @xmath8 into at most @xmath263 discs .
we can add a @xmath149-handle to @xmath254 for each arc , and then a @xmath13-handle for each disc , to produce a manifold @xmath264 which is homeomorphic to @xmath254 union a regular neighborhood of @xmath265 .
we may then form @xmath256 by adding at most @xmath11 @xmath13-handles .
the total number of @xmath149- and @xmath13-handles required is at most @xmath266 .
let @xmath76 be the voronoi cell corresponding to the single cell splitter @xmath43 contained in the interval @xmath267 $ ] .
the surfaces @xmath246 and @xmath247 are parallel everywhere , except in a regular neighborhood of @xmath76 .
if the voronoi cell @xmath76 is disjoint from @xmath199 , then it is a ball , and is attached to @xmath250 along a subsurface consisting of a union of faces of @xmath76
. therefore the number of boundary components of the attaching surface is at most @xmath101 , where @xmath101 is the constant from proposition [ prop : bound ] , so by proposition [ bounded_handles3 ] , @xmath251 is obtained from @xmath250 by attaching at most @xmath268 handles .
if the voronoi cell @xmath76 intersects @xmath199 , then @xmath251 is obtained from @xmath250 by adding regions of @xmath269 , which we shall refer to as the _ complementary regions _ , together with regions of @xmath270 .
the complementary regions may not be topological balls , but kobayashi and rieck @xcite show that they are handlebodies of bounded genus .
@xcite let @xmath55 be a margulis constant for @xmath145 , @xmath1 be a finite volume hyperbolic @xmath0-manifold , let @xmath271 , and let @xmath32 be a regular voronoi decomposition of @xmath1 arising from a maximal collection of @xmath54-separated points .
then there is a number @xmath272 , depending only on @xmath55 and @xmath54 , such that for any voronoi cell @xmath33 , there are at most @xmath272 connected components of @xmath273 , each of which is a handlebody of genus at most @xmath272 , attached to @xmath199 by a surface with at most @xmath272 boundary components .
we state a simplified version of their result which suffices for our purposes .
their stated result involves extra parameters @xmath274 and @xmath256 , but if @xmath274 is chosen close to @xmath275 , then @xmath256 is close to @xmath55 , and we obtain the result above .
their proof involves showing that in the universal cover , for any point @xmath276 in @xmath277 , projection to @xmath208 along geodesic rays based at @xmath276 gives a topological product structure to @xmath273 as @xmath278 .
an examination of their proof shows that we may choose @xmath279 , where @xmath101 is the constant from proposition [ prop : bound ] .
then proposition [ bounded_handles3 ] implies that adding the complementary regions of a voronoi cell which intersects @xmath280 requires at most @xmath281 handles . however , if the voronoi cell intersects a solid torus component @xmath93 of @xmath199 , then the surfaces @xmath246 and @xmath247 need not be parallel inside @xmath93 , and so we now bound the number of handles needed to add the region corresponding to @xmath282 . if @xmath283 is equal to all of @xmath229 , then the additional region is a solid torus attached along @xmath284 , so adding this region requires at most @xmath285 handles , by proposition [ bounded_handles ] . if @xmath283 is not equal to all of @xmath229 , then this region is homeomorphic to @xmath286 ) \setminus ( u_1 \cross [ 0 , \tfrac{1}{2}])$ ] , and so is homeomorphic to @xmath287 , which is a handlebody of genus at most @xmath288 . the region is attached along @xmath283 , so adding this region requires at most @xmath289 handles , and so in either case , at most @xmath290 are required .
therefore @xmath251 may be constructed from @xmath250 by adding at most @xmath291 handles , as required .
the manifold @xmath1 may be constructed by adding the voronoi cells in the order arising from the cell splitters @xmath39 in @xmath4 .
choose a finite collection of generic points @xmath214 , so that each pair of adjacent cell splitters is separated by one of the @xmath223 , and let @xmath292 be the corresponding collection of @xmath93-capped surfaces .
the linear width is at most the largest genus of any surface in the collection @xmath293 , plus the maximum number of handles added by attaching a single voronoi cell .
therefore the bounds from propositions [ prop : bounded_genus ] and [ bounded_handles ] imply @xmath294 as @xmath101 is at least @xmath149 , this gives @xmath295 the constants @xmath101 and @xmath30 only depend on the choice of mv - constants , which may be chosen independently of the hyperbolic @xmath0-manifold @xmath1 , and so this completes the proof of the left hand bound of theorem [ theorem : morse ] .
in this section we will show linear bounds for morse area in terms of scharlemann - thompson width , assuming the pitts and rubinstein result theorem [ conjecture : pr ] , i.e. we will show the right hand bound of theorem [ theorem : morse ] .
this result is due to gabai and colding @xcite*appendix a , using recent work of colding and minicozzi @xcite , but we give a brief description for the convenience of the reader , as they do not state this result explicitly .
we will use properties of a refinement of linear width , known as _ thin position _ , which we now describe .
let @xmath296 be the collection of upper boundaries of compression bodies in the linear splitting , and let @xmath27 be the complexity of the surface @xmath26 , i.e. the sum of the genera of its connected components .
we say that the complexity of the linear splitting is the collection of integers @xmath297 , arranged in decreasing order .
a linear splitting which gives the minimum complexity of all possible linear splittings in the lexicographic ordering on sets of integers is called a _ thin position _ linear splitting .
scharlemann and thompson @xcite showed that thin position linear splittings have the following property .
@xcite let @xmath298 be a linear splitting that is in thin position
. then every even surface is incompressible in @xmath1 and the odd surfaces form strongly irreducible heegaard surfaces for the components of @xmath1 cut along the even surfaces . if follows from freedman , hass and scott @xcite that the incompressible surfaces may be chosen to be disjoint least area minimal surfaces , and in fact the off surfaces may also be chosen to be disjoint minimal surfaces , see for example lackenby @xcite or renard @xcite . in a hyperbolic manifold
the intrinsic curvature of a minimal surface is at most @xmath299 , so the gauss - bonnet formula gives an upper bound for the area of the minimal surface .
therefore the area of a minimal surface of genus @xmath11 is at most @xmath300 .
we say that a hyperbolic @xmath0-manifold @xmath1 has _
least area boundary _ if its boundary components are ( possibly empty ) least area minimal surfaces , and we say that a heegaard splitting @xmath298 for @xmath1 is _ minimal _ if it is isotopic to an unstable minimal surface .
the right hand bound of theorem [ theorem : morse ] is a consequence of the following result of colding and gabai @xcite , which constructs bounded area foliations for a pair of compressions bodies with least area lower boundaries , sharing a common minimal heegaard splitting surface .
@xcite[theorem : minimal ] let @xmath1 be a hyperbolic manifold , with ( possibly empty ) least area boundary , with a minimal heegaard splitting @xmath298 of genus @xmath11 . then
, assuming theorem [ conjecture : pr ] , the manifold @xmath1 has a ( possibly singular ) foliation by compact leaves , containing the boundary surfaces as leaves , such that each leaf has area at most @xmath301 . in a @xmath0-manifold
a foliation with singularities of `` standard type '' means that almost all leaves are completely smooth ( i.e. , without any singularities ) .
in particular , any connected subset of the singular set is completely contained in a leaf . furthermore , the entire singular set is contained in finitely many ( compact ) embedded lipschitz curves with cylinder singularities together with a countable set of spherical singularities .
@xcite*appendix a let @xmath302 be an unstable minimal surface in a hyperbolic manifold @xmath1 .
then there is a regular neighborhood of @xmath302 with a smooth mean convex product foliation @xmath303 , @xmath304 $ ] , with non - minimal boundary leaves @xmath305 and @xmath306 .
in particular , each leaf in the foliation has area at most @xmath301 .
as the boundary leaves @xmath305 and @xmath306 are non - minimal mean convex surfaces , we may apply the mean curvature flow results of colding and minicozzi @xcite , which show that the mean curvature flow gives rise to a mean convex foliation with standard singularities . as the mean curvature flow gives a foliation by surfaces of decreasing area , the only possible singularities which may arise are disc compressions , @xmath13-spheres collapsing to a point or tori collapsing to circles .
in particular , each non - singular leaf bounds a compression body in the interior of the compression body it is contained in .
if all leaves eventually collapse , then the compression body has empty lower boundary , i.e. it is a handlebody , and this gives a mean convex foliation , and hence area decreasing foliation , of the handlebody .
otherwise , the mean curvature flow limits to a stable minimal surface @xmath148 whose components bound compression bodies together with the lower boundary of the original compression body . if the stable minimal surface @xmath148 is not equal to the stable boundary of the compression body , then it bounds a sub - compression body with stable boundary , whose standard heegaard splitting is strongly irreducible , so we may apply the argument again .
anderson @xcite and white @xcite showed that there are only finitely many minimal surfaces of bounded genus in a compact riemannian manifold , and so this process may occur only finitely many times , resulting in a foliation of the entire compression body .
this completes the proof of theorem [ theorem : minimal ] . by conjecture [ conjecture : pr ] ,
the irreducible heegaard surface for a hyperbolic @xmath0-manifold @xmath1 with stable boundary is either isotopic to an unstable minimal surface @xmath302 , to which we may apply theorem [ theorem : minimal ] directly , or isotopic to a regular neighborhood of a one - sided stable minimal surface union a small tube parallel to one of the normal fibers . in the latter case
, the heegaard surface bounds a handlebody on at least one side , and cutting along the stable one - sided surface leaves a compression body homeomorphic to the heegaard surface cut along the disc corresponding to the tube , where all boundary components are stable minimal surfaces . as the standard heegaard splitting of a compression body is strongly irreducible , we may now apply theorem [ theorem : minimal ] in this case as well . | scharlemann and thompson define a numerical complexity for a @xmath0-manifold using handle decompositions of the manifold .
we show that for compact hyperbolic @xmath0-manifolds this is linearly related to a definition of metric complexity in terms of the areas of level sets of morse functions . |
cylindrical structures are ubiquitous in nature and technology over a range of length scales , from carbon nanotubes and cytoskeletal microtubules to aircraft fuselages .
the separation of geometric scales inherent in these structures gives them their large specific stiffness ( per unit weight ) and endows them with strongly geometrical modes of deformation that involve faceting and other localized modes when subject to pressure , compression , twisting and bending .
here we focus on a peculiar global mode of deformation that is also ubiquitous .
our starting point is the simple observation that when a paper or plastic drinking straw of length @xmath0 , radius @xmath1 and thickness @xmath2 ( @xmath3 ) is pinched at an end so that it becomes elliptical locally , as shown in figure [ fig1 ] ( try it yourself ) ; the deformation of the straw persists over a length that is much larger than the radius of the straw .
this raises a natural question : what is the persistence length of a pinch ? to analyze this , we start with the equations of equilibrium for a shallow cylindrical shell , the von karman- donnell equations
@xcite : @xmath4 here @xmath5 , @xmath6 is the deflection of the cylinder relative to its naturally curved state parametrized in terms of the azimuthal coordinate @xmath7 and the axial coordinate @xmath8 , @xmath9 is the airy stress function whose derivatives are the components of the in - plane stress tensor , @xmath10 etc . , and @xmath11 is the bending stiffness of the sheet of material with young s modulus @xmath12 and poisson s ratio @xmath13 .
the first equation quantifies the balance of forces perpendicular to the cylinder surface , while the second is a geometric compatibility relation involving the in - plane strains .
we note that the karman - donnell system is one of the simplest of a class of approximate equations of increasing sophistication for the deformations of elastic shells @xcite ; however they are sufficient to explain the phenomena at hand as we shall see .
when one end of the cylinder is pinched locally exciting the first elliptical mode in the azimuthal direction , we assume that the axial variations induced are on length scales large compared to the radius . thus suggests a solution of the form @xmath14 . substituting into eq .
( [ vonkarmaneqn ] ) , and keeping only the leading order terms , we find that @xmath15 where @xmath16 , and we have dropped the nonlinear terms in ( [ vonkarmaneqn ] ) . eliminating the function @xmath17 from the above equations , we find that the amplitude of the deformation @xmath18 satisfies the linear equation @xmath19 it is useful at this point to recall that the balance of forces on an elastic beam supported on an elastic foundation of stiffness @xmath20 leads to a fourth order equation @xcite @xmath21 . indeed , for axisymmetric deformations of a cylindrical shell , @xmath22 , so that @xmath23 , which while superficially similar to ( [ wfinaleqn ] ) is qualitatively different ; in the limit @xmath24 , the former equation is of a singularly perturbed type exhibiting boundary - layer like regions where the solution changes rapidly , quite unlike ( [ wfinaleqn ] ) . indeed , we can see this immediately by noting that the general solution to ( [ wfinaleqn ] ) is @xmath25^{1/4}. \label{scaling}\ ] ] where the amplitude @xmath26 and the phase @xmath27 are determined by the boundary conditions at the end where the pipe is pinched .
thus an applied pinch of amplitude @xmath26 decays as an oscillatory exponential with a characteristic persistence length @xmath28 , for typical materials ( @xmath29 $ ] ) , a scaling hinted at in @xcite with no derivation .
the ellipticity induced by the pinch rotates slowly as one progresses along the pipe , analogous to the polarization of a wave . since @xmath30
, our assumption that the shape and stress have variations with gradients that are smaller in the axial direction than in the azimuthal direction is justified .
we also observe that the persistence length is determined primarily by the geometry of the tube ; the dependence on the material properties via poisson s ratio is very weak since @xmath31 $ ] for isotropic materials .
even more surprising is the anomalous and counter - intuitive divergence of the persistence length as the thickness of the tube vanishes or as the radius of the cylinder diverges , i.e. as the tube is flattened .
this geometric amplification is a result in sharp contrast with the focusing that leads to the generation of fine scales and singularities in the inhomogeneous deformation of thin films in such instances as wrinkling and crumpling .
the above linearized analysis is valid only for small deformations in light of the approximate nature of the von - karman - donnell equations and our asymptotic analysis . to check the validity of our analysis
, we carried out numerical simulations using the finite element method .
the pinch was applied incrementally and a commercial package abaqus was used to minimize the elastic energy of a linear isotropic solid in a slender geometry . for very thin shells ,
this elastic energy density that is minimized is approximately : @xmath32 \nonumber \\ & + & \frac{et^3}{24(1-\sigma^2 ) } \left[({\kappa_1}+ { \kappa_2})^2 - 2 ( 1-\sigma ) ( \kappa_1 \kappa_2 - \tau^2 ) \right ] .
\end{aligned}\ ] ] here @xmath33 are the in - plane extensional and shear strains and @xmath34 are the curvatures and twist relative to the undeformed state of the tube .
the first term in the expression accounts for in - plane deformations and the second term accounts for out - of - plane deformations .
four - node , quadrilateral shell elements with reduced integration and a large - deformation formulation to account for the finite curvatures of the shell were used in calculations .
a sensitivity study was conducted to ensure the independence of the results on the computational mesh .
we considered very thin isotropic and homogeneous elastic tubes which were pinched symmetrically at one end with a prescribed indentation deformation @xmath35 .
the tubes were assumed to be made of linearly elastic material ( with young s modulus @xmath36 mpa and poisson s ratio @xmath37 ) of varying @xmath2 in our simulations , with @xmath38 . in figure [ fig2 ]
we see that the tube responds with an oscillatory deformation of varying ellipticity consistent with our simple analytical predictions . in fig . , [ fig3 ]
, we show that the persistence length measured numerically by following the period of oscillation of the ellipticity also follows our analytical predictions ; the numerical experiments were carried out for tubes of various @xmath39 keeping the applied pinch amplitude @xmath26 and the tube radius @xmath1 constant . in the inset
we show that the amplitude of the deformation of the tube decays exponentially away from the location of the pinch .
however , there is a discrepancy between the prefactor of the wavelength determined numerically and that in ( [ scaling ] ) which may be attributed to the approximate nature of the von karman - donnell equations and our linearized analysis thereof .
we also carried out experiments with long straws and tubes of various thickness ( ranging from 0.1 mm to 0.3 mm ) and radii ( ranging from 3 mm to 12.5 mm ) .
the images of the pinched straws were taken with a digital camera and were analyzed using the canny edge - detection - algorithm in matlab to extract the information on boundaries of the deformed profile .
the thickness and radius of the tubes were measured using slide calipers . in fig .
[ fig3 ] , we show that the experimentally measured values of the persistence length match the numerical simulations and are consistent with the scaling derived analytically .
before we discuss our results , it is worthwhile to compare and contrast our results for the persistence of a pinch in cylinder with that of a pinch in a narrow flat plate of width @xmath40 and length @xmath41 ( equivalently , we could consider periodically pinched sheet such as a drape ) .
when such a sheet is pinched at an end to make it slightly narrower so that the amplitude of the pinch is @xmath42 , casual observations show that the persistence of the pinch is much larger than the width of the strip . to understand this persistence
, we note that the dominant component of the curvature scales as as @xmath43 so that bending energy in the sheet is roughly @xmath44 , where @xmath45 is the persistence length of a pinch . over this length scale ,
the sheet is weakly curved in two directions so that it must be stretched ( since the gauss curvature of a doubly curved surface is non - zero , it follows from gauss s theorema egregium that it must have been stretched ) , with a typical strain that scales as @xmath46 , costing an energy that scales as @xmath47 . minimizing the sum of the bending and stretching energy , we see that the persistence length of a pinch scales as @xmath48 @xcite , showing a clear dependence on the amplitude of the pinch . however , for a pinched pipe , when @xmath49 , we find that @xmath50 consistent with the result obtained here .
some immediate implications follow from our results : in molecular carbon nanotubes @xcite or cytoskeletal microtubules @xcite our result suggest how mechanical strains may be transmitted over long distances along tubes . since the energy associated with pinching the pipe scales as @xmath51 , the stiffness associated with a pinching deformation scales as @xmath52 is quite small , and raises the question of whether nature might already use this mode .
on mesoscales , a dramatic example of the failure of tubes is afforded by the flip - flop buckle propagation in submarine pipelines @xcite , wherein a pipe flattens for a while in one direction and then in an orthogonal direction , with a characteristic length scale of the flattened regions .
our result complements numerical simulations of the process @xcite by providing a simple explanation for the flip - flop which we interpret as a simple consequence of a globally deformed tubular shell induced for example by a local pinch .
then , an oscillatory elliptic mode follows naturally , and sets the stage for dynamic and plastic buckling that will likely lead to a flip - flop mode of propagation .
thus our calculation immediately suggests a criterion for the design of buckle arrestors - which might be spaced at a distance comparable to @xmath53 .
finally , on planetary length scales , the curved nature of continental and oceanic plates together with our calculation suggests a possible mechanism for the very long scale persistence of deformations induced by the curvature of the crust that would otherwise sink far more quickly under its own weight @xcite . | the response of low - dimensional solid objects combines geometry and physics in unusual ways , exemplified in structures of great utility such as a thin - walled tube that is ubiquitous in nature and technology .
here we provide a particularly surprising consequence of this confluence of geometry and physics in tubular structures : the anomalously large persistence of a localized pinch in an elastic pipe whose effect decays very slowly as an oscillatory exponential with a persistence length that diverges as the thickness of the tube vanishes , which we confirm experimentally .
the result is more a consequence of geometry than material properties , and is thus equally applicable to carbon nanotubes as it is to oil pipelines . |
suppression of @xmath0 production due to screening of the binding potential of charm and anti - charm quarks in hot and dense matter is a classic signature of the quark gluon plasma formation @xcite and has been intensively studied in previous years .
the na50@xcite and na60@xcite experiments at cern - sps ( at @xmath2 = 17.3 gev ) observed a strong suppression of @xmath0 as a function of centrality . the results from the phenix collaboration at rhic show that the @xmath0 suppression for @xmath3 5 gev / c at mid - rapidity is similar to that observed at sps energies@xcite , although the energy density and temperature reached at rhic are much higher than at sps . on the other hand , the suppression at forward rapidity is stronger than at mid - rapidity .
such a pattern suggests that additional processes other than @xmath0 dissociation take place , such as recombination of charm quarks , sequential dissociation of exited charmonia states or feeddown from b - mesons , may compensate for @xmath0 dissociation .
one of the interesting developments is the so called hot wind dissociation@xcite .
using ads / cft duality , it was predicted that the @xmath0 effective dissociation temperature decreases with increasing @xmath0 velocity . at rhic energies ,
the expectation from the hot wind model is that the system temperature should be below the @xmath0 effective dissociation temperature for @xmath0 with @xmath4 5 gev / c , and above for higher @xmath5 , and therefore the suppression is expected to be stronger at high @xmath5 .
these predictions can be tested with @xmath0 measurements at high transverse momentum .
in addition to the study of @xmath0 interactions with the medium produced in nuclear collisions , charmonium production in proton collisions is of interest in its own right . the production is not well understood as several mechanisms can play a role .
these include direct production via gluon fusion , whether the state is formed through color singlet or octet , the contributions from parton fragmentation , and feeddown from other charmonium states and b - mesons . at the moment
no model explains the major features of the existing data@xcite .
the measurement of high-@xmath5 @xmath0 in nucleus - nucleus collisions , together with high-@xmath5 @xmath0 p+p data , might shed some light on charmonium production . in this paper , the recent studies of @xmath0 production in au+au and cu+cu from the star experiment are presented .
we report and discuss the measurements of the nuclear modification factor as a function of transverse momentum and centrality .
the study of low-@xmath5 @xmath0 production ( @xmath6 5 gev / c ) has been done using minimum bias data in cu+cu ( year 2005 ) and au+au ( year 2007 ) collisions at @xmath2 = 200 gev while the high-@xmath5 @xmath0 were studied using triggered data collected in cu+cu collisions . in the analysis reported here
, the @xmath0 were reconstructed through the di - electron decay channel ( branching ratio = 5.9% ) .
star is well suited for such studies , as the large acceptance time projection chamber ( tpc ) provides electron identification at moderate @xmath5 ( via de / dx measurements ) , and at high-@xmath5 the barrel electromagnetic calorimeter ( bemc ) is a very effective tool to identify electrons .
the bemc can be also used for fast online triggering to enrich the electron sample . in the case of minimum - bias cu+cu collisions ( 0 - 60% of total cross - section , 27 m events )
only de / dx from the star tpc was used to identify electrons and reject hadrons . for minimum - bias au+au collisions ( 0 - 80% of total cross - section , 64 m events ) ,
the bemc provided an additional cut on the ratio of track momentum to energy . to improve the signal - to - background ratio ,
a cut of @xmath7 1.1 gev / c in cu+cu ( @xmath7 1.2 gev / c in au+au ) was applied for each electron selected in the studies . for the high-@xmath5 @xmath0 analysis ,
the bemc triggered data for cu+cu collisions ( year 2005 ) were used with transverse energy threshold @xmath83.75 gev ( integrated luminosity @xmath9 @xmath10 ) .
the higher-@xmath5 electron is identified using combined information from the bemc ( tower energy ) , the shower max detector embedded in bemc ( shower shape ) and de / dx measured in tpc , while only tpc is used to identify the second electron . to ensure a clean @xmath0 identification , @xmath5 cuts of 3.5 gev / c for the higher-@xmath5 electron and 1.5 gev / c for the lower-@xmath5
electron were applied .
figure [ jpsisignal ] shows the @xmath11 invariant mass distributions in au+au at @xmath2 = 200 gev . on the left plot ,
the signal and background distribution before background subtraction are presented for au+au , while on the right plot the @xmath0 invariant mass after background subtraction is shown .
the background in the case of au+au , as well as for cu+cu collisions , was estimated by the event mixing technique and normalized to the total number of like -sign pairs .
the shape of the @xmath0 mass peak is well described by a crystal ball function@xcite . despite rather high background level ( signal - to - background ratio = 1/13 for cu+cu and 1/28 for au+au )
we have observed a prominent @xmath0 signal with significance of 4.5@xmath12 in cu+cu and 6.5@xmath12 in au+au collisions .
the @xmath0 signal in au+au is promising and the studies are underway .
the @xmath0 @xmath5 spectrum presented on fig .
[ ptspectrum ] shows star s ability to measure the production of low and high @xmath5 @xmath0 .
the recent star upgrades , namely new time - of - flight detector and fast tpc daq system , will greatly improve the low-@xmath5 @xmath0 measurements .
the nuclear modification factor ( @xmath13 ) for @xmath0 as a function of centrality ( represented by number of participants ) ( left panel ) and @xmath5 ( right panel ) for cu+cu at @xmath2 = 200 gev are presented on fig . [ raa ] and compared to phenix results@xcite .
the @xmath13 vs @xmath14 was obtained using minimum - bias cu+cu collisions only .
the results are consistent with previously published data but , due to relatively modest statistics , the results have limited discrimination power . the @xmath13 vs @xmath5 for high-@xmath5 @xmath0
is presented for top @xmath15 and @xmath16 cu+cu collisions@xcite .
the @xmath13 seems to increase with @xmath5 and it is consistent with unity for @xmath5 @xmath17 5 gev / c : the average of the two star @xmath18 data points gives @xmath13 = 1.4 @xmath19 0.4 ( stat . ) @xmath19 0.2 ( syst . ) .
the dashed curve shows the predictions of an ads / cft - based calculation embedded in a hydrodynamic model with hot wind dissociation incorporated - the data contradict the predicted @xmath5 dependence .
the dotted line represents calculations of the so called two component model@xcite which includes @xmath0 dissociation , statistical @xmath20 coalescence , @xmath0 formation time effects and b - meson feeddown .
the model describes the overall trend of the data fairly well . the same calculations but without bottom feeddown and finite formation time , predict decrease of @xmath13 with @xmath5 which may suggest the important role of these effects .
on the other hand the recent star results show the b - meson feeddown contribution to @xmath0 production at high @xmath5 is relatively small , on the level of 13%@xcite .
the solid and dash - dot lines provide a comparison to open charm @xmath13 .
the solid line is based on the whdg model@xcite for charm energy loss for @xmath18 cu+cu .
the calculations include the elastic as well as radiative parton energy loss and the predictions were obtained with medium gluon density @xmath21 .
the dash - dotted line are results of the glv model@xcite for d - meson energy loss . in this approach
the heavy meson suppression is described by collisional dissociation in quark - gluon plasma with @xmath22 .
both calculations correctly describe heavy flavor suppression in au+au , thus one would expect a strong suppression for @xmath0 as well for @xmath5 @xmath17 5 gev / c when they traverse through the medium carrying color charges , which is in contrast to the data .
these results may suggest the high-@xmath5 @xmath0 production is dominated by the color singlet channel but other processes , like bottom feeddown or finite formation time , can compensate for the predicted suppression .
it is worth to mention that the finite formation time , required to build the @xmath0 wave function , is expected to reduce the dissociation cross section for @xmath23 compared to fully formed charmonium . in case of longer formation time
, some portion of @xmath0 is produced outside the medium and therefore it should be sensitive to gluons and charm quark energy loss in the quark - gluon plasma .
we reported recent studies of @xmath0 production in nucleus - nucleus collisions at star .
the @xmath0 @xmath13 seems to increase with @xmath5 and we observed no suppression for @xmath5 @xmath17 5 gev / c .
the lack of @xmath0 suppression at high-@xmath5 is in contrast to the predictions for substantial suppression of charm quarks and open charm@xcite at similar @xmath5 , and this may suggest that @xmath0 in this @xmath5 range is produced mostly via the color single channel .
00 t. matsui and h. satz , _ phys .
b _ * 178 * ( 1986 ) 416 .
m. c. abreu it et al .
[ na50 collaboration ] , _ phys .
b _ * 410 * ( 1997 ) 327 .
b. alessandro et al .
[ na50 collaboration ] , _ eur .
j. c _ * 39 * ( 2005 ) 335 [ arxiv:0412036 ] .
r. arnaldi et al .
[ na60 collaboration ] , _ phys .
* 99 * ( 2007 ) 132302 .
a. adare et al .
[ phenix collaboration ] , _ phys .
* 98 * ( 2007 ) 232301 [ arxiv:0611020 ] .
h. liu , k. rajagopal and u. a. wiedmann , _ phys .
* 98 * ( 2007 ) 182301 .
j. p. lansberg , _ int .
j. mod .
a _ * 21 * ( 2006 ) 3857 [ arxiv:0602091 ] .
j. p. lansberg , [ arxiv:0811.4005 ] .
j. gaiser,``charmonium spectroscopy from radiative decays of the j / psi and psi - prime '' , slac - r-255 a. adare et al .
[ phenix collaboration ] , _ phys .
lett . _ * 101 * ( 2008 ) 122301 [ arxiv:0801.0220 ] .
b. i. abelev et al .
[ star collaboration ] , [ arxiv:0904.0439 ] .
x. zhao and r. rapp , [ arxiv:0806.1239v1 ] s. wicks et al .
a _ * 784 * ( 2007 ) 426 , and w. a. horowitz private communication .
a. adil and i. vitev , _ phys .
b _ * 649 * ( 2007 ) 139 , and i. vitev private communication . | @xmath0 production is considered to be a sensitive probe of the properties of quark gluon plasma created in nucleus+nucleus collisions at rhic . in this article , the recent analysis of mid - rapidity ( @xmath1 ) @xmath0 production via the dielectron decay channel in au+au ( year 2007 ) and cu+cu ( year 2005 ) collisions at @xmath2 = 200 gev at star is reported .
it is compared to star p+p results in order to study the nuclear modification factor as a function of transverse momentum and centrality .
the results are compared to previously published data and available theoretical models . |
a measurement is by definition an operation that acts on a system and returns some classical information , i.e. a label `` @xmath4 '' identified by an index @xmath5 . this operation typically , but not always , changes the state of the measured system ( _ wave - function collapse _
mechanism ) . when measuring classical objects , the state change is only due to the change in the information we have on the system : the uncertainty in the state is usually reduced .
when measuring quantum objects , the state change can have a dynamical nature that perturbs the system . the rigorous description of these mechanisms results from the kraus decomposition of the measurement apparatus @xcite , which identifies the completely positive ( cp ) map of its evolution .
when the system is initially in a state @xmath6 , the @xmath5th measurement outcome occurs with a probability @xmath7tr@xmath8 $ ] , where @xmath9 is the apparatus povm ( positive operator - valued measure ) , a set of positive operators normalized so that @xmath10 . after the outcome @xmath4
is obtained , the state is changed to @xmath11 where @xmath12 is a set of indices @xmath13 and @xmath14 are the apparatus kraus operators .
the povm in terms of these is given by @xmath15 .
on the basis of the above description of the measurement , we can define the following information and disturbance measures . _ information : _ the outcome @xmath4 provides the experimenter with some information @xmath2 .
we are , obviously , interested in the case in which @xmath4 provides some information on the measured system , and it is not just independently generated by the apparatus .
thus , a good measure for @xmath2 is the mutual information between the measurement results and some property of the system state .
a significant ( basis - independent ) property is the spectrum of the state @xmath6 , i.e. the probability distribution of its eigenvalues @xmath16 .
of course , the ensemble composed by the eigenvectors of @xmath6 weighted by the corresponding eigenvalues is not the only one that originates from the state @xmath6 .
however , it is easy to see that it is the one that allows to recover the maximal accessible information ( it saturates the holevo bound ) @xcite .
it is thus the appropriate choice since we ultimately want to bound the information from above .
we then use @xmath17 the mutual information between the label @xmath18 identifying the eigenvalue of @xmath6 and the label @xmath5 identifying the measurement result : @xmath2 is the number of bits that the experimenter gains from the result @xmath4 on which eigenvector of @xmath6 the system was in ( before the measurement ) .
it is maximal when all eigenvalues are equal ( the experimenter has no prior info on the state ) , and it is null when the state is pure ( the experimenter already has total knowledge of the state , and can not gain any more information on it ) . [ note that @xmath6 here refers to the state from the experimenter s point of view : it reflects his prior knowledge of the system state , i.e. that each eigenvector has a prior probability @xmath16 .
the ` true ' state ( i.e. from the point of view of who is preparing the system ) will be in general purer . the knowledge that can be acquired by the experimenter is upper bounded by the difference in entropy between these two representations of the state . ]
if the hilbert space of the system has finite dimension @xmath19 , we can normalize @xmath2 by dividing it with its maximum value @xmath20 , so to have @xmath21 .
_ disturbance : _ a disturbance is an irreversible change in the state of the system , caused by a cp - map evolution ( such as the dynamical disturbance caused by quantum correlations that leak out to the environment and are lost ) .
thus , any quantity @xmath0 that measures disturbance should satisfy the following requirements , inspired by ref .
@xcite : * @xmath0 should be a function only of the input state @xmath6 and of the apparatus , identified through its kraus operators @xmath22 , i.e. @xmath23 .
* @xmath0 should be null if and only if the transformation @xmath22 is invertible on @xmath6 is a purification of @xmath6 , there exists an `` inversion '' cp - map identified by the kraus operators @xmath24 such that @xmath25 , where the identity @xmath26 acts on the purification space . ] . in this case
the state change can be undone , and such transformation is not disturbing the system .
* once the state has been disturbed , it should not be possible to decrease @xmath0 with any successive transformation .
this means that @xmath0 should be monotonically non - decreasing for successive applications of cp - maps @xcite ( i.e. it should satisfy a sort of pipeline inequality ) .
this requirement captures the notion that a disturbance should be irreversible , and is connected with the concept of `` cleanness '' @xcite .
* @xmath0 should be continuous : maps and input states which do not differ too much should give similar values of @xmath0 .
the above requirements , which define the disturbance axiomatically , have nothing to do with the information the measurement provides . as such
, there is no obvious _ a priori _ reason why an information disturbance tradeoff should hold .
definitions of disturbance are customarily based on the fidelity or the bures distance @xcite between input and output states .
even though valid information disturbance relations can be found @xcite , these definitions do not seem to appropriately gauge the disturbance , intended as an irreversible evolution .
in fact , even though a unitary transformation is perfectly reversible , it can rotate a state to an orthogonal configuration , generating the maximum possible fidelity - based disturbance .
these quantities do not satisfy the requirements _ ii ) _ and _ iii)_. analogous considerations apply also if we use the entanglement fidelity @xcite in place of the fidelity , where @xmath27 is the fidelity , @xmath6 and @xmath28 are the input and output states , and the maximization runs over all unitaries @xmath29 .
also this definition is inadequate , since @xmath30 is null if @xmath6 and @xmath28 have the same eigenvalues , which does not necessarily entail that the transformation is invertible , i.e. requirement _ ii ) _ still does not hold . ] .
a definition of disturbance @xmath0 that satisfies all the above requirements can be found by recalling that a cp - map @xmath31 is invertible if and only if @xcite the map s coherent information @xmath32 is equal to the von neumann entropy @xmath33tr@xmath34 $ ] of the input state @xmath6 .
the coherent information @xcite is defined as @xmath35 , where @xmath36 is a purification of @xmath6 and the map @xmath37 acts with @xmath31 on the system space and with the identity @xmath26 on the purification space .
the quantity @xmath38 is non - increasing for application of cp - maps ( data - processing inequality ) @xcite .
namely , for any two maps @xmath31 and @xmath39 , we have @xmath40 , where @xmath41 denotes composition of maps .
thus , a disturbance measure that satisfies requirements _ ii)-iii ) _ must be a function @xmath42 of @xmath43 , with @xmath42 non - decreasing and null when its argument is : @xmath44 .
we then define @xmath45 which , in addition to _
ii)-iii ) _ , also satisfies requirements _ i ) _ and _ iv ) _ since it is continuous ( see the appendix ) .
analogously to @xmath2 , also @xmath0 can be normalized in @xmath19-dimensional hilbert spaces by dividing it by @xmath20 , so that @xmath46 . from the postulates of quantum mechanics
it follows that the system state @xmath6 describes the information the experimenter possesses on the system .
hence , there are two mechanisms that lead to a state change : the system dynamics and the acquisition of new information .
heisenberg , in his uncertainty principle , was considering the former mechanism only . to exemplify the latter ,
suppose i acquire a qubit in an unknown state .
initially , i will assign to it the state @xmath47 , but as soon as the preparer tells me that the qubit was in the state @xmath48 , from my point of view it undergoes a state change ( even though i may have not interacted with it ) described by the map @xmath49 $ ] .
we can call this a `` purely informational '' state change .
since both the dynamical and the informational state changes are described by cp - maps , they both fall in the general framework described above .
[ in this sense it may be interpreted as a generalization of heisenberg s intuition ] . is it possible to weight the contribution of these two mechanisms in each measurement apparatus ? yes : since the set of cp - maps is a convex set , the apparatus cp - map @xmath31 ( identified by the kraus operators @xmath22 ) can always be written as a convex combination of the purely informational map @xmath50 and of a `` dynamical '' map @xmath51 as @xmath52 \labell{sdas}\;.\end{aligned}\ ] ] the kraus operators of the map @xmath50 are @xmath53 , where @xmath54 and @xmath55 are the eigenvalues and eigenvectors of the ` true ' ( i.e. from the point of view of the preparer ) state @xmath48 : the action of @xmath50 must not change the true state , @xmath56=\rho_p$ ] .
[ note that in the case of a degenerate state @xmath48 , the map @xmath57 might not be univocally defined ] .
the povm of a purely informational measurement is then @xmath58 ( the projectors on the eigenspaces of @xmath48 ) . in this respect ,
the _ truly quantum _ contribution to the disturbance in a measurement is related to the map @xmath51 : it is present in those measurement apparatuses with @xmath59 .
we now prove the information disturbance tradeoff @xmath60 any measurement apparatus can be decomposed into a unitary evolution @xmath61 ( the stinespring dilation of the apparatus ) followed by a von neumann projective measurement on a probe p , the so - called _ indirect measurement model _
@xcite , see fig .
[ f : imm ] .
the unitary @xmath61 couples q with the probe p in the apparatus yielding q@xmath62 and p@xmath62
. the joint evolution of probe and system pq in the apparatus can be seen as composed by two complementary quantum channels : a channel q@xmath63p@xmath62 that describes the transfer of information from the system to the state of the probe in p@xmath62 which is then measured yielding the measurement result @xmath4 , and a channel q@xmath63q@xmath62 that evolves the system before the measurement into the system after the measurement conditioned on its result @xmath4 .
these two channels are respectively described by the cp - maps @xmath64\equiv$]tr@xmath65 $ ] and @xmath66\equiv$]tr@xmath67/p_k$ ] , where @xmath68 is the initial pure state of the probe , @xmath69 is the basis representing the projective measurement on the probe , and @xmath70 is the probability of the @xmath5th result .
since the final state of the system in q@xmath62 is given by @xmath71 $ ] with probability @xmath70 , it can be written as @xmath72tr@xmath73\equiv{\cal q}[\rho]$ ] , where the kraus operators of the map @xmath31 can be immediately obtained from the ones of the maps @xmath74 .
the map @xmath31 describes the unitary coupling of system and probe in the apparatus and the successive trace on the probe space , which yields the unconditioned output state .
the system s initial state @xmath6 , expanded on its eigenvectors @xmath75 is given by @xmath76 .
we use the holevo - schumacher - westmoreland theorem @xcite with an alphabet composed by @xmath77 with probability @xmath16 flowing through a channel described by @xmath78 .
such theorem implies that the mutual information @xmath79 between the variable @xmath18 and the measurement results @xmath5 ( whatever measurement strategy is employed ) is upper bounded as @xmath80)-\sum_j\lambda_j s\big({\cal p}\big[|j\rangle\langle j|\big]\big ) \;\labell{hsw}.\end{aligned}\ ] ] the system in space q can be purified by adding an auxiliary reference space r , so that the system in qr is initially in a pure state @xmath36 .
the entropy @xmath81)$ ] of the probe just before the final von neumann measurement is then equal to the entropy @xmath82q@xmath62r@xmath83 of the joint state in q@xmath62r@xmath62 of the system and the reference after the interaction @xmath61 .
in fact , the initial state in pqr is pure and it is evolved into p@xmath62q@xmath62r@xmath62 by a unitary evolution .
thus , @xmath84)=s(\mbox{q}'\mbox{r}')=s_e(\rho,{\cal q } ) \;\labell{ppp1},\end{aligned}\ ] ] where @xmath85 is the exchange entropy @xcite of the map @xmath31 .
it is defined as the entropy of the joint q@xmath62r@xmath62 output state of system and purification - reference , i.e. @xmath86\big)$ ] . moreover , the entropy in q@xmath62 satisfies @xmath87 ) \labell{ppp2}\\\nonumber&&= s\big(\sum_j\lambda_j{\cal q}\big[|j\rangle\langle j|\big]\big)\leqslant \sum_j\lambda_j s\big({\cal q}\big[|j\rangle\langle j|\big]\big)+h(\lambda_j ) , \end{aligned}\ ] ] where @xmath88 is the shannon entropy of the probability distribution @xmath16 , and where the inequality @xmath89 ( valid for all probabilities @xmath90 and states @xmath91 ) has been used @xcite .
notice that if the system in q is initially in a pure state @xmath75 , the entropy of the output of the two channels @xmath78 and @xmath31 coincides since the entropy of the joint system pq is initially null .
hence , @xmath92\big)=s\big({\cal p}\big[|j\rangle\langle j|\big]\big)$ ] , so that eq .
( [ ppp2 ] ) implies @xmath93\big)\geqslant s({\cal q}[\rho])-s(\rho ) \;\labell{ppp3}.\end{aligned}\ ] ] joining eqs .
( [ hsw ] ) , ( [ ppp1 ] ) and ( [ ppp3 ] ) , we find @xmath94)+s(\rho)=s(\rho)-i_c(\rho,{\cal q } ) = d$ ] , thus proving eq .
( [ infodis ] ) .
notice that such proof works also in the case in which the input and output hilbert spaces q and q@xmath62 do not coincide , and in the case of infinite dimensional hilbert spaces @xcite .
there is a simple , not - very - rigorous , intuition behind the preceding proof .
the total quantum information of the initial state @xmath6 can be quantified by @xmath95 qubits .
the unitary evolution @xmath61 transfers @xmath96 of them to the probe space p , where the projective measurement can return a number of bits @xmath97 , due to the holevo bound @xcite . the remaining @xmath98 qubits constitute an upper bound to the quantum capacity to transfer the quantum information in the initial state through the channel q@xmath63q@xmath62 consisting of the measurement apparatus .
the quantum capacity is measured by the coherent information @xcite , so that @xmath99 .
@xmath100 we now deduce the equality conditions for the information disturbance bound @xmath1 , showing that it is achievable .
the equality in the holevo - schumacher - westmoreland relation of eq .
( [ hsw ] ) is achieved if the alphabet states @xmath101 $ ] commute @xcite .
moreover , the equality in the relation @xmath89 , which was employed in eq .
( [ ppp2 ] ) , is achieved if and only if the states @xmath91 have support on orthogonal subspaces @xcite ( which implies that they commute ) .
thus , we have equality @xmath3 if and only if the channel @xmath78 maps different eigenvectors @xmath75 of the initial state @xmath6 into orthogonal subspaces .
a typical example is a projective measurement whose kraus operators are projectors on the basis @xmath75 .
it is a purely informational measurement , where the only uncertainty derives from classical probability .
[ notice that , in a @xmath19-dimensional hilbert space , the converse also holds for measurements with @xmath19 outcomes : if @xmath3 and the measurement povm has @xmath19 elements , then the measurement is a von neumann - type projection , i.e. its kraus operators are of the form @xmath102 where @xmath103 is a fixed unitary and @xmath104 is a basis .
the proof of this assertion follows immediately from the fact that the measurement must map a basis @xmath75 of @xmath19 elements into @xmath19 orthogonal subspaces , which , in a @xmath19-dimensional hilbert space , must then be one - dimensional ] .
the definitions we used for information @xmath79 and disturbance @xmath105 are explicitly dependent both on the input state @xmath6 and on the apparatus . we can forgo the state dependence by averaging on all possible input states with equal weights ( for symmetry reasons ) , i.e. by using a state @xmath106 in a @xmath19-dimensional hilbert space . [ in infinite - dimensional spaces , additional requirements are also necessary , such as using states with upper - bounded energy . ]
thus , we can define a state independent information as @xmath107 and a state
independent disturbance as @xmath108 .
the quantity @xmath109 measures the percentage of the maximum retrievable information that is achieved by the apparatus .
the quantity @xmath110 measures the disturbance the apparatus causes to a completely unknown state .
notice that @xmath111 if and only if the apparatus acts on the state with a unitary transformation ( i.e. it yields no information on the state ) .
in fact , @xmath112 if and only if the map @xmath31 is invertible on all the pure states in the support of @xmath6 @xcite , which for @xmath106 implies that the map is unitary .
we have introduced a new measure @xmath105 of the disturbance ( intended as an irreversible state change ) that a map with kraus operators @xmath22 induces on a system in a state @xmath6 .
we have derived an information disturbance tradeoff for such a quantity in the form @xmath1 , where @xmath2 is the classical info the map @xmath22 returns on the state @xmath6 to the experimenter .
the equality conditions for this bound have been also derived . moreover , a state
independent tradeoff @xmath113 was obtained , which bounds the percentage of the maximum achievable information @xmath114 with the disturbance @xmath110 caused to a completely unknown input state .
even though we found a valid information disturbance tradeoff , we have to conclude that the disturbance definition @xmath0 we used here may not be the appropriate definition to capture the true spirit of heisenberg s intuition ( the `` uncertainty principle '' ) .
in fact , a `` classical '' measurement ( or a purely informational state change ) where one gains information on which , out of a set of orthogonal configurations , our state is in , will perturb the state in an irreversible manner even though there is no dynamical interaction on the system .
heisenberg , on the other hand , analyzed only the irreversible state changes induced by dynamical actions ( measurements ) on the system . in this sense , the information disturbance relation derived here might be considered as an extension of heisenberg s intuition on the disturbance a measurement induces .
here we prove that the entropic disturbance @xmath0 is continuous .
more rigorously , we prove the following two statements : _
i ) _ @xmath115 , i.e. @xmath116 , implies @xmath117 , where the trace distance @xmath118 is defined as @xmath119tr@xmath120/2 $ ] ; _ ii ) _
@xmath121 , i.e. @xmath122,{\cal q}'[\rho])\to 0 $ ] , implies @xmath123 .
proof of _ i ) _ : start from @xmath124\labell{st}\\\nonumber & & -[s({\cal q}(\rho))-s({\cal q}(\rho'))]+[s_e(\rho,{\cal q})-s_e(\rho',{\cal q } ) ] \;.\end{aligned}\ ] ] the first bracket in eq .
( [ st ] ) tends to zero for @xmath115 thanks to the continuity of the entropy .
it derives from fannes inequality @xcite , according to which @xmath125 with the function @xmath126 for @xmath127 . the second bracket in eq .
( [ st ] ) analogously goes to zero since it is bounded by the first : the contractivity of cp - maps @xcite implies that @xmath128 . to show that also the last bracket tends to zero , recall that the exchange entropy can be written as @xmath129 with the matrix @xmath130 defined by @xmath131tr@xmath132 $ ] , @xmath14 being the kraus operators of @xmath31 @xcite .
thus , fannes inequality implies @xmath133 , and for @xmath115 , we have @xmath134 .
in fact , since @xmath130 is hermitian , @xmath135\mbox{tr}[a_i^\dag a_j(\rho-\rho')]\nonumber\\&\leqslant & \mbox{tr}[a_j^\dag a_ia_i^\dag a_j]\mbox{tr}[(\rho-\rho')^2 ] \;\labell{fin},\end{aligned}\ ] ] where we used the schwarz inequality for the hilbert - schmidt scalar product of operators : @xmath136tr@xmath137 $ ] .
w. heisenberg , zeitsch . phys . * 43 * , 172 ( 1927 ) , english translation : j. a. wheeler and h. zurek , _ quantum theory and measurement _ ( princeton univ . press , 1983 ) , pg . 62 .
h. p. robertson , phys . rev . * 34 * , 163 ( 1929 ) .
a. peres _ quantum theory : concepts and methods _
( kluwer ac .
dordrecht , 1993 ) .
g. m. dariano , fortschr .
phys . * 51 * , 318 ( 2003 ) . c. a. fuchs and a. peres , phys .
a * 53 * , 2038 ( 1996 ) ; c. a. fuchs , fortschr . phys . * 46 * , 535 ( 1998 ) ; h. barnum , quant - ph/0205155 ( 2002 ) ; l. maccone , phys . rev .
a * 73 * , 042307 ( 2006 ) .
m. ozawa , ann . phys . * 311 * , 350 ( 2004 ) . c. h. bennett , g. brassard , and n. d. mermin , phys .
lett . * 68 * , 557 ( 1992 ) ; m. ban , j. phys .
a : math . gen . * 32 * , 6527 ( 1999 ) ; k. banaszek , phys . rev . lett . *
86 * , 1366 ( 2001 ) ; p. arrighi , int . j. quantum inf . * 2 * , 341 ( 2004 ) ; f. sciarrino , m. ricci , f. de martini , r. filip , and l. mita , jr .
lett . * 96 * , 020408 ( 2006 ) ; m. christandl and a. winter , quant - ph/0501090 ( 2005 ) . s. d. bartlett , t. rudolph , r. w. spekkens , p. s. turner , new j. phys . * 8 * , 58 ( 2006 ) .
k. kraus _ states effects and operations _ ( springer - verlag , berlin , 1983 ) ; k. kraus , phys .
d * 35 * , 3070 ( 1987 ) .
r. jozsa , d. robb , and w. k. wootters , phys .
a * 49 * , 668 ( 1994 ) .
f. buscemi , g. m. dariano , m. keyl , p. perinotti , and r. f. werner , j. of math .
* 46 * , 082109 ( 2005 ) .
m. a. nielsen and i. l. chuang , _ quantum computation and quantum information _
( cambridge univ . press ,
cambridge , 2000 ) .
b. schumacher , phys .
a * 54 * , 2614 ( 1996 ) .
b. schumacher and m. a. nielsen , phys .
a * 54 * , 2629 ( 1996 ) .
s. lloyd , phys .
a * 55 * , 1613 ( 1997 ) .
a. s. holevo , ieee trans .
theory * 44 * , 269 ( 1998 ) ; b. schumacher and m. d. westmoreland , phys .
a * 56 * , 131 ( 1997 ) .
h. p. yuen , m. ozawa , phys .
70 * , 363 ( 1992 ) .
h. barnum , m. a. nielsen , and b. schumacher , phys .
a * 57 * , 4153 ( 1998 ) ; i. devetak , ieee trans .
theory * 51 * , 44 ( 2005 ) .
m. j. w. hall , phys .
a * 55 * , 100 ( 1997 ) .
m. fannes , commun .
* 31 * , 291 ( 1973 ) . m. b. ruskai , rev . math
* 6 * , 1147 ( 1994 ) . | we show the flaws found in the customary fidelity - based definitions of disturbance in quantum measurements and evolutions .
we introduce the `` entropic disturbance '' @xmath0 and show that it adequately measures the degree of disturbance , intended essentially as an irreversible change in the state of the system .
we also find that it complies with an information disturbance tradeoff , namely the mutual information between the eigenvalues of the initial state and the measurement results is less than or equal to @xmath0 .
the correct interpretation of the heisenberg uncertainty relations @xcite connects the uncertainty ( or spread ) in the measurement results of one observable to the spread of another observable in the initial state of the system @xcite . the true spirit of heisenberg s intuition @xcite , however , was that any measurement ( i.e. extraction of classical information from a quantum system ) necessarily entails some kind of disturbance to the measured system .
this intuition has yet to be captured in a truly general information disturbance tradeoff relation .
nonetheless , several of these relations have been put forth and cover many conceivable situations ( e.g. see @xcite ) .
the main problem in deriving a general version of such a tradeoff lies in the identification of an appropriate definition for the disturbance . in this paper
we explore the consequences of defining the disturbance as an irreversible change in the state of the system .
we proceed axiomatically , by enumerating the abstract properties that a disturbance measure should possess , and we show that the customarily used fidelity - based disturbance measures do not satisfy them .
we introduce one that does , namely the `` entropic disturbance '' @xmath0 .
a simple , general information disturbance tradeoff is then derived , namely @xmath1 ( @xmath2 being the mutual information between the eigenvalues of the initial state and the measurement results ) .
its interpretation is straightforward : every time an amount @xmath2 of information is obtained from the measurement apparatus , a disturbance @xmath0 at least as big is introduced on the system , but , obviously , the system can be disturbed by a process that returns little information .
we also derive the equality conditions @xmath3 showing that the @xmath1 bound is tight , and proving ( as is to be expected @xcite ) that von neumann - type measurements are among the least disturbing amid all measurements that retrieve the same information .
finally , we give a state independent tradeoff by averaging over all possible input states .
this provides a measure of the global disturbance introduced by the apparatus or by a generic evolution . |
no hydromagnetic dynamo can exist driven only by differential rotation @xcite , but it is known that such dynamos can exist if the turbulence is helical in the sense that its kinetic helicity @xmath4 and/or its current helicity @xmath5 do not vanish .
here @xmath6 and @xmath7 are the fluctuating parts of the flow @xmath8 and magnetic field @xmath9 .
this condition of non - vanishing helicity is clearly fulfilled if the turbulence is rotating and stratified . in such turbulence a pseudo - scalar exists which allows the pseudo - scalars ( [ 1 ] ) and ( [ 2 ] ) to take finite values .
the same is true for linear shear flows where the stratified turbulence in the presence of the shear also can form a kinetic helicity ( see rdiger & kitchatinov 2006 ) .
the simplest pseudo - scalar is the scalar product @xmath10 with @xmath11 as the gradient vector of the turbulence ( or the fluid density ) and @xmath12 the rotation vector . in spheres
the gradient vector @xmath11 is mainly radial so the pseudo - scalar @xmath10 has opposite signs in the two hemispheres , and vanishes at the equator .
because of the close relationship of the helicity to the @xmath0-effect in the mean - field electrodynamics of turbulent media , @xmath13 ( the dots represent higher derivatives of @xmath14 ) the above - mentioned sign rules are also the sign rules of the @xmath0-effect i.e. @xmath15 .
this sort of @xmath0-effect only exists for inhomogeneous turbulence . in planetary cores , however , and also in laboratory experiments the only inhomogeneities result from boundary conditions , as the density gradients are negligible .
one can show that under the presence of magnetic background fields other inhomogeneities also form pseudo - scalars and , as a consequence , lead to new mechanisms for an @xmath0-effect ( e.g. gellert , rdiger & elstner 2008 ) . in the present study
we demonstrate that instabilities due to inhomogeneous magnetic fields also lead to finite values of the helicities ( [ 1 ] ) and ( [ 2 ] ) , and in accord with ( [ 3 ] ) also to finite values of @xmath0 .
indeed , in the presence of electric currents the simplest existing pseudo - scalar is @xmath16 which does not vanish for helical field geometries .
we show that for such background fields the small - scale helicities obtain final values with the opposite sign as the helicity of the background field .
according to the rayleigh criterion , in the absence of mhd effects an ideal flow is stable against axisymmetric perturbations whenever the specific angular momentum increases outward ( r^2)^2 > 0 , [ ray ] where @xmath17 is the angular velocity , and ( @xmath18 , @xmath19 , @xmath20 ) are cylindrical coordinates in a right - handed system . in the presence of an azimuthal magnetic field
@xmath21 , this criterion is modified to ( r^2)^2- ( ) ^2 > 0 , [ mich ] where @xmath22 is the permeability and @xmath23 the density @xcite .
note also that this criterion is both necessary and sufficient for ( axisymmetric ) stability .
in particular , _ all _ ideal flows can thus be destabilized , by azimuthal magnetic fields with the right profiles ( steeply increasing outwards ) and amplitudes . on the other hand , for nonaxisymmetric modes one has ( r b_0,^2 ) < 0 [ tay ] as the necessary and sufficient condition for stability of an ideal fluid at rest @xcite .
again , outwardly increasing fields are thus unstable .
if ( [ tay ] ) is violated , the most unstable mode has azimuthal wave numbers of @xmath24 .
we are interested in the stability of the background field @xmath25 , with @xmath26 , and the flow @xmath27 .
the governing equations for the flow @xmath8 and the field @xmath9 are @xmath28 @xmath29 and @xmath30 where @xmath31 is the pressure , @xmath32 the kinematic viscosity and @xmath33 the magnetic diffusivity .
their ratio is the magnetic prandtl number = .
[ pm ] from now on we drop the subscripts from the large - scale values so that the total flow is @xmath34 and the total field is @xmath35 . the stationary background solution is = a_+ , b_=a_b r+ , [ basic ] where @xmath36 , @xmath37 , @xmath38 and @xmath39 are constants defined by a_=_in , b_=_in r_in^2 , + a_b= , b_b = b_inr _ in , [ ab ] with @xmath40 @xmath41 and @xmath42 are the radii of the inner and outer cylinders , @xmath43 and @xmath44 are their rotation rates , and @xmath45 and @xmath46 are the azimuthal magnetic fields at the inner and outer cylinders .
a field of the form @xmath47 is generated by running an axial current only through the inner region @xmath48 , whereas a field of the form @xmath49 is generated by running a uniform axial current through the entire region @xmath50 , including the fluid .
given the @xmath20-component of the electric current , @xmath51 , one finds for the current helicity of the background field @xmath52 which may be positive , negative , or zero .
the inner value @xmath53 is normalized by the uniform vertical field , i.e. @xmath54 for our standard profile @xmath55 we have @xmath56 for the helicity of the background field . for fixed toroidal field amplitude
this quantity scales as @xmath57 : @xmath58 the sign of @xmath59 determines the sign of the current helicity . if the toroidal field is due to the interaction of a poloidal field with a differential rotation with negative shear then @xmath59 is negative and vice versa .
interchanging @xmath60 simply interchanges left and right spirals , @xmath61 .
as usual , the toroidal field amplitude is measured by the hartmann number and the global rotation by the reynolds number , i.e. = , re=. [ ha ] @xmath62 is used as the unit of length , @xmath63 as the unit of velocity and @xmath53 as the unit of the azimuthal fields .
frequencies , including the rotation @xmath64 , are normalized with the inner rotation rate @xmath65 . the lundquist number @xmath66 is defined by @xmath67
. the magnetic - diffusion frequency is @xmath68 and then the alfvn frequency @xmath69 is _ a = .
[ oa ] throughout the whole paper numerical values of helicities are given in units of @xmath70 .
in this notation the helicity ( [ heli ] ) of the background field can be written as _ 0 _ 0 = _
a^2 d , [ heli2 ] the boundary conditions associated with the perturbation equations are no - slip for @xmath6 and perfectly conducting for @xmath7 , at both @xmath71 and @xmath72 , where we fix @xmath73 , i.e. @xmath74 . the computational domain is periodic in @xmath20 .
the nonlinear mhd code used for the solution of eqs .
( [ mhd ] ) and ( [ mhd1 ] ) has been described in detail by @xcite ( see also fournier et al .
2004 , 2005 ) .
fig . [ growth ] shows the growth rates for a purely toroidal field ( @xmath75 ) , and no rotation .
we see that beyond the critical lundquist number , the growth rate is essentially linear , i.e. _ gr _ a , [ ogr ] where @xmath76 has the steepest slope , and is thus more unstable than both @xmath77 and @xmath78 .
growth rate curves for various magnetic prandtl numbers , for stationary cylinders and a purely toroidal field , with @xmath55 .
the growth rate and the magnetic field are normalized with the magnetic - dissipation frequency @xmath79.,width=321,height=226 ] the azimuthal wavenumber of the modes shown in fig .
[ growth ] is @xmath80 . for a purely toroidal field , @xmath81 , corresponding to left- and right - handed spirals , are degenerate , and necessarily have exactly the same growth rate curves .
see also @xcite , who obtained the same effect in magnetorotational instabilities , and @xcite , who consider instabilities of toroidal fields in spheres .
we next consider the nonlinear equilibration of these modes . as fig .
[ heltor ] shows , even though @xmath80 are degenerate , the equilibrated solutions do not consist of equal mixtures of both modes . instead , either the left or the right mode wins out , and completely suppresses the other .
which mode one obtains depends on the precise initial conditions .
if these already favor one mode , then ( not surprisingly ) that one wins , but if the initial condition is evenly balanced between the two modes , it is ultimately just numerical noise that determines which mode wins . eventually though one mode always wins ; the solution consisting of an equal mixture of both is unstable .
spontaneous parity - breaking bifurcations of this type are well known in classical , non - magnetic taylor - couette flow ( e.g. hoffmann et al .
2009 , altmeier et al .
2010 and reference therein ) , but are almost unknown in magnetohydrodynamic problems . to the best of our knowledge ,
the only other example is in the very recent work by @xcite . given the importance of helicity in mean - field dynamics , any effect that generates helicity from an underlying basic state without helicity could be significant .
the turbulence intensities for flow ( top ) and field ( bottom ) fluctuations in units of @xmath82 .
note that once @xmath59 exceeds @xmath83 it has virtually no further influence .
@xmath55 , @xmath84 , @xmath76.,title="fig:",width=321,height=226 ] the turbulence intensities for flow ( top ) and field ( bottom ) fluctuations in units of @xmath82 . note that once @xmath59 exceeds @xmath83 it has virtually no further influence .
@xmath55 , @xmath84 , @xmath76.,title="fig:",width=321,height=226 ] we next present two series of solutions where @xmath14 includes an axial component ( @xmath85 ) .
in contrast to figs . [ growth ] and
[ heltor ] , we also include a differential rotation here .
the profiles of the basic state field and flow are fixed at @xmath55 and @xmath84 .
their amplitudes are @xmath86 and @xmath87 for the first series , and @xmath88 , @xmath89 for the second .
the first series is thus rotationally dominated ( @xmath90 ) , whereas the second is magnetically dominated ( @xmath91 ) .
the astrophysically relevant case is rotationally dominated , which is _ _ n__ot the classical realization of the tayler instability .
we have called this instability the azimuthal magnetorotational instability ( amri , see hollerbach , teeluck & rdiger 2009 ) . for both series of runs fig .
[ energy ] shows the kinetic and magnetic turbulence intensities @xmath92 and @xmath93 . for sufficiently large @xmath59
its influence is very small ; the axial component of @xmath94 is then so weak that it has no further influence .
this is not true for small @xmath59 , where the axial field starts to dominate . for @xmath95 the kink - instability
is strongly stabilized @xcite and the resulting energies of the perturbations are reduced .
the kinetic and current helicities of the nonaxisymmetric perturbations as functions of @xmath59 .
top : for @xmath90 ( @xmath86 , @xmath87 ) .
bottom : for @xmath91 ( @xmath88 , @xmath89 ) .
the dash - dotted lines indicate the limits @xmath96 of the kinetic helicity of the left and right modes in fig .
[ heltor].,title="fig:",width=321,height=226 ] the kinetic and current helicities of the nonaxisymmetric perturbations as functions of @xmath59 .
top : for @xmath90 ( @xmath86 , @xmath87 ) .
bottom : for @xmath91 ( @xmath88 , @xmath89 ) .
the dash - dotted lines indicate the limits @xmath96 of the kinetic helicity of the left and right modes in fig .
[ heltor].,title="fig:",width=321,height=226 ] fig .
[ helk ] shows the kinetic ( [ 1 ] ) and current ( [ 2 ] ) helicities for the two series of runs .
for both series , both helicities have the opposite sign as @xmath59 ( see also kpyl & brandenburg 2009 ) for comparison ) . in stating this result , it is important though to specify carefully the nature of the initial conditions used in each run . for @xmath97 ,
the basic state has a sufficiently strong handedness that it forces the instabilities to have a particular parity as well , which as indicated turns out to be opposite to that of the basic state .
if one then gradually increases @xmath59 , each time using the previous solution as the new initial condition , this parity of the instabilities is preserved all way to @xmath98 , where the basic state no longer has a handedness , and both left and right instabilities could exist equally well , as in fig .
[ heltor ] . that is , by the time one reaches @xmath99 , say , the basic state makes sufficiently little distinction between left and right modes that both could exist , but because of the way we have reached @xmath99 , we consistently obtain the right mode .
however , suppose one does the following experiment now : take the right mode at @xmath99 , swap its parity to be left , and use that as a new initial condition for a series of runs where @xmath59 is now gradually reduced .
eventually there comes a point where the basic state s handedness is sufficiently great that it no longer allows the instability to have the ` wrong ' parity , and the solution reverts back to the right mode .
this feature that both left and right modes are allowed for sufficiently large @xmath59 ( where the degeneracy between the two modes is only weakly broken ) but not for smaller @xmath59 ( where the degeneracy is strongly broken ) is in many ways analogous to an imperfect pitchfork bifurcation .
we have also calculated the @xmath0-effect in ( [ 3 ] ) with the same averaging procedure over the azimuth . because of the complex structure of the background field it is even possible to determine parts of the tensorial structure of the @xmath0-tensor .
in particular we are interested in the signs and amplitudes of the @xmath0-effect in both azimuthal and axial directions . according to the general rule that the azimuthal @xmath0-effect is anticorrelated with the ( kinetic ) helicity we expect the azimuthal @xmath0-effect to be positive for @xmath100 .
the expected sign of the axial @xmath0-effect is not clear .
there are theories and simulations leading to @xmath1 and @xmath2 with opposite signs ( see rdiger & hollerbach 2004 for an overview ) .
we should not be surprised to find a similar behavior in the present simulations .
it also means that any dynamo with very weak differential rotation can not be treated with a scalar @xmath0-effect .
the @xmath0-effect in the slowly rotating case , @xmath101 , @xmath89 , @xmath88 .
the top shows @xmath1 , the bottom @xmath2 .
the other parameters are @xmath102 , @xmath84 , @xmath55 , @xmath76.,title="fig:",width=321,height=226 ] the @xmath0-effect in the slowly rotating case , @xmath101 , @xmath89 , @xmath88 .
the top shows @xmath1 , the bottom @xmath2 .
the other parameters are @xmath102 , @xmath84 , @xmath55 , @xmath76.,title="fig:",width=321,height=226 ] the @xmath0-effect in the rapidly rotating case , @xmath90 , @xmath87 , @xmath86 .
the top shows @xmath1 , the bottom @xmath2 .
the other parameters are @xmath102 , @xmath84 , @xmath55 , @xmath76.,title="fig:",width=321,height=226 ] the @xmath0-effect in the rapidly rotating case , @xmath90 , @xmath87 , @xmath86 .
the top shows @xmath1 , the bottom @xmath2 .
the other parameters are @xmath102 , @xmath84 , @xmath55 , @xmath76.,title="fig:",width=321,height=226 ] figs .
[ alpha1 ] and [ alpha2 ] give the results for slow and rapid rotation . on the basis of eq .
( [ 3 ] ) the dimensionless @xmath0-effect in the form @xmath103 is plotted for the components @xmath1 and @xmath2 . in both cases ,
these two components have opposite signs , with @xmath104 and @xmath105 almost everywhere in the meridional plane . this anti - correlation between the two components
is also strongest in the center of the gap , and weakest near the boundaries .
it is therefore not caused by the boundaries .
that figs .
[ alpha1 ] and [ alpha2 ] show such similar results is surprising , and is one of the basic results of this paper .
the influence of rotation on @xmath0 is evidently rather weak .
the fact that contrary to previous results for rotating convection
@xmath1 is actually smaller for rapid rotation than for slow rotation illustrates just how different these magnetic - induced helicities are from some of the previous results .
finally , fig .
[ alpha ] shows how the amplitudes of @xmath1 vary with @xmath59 , being roughly inversely proportional in both cases .
the dimensionless dynamo number @xmath106 of the azimuthal @xmath0-effect for @xmath101 and @xmath90 . for higher @xmath59 ( smaller @xmath107 ) the @xmath0-effect decreases like @xmath108 with @xmath109 .
@xmath84 , @xmath55 , @xmath76.,width=321,height=226 ] to consider some possible astrophysical implications of these results , imagine a disk dynamo with dominant field components @xmath21 and @xmath110 .
dynamo waves of @xmath3-type require for self - excitation that the product of ( [ calfa ] ) and @xmath111 exceeds a critical value of order unity , i.e. @xmath112 . the ratio of the amplitudes of the field components @xmath21 and @xmath110 follows the simple rule @xmath113 so that dynamo excitation requires @xmath114 we know from fig .
[ alpha ] that @xmath115 with @xmath116 , so that ( [ bfr1 ] ) gives , at least as an order - of - magnitude estimate , the condition @xmath117 for self - excitation of a dynamo with differential rotation and current - driven @xmath0-effect . for disk dynamos @xmath110 dominates @xmath107 , and for spherical dynamos @xmath110 is comparable to @xmath107 . in both cases
the condition for self - excitation becomes @xmath118 , which can not be fulfilled according to fig .
[ alpha ] , which suggests instead that @xmath119 .
the @xmath0-effect due to the current helicity of the background field appears as much too small to allow the operation of an @xmath3-dynamo .
another argument concerns the growth rate of such a dynamo ( if it exists at all ) in relation to the very long magnetic diffusion times in radiative zones .
assume that for self - excitation @xmath120 , then the growth rate @xmath121 is given by @xmath122 hence , only for @xmath118 the growth time of the dynamo would be _ shorter _ than the magnetic diffusion time @xmath123 , which is known to be of order gyr for the radiative interior of stars .
one can also argue as follows .
the relation ( [ omgr ] ) also reads @xmath124 independent of the magnetic diffusivity . on the other hand , for given @xmath106 ( [ omgr1 ] ) states @xmath125 which for the computed value @xmath126 taken from fig .
[ alpha ] and @xmath127 @xmath128/s for the solar core leads to values of order @xmath129 s@xmath130 , i.e. to growth times of order 10 myr . as it is typical for @xmath3-dynamos their growth times are only slightly shorter than the basic magnetic decay time .
we have shown that the current - driven instability of helical large - scale fields does produce small - scale helicity ( kinetic plus current helicity ) and even @xmath0-effects , but the resulting numerical values seem to be too small for the operation of large - scale dynamos in radiative zones of early - type stars . | helical magnetic background fields with adjustable pitch angle are imposed on a conducting fluid in a differentially rotating cylindrical container .
the small - scale kinetic and current helicities are calculated for various field geometries , and shown to have the opposite sign as the helicity of the large - scale field .
these helicities and also the corresponding @xmath0-effect scale with the current helicity of the background field .
the @xmath0-tensor is highly anisotropic as the components @xmath1 and @xmath2 have opposite signs .
the amplitudes of the azimuthal @xmath0-effect computed with the cylindrical 3d mhd code are so small that the operation of an @xmath3 dynamo on the basis of the current - driven , kink - type instabilities of toroidal fields is highly questionable . in any case the low value of the @xmath0-effect would lead to very long growth times of a dynamo in the radiation zone of the sun and early - type stars of the order of mega - years .
[ firstpage ] magnetic fields - instabilities - stars : magnetic field - dynamo |
a large of neurons is involved in any computation , and the presence of non - trivial correlations makes understanding the mechanisms of computation in brain a difficult challenge @xcite . the simplest model for describing multi - neuron spike statistics is the pairwise ising model @xcite .
the inference of the couplings of an ising model from data , the _ inverse ising problem _ , has recently attracted attention , see @xcite where data from a simulated cortical network were considered ; the general idea is to find an ising model with the same means and pairwise correlations as the data .
several approximate methods can be used , like that by sessak and monasson @xcite or the inversion of tap equations @xcite : equal - time correlations from data are used in those methods @xcite .
the flow of information between spins is related , instead , to correlations at different times between variables : it can be expected that measuring the information flow between spins one may improve the estimate of couplings from data .
two major approaches are commonly used to estimate the information flow between variables , transfer entropy @xcite and granger causality @xcite .
recently it has been shown that for gaussian variables granger causality and transfer entropy are entirely equivalent as the following relation holds : _ granger causality = 2 transfer entropy_. this result provides a bridge between autoregressive and information - theoretic approaches to causal inference from data @xcite . the purpose of this work is to explore the use of granger causality to learn ising models from data .
the _ inverse ising problem _ is here seen as belonging to the more general frame of the inference of dynamical networks from data , a topic which has been studied in recent papers @xcite : its relevance is due to the fact that dynamical networks @xcite model physical and biological behavior in many applications @xcite .
we show that for weak couplings , the linear granger causality of each link is two times the corresponding transfer entropy , also for ising models : this occurrence justifies the use of autoregressive approaches to the inverse ising problem . in the same limit , for each link , the following relation exists between the coupling ( @xmath1 ) and the causality ( @xmath2 ) : @xmath3 . in the case of limited samples ,
granger causality gives poor results : almost all the connections are not assessed as significative for low number of samples . in these cases ,
we propose the use of @xmath0 least squares method @xcite , a penalized autoregressive approach tailored to embody the sparsity assumption , to recover the non -vanishing connections of a sparse ising model ; as expected the @xmath0 approach outperforms granger causality in this case .
finally we show that nonlinear granger causality is related to multi - spin interactions .
the paper is organized as follows . in the next section
we briefly recall the notions of granger causality and transfer entropy , and we also describe the ising models that we use for simulations . in section iii
we describe our results on fully connected models , sparse ising models and models with higher order spin interactions .
some conclusions are drawn in section iv .
in this section we review the notions of granger causality analysis and transfer entropy .
we also discuss the application of these methods to binary time series arising from ising models .
granger causality has become the method of choice to determine whether and how two time series exert causal influences on each other @xcite .
it is based on prediction : if the prediction error of the first time series is reduced by including measurements from the second one in the linear regression model , then the second time series is said to have a causal influence on the first one @xcite .
the estimation of linear granger causality from fourier and wavelet transforms of time series data has been recently addressed @xcite .
the nonlinear generalization of granger causality has been developed by a kernel algorithm which embeds data into a hilbert space , and searches for linear granger causality in that space @xcite ; the embedding is performed implicitly , by specifying the inner product between pairs of points @xcite , and a statistical procedure is used to avoid over - fitting .
quantitatively , let us consider @xmath4 time series @xmath5 @xcite ; the lagged state vectors are denoted @xmath6 @xmath7 being the window length .
let @xmath8 be the mean squared error prediction of @xmath9 on the basis of all the vectors @xmath10 ( corresponding to the kernel approach described in @xcite ) : @xmath11 is equal to @xmath12 , where @xmath13 , the predicted values of @xmath9 using @xmath10 , is the projection of @xmath9 on a suitable space @xmath14 .
the prediction of @xmath9 on the basis of all the variables but @xmath15 , @xmath16 , corresponds instead to the projection on a space @xmath17 with @xmath18 .
@xmath19 represents the information that one gains from the knowledge of @xmath15 .
the multivariate granger causality index @xmath20 is defined as the ( normalized ) variation of the error in the two conditions , i.e. @xmath21 note that the numerator , in the equation above , coincides with the projection of @xmath9 on @xmath19 : as described in @xcite , one may write @xmath22 where @xmath23 are suitable pearson s correlations . by summing , in equation ( [ somma ] ) , only over significative correlations ,
a _ filtered _ linear granger causality index is obtained which measures the causality without further statistical test . in @xcite it has been shown that not all the kernels are suitable to estimate causality .
two important classes of kernels which can be used to construct nonlinear causality measures are the _ inhomogeneous polynomial kernel _ ( whose features are all the monomials , in the input variables , up to the @xmath24 degree ; @xmath25 corresponds to linear granger causality ) and the _ gaussian kernel_. note that in @xcite a different index of causality is adopted : @xmath26};\ ] ] @xmath27 and @xmath2 coincide at small @xmath2 .
the choice of the window length @xmath7 is usually done using the standard cross - validation scheme @xcite ; as in this work we know how data are generated , here we use @xmath28 .
the formalism of granger causality is constructed under the hypothesis that time series assume continuous values @xmath29 . in recent papers
the application of granger causality to data in the form of phases has been considered @xcite .
even though there is not theoretical justification in the case of binary variables , here we apply the formalism of granger casuality to @xmath4 binary time series @xmath30 , by substituting @xmath31 and @xmath32 in this work we justify the application of granger causality to binary time series in terms of its relation with transfer entropy . using the same notation as in the previous subsection , the transfer entropy index @xmath33 is given by @xcite @xmath34 and measures the flow of information from @xmath35 to @xmath36 . for gaussian variables
it has been shown @xcite that causality is determined by the transfer entropy and @xmath37 ; hence @xmath38 for gaussian variables .
the probabilities @xmath39 s , in ( [ tau ] ) , must be estimated from data using techniques of nonlinear time series analysis @xcite . in the case of binary variables @xmath40 , the number of configurations is finite and the integrals in ( [ tau ] ) become sums over configurations ; the probabilities can be estimated as frequencies in the data - set at hand .
therefore @xmath41 where @xmath42 is the fraction of times that the configuration @xmath43 is observed in the data set , and similar definitions hold for the other probabilities .
we remark that the number of configurations increases exponentially as the number of spins grows , hence the direct evaluation of ( [ te ] ) is feasible only for systems of small size .
the binary time series analyzed in this work are generated by parallel updating of ising variables @xmath44 : @xmath45 where the local fields are given by @xmath46 with couplings @xmath47 .
starting from a random initial configuration of the spin , equations ( [ update ] ) are iterated and , after discarding the initial transient regime , @xmath48 consecutive samples of the system are stored for further analysis .
in order to generalize granger causality to discrete variables , we consider the regression function of ( [ update ] ) . for weak couplings the conditional expectation ( which coincides with the regression function @xcite ) can be written @xmath49 and is a linear function of the couplings .
therefore the linear causality is @xmath50 at the lowest order in @xmath1 s .
analogously , expanding eq.([te ] ) at the lowest order , the transfer entropy reads @xmath51 .
this means that , for low couplings , the value of @xmath1 , for any given link , determines both the transfer entropy and the linear granger causality for that link , and the two quantities differ only by the factor @xmath52 .
the same relation , proved in the gaussian case , holds also for ising models at weak couplings .
being related to the transfer entropy , granger causality thus measures the information flow for these systems , and this justifies the use of granger causality methods for ising systems .
synaptic couplings are directed , so @xmath47 is not in general equal to @xmath53 ( the equilibrium ising model requires symmetric couplings @xcite ) .
therefore we consider an asymmetric system of spins with couplings @xmath47 chosen at random from a normal distribution with zero mean and standard deviation @xmath54 ; no self interactions are assumed ( @xmath55 ) . in figure ( [ f1 ] )
we report the plot of the numerical estimates of linear causality and transfer entropy , as a function of @xmath1 , for several realizations of the couplings and for some values of @xmath54 ; the simulations confirm that for low couplings a one - one correspondence exists between causality and transfer entropy . in figure ( [ f2 ] )
we depict , as a function of the coupling @xmath1 , both the linear causality and the transfer entropy in a typical asymmetric model of six spins with @xmath56 and @xmath57 samples , and @xmath58 . in figure ( [ f3 ] )
we depict , as a function of the number of samples @xmath48 , the difference between the values of transfer entropy and causality ( as estimated on @xmath59 samples ) and their estimates based on @xmath48 samples ( the difference is averaged over 1000 runs of the ising system ) : at low @xmath48 the estimates of causalities are more reliable than those of transfer entropy .
moreover , the computational complexity of the estimation of transfer entropy is much higher than those corresponding to the evaluation of granger causality .
another interesting situation is all the couplings being equal to a positive quantity @xmath1 ( still , without self - interactions ) . in figure ( [ f4 ] )
we depict the granger causality and the transfer entropy as a function of @xmath1 ( these quantities are the same for all the links , due to symmetry ) . for small values of @xmath1
both quantities coincide with the values that correspond to @xmath1 in the asymmetric model , and the relation @xmath60 holds . on the other hand ,
as @xmath1 increases , this relation is more and more violated .
the departure of the ratio @xmath61 from 2 is connected to the emergence of feedback effects in the system , see e.g. the dependency , on @xmath1 , of the auto - correlation time of the magnetization @xmath62 in fig.([f5 ] ) . in many applications one may hypothesize that the connections among variables are sparse .
the main goal , in those cases , is to infer the couplings which are not vanishing , independently of their strengths , in particular when the number of samples is low .
moreover , in the case of limited data , granger causality gives poor results ; indeed almost all connections would not be assessed as significative ( for a given amount a data , only couplings stronger than a critical value can be recognized by granger causality @xcite ) .
a major approach to sparse signal reconstruction is the @xmath0 regularized least squares method @xcite .
although it has been developed to handle continuous variables , we will apply this method to the configurations of ising models . for each target
spin @xmath63 , the vector of couplings @xmath64 , with @xmath65 , is sought for as the minimizer of @xmath66 where @xmath67 is a regularization parameter and @xmath68 is the @xmath0 norm of the vector of couplings .
as @xmath69 is increased , the number of vanishing couplings in the minimizers increases : @xmath69 controls the sparsity of the solution @xcite .
the strategy to fix the value of @xmath69 we use here is 10-fold cross - validation @xcite : the original sample is randomly partitioned into 10 subsamples and , out of the 10 subsamples , a single subsample is retained as the validation data .
the remaining 9 subsamples are used in ( [ elleuno ] ) to determine the couplings @xmath70 ; the quality of this solution is evaluated as the average number of errors on the validation data .
the cross - validation process is then repeated 10 times ( the folds ) , with each of the 10 subsamples used exactly once as the validation data , and the error on the validation data is averaged over the 10 folds .
the whole procedure is then repeated as @xmath69 is varied .
the optimal value of @xmath69 is chosen as the one leading to the smallest average error on the validation data .
as an example , we simulate a system made of 30 spins constituted by ten modules of three spins each . the non - vanishing couplings of the ising model are given by : @xmath71 for @xmath72 . after evaluating the couplings @xmath73 , using the algorithm described in @xcite
, we calculate the sensitivity ( fraction of non - vanishing connections @xmath1 leading to non - vanishing couplings @xmath70 ) and the specificity ( fraction of vanishing connections @xmath1 leading to vanishing couplings @xmath70 ) as a function of @xmath69 . the roc curves ( as @xmath69 is varied , the roc curve is sensitivity as a function of 1-specificity @xcite ) we obtain , in correspondence to three values of the number of samples n ( 100 , 250 and 500 ) , are depicted in fig .
( [ f6 ] ) .
the stars on the curves represent the points corresponding to the value of @xmath69 found by ten - fold cross validation ; these points correspond to a good compromise between specificity and sensitivity .
the empty symbols , instead , represent the values of sensitivity and specificity obtained using granger causality in the three cases ; the specificity by granger causality is nearly one in all cases , while the sensitivity is strongly dependent on the number of samples and goes to zero as @xmath48 decreases . to conclude this subsection ,
we have shown that in the case of low number of samples and sparse connections the @xmath0 regularized least squares method can be used to infer the connections in ising models and outperforms granger causality in these situations .
we remark that direct evaluation of the transfer entropy in these cases is unfeasible .
the case of higher order spin interactions requires use of nonlinear granger causality : in the presence of @xmath74spins interactions , the kernel approach @xcite with the polynomial kernel of at least @xmath75 degree is needed . as an example
, we consider a system of three spins with local fields given by : @xmath76 in figure ( [ f7 ] ) we depict the causalities @xmath77 and @xmath78 , as a function of @xmath1 , using the linear kernel and for the approach with the @xmath79 polynomial kernel .
note that , due to symmetry , @xmath80 and @xmath81 ; all the other causalities are vanishing . the linear approach is not able to detect the three spins interaction , while using the nonlinear approach the interaction is correctly inferred .
we stress that the presence of multispin interactions is connected to the presence of synergetic variables , see @xcite for a discussion about the notions of redundancy and synergy in the frame of causality .
it is interesting to show the performance by transfer entropy on the same problem , see fig.([f8 ] ) : it correctly detects all the interactions , and the value of the transfer entropy is again very close to be half of those from nonlinear granger causality .
we stress that transfer entropy can be applied without prior assumptions about the order of the spins interactions .
a major problem in the inference of dynamical networks is the selection of an appropriate model ; in the case of transfer entropy this issue does not arise , although this advantage may be offset by problems associated with reliable estimation of entropies in sample .
we have proposed the use of autoregressive methods to learn ising models from data .
commonly , the formulation of the inverse ising problem assumes symmetric interactions and is solved by exploiting the relations that exist , at equilibrium , between the pairwise correlations ( at equal times ) and the matrix of couplings . in the general case of asymmetric couplings , no equilibrium is reached and also time delayed correlations among spins should be used to infer the connections .
we have shown that autoregressive approaches can solve the inverse ising problem for weak couplings : for each link @xmath82 , whilst the sign of @xmath1 coincides with the sign of the linear correlation between @xmath63 and @xmath83 .
for weak couplings , granger causality is proportional to the transfer entropy and requires less samples , than transfer entropy , to provide a reliable estimate of the information flow . for sparse connections and low number of samples ,
the @xmath0 regularized least squares method is preferable to granger causality ; nonlinear granger causality is related to multispin interactions .
99 f. rieke , d. warland , r.r .
de ruyter van steveninck , and w. bialek , _ spikes : exploring the neural code _
( mit press , cambrige , ma , 1997 ) .
e. schneidman , m.j .
berry , r. segev , w. bialek , nature * 440 * , 1007 ( 2006 ) .
j. shlens , g.d .
field , j.l .
gauthier , m.i .
grivich , d. petrusca , a. sher , a.m. litke , e.j .
chichilnisky , j. neurosci .
* 28 * 505 ( 2008 ) .
y. roudi , j. tyrcha , j. hertz , physical review e * 79 * , 051915 ( 2009 ) .
v. sessak and r. monasson , j. phys .
a * 42 * , 055001 ( 2009 )
kappen and f.b .
rodriguez , neural computation * 10 * , 1137 ( 1998 ) .
y. roudi , e. aurell and j.a .
hertz , front .
. neurosci . * 3 * , 22 ( 2009 ) .
t. schreiber , phys .
lett . * 85 * , 461 ( 2000 ) c.w.j .
granger , econometrica * 37 * , 424 ( 1969 ) .
l. barnett , a.b .
barrett , and a.k .
seth , phys .
* 103 * , 238701 ( 2009 ) .
d. yu , m. righero , l. kocarev , phys.rev.lett . * 97 * , 188701 ( 2006 ) .
d. napoletani , t. sauer , phys . rev . * e 77 * , 26103 ( 2008 ) . d. marinazzo , m. pellicoro and s. stramaglia , phys . rev .
e * 77 * , 056215 ( 2008 ) .
d. materassi , g. innocenti , physica a * 388 * , 3866 ( 2009 ) .
barabasi , _ linked : the new science of networks_. ( perseus publishing , cambridge mass . , 2002 ) .
s. boccaletti , v. latora , y. moreno , m. chavez and d .- u .
hwang , phys .
rep . * 424 * , 175 ( 2006 ) .
r. tibshirani , j. roy .
* b 58 * , 267 ( 1996 ) . k. hlavackova - schindler , m. palus , m. vejmelka , j. bhattacharya , physics reports * 441 * , 1 ( 2007 ) . y. chen , g. rangarajan , j. feng , and m. ding , phys . lett . * a 324 * , 26 ( 2004 ) . m. dhamala , g. rangarajan , m. ding , phys.rev.lett . * 100 * , 18701 ( 2008 ) .
d. marinazzo , m. pellicoro , s. stramaglia , phys .
lett . * 100 * , 144103 ( 2008 ) .
j. shawe - taylor and n. cristianini , _ kernel methods for pattern analysis_. ( cambridge university press , london , 2004 ) . after a linear transformation
, we may assume all the time series to have zero mean and unit variance .
m. stone , j. roy .
* b 36 * , 111 ( 1974 ) .
n. ancona and s. stramaglia , neural comput . * 18 * , 749 ( 2006 ) .
l. angelini , m. pellicoro and s. stramaglia , phys . lett . * a 373 * , 2467 ( 2009 ) .
d. smirnov and b. bezruchko , phys .
e 79 * , 046204 ( 2009 ) h. kantz , t. schreiber , _ nonlinear time series analysis _
( cambridge university press , cambridge , 1997 ) .
kadanoff , _ statistical physics _ ( world scientific , singapore , 2000 ) .
a. papoulis , _ probability , random variables and stochastic processes mcgraw hill , new york , 1965_. s.j .
kim , k. koh , m. lustig , s. boyd , d. gorinevsky , ieee j. sel .
top . in sig
. process .
* 1 * , 606 ( 2007 ) .
m.h . zweig and g. campbell ,
clin . chem .
* 39 * , 561 ( 1993 )
. l. angelini et al . , phys .
e 81 * , 037201 ( 2010 ) . .
the points , corresponding to 15 realizations of the couplings of six - spins ising systems with @xmath54 ranging in @xmath84 $ ] , are displayed ( the two quantities are estimated over samples of @xmath85 length ) .
the curves are the quadratic expansions at weak coupling : @xmath86 ( dashed - dotted line ) and @xmath87 ( continuous line ) .
( bottom ) the same points are displayed in the @xmath88 plane , showing that @xmath89 at weak coupling .
[ f1],width=377 ] , the linear granger causality ( right ) and the transfer entropy ( left ) are depicted for each link , as a function of its coupling @xmath1 , for @xmath90 ( top ) , @xmath91 ( middle ) and @xmath57 ( bottom ) samples .
the continuous curves represent the _ true _ values ( obtained by fitting the points in fig .
1).[f2],width=377 ] the true value of the transfer entropy and @xmath92 its estimate based on @xmath48 samples , averaged over 1000 runs of the ising system , we define @xmath93 .
the quantity e , thus obtained , is here plotted versus n ( empty circles ) .
a similar quantity e , concerning granger causality , is also plotted ( stars).[f3],width=377 ] and uniform couplings @xmath94 , is considered . as a function of @xmath1 , the linear granger causality
( stars ) and the transfer entropy ( empty circles ) are depicted versus @xmath1 .
both quantities are the same for all links , due to symmetry .
the two curves are the relations between the coupling and transfer entropy ( and between coupling and causality ) which hold for the asymmetric ising model ( obtained by fitting the points of fig .
1 ) . [ f4],width=377 ] , 100 ( dashed line ) , 250 ( dotted line ) and 500 ( continuous line ) .
the stars on these curves represent the points found by ten - fold cross validation .
the other three symbols are the performances by granger causality on @xmath90 ( empty diamond ) , @xmath95 ( empty circle ) , @xmath96 ( empty square ) .
[ f6],width=377 ] and @xmath97 are depicted as a function of @xmath1 for the three spins system described in the text .
causalities are estimated using the linear kernel ( top ) and the @xmath79 polynomial kernel ( bottom ) .
[ f7],width=377 ] | the inference of the couplings of an ising model with given means and correlations is called _ inverse ising problem_. this approach has received a lot of attention as a tool to analyze neural data . we show that autoregressive methods may be used to learn the couplings of an ising model , also in the case of asymmetric connections and for multi - spin interactions .
we find that , for each link , the linear granger causality is two times the corresponding transfer entropy ( i.e. the information flow on that link ) in the weak coupling limit . for sparse connections and a low number of samples ,
the @xmath0 regularized least squares method is used to detect the interacting pairs of spins .
nonlinear granger causality is related to multispin interactions . |
casimir effect is a quantum effect that can not be ignored in the realm of nanotechnology .
it can cause malfunctions of nano devices due to stiction @xcite . in the last decade ,
intensive research have been carried out to determine the exact analytic formula for the casimir effect between two nonplanar objects and its effective numerical computations ( see , for example , the references cited in @xcite ) . prior to this , one can only rely on the proximity force approximation ( pfa ) to compute an approximation for the casimir interaction , and there is no way to determine the magnitude of the error in such an approximation . in the case of the sphere - plate setup , the most popular configuration used in casimir experiments , there is only one curvature parameter given by the radius of the sphere , @xmath1 . hence it is expected that as @xmath0 , the distance from the sphere to the plate , is much smaller than @xmath1 , the casimir interaction energy has an asymptotic expansion of the form @xmath4where @xmath5 is the proximity force approximation to the casimir interaction energy .
it follows that for the casimir force @xmath6 and force gradient @xmath7 , one also has expansions of the form @xmath8 a few years ago , experiments have been set up to measure @xmath2 using a micromachined torsional oscillator @xcite .
this gives a more ernest reason for the theoretical computation of the next - to - leading order terms of the casimir interaction .
one of the breakthroughs in casimir research brought by the achievement in explicit functional representation of the casimir interaction is that it becomes possible to compute analytically the next - to - leading terms , as has been shown in @xcite for the cylinder - plate configuration , in @xcite for the sphere - plate configuration , in @xcite for the cylinder - cylinder configuration , and in @xcite for the sphere - sphere configuration .
however , except for @xcite , all the other works only deal with ideal or non - physical boundary conditions , i.e. , dirichlet , neumann , perfectly conducting , infinitely permeable or robin boundary conditions .
so far no work has discussed the exact analytical computation of the next - to - leading order term in the casimir interaction between a sphere and a plate when both of these objects are made of real materials , and this is the goal of the current work to deal with this problem .
it should be mentioned that there has been an attempt to compute the material dependent next - to - leading order term in the casimir interaction between a sphere and a plate carried out by bimonte , emig and kadar @xcite , which used the method of derivative expansion postulated in @xcite , which in turn is inspired by the work @xcite .
however , it is still desirable to check the validity of the postulate in @xcite by computing the next - to - leading order terms from the exact formula for the casimir interaction
. therefore , the results of our current work complement those obtained in @xcite .
in this article , we recall the formula for the casimir interaction between a sphere and a plate .
assume that the sphere has relative permittivity @xmath9 , and the plate has relative permittivity @xmath10 . when the thicknesses of the sphere and the plate are larger than their respective skin - depths , we can model this configuration by a ball and a semi - infinite space .
let @xmath0 be the distance from the sphere to the plate , and let @xmath11 , where @xmath1 is the radius of the ball .
as shown in @xcite , the electromagnetic casimir interaction energy of this sphere - plate configuration is given by @xmath12 where the trace tr is @xmath13 with tr being the trace over @xmath14 matrices .
the matrix elements of @xmath15 are given by @xmath16where @xmath17is a diagonal matrix with elements @xmath18 and @xmath19is a diagonal matrix with elements @xmath20here @xmath21 and @xmath22 are the associated legendre functions given by @xmath23 when @xmath24 , and @xmath25 direct numerical computations of the casimir interaction energy from the formula have been performed in a few works , for example , in @xcite . in numerical computations ,
the infinite matrix @xmath15 has to be truncated to a matrix of finite size .
a drawback of this direct numerical computation is that when @xmath3 gets smaller , one has to use a truncated matrix of larger size for accuracy , and this is subjected to the capacity of the computer .
currently , numerical computations are limited to @xmath26 . however , in experiments , we usually have @xmath27 .
hence , analytical computation of the casimir interaction energy becomes desirable .
in this section , we want to derive analytically the small separation asymptotic expansion of the casimir interaction energy , casimir force and the force gradient up to the next - to - leading order term
. one of the technical issues in the analytical computation of the casimir interaction energy is the appearance of the associated legendre functions @xmath22 .
first notice that because of the relation and @xmath28 the matrix element @xmath29 is equal to that when @xmath30 is changed to @xmath31 .
hence , it is sufficient to consider nonnegative @xmath30 . in this case
, one can show that @xmath32differentiating with respect to @xmath33 gives@xmath34therefore , @xmath35 making a change of variables @xmath36 and expanding the logarithm in , we have @xmath37where @xmath38with @xmath39here @xmath40 = te or tm , and @xmath41 , @xmath42 . the minus signs on @xmath43 and @xmath44 in come from the two matrices @xmath45 in .
let @xmath46 in the following , we make a shift of parameters @xmath47 where @xmath48when @xmath49 , the main contributions to the casimir interaction energy come from the terms with @xmath50 in the small @xmath51 expansion below , we will count the order of @xmath52 and @xmath53 as @xmath54 and @xmath51 respectively . making a change of variables
@xmath55we have @xmath56where @xmath57 and @xmath58 by default .
the integration over @xmath33 is from @xmath59 to @xmath60 which can be approximated by an integration from @xmath61 to @xmath60 , since @xmath33 is of order @xmath51 and @xmath62 is of order @xmath63 .
now we perform the small @xmath51 expansion of . writing @xmath64 as @xmath65 and using the fact that @xmath66 we have the following small @xmath51 expansion : @xmath67 in the second line
, we have performed a rescaling @xmath68 so that the main contribution to the integration over @xmath69 comes from @xmath69 that are @xmath70 . here and in the following , for any @xmath71 , @xmath72 and @xmath73 are , respectively , terms of order @xmath74 and @xmath51 . when these terms do not depend on @xmath75 ,
@xmath75 would be omitted .
changing @xmath76 to @xmath77 and @xmath78 to @xmath79 in , we obtain a similar expansion : @xmath80next , we can use stirling s formula @xmath81to obtain an expansion @xmath82on the other hand , we have @xmath83 for the terms involving @xmath33 , expanding in small @xmath51 gives @xmath84 @xmath85 @xmath86 here @xmath87 is of order @xmath74 .
we do not need the term that is of order @xmath51 for the off - diagonal terms of these matrices as they wo nt contribute to the next - to - leading order term of the casimir interaction energy . finally , the small @xmath51 expansions of @xmath88 is the same as the small @xmath33 expansions : @xmath89 where @xmath90 , @xmath91 , @xmath92 , @xmath93notice that @xmath94 only depend on @xmath10 and @xmath95 .
they are independent of @xmath76 and @xmath51 .
gathering the expansions obtained above , we can write @xmath96where @xmath97notice that @xmath98 are independent of @xmath99 and @xmath33 . performing the summation over @xmath78 and @xmath79 using the formulas
@xmath100we obtain an expansion of the form @xmath101 the @xmath102 term comes from @xmath103 and @xmath104 .
the gaussian integrations over @xmath69 and @xmath105 can be performed straightforwardly and give @xmath106@xmath107 comes from @xmath102 and it is independent of @xmath33 . before performing integration over @xmath33 ,
one is supposed to multiply @xmath108 into the matrix after it . up to the terms of order @xmath51
, we can write @xmath109 and only multiply @xmath110 into the matrix . on the other hand , up to the terms of order @xmath51 , we can extract the term @xmath107 of order @xmath51 out from the matrix .
these give @xmath111performing the gaussian integration over @xmath33 , we have @xmath112where @xmath113 next we consider the small @xmath51 expansions of @xmath114 . debye asymptotic expansions of modified bessel functions say that : @xmath115where @xmath116 let @xmath117 then we find that up to terms of order @xmath51 , we have @xmath118 in small @xmath51 expansion , @xmath119 therefore , we have an expansion of the form @xmath120 where @xmath121with @xmath122notice that @xmath123 only depend on @xmath9 and @xmath95
. they are independent of @xmath76 and @xmath51 . substituting into
, we have an expansion of the form : @xmath124where @xmath125 substituting into , and extracting terms up to order @xmath51 , we have @xmath126^{s+1 } \left(1+\sum_{i=0}^{s}\sum_{j=0}^s\mathcal{z}_{i,1}\mathcal{z}_{j,1}+\sum_{i=0}^s\mathcal{z}_{i,2}+(s+1)\mathcal{u}_2\right)+ x\mathcal{m}_1 ^ 2 \right.\\ & + \sum _ { * = \text{te } , \text{tm}}\left[t^{*}_{0}\widetilde{t}^{*}_{0}\right]^{s+1 } \left ( \sum_{i=0}^s\sum_{j=0}^s\mathcal{z}_{i,1}\mathcal{j}_{j,1}^{*}+\sum_{i=0}^s\sum_{j\neq i}\mathcal{z}_{i,1}\mathcal{v}_{j,1}^{*}+ \sum_{i=0}^s\mathcal{t}_{i,1}\mathcal{v}_{i,1}^{*}+\sum_{i=0}^s\widetilde{\mathcal{s}}_{i,2}\mathcal{k}_1^ { * } \right.\\&\left.\left.\hspace{3cm}\sum_{i=0}^s\sum_{j = i+1}^s \mathcal{j}^{*}_{i,1}\mathcal{j}^{*}_{j,1 } + \sum_{i=0}^s\sum_{j = i+1}^s\mathcal{v}^{*}_{i,1}\mathcal{v}^{*}_{j,1 } + \sum_{i=0}^s \mathcal{j}^{*}_{i,1}\sum_{j= 0}^s\mathcal{v}^{*}_{j,1}+\sum_{i=0}^s \mathcal{j}^{*}_{i,2}+ \sum_{j=0}^s\mathcal{v}^{*}_{i,2}\right)\right\ } , \end{split}\]]where @xmath127and @xmath128^{s+1 } -\left[t^{\text{tm } } _ { 0}\widetilde{t}^{\text{tm } } _ { 0}\right]^{s+1 } } { t_0^{\text{te}}\widetilde{t}_0^{\text{te}}- t_0^{\text{tm}}\widetilde{t}_0^{\text{tm}}}+2t^{\text{te } } _ { 0}\widetilde{t}_0^{\text{te}}t^{\text{tm } } _ { 0}\widetilde{t}_0^{\text{tm } } \frac{\left[t^{\text{te } } _ { 0}\widetilde{t}^{\text{te } } _ { 0}\right]^{s } -\left[t^{\text{tm } } _ { 0}\widetilde{t}^{\text{tm } } _ { 0}\right]^{s } } { t_0^{\text{te}}\widetilde{t}_0^{\text{te}}- t_0^{\text{tm}}\widetilde{t}_0^{\text{tm}}}\right\}.\end{aligned}\]]we have omitted those terms of order @xmath74 since they are odd in one of the @xmath76 and thus would give zero after integration with respect to @xmath76 .
it follows that @xmath129@xmath130 is the leading order term that comes from those terms of order @xmath131 , and @xmath132 is the next - to - leading order term that comes from those terms of order @xmath51 . @xmath9 and @xmath10 are functions of @xmath133they are independent of @xmath76 and @xmath30 . performing the gaussian integration over @xmath76 , @xmath134 , and @xmath30 , we find that @xmath135^{s+1 } , \end{split}\]]@xmath136^{s+1 } \left(\mathscr{a}+\mathscr{c}^{*}+\mathscr{d}^{*}\right ) + x\mathscr{b } \biggr\}. \end{split}\]]the explicit formulas for @xmath137 , @xmath138 , @xmath139 and @xmath140 are given by @xmath141with @xmath142 using the fact that @xmath143 are independent of @xmath51 , it is straightforward to take derivative with respect to @xmath0 . for the casimir force @xmath144
we find that @xmath145where @xmath146 and @xmath147 are respectively the leading order and next - to - leading order terms with @xmath148^{s+1},\end{aligned}\]]@xmath149^{s+1 } \left(\widetilde{\mathscr{a}}+\widetilde{\mathscr{c}}^{*}+\mathscr{d}^{*}\right ) + x\mathscr{b } \biggr\}.\end{aligned}\]]here @xmath150 for the force gradient @xmath7 , the leading order and next - to - leading order terms are @xmath151^{s+1},\end{aligned}\]]@xmath152^{s+1 } \left(\widehat{\mathscr{a}}+\widehat{\mathscr{c}}^{*}+\mathscr{d}^{*}\right ) + x\mathscr{b } \biggr\},\end{aligned}\]]where @xmath153 let us compare the leading order term to the proximity force approximation .
the casimir energy density between a pair of parallel dielectric plates with relative permittivities @xmath9 and @xmath10 is given by the lifshitz s formula @xcite : @xmath154 where @xmath155 the proximity force approximation to the casimir interaction energy between a sphere and a plate with relative permittivities @xmath9 and @xmath10 is given by @xmath156 expanding the logarithm in and substitute into , we find that @xmath157^{s+1}\\ = & -\frac{\hbar c r}{4\pi}\sum_{s=0}^{\infty}\frac{1}{(s+1)^2 } \int_0^{\infty } d\kappa \int_{\kappa}^{\infty } dq e^{-2q(s+1)d}\sum _ { * = \text{te } , \text{tm}}\left[r_1^{*}r_2^ { * } \right]^{s+1}. \end{split}\ ] ] now making a change of variables @xmath158we finally obtain @xmath159^{s+1 } , \end{split}\]]where @xmath160 compare to , we find that our result for the leading order term agrees completely with the proximity force approximation .
in this section , we consider the special case where the dielectric permittivities of the sphere and the plate are described by the plasma model : @xmath161 where @xmath162 is the plasma frequency of the material .
let @xmath163 in terms of the variables @xmath164 and @xmath95 , we have @xmath165 .[t1 ] the coefficients @xmath166 . [ cols="^,^,^",options="header " , ] first consider the case @xmath167 the limit where @xmath168 is the perfect conductor limit .
we can compute analytically the asymptotic expansion of the leading and next - to - leading order terms in the small parameters @xmath169 specifically , we have @xmath170where @xmath171 and @xmath172is the leading order approximation to the casimir interaction energy between a perfectly conducting sphere and a perfectly conducting plate .
the exact values of @xmath166 and @xmath173 for @xmath174 are listed in table [ t1 ] and table [ t2 ] . from
, it is obvious that the leading term is symmetric when we interchange @xmath9 with @xmath10 .
it follows that @xmath175hence , we only list the coefficients of @xmath166 when @xmath176 in table [ t1 ] .
from , we have @xmath177 @xmath178 here @xmath179and @xmath180 are respectively the leading order approximations to the casimir force and force gradient between a perfectly conducting sphere and a perfectly conducting plate . setting @xmath181 in , and
, we obtain @xmath182+\ldots\right ) , \\
f_{\text{cas } } \approx & f_{\text{cas}}^{\text{pfa , pc}}\left(1+\frac{d}{r}\left[\frac{1}{6}-\frac{10}{\pi^2}\right]+\ldots\right),\\ \frac{{\partial}f_{\text{cas}}}{{\partial}d } \approx & \frac{{\partial}f_{\text{cas}}^{\text{pfa , pc}}}{{\partial}d}\left(1+\frac{d}{r}\left[\frac{1}{9}-\frac{20}{3\pi^2}\right]+\ldots\right),\end{aligned}\]]which are well - known results for the leading and next - to - leading order terms of the perfectly conducting sphere - plate configuration @xcite .
next we consider numerical results with @xmath183 9ev , which is the plasma frequency for gold @xcite .
the radius of the sphere @xmath1 is taken to be 1 mm . substituting into the formulas obtained in the previous section , we can compute numerically the leading order term ( the proximity force approximation ) and the next - to - leading order term of the casimir interaction . in figs .
[ f1 ] , [ f2 ] and [ f3 ] , we plot the leading order term , the sum of the leading order and next - to - leading order terms of the casimir interaction energy , casimir force and force gradient , normalized respectively by @xmath184 and @xmath185 , as a function of @xmath0 for @xmath0 between 10 nm and 100@xmath186 m . from the figures , we notice that when @xmath187 , the corrections to pfa become significant and they would contribute corrections of about 10% .
= 0.45 = 0.45 = 0.45 to have a better picture about the corrections to the proximity force approximations , define @xmath188 , @xmath189 and @xmath190 by @xmath191so that and hold .
= 0.32 = 0.32 = 0.32 in figs .
[ f4 ] , we plot @xmath188 , @xmath189 and @xmath190 as functions of @xmath0 for @xmath0 between 10 nm and 100@xmath186 m . as @xmath0 increases to 100@xmath186 m , we find that @xmath188 , @xmath189 and @xmath190 tend respectively to the values @xmath192 , @xmath193 and @xmath194 , which are corresponding values for perfect conductors .
this is not surprising since @xmath195 is approximately equal to 5000 when @xmath196 .
when @xmath0 is small , the deviations from the limiting values for perfect conductors are very significant . on the other hand
, we also notice that @xmath189 and @xmath2 are bounded below .
@xmath2 is a quantity that can be measured experimentally @xcite . from fig .
[ f4 ] , we find that it is bounded below by @xmath197 .
the drude dielectric function is given by @xmath198 where @xmath199 is the relaxation frequency of the material .
in the limit where @xmath200 , the drude dielectric function becomes the plasma dielectric function .
let @xmath201 in terms of the variables @xmath202 and @xmath95 , we have @xmath203 substituting this into the results obtained in section [ ae ] , we can compute numerically the leading order and next - to - leading order terms of the casimir interaction for drude models .
let us consider the case where @xmath183 9ev and @xmath204 ev , which are the conventional values used for gold @xcite .
= 0.45 = 0.45 = 0.45 in figs .
[ f6 ] , [ f7 ] and [ f8 ] , we plot the leading order term , the sum of the leading order and next - to - leading order terms of the casimir interaction energy , casimir force and force gradient , normalized respectively by @xmath184 and @xmath185 , as a function of @xmath0 for @xmath0 between 10 nm and 100@xmath186 m . both the drude model and the plasma model are plotted on the same graph to show the comparison . to get a better picture
, we plot the ratio of the plasma model to the drude model for the sum of the leading order and next - to - leading order terms in fig .
[ f11 ] . from the figure
, we notice that if the plasma model is used instead of the drude model , the error is at most 2% .
= 0.45 in figs .
[ f9 ] , we plot @xmath205 and @xmath2 for drude model and compare to that for plasma model . as for the plasma model
, we notice that for the drude model , as @xmath0 increases , @xmath188 , @xmath189 and @xmath190 tend respectively to the limiting values @xmath192 , @xmath193 and @xmath194 , the corresponding values for perfect conductors .
when @xmath0 is small , the deviations from these limiting values are very significant . on the other hand , @xmath189 and @xmath2 are also bounded from below .
the ratios of the plasma model to the drude model for @xmath205 and @xmath2 are plotted in fig .
[ f12 ] . from the figure
, we find that if the plasma model is used instead of the plasma model , the error is not more than 4.5% .
= 0.32 = 0.32 = 0.32 = 0.45
starting from the functional determinant representation of the casimir interaction energy , we have used the perturbation method to obtain analytically the leading order and next - to - leading order terms of the casimir interaction energy , casimir force and force gradient for the interaction between a sphere and a plate .
the results are written as double integrals over functions of the dielectric permittivities of the objects , and are hence general .
the leading order terms are shown to equal to that predicted by proximity force approximation .
the results on the next - to - leading order terms are new , and they complement those obtained in @xcite using derivative expansion . with given dielectric permittivities of the sphere and the plate , the double integrals representing the leading order and next - to - leading order terms can be computed numerically , and this is demonstrated for a gold sphere in front of a gold plate , where both plasma and drude models are used for the dielectric functions of gold . it is observed that even at @xmath187 , the next - to - leading order term would contribute a correction to the leading order term of about 10% . of particular interest is the ratio of the next - to - leading order term divided by @xmath3 to the leading order term , denoted by @xmath2 .
it is found that when @xmath206 is large enough , @xmath207 tends to the corresponding limiting value for perfect conductors .
however , when @xmath206 is small , the deviation from the limiting perfect conductor value is significant .
this signifies that in the nano range , we can not model real metals by perfect conductors .
a comparison between plasma model and drude model shows that their difference is below 2% for the sum of the first two leading order terms and below 4.5% for the values of @xmath207 .
in fact , this small difference is expected at zero temperature . in this work
, we havent considered the thermal effect . when @xmath0 is small enough such that @xmath208 is @xmath209 , thermal effect can be neglected .
for example , when @xmath210k , thermal effect can be neglected when @xmath211 1@xmath186 m .
nevertheless , it would be interesting to examine the behavior of the casimir interaction at room temperature and its interplay with material properties .
this would be addressed in a forthcoming work .
10 f. m. serry , d. walliser and g. j. maclay , j. microelectromech . syst . * 4 * , 193 ( 1995 )
. f. m. serry , d. walliser and g. j. maclay , j. appl . phys . * 84 * , 2501 ( 1998 ) .
m. rose , photonics spectra * 42 * , 77 ( 2008 ) .
l. p. teo , int . j. mod
. phys . a * 27 * , 1230021 ( 2012 ) .
d. e. krause , r. s. decca , d. lopez and e. fischbach , phys .
98 * , 050403 ( 2007 ) .
m. bordag , phys .
d * 73 * , 125018 ( 2006 ) .
m. bordag , phys .
d * 75 * , 065003 ( 2007 ) .
l. p. teo , phys .
d * 84 * , 025022 ( 2011 ) .
m. bordag and v. nikolaev , j. phys . a : math
. theor . * 41 * , 164002 ( 2008 ) .
m. bordag and v. nikolaev , phys . rev .
d * 81 * , 065011 ( 2010 ) .
l. p. teo , m. bordag and v. nikolaev , phys .
d * 84 * , 125037 ( 2011 ) .
l. p. teo , phys .
d * 84 * , 065027 ( 2011 ) .
l. p. teo , phys .
d * 85 * , 045027 ( 2012 ) .
g. bimonte , t. emig and m. kadar , appl .
. lett . * 100 * , 074110 ( 2012 ) .
g. bimonte , t. emig , r. l. jaffe and m. kadar , europhys . lett . *
97 * , 50001 ( 2012 ) . c. d. fosco , f. c. lombardo and f. d. mazzitelli , phys .
d * 84 * , 105031 ( 2011 ) .
s. j. rahi , t. emig , n. graham , r. l. jaffe and m. kadar , phys , rev .
d * 80 * , 085021 ( 2009 ) .
l. p. teo , int . j. mod .
a * 27 * , 1230021 ( 2012 ) .
a. canaguier - durand , p. a. m. neto , a. lambrecht and s. reynaud , phys .
lett . * 104 * , 040403 ( 2010 ) .
a. canaguier - durand , p. a. m. neto , a. lambrecht and s. reynaud , phys .
a * 82 * , 012511 ( 2010 ) . | we derive analytically the asymptotic behavior of the casimir interaction between a sphere and a plate when the distance between them , @xmath0 , is much smaller than the radius of the sphere , @xmath1 .
the leading order and next - to - leading order terms are derived from the exact formula for the casimir interaction energy .
they are found to depend nontrivially on the dielectric functions of the objects . as expected , the leading order term coincides with that derived using the proximity force approximation .
the result on the next - to - leading order term complements that found by bimonte , emig and kardar [ appl .
. lett . * 100 * , 074110 ( 2012 ) ] using derivative expansion .
numerical results are presented when the dielectric functions are given by the plasma model or the drude model , with the plasma frequency ( for plasma and drude models ) and relaxation frequency ( for drude model ) given respectively by 9ev and 0.035ev , the conventional values used for gold metal .
it is found that if plasma model is used instead of drude model , the error in the sum of the first two leading terms is at most 2% , while the error in @xmath2 , the ratio of the next - to - leading order term divided by @xmath3 to the leading order term , can go up to 4.5% . |
until is a notoriously difficult temporal operator . this is because of its `` ambivalent '' nature of being an operator that is both existential and universal at the same time : @xmath0 holds at the current time instant ( sometimes `` world '' or `` state ''
is used in place of `` time instant '' ) @xmath1 iff either @xmath2 holds at @xmath1 or there _ exists _ a time instant @xmath3 in the future at which @xmath2 holds and such that @xmath4 holds in _ all _ the time instants between the current one and @xmath3 .
the words in emphasis highlight the dual existential and universal nature of @xmath7 , which poses a significant challenge when attempting to give deduction rules for until , so that deduction systems for temporal logics either deliberately exclude until from the set of operators considered or devise clever ways to formalize reasoning about until . and even if one manages to give rules , these often come at the price of additional difficulties for , or even the impossibility of , proving useful metatheoretic properties , such as normalization or the subformula property .
( this is even more so in the case of hilbert - style axiomatizations , which provide axioms for until , but are not easily usable for proof construction . )
see , for instance , @xcite , where techniques for formalizing suitable inference rules include introducing additional information ( such as the use of a skolem function @xmath8 to name the time instant where @xmath2 begins to hold ) , or exploiting the standard recursive unfolding of until @xmath9 which says that @xmath0 iff either @xmath2 holds or @xmath4 holds and in the successor time instant ( as expressed by the _ next _
operator @xmath10 ) we have again @xmath0 . in this paper , in contrast , we make explicit the duality of until by introducing a new temporal operator @xmath5 that allows us to formalize the `` history '' of until , i.e. , the fact that when we have @xmath11 the formula @xmath4 holds in all the time instants between the current one and the one where @xmath2 holds .
we express this `` historic '' universal quantification by means of @xmath5 with respect to the following intuitive translation : @xmath12 that is : @xmath0 iff either @xmath2 holds or there exists a time instant @xmath3 in the future ( as expressed by the _ sometime in the future _
operator @xmath13 ) such that * @xmath2 holds in the successor time instant , and * @xmath4 holds in all the time instants between the current one and @xmath3 ( included ) .
the latter conjunct is precisely what the _ history _ operator @xmath5 expresses that we achieve with @xmath5 is precisely what allows us to give well - behaved ( in a sense made clearer below ) natural deduction rules . ]
this is better seen when introducing labeling : since @xmath5 actually quantifies over the time instants in an interval ( delimited by the current instant and the one where the @xmath2 of the until holds ) , we adopt a labeling discipline that is slightly different from the more customary one of labeled deduction .
the framework of _ labeled deduction _ has been successfully employed for several non - classical , and in particular modal and temporal , logics , e.g. , @xcite , since labeling provides a clean and effective way of dealing with modalities and gives rise to deduction systems with good proof - theoretical properties .
the basic idea is that labels allow one to explicitly encode additional information , of a semantic or proof - theoretical nature , that is otherwise implicit in the logic one wants to capture .
so , for instance , instead of a formula @xmath4 , one can consider the _ labeled formula _
@xmath14 , which intuitively means that @xmath4 holds at the time instant denoted by @xmath15 within the underlying kripke semantics .
one can also use labels to specify how time instants are related , e.g. , the _ relational formula _
@xmath16 states that the time instant @xmath17 is accessible from @xmath15 . considering labels that consist of a single time
instant is not enough for @xmath5 , as the operator is explicitly designed to speak about a sequence of time instants ( namely , the ones constituting the history of the corresponding until , if indeed @xmath5 results from the translation of an @xmath7 ) .
we thus consider labels that are built out of a sequence of time instants , so that we can write @xmath18 to express , intuitively , that @xmath4 holds in the interval between time instants @xmath19 and @xmath20 , which together with the sub - sequence @xmath21 constitute a sequence of time instants @xmath22 .
this allows us to give the natural deduction elimination rule @xmath23{\alpha b_1 b_2 : a}{\alpha b_1 b_3 : { \nabla}a & b_1 { \leqslant}b_2 & b_2 { \leqslant}b_3}\ ] ] that says that if @xmath24 holds at time instant @xmath20 at the end of the sequence @xmath25 and if @xmath26 is in - between @xmath19 and @xmath20 , as expressed by the relational formulas with the accessibility relation @xmath27 , then we can conclude that @xmath4 holds at @xmath26 .
dually , we can introduce @xmath24 at time instant @xmath20 at the end of the sequence @xmath25 whenever from the assumptions @xmath28 and @xmath29 for a _ fresh _
@xmath26 we can infer @xmath30 , i.e. is fresh means that @xmath26 is different from @xmath19 and @xmath20 , and does not occur in any assumption on which @xmath30 depends other than the discarded assumptions @xmath31 and @xmath32 .
] , : @xmath33{\alpha b_1 b_3 : { \nabla}a}{\infer*{\alpha b_1 b_2:a}{[b_1 { \leqslant}b_2]\ , [ b_2{\leqslant}b_3]}}\ ] ] the adoption of time instant sequences for labels has thus allowed us to give rules for @xmath5 that are well - behaved in the spirit of natural deduction @xcite : there is precisely one introduction and one elimination rule for @xmath5 , as well as for the other connectives and temporal operators ( @xmath34 , @xmath35 , and @xmath10 ) .
this paves the way to a proof - theoretical analysis of the resulting natural deduction systems , e.g. , to show proof normalization and other useful meta - theoretical analysis , which we are tackling in current work .
moreover , the rules @xmath36 and @xmath37 provide a clean - cut way of reasoning about until , according to the translation , provided that we also give rules for @xmath13 and @xmath10 .
these operators have a local nature , in the sense that they speak not about sequences of time instants but about single time instants .
still , we can easily give natural deduction rules for them by generalizing the more standard `` single - time instant '' rules ( e.g. , @xcite ) using our labeling with sequences of time instants . as we will discuss in more detail below , if we collapse the sequences of time instants to consider only the final time instant in the sequence ( or , equivalently , if we simply ignore all the instants in a sequence but the last ) , then these rules reduce to the standard ones .
for instance , for the _ always in the future _
operator @xmath35 ( the dual of @xmath13 ) and @xmath10 , with the corresponding _ successor relation _
@xmath38 , we can give the elimination rules @xmath39{\alpha b_1 b_2 : a}{\alpha b_1 : { \mathsf{g}}a & b_1 { \leqslant}b_2 } \qquad \text{and } \qquad \infer[{{\mathsf{x}}e}]{\alpha b_1b_2 : a}{\alpha b_1 : { \mathsf{x}}a & b_1 { \lhd}b_2}\ ] ] the rule @xmath40 says that if @xmath41 holds at time instant @xmath19 , which is the last in the sequence @xmath42 and @xmath26 is @xmath27-accessible from @xmath19 ( i.e. , @xmath31 ) , then we can conclude that @xmath4 holds for the sequence @xmath43 .
the rule @xmath44 is justified similarly ( via @xmath38 ) .
the corresponding introduction rules are given in section [ sec : system ] , together with rules for @xmath45 and the connective @xmath34 , as well as a rule for induction on the underlying linear ordering .
as we will see , we also need rules expressing the properties of the relations @xmath27 and @xmath38 .
moreover , the fact that we consider sequences of time instants as labels requires us to consider some structural rules to express properties of such sequences ( with respect to formulas ) .
this approach thus provides the basis for formalizing deduction systems for temporal logics endowed with the until operator . for concreteness , we give here a labeled natural deduction system for a linear - time logic endowed with the new history operator @xmath5 and show that , via a proper translation , such a system is also sound and complete with respect to the linear temporal logic @xmath6 with until .
( we do not consider past explicitly here , but adding operators and rules for it should be unproblematic , e.g. , as in @xcite . )
we proceed as follows . in section [ sec : ltl ] , we briefly recall the syntax and semantics of @xmath6 , and an axiomatization for it . in section [ sec : ltlh ] , we define @xmath46 , the logic that is obtained from @xmath6 by replacing @xmath7 with the history @xmath5 , and give a validity - preserving translation , based on , from @xmath6 into @xmath46 . in section [ sec :
system ] , we give a labeled natural deduction system @xmath47 that it is sound with respect to the semantics of @xmath46 . by focusing only on those derivations
whose conclusion and open assumptions correspond to the translation of @xmath6-formulas , we show that @xmath47 can be used to capture reasoning in @xmath6 and that it is in fact sound and complete with respect to the semantics of @xmath6 . in section [ sec : discussion ] , we draw conclusions and illustrate directions of current and future work . due to lack of space ,
some of the proofs are only sketched here .
full proofs are given in * ? ? ?
full proofs are given in the appendix .
we recall the syntax and semantics of @xmath6 and an axiomatization for it .
_ given a set @xmath48 of propositional symbols , the set of _ ( well - formed ) @xmath6-formulas _ is defined by the grammar @xmath49 where @xmath50 .
the set of @xmath6-_atomic formulas _ is @xmath51 .
the _ complexity _ of an @xmath6-formula is the number of occurrences of the connective @xmath34 and of the temporal operators @xmath35 , @xmath10 , and @xmath7 . _
the intuitive meaning of @xmath35 , @xmath10 , and @xmath7 is the standard one : @xmath41 states that @xmath4 holds always in the future , @xmath52 states that @xmath4 holds in the next time instant , and @xmath0 states that @xmath2 holds at the current time instant or there is a time instant @xmath1 in the future such that @xmath2 holds in @xmath1 and @xmath4 holds in all the time instants between the current one and @xmath1 . as usual
, we can introduce abbreviations and use , e.g. , @xmath53 , @xmath54 and @xmath55 for negation , disjunction , and conjunction , respectively : @xmath56 , @xmath57 , and @xmath58 .
we can also define other temporal operators , e.g. , @xmath59 to express that @xmath4 holds sometime in the future .
we write @xmath60 to denote a _ set of @xmath6-formulas_. [ def : ltl - truth]_let @xmath61 be the standard structure of natural numbers , where @xmath62 and @xmath63 are respectively the successor function and the total ( reflexive ) order relation .
@xmath6-model _ is a pair @xmath64 where @xmath65 .
_ truth _ for an @xmath6-formula at a point @xmath66 in an @xmath6-model @xmath67 is the smallest relation @xmath68 satisfying : _ @xmath69 note that @xmath70 for every @xmath71 and @xmath72 .
by extension , we write : @xmath73 we now present a sound and complete hilbert - style axiomatization , which we call @xmath74 , for @xmath6 ( see , e.g. , @xcite ) .
@xmath74 consists of the axioms @xmath75 where we denote with @xmath76 the double implication , and of the rules of inference @xmath77 the set of theorems of @xmath74 is the smallest set containing these axioms and closed with respect to these rules of inference .
in this section , we give the linear temporal logic @xmath46 , which is obtained from @xmath6 by replacing the operator @xmath7 with a new unary temporal operator @xmath5 , called _
the definition of the semantics of @xmath46 requires a notion of truth given with respect to sequences of time instants rather than just to time instants .
we will then provide a translation from the language of @xmath6 into the language of @xmath46 and show some properties of such a translation .
_ given a set @xmath48 of propositional symbols , the set of _ ( well - formed ) @xmath46-formulas _ is defined by the grammar @xmath78 where @xmath50 .
the set of @xmath46-_atomic formulas _ is @xmath51 .
the _ complexity _ of an @xmath46-formula is the number of occurrences of the connective @xmath34 and of the temporal operators @xmath10 , @xmath35 , and @xmath5 . _ the intuitive meaning of the operators @xmath35 and @xmath10 is the same as for @xmath6 , while @xmath24 intuitively states that @xmath4 holds at any instant of a particular time interval ( but here we see that we need sequences of time instants to formalize the semantics of the history operator , as we anticipated in the introduction ) .
again , we can define other connectives and operators as abbreviations , e.g. , @xmath53 , @xmath54 , @xmath55 , @xmath13 and so on .
we write @xmath79 to denote a _ set of @xmath46-formulas_. to define a labeled deduction system for the logic @xmath46 , we extend the language with a set of labels and finite sequences of labels , and introduce the notions of labeled formula and relational formula . _ _
let @xmath80 be a set of labels .
a finite non - empty sequence of labels ( namely , an element of @xmath81 ) is called a _
sequence_. if @xmath4 is an @xmath46-formula and @xmath21 is a sequence , then @xmath82 is a _ labeled ( well - formed ) formula _ ( _ lwff _ for short ) .
the set of _ relational ( well - formed ) formulas _ ( _ rwffs _ for short ) is the set of expressions of the form @xmath83 or @xmath84 , where @xmath15 and @xmath17 are labels . in the rest of the paper , we will assume given a fixed denumerable set @xmath80 of labels and we will use @xmath85 to denote labels , @xmath86 to denote finite sequences of labels to denote possibly empty subsequences and thus write @xmath87 ( for @xmath88 ) to denote a sequence where @xmath21 may be empty . ]
( e.g. , @xmath89 or just @xmath15 in the case of a sequence consisting of only one time instant ) , @xmath90 to denote a _ generic formula _
( either labeled or relational ) and @xmath91 to denote a _ set of generic formulas_. [ def : ltlh - truth ] _ _
an _ observation sequence _ is a non - empty sequence @xmath92 @xmath93 @xmath94 $ ] of natural numbers .
_ truth _ for an @xmath46-formula at an observation sequence @xmath95 in an @xmath6-model @xmath67 is the smallest relation @xmath96 satisfying : @xmath97 { \models_{_{\nabla}}}p & \quad \text{iff } \quad & p \in { \mathcal{v}}(n_k ) \\ { \mathcal{m}},[n_0,\ldots , n_k ] { \models_{_{\nabla}}}a { \supset}b & \quad \text{iff } \quad & { \mathcal{m}},[n_0,\ldots , n_k ] { \models_{_{\nabla}}}a \text { implies } \\ & & { \mathcal{m}},[n_0,\ldots , n_k ] { \models_{_{\nabla}}}b \\ { \mathcal{m}},[n_0,\ldots , n_k]{\models_{_{\nabla}}}{\mathsf{g}}a & \quad \text{iff } \quad & { \mathcal{m}},[n_0,\ldots , n_k , m ] { \models_{_{\nabla}}}a \text { for all } m \ge n_k\\ { \mathcal{m}},[n_0,\ldots , n_k]{\models_{_{\nabla}}}{\mathsf{x}}a & \quad \text{iff } \quad & { \mathcal{m}},[n_0,\ldots , n_k , n_k+1]{\models_{_{\nabla}}}a \\ { \mathcal{m}},[n_0,\ldots , n_{k-1},n_k ] { \models_{_{\nabla}}}{\nabla}a & \quad \text{iff } \quad & { \mathcal{m}},[n_0,\ldots , n_{k-1},m]{\models_{_{\nabla}}}a\\ & \quad & \text{for all } n_{k-1 } \le m \le n_k \text { ( if } 0 < k\text{)}\end{aligned}\ ] ] @xmath98{\models_{_{\nabla}}}{\nabla}a & \quad \text{iff } \quad & { \mathcal{m}},[n_0]{\models_{_{\nabla}}}a \end{aligned}\ ] ] by extension , we write : @xmath99 given an @xmath6-model @xmath71 , a _ structure _ is a pair @xmath100 where @xmath101 .
let @xmath102 be the set of observation sequences and @xmath103 the extension of @xmath104 to sequences , i.e. , @xmath105 @xmath106 $ ] .
_ truth _ for a generic formula @xmath90 in a structure @xmath107 is the smallest relation @xmath96 satisfying : @xmath108 note that @xmath109 and @xmath110 for every @xmath71 , @xmath95 and @xmath104 . given a set @xmath91 of generic formulas and a generic formula @xmath90 : @xmath111 @xmath6 and @xmath46 are , obviously , related logics .
in fact , below we will define a validity - preserving translation @xmath112 from @xmath6 into @xmath46 .
then , in lemma [ lem : last ] , we will show that if an @xmath46-formula corresponds to the translation of some @xmath6-formula , then it can be interpreted `` locally '' , i.e. , its truth value with respect to an observation sequence depends only on the last element of the sequence .
finally , in lemma [ lem : translation ] and theorem [ th : translation ] , we will use this result to prove that the translation preserves the validity of formulas .
this property allows us to use the deduction system for @xmath46 , which will be presented in section [ sec : system ] , for reasoning on @xmath6 too , as we will show in section [ sub : sandc ] , when discussing soundness and completeness of the system .
_ we define the translation @xmath112 from the language of @xmath6 into the language of @xmath46 inductively as follows : + _ [ cols="<,^ , < , < ,
< , < " , ] we extend @xmath112 to sets of formulas in the obvious way : @xmath113 .
[ lem : last ] let @xmath71 be an @xmath6-model , @xmath114 $ ] an observation sequence , and @xmath4 an @xmath6-formula . then @xmath115 { \models_{_{\nabla}}}{a^ * } \;$ ] iff @xmath116 $ ] @xmath117 for every sequence @xmath118 .
[ cor : last ] let @xmath71 be an @xmath6-model , @xmath114 $ ] an observation sequence , and @xmath4 an @xmath6-formula . then @xmath119 { \models_{_{\nabla}}}{a^*}$ ] iff @xmath120 { \models_{_{\nabla}}}{a^*}$ ]
[ lem : translation ] let @xmath71 be an @xmath6-model , @xmath72 a natural number , and @xmath4 an @xmath6-formula
. then @xmath121 iff @xmath122 { \models_{_{\nabla}}}{a^*}$ ] .
[ th : translation ] let @xmath60 be a set of @xmath6-formulas , @xmath4 an @xmath6-formula and @xmath123 .
then @xmath124 iff @xmath125 by definition [ def : ltl - truth ] , @xmath126 iff @xmath127 implies @xmath128 iff @xmath129 implies @xmath130 iff ( by lemma [ lem : translation ] ) @xmath131 { \models_{_{\nabla}}}{b^*}$ ] implies @xmath132 { \models_{_{\nabla}}}{a^ * } \,)\,$ ] iff ( by lemma [ lem : last ] ) @xmath133 implies @xmath134 iff ( by definition [ def : ltlh - truth ] ) @xmath135 implies @xmath136 iff @xmath137 implies @xmath136 iff @xmath138 .
in this section , we will first define a labeled natural deduction system @xmath47 that is sound with respect to the semantics of @xmath46 .
then , by considering a restriction of the set of @xmath47-derivations and by using the translation @xmath112 and the related results , we will show that @xmath47 can be also used for reasoning on @xmath6 : we will prove soundness with respect to the semantics of @xmath6 and we will give a proof of weak completeness with respect to @xmath6 , by exploiting the hilbert - style axiomatization @xmath74 .
@xmath139{\alpha_1 : a}{\infer*{\alpha_2:{\perp}}{[\alpha_1 : a { \supset}{\perp } ] } } \qquad \infer[{{\supset}\!\!i}]{\alpha : a { \supset}b}{\infer*{\alpha : b}{[\alpha : a ] } } \qquad \infer[{{\supset}\!\!e}]{\alpha : b}{\alpha : a { \supset}b & \alpha : a } \\ \infer[{{\mathsf{g}}i}]{\alpha b_1 : { \mathsf{g}}a}{\infer*{\alpha b_1b_2:a}{[b_1 { \leqslant}b_2 ] } } \qquad \infer[{{\mathsf{g}}e}]{\alpha b_1 b_2 : a}{\alpha b_1 : { \mathsf{g}}a & b_1 { \leqslant}b_2 } \qquad \infer[{\mathit{ser}_{\lhd}}]{\alpha : a}{\infer*{\alpha : a}{[b_1 { \lhd}b_2 ] } } \qquad \infer[{\mathit{lin}_{\lhd}}]{\alpha : a}{b_1 { \lhd}b_2 & b_1 { \lhd}b_3 & \varphi & \infer*{\alpha : a}{[\varphi[b_3/b_2 ] ] } } \\ \infer[{{\mathsf{x}}i}]{\alpha b_1 : { \mathsf{x}}a}{\infer*{\alpha b_1b_2:a}{[b_1 { \lhd}b_2 ] } }
\qquad \infer[{{\mathsf{x}}e}]{\alpha b_1b_2 : a}{\alpha b_1 : { \mathsf{x}}a & b_1 { \lhd}b_2 } \qquad \infer[{\mathit{refl}_{\leqslant}}]{\alpha : a}{\infer*{\alpha : a}{[b_1 { \leqslant}b_1 ] } } \qquad \infer[{\mathit{trans}_{\leqslant}}]{\alpha : a } { b_1 { \leqslant}b_2 & b_2 { \leqslant}b_3 & \infer*{\alpha : a}{[b_1 { \leqslant}b_3 ] } } \\ \infer[{{\nabla}i}]{\alpha b_1 b_3 : { \nabla}a}{\infer*{\alpha b_1 b_2:a}{[b_1 { \leqslant}b_2]\ , [ b_2{\leqslant}b_3 ] } } \qquad \infer[{{\nabla}e}]{\alpha b_1 b_2 : a}{\alpha b_1 b_3 : { \nabla}a & b_1 { \leqslant}b_2 & b_2 { \leqslant}b_3 } \qquad \infer[{\mathit{last}}]{\alpha b : { { a}^l}}{\beta b:{{a}^l } } \qquad \infer[{\mathit{eq}_{\leqslant}}]{\alpha b_2:a}{b_1{\leqslant}b_2 & b_2 { \leqslant}b_1 & \alpha b_1:a } \\ \infer[{\mathit{split}_{\leqslant}}]{\alpha : a}{b_1 { \leqslant}b_2 & \varphi & \infer*{\alpha : a}{[\varphi[b_2/b_1 ] ] } & \infer*{\alpha : a}{[b_1 { \lhd}b ' ] \ , [ b ' { \leqslant}b_2 ] } } \qquad \infer[{\mathit{base}_{\leqslant}}]{\alpha : a}{b_1 { \lhd}b_2 & \infer*{\alpha : a}{[b_1 { \leqslant}b_2 ] } } \qquad \infer[{\mathit{ind}}]{\alpha b : a}{\alpha b_0:a & b_0 { \leqslant}b & \infer*{\alpha b_j : a}{[b_0 { \leqslant}b_i ] \ , [ b_i { \lhd}b_j ] \ , [ \alpha b_i : a ] } } \end{array}\ ] ] the rules have the following side conditions : * in @xmath140 ( @xmath141 ) , @xmath26 is _ fresh _ , i.e. , it is different from @xmath19 and does not occur in any assumption on which @xmath30 depends other than the discarded assumption @xmath142 ( @xmath28 ) . * in @xmath36
, @xmath26 is _ fresh _
, i.e. , it is different from @xmath19 and @xmath20 , and does not occur in any assumption on which @xmath143 depends other than the discarded assumptions @xmath31 and @xmath32 .
* in @xmath144 , the formula must be of the form @xmath145 , as defined in . * in @xmath146
, @xmath26 is fresh , i.e. , it is different from @xmath15 and does not occur in any assumption on which @xmath82 depends other than the discarded assumption @xmath142 . * in @xmath147
, @xmath148 is fresh , i.e. , it is different from @xmath19 and @xmath26 and does not occur in any assumption on which @xmath82 depends other than the discarded assumptions @xmath149 and @xmath150 . * in @xmath151 , @xmath152 and @xmath153
are fresh , i.e. , they are different from each other and from @xmath15 and @xmath154 , and do not occur in any assumption on which @xmath155 depends other than the discarded assumptions of the rule .
the rules of @xmath47 are given in figure [ fig : rules ] .
in @xmath47 we do not make use of a proper relational labeling algebra ( as , e.g. , in @xcite ) that contains rules that derive rwffs from other rwffs or even lwffs .
since we are mainly interested in the derivation of logical formulas , we rather follow an approach that aims at simplifying the system : we use rwffs only as assumptions for the derivation of lwffs ( as in simpson s system for intuitionistic modal logic @xcite ) .
thus , in @xmath47 there are no rules whose conclusion is an rwff .
the rules @xmath156 and @xmath157 are just the labeled version of the standard @xcite natural deduction rules for implication introduction and elimination , where the notion of _ discharged / open assumption _ is also standard ; e.g. , @xmath158 $ ] means that the formula is discharged in @xmath156 .
the rule @xmath159 is a labeled version of _ reductio ad absurdum _
, where we do not constrain the time instant sequence ( @xmath160 ) in which we derive a contradiction to be the same ( @xmath161 ) as in the assumption .
the rules for the introduction and the elimination of @xmath35 and @xmath10 share the same structure since they both have a `` universal '' formulation .
consider , for instance , @xmath35 and the corresponding relation @xmath27 .
the idea underlying the introduction rule @xmath141 is that the meaning of @xmath162 is given by the metalevel implication @xmath163 for an arbitrary @xmath26 @xmath27-accessible from @xmath19 ( where the arbitrariness of @xmath26 is ensured by the side - condition on the rule ) . as we remarked above , the operators @xmath35 and @xmath10 have a local nature , in that when we write @xmath164 ( and similarly for @xmath165 ) we are stating that @xmath41 holds at time instant @xmath19 , which is the last in the sequence @xmath42 .
hence , the elimination rule @xmath40 says that if @xmath26 is @xmath27-accessible from @xmath19 ( i.e. , @xmath31 ) , then we can conclude that @xmath4 holds for the sequence @xmath43 .
similar observations hold for @xmath10 and the corresponding relation @xmath38 .
the rule @xmath146 models the fact that every time instant has an immediate successor , while the rule @xmath166 specifies that such a successor must be unique . @xmath146 tells us that if assuming @xmath142 we can derive @xmath82 , then we can discharge the assumption and conclude that indeed @xmath82 .
@xmath166 is slightly more complex : assume that @xmath19 had two different immediate successors @xmath26 and @xmath20 ( which we know can not be ) and assume that the generic formula @xmath90 holds ; if by substituting @xmath20 for @xmath26 in @xmath90 we obtain @xmath82 , then we can discharge the assumption and conclude that indeed @xmath82 . similarly , the rules @xmath167 and @xmath168 state the reflexivity and transitivity of @xmath27 , while @xmath169 captures substitution of equals.$ ] from @xmath90 , @xmath170 and @xmath171 . ] the rule @xmath147 states that if @xmath31 , then either @xmath172 or @xmath173 .
the rule thus works in the style of a disjunction elimination : if by assuming either of the two cases , we can derive a formula @xmath82 , then we can discharge the assumptions and conclude @xmath82 . since we do not use @xmath174 and @xmath175 explicitly in our syntax , we express such relations in an indirect way : the equality of @xmath19 and @xmath26
is expressed by replacing one with the other in a generic formula @xmath90 , @xmath175 by the composition of @xmath38 and @xmath27 .
the rule @xmath176 expresses the fact that @xmath27 contains @xmath38 , while the rule @xmath151 models the induction principle underlying the relation between @xmath38 and @xmath27 .
if ( base case ) @xmath4 holds in @xmath177 and if ( inductive step ) by assuming that @xmath4 holds in @xmath178 for an arbitrary @xmath152 @xmath27-accessible from @xmath154 , we can derive that @xmath4 holds also in @xmath179 , where @xmath153 is the immediate successor of @xmath152 , then we can conclude that @xmath4 holds in every @xmath180 such that @xmath15 is @xmath27-accessible from @xmath154 . finally , we have three rules that speak about the history and the label sequences : the rules @xmath36 and @xmath37 , which we already described in the introduction , and @xmath144 .
this rule expresses what we also anticipated in the introduction : the standard operators ( and connectives ) of @xmath6 speak not about sequences of time instants but about single time instants , and thus if a formula @xmath4 whose outermost operator is not @xmath5 holds at @xmath181 , then we can safely replace @xmath182 by any other sequence @xmath21 and conclude that @xmath4 holds at @xmath180 . to formalize this
, we define the set of ( well - formed ) _
@xmath183-formulas _ ( denoted by @xmath145 ) by means of the grammar @xmath184 where @xmath185 is a propositional symbol . hence
, in a formula @xmath145 , the history operator @xmath5 can only appear in the scope of a temporal operator @xmath35 ( and thus of @xmath13 as in the translation ) or @xmath10 .
the rule @xmath144 applies to these formulas only ; in fact , the `` @xmath186 '' in @xmath145 stands for `` last '' , but it also conveniently evokes both `` local '' and `` @xmath6 '' . for formulas @xmath24 whose outermost operator is the history operator @xmath5 , such a rule does not make sense ( and in fact is not sound ) as it would mean changing the interval over which @xmath4 holds .
such considerations are formalized in the following lemma , where we prove , for @xmath183-formulas , a result that is the analogous of the one given in lemma [ lem : last ] with respect to the translation of @xmath6-formulas .
is a direct consequence of lemma [ lem : last - al ] and of lemma [ lem : tr - ltl - is - ltllocal ] below .
] at the same time , we also prove that if @xmath4 is a formula belonging to the syntactic category @xmath187 of the grammar ( we will call such formulas _ @xmath188-formulas _ ) , then the truth value of @xmath4 depends on at most the last two elements of an observation sequence .
[ lem : last - al ] let @xmath71 be an @xmath6-model , @xmath114 $ ] an observation sequence , @xmath145 an @xmath183-formula and @xmath187 an @xmath188-formula .
then : @xmath189 @xmath115 { \models_{_{\nabla}}}{{a}^l}$ ] iff @xmath116 { \models_{_{\nabla}}}{{a}^l}$ ] for every sequence @xmath190 and @xmath191 @xmath192 @xmath193 @xmath94 $ ] @xmath194 iff @xmath195 { \models_{_{\nabla}}}{{a}^{l{\nabla}}}$ ] for every sequence @xmath118 . given the rules in fig .
[ fig : rules ] , the notions of _ derivation _ , _ assumption _ ( _ open _ or _ discharged _ , as we remarked ) and _ conclusion _ are the standard ones for natural deduction systems @xcite .
we write @xmath196 to say that there exists a derivation of @xmath82 in the system @xmath47 whose open assumptions are all contained in the set of formulas @xmath91 .
a derivation of @xmath82 in @xmath47 where all the assumptions are discharged is a _ proof _ of @xmath82 in @xmath47 and we then say that @xmath82 is a _ theorem _ of @xmath47 and write @xmath197 . to denote that @xmath198 is a derivation of @xmath82 whose set of assumptions may contain the formulas @xmath199 , we write @xmath200
if we are interested in @xmath6-reasoning , then we can restrict our attention to a subset of the @xmath47-derivations , namely , to the derivations where the conclusion and all the open assumptions correspond to the translations of @xmath6-formulas .
_ [ def : vdashltl ]
let @xmath198 be a derivation in @xmath47 and @xmath91 the set containing the conclusion and the open assumptions of @xmath198 .
we say that @xmath198 is an _
iff there exists a label @xmath15 such that for every @xmath90 in @xmath91 there exists an @xmath6-formula @xmath4 such that @xmath201 .
we write @xmath202 to denote that in @xmath47 there exists an @xmath6-derivation of @xmath203 from open assumptions in a set @xmath91 , where @xmath204 . _ in definition
[ def : vdashltl ] , we require all the open assumptions and the conclusion of an @xmath6-derivation to be lwffs labeled by the same single label @xmath15 .
note that , as a consequence of corollary [ cor : last ] , we would obtain the same notion of @xmath6-derivation by requiring instead that such formulas were labeled by the same sequence @xmath21 . in section [ sub : sandc ]
, we will show that @xmath47 is sound with respect to the semantics of @xmath46 and , by considering the notion of @xmath6-derivability @xmath205 , that it is sound and weakly complete with respect to @xmath6 .
an investigation of completeness with respect to @xmath46 is left for future work , together with the formalization of an axiomatization of @xmath46 .
related to this , it is important to understand what exactly is the relationship of the class of @xmath183-formulas and the class of @xmath6-formulas , in particular with respect to the translation @xmath112 .
it is not difficult to see that the co - domain of the translation is included in @xmath183 by construction of @xmath112 , i.e. , by induction on the formula complexity it follows that : [ lem : tr - ltl - is - ltllocal ] if @xmath4 is an @xmath6-formula , then @xmath206 is an @xmath183-formula .
the other direction is trickier , as it basically amounts to defining an inverse translation . to solve this problem , we have been considering normal forms of @xmath183-formulas and we conjecture that the following fact indeed holds .
if @xmath4 is an @xmath183-formula , then there exists an @xmath6-formula @xmath2 such that @xmath207 is semantically equivalent to @xmath4 .
[ th : ltlh - soundness ] for every set @xmath91 of labeled and relational formulas and every labeled formula @xmath82 , if @xmath196 , then @xmath208 .
the proof proceeds by induction on the structure of the derivation of @xmath82 .
the base case is when @xmath209 and is trivial .
there is one step case for every rule and we show here only the two representative cases @xmath210{\beta b_1 b_3:{\nabla}b } { \deduce{\beta b_1 b_2:b}{\deduce{\pi}{[b_1 { \leqslant}b_2 ] & [ b_2 { \leqslant}b_3 ] } } } \qquad \quad { \normalsize \text{and } } \qquad \quad \infer[{\mathit{last}}]{\beta b : a } { \deduce{\beta ' b : a } { \pi } } \ ] ] some more cases are in * ? ? ?
some more cases are in appendix [ ap : ltlh - soundness ] .
first , consider the case in which the last rule application is a @xmath36 , where @xmath211 , @xmath212 , and @xmath198 is a proof of @xmath213 from hypotheses in @xmath214 , with @xmath26 fresh and with @xmath215 . by the induction hypothesis , for every interpretation @xmath104 , if @xmath216 , then @xmath217 . we let @xmath104 be any interpretation such that @xmath218 , and show that @xmath219 .
let @xmath220 , @xmath221 and @xmath222 $ ] . since @xmath26 is fresh , we can extend @xmath104 to an interpretation ( still called @xmath104 for simplicity ) such that @xmath223 for an arbitrary @xmath224 .
the induction hypothesis yields @xmath217 , i.e. , @xmath225{\models_{_{\nabla}}}b$ ] , and thus , since @xmath226 is an arbitrary point between @xmath227 and @xmath228 , we obtain @xmath229 { \models_{_{\nabla}}}{\nabla}b$ ] .
it follows @xmath230 .
now consider the case in which the last rule applied is @xmath144 and @xmath231 , where @xmath198 is a proof of @xmath232 from hypotheses in @xmath91 . by applying the induction hypothesis on @xmath198
, we have @xmath233
. we proceed by considering a generic @xmath6-model @xmath71 and a generic interpretation @xmath104 on it such that @xmath218 and showing that this entails @xmath234 . by the induction hypothesis , @xmath235 ,
i.e. , @xmath236 by definition [ def : ltlh - truth ] .
since @xmath4 is an @xmath183-formula by the side condition of the rule and the two observation sequences @xmath237 and @xmath238 share the same last element @xmath239 , we can apply lemma [ lem : last - al ] and obtain @xmath240 , i.e. , @xmath234 by definition [ def : ltlh - truth ] . by exploiting the translation of section [ sub : translation ] and the notion of @xmath6-derivation of definition [ def : vdashltl ]
, we also prove a result of soundness with respect to @xmath6 .
[ th : ltl - soundness ] for every set @xmath60 of @xmath6-formulas and every @xmath6-formula @xmath4 , if @xmath202 , then @xmath241 . by definition of @xmath205 , for a given label @xmath15 ,
there exists a derivation in @xmath47 of @xmath203 from open assumptions in @xmath242 . by theorem
[ th : ltlh - soundness ] , @xmath243 implies @xmath244 .
since @xmath15 is generic , we have that for every @xmath6-model @xmath71 and every interpretation @xmath104 , @xmath245 implies @xmath246 iff for every natural number @xmath72 , @xmath122 { \models_{_{\nabla}}}{{\lambda}^*}$ ] implies @xmath247 { \models_{_{\nabla}}}{a^*}$ ] , where @xmath248 . by lemma [ lem : last ] , we infer that for every observation sequence @xmath95 , @xmath249 implies @xmath250 . by definition [ def : ltlh - truth ] , @xmath251 and thus , by theorem [ th : translation ] , we conclude @xmath241 . as we anticipated , an analysis of the completeness of @xmath47 with respect to @xmath46 is left for future work . here
we discuss completeness with respect to @xmath6 .
the proposed natural deduction system consists of only finitary rules ; consequently , it can not be strongly complete for @xmath6 .
and @xmath35 have such a defect ; see , e.g. , ( * ? ? ?
* ch . 6 ) . ]
nevertheless , by using the axiomatization @xmath74 and the translation @xmath112 , we can give a proof of weak completeness for it ; namely : [ th : ltl - completeness ] for every @xmath6-formula @xmath4 , if @xmath252 , then @xmath253 .
we can prove the theorem by showing that @xmath47 is complete with respect to the axiomatization @xmath74 given in section [ sec : ltl ] , which is sound and complete for the logic @xmath6 .
that is , we need to prove that : ( i ) the translation , via @xmath112 , of every axiom of @xmath74 is provable in @xmath47 by means of an @xmath6-derivation , and ( ii ) the notion of @xmath205 is closed under the ( labeled equivalent of the ) rules of inference of @xmath74 . showing ( ii ) is straightforward and we omit it here . as an example for ( i ) , we give here a derivation of the translation of @xmath254 .
the other cases are presented in * ? ? ?
the other cases are presented in appendix [ ap : ltl - completeness ] .
@xmath255 { b : { \mathsf{g } } ( a { \supset}{\mathsf{x}}a ) { \supset } ( a { \supset}{\mathsf{g}}a ) } { \infer[{{\supset}\!\!i}^2 ] { b : a { \supset}{\mathsf{g}}a } { \infer[{\supset}^3 ] { b : { \mathsf{g}}a } { \infer[{\mathit{last } } ] { bc : a } { \infer[{\mathit{ind}}^4 ] { c : a } { [ b : a]^2 & [ b{\leqslant}c]^3 & \infer[{\mathit{last } } ] { b_j : a } { \infer[{{\mathsf{x}}e } ] { b b_i b_j : a } { \infer[{{\supset}\!\!e } ] { b b_i : { \mathsf{x}}a } { \infer[{{\mathsf{g}}e } ] { b b_i : a { \supset}{\mathsf{x}}a } { [ b:{\mathsf{g } } ( a { \supset}{\mathsf{x}}a)]^1 & [ b { \leqslant}b_i]^4 } & \infer[{\mathit{last } } ] { b b_i : a } { [ b_i : a]^4 } } & [ b_i { \lhd}b_j]^4 } } } } } } } } \ ] ]
the introduction of the operator @xmath5 has allowed us to formalize the `` history '' of until and thus , via a proper translation , to give a labeled natural deduction system for a linear time logic endowed with @xmath5 that is also sound and complete with respect to @xmath6 with until . as we remarked above ,
we see this work as spawning several different directions for future research .
first , the `` recipe '' for dealing with until that we gave here is abstract and general , and thus provides the basis for formalizing deduction systems for temporal logics endowed with @xmath7 , both linear and branching time .
we are currently considering @xmath256 and its sublogics as in @xcite and are also working at a formal characterization of the class of logics that can be captured with our approach .
second , the well - behaved nature of our approach , where each connective and operator has one introduction and one elimination rule , paves the way to a proof - theoretical analysis of the resulting natural deduction systems , e.g. , to show proof normalization and other useful meta - theoretical properties , which we are tackling in current work .
moreover , we are also considering different optimizations of the rules .
in particular , along the lines of the discussion about the rule @xmath144 ( and corollary [ cor : last ] and definition [ def : vdashltl ] ) , we are investigating to what extent we can use sequences as labels only when they are really needed , which would also simplify the proofs of normalization and other meta - properties , i.e. , considering @xmath145 and @xmath187 , is closely related , at least in spirit , to the focus on _ persistent _ formulas when combining intuitionistic and classical logic so as to avoid the collapse of the two logics into one , see @xcite but also @xcite .
we are , after all , considering here formulas stemming from two classes ( if not two logics altogether ) , and it makes thus sense that they require different labeling ( single instants and sequences ) . ] .
this is closely related to the formalization of the relationship between the class of @xmath183-formulas and that of @xmath6-formulas , which in turn will allow us to reason about the completeness of @xmath47 with respect to the semantics of @xmath46 and also to provide an axiomatization of @xmath46 ( thus treating it as a full - fledged logic as opposed to as a `` service '' logic for @xmath6 as we did here ) .
finally , it is worth observing that several works have considered _ interval temporal logics _ , e.g. , @xcite .
while these works consider intervals explicitly , we have used them somehow implicitly here , as a means to formalize the dual nature of until via the history @xmath5 , and this is another reason why it is interesting to reduce the use of label sequences as much as possible .
a more detailed comparison of our approach with these works is left for future work . *
acknowledgments * this work was partially supported by the prin projects `` concerto '' and `` soft '' .
10url # 1`#1`urlprefix basin , d. , c. caleiro , j. ramos and l. vigan , _ labeled tableaux for distributed temporal logic _
, j. logic and computation ( 2009 , doi : 10.1093/logcom / exp022 ) .
+ bolotov , a. , o. grigoriev and v. shangin , _ automated natural deduction for propositional linear - time temporal logic .
_ , in : _ proc .
time07 _ ( 2007 ) , pp .
+ bowman , h. and s. thompson , _ a decision procedure and complete axiomatization of finite interval temporal logic with projection _ ,
j. logic and computation * 13 * ( 2003 ) .
+ caleiro , c. and j. ramos , _ combining classical and intuitionistic implications _
, in : _ proc .
frocos07 _ , lncs 4720 , 2007 , pp .
+ cerrito , s. and m. cialdea mayer , _ labelled tableaux for linear time temporal logic over finite time frames _ , in : _ labelled deduction _ ,
kluwer , 2000 , pp .
+ farias del cerro , l. and a. herzig , _ combining classical and intuitionistic logic _ , in : _ proc .
frocos96 _ , 1996 , pp . 93102 .
+ fisher , m. , d. m. gabbay and l. vila , editors , `` handbook of temporal reasoning in artificial intelligence i , '' elsevier , 2005 .
+ gabbay , d. m. , `` labelled deductive systems , '' clarendon press , 1996 .
+ gabbay , d. m. , _ an overview of fibred semantics and the combination of logics _ , in : _ proc .
frocos07 _ , lncs 4720 , 2007 , pp .
+ goldblatt , r. i. , `` logics of time and computation , '' csli lecture notes , 1987 .
+ goranko , v. , a. montanari , p. sala and g. sciavicco , _ a general tableau method for propositional interval temporal logics : theory and implementation _ , j. applied logic * 4 * ( 2006 ) , pp .
+ gore , r. , _ tableau methods for modal and temporal logics _ , in : _ handbook of tableau methods _ , kluwer , 1999 .
+ gough , g. d. , _ decision procedures for temporal logic _ , technical report umcs-89 - 10 - 1 , department of computer science , university of manchester ( 1984 ) .
+ halpern , j. and y. shoham , _ a propositional modal logic of time intervals _ , jacm * 38 * ( 1991 ) , pp .
+ krger , f. , `` temporal logic of programs , '' springer - verlag , 1987 .
+ masini , a. , l. vigan and m. volpe , _ a labeled natural deduction system for a fragment of @xmath257 _ , in : _ proc .
lfcs09 _ , lncs 5407 ( 2009 ) , pp .
+ prawitz , d. , `` natural deduction , ' ' almquist and wiskell , 1965 .
+ reynolds , m. , _ a tableau for bundled ctl * _ , j. logic and computation * 17 * ( 2007 ) .
+ schmitt , p. h. and j. goubault - larrecq , _ a tableau system for linear - time temporal logic _ , in : _ proc .
tacas97 _ , lncs 1217 ( 1997 ) , pp .
130144 , see also the unpublished manuscript _ a tableau system for full linear temporal logic_. + schwendimann , s. , _ a new one - pass tableau calculus for pltl _ , in : _ proc .
tableaux98 _ , lnai 1397 ( 1998 ) , pp .
+ simpson , a. , `` the proof theory and semantics of intuitionistic modal logic , '' ph.d .
thesis , school of informatics , university of edinburgh ( 1994 ) .
+ vigan , l. , `` labelled non - classical logics , '' kluwer , 2000 .
+ vigan , l. and m. volpe , _ labeled natural deduction systems for a family of tense logics _ , in : _ proc .
time08 _ ( 2008 ) , pp .
@xmath260 . : : then the translation of @xmath4 is @xmath261 . by definition [ def : ltlh - truth ] , we obtain @xmath115 { \models_{_{\nabla}}}{b^ * } { \supset}{c^*}$ ] iff @xmath115 { \models_{_{\nabla}}}{b^ * } \,$ ] implies @xmath262 { \models_{_{\nabla}}}{c^*}.$ ] by the induction hypothesis , we see that this holds iff @xmath116 { \models_{_{\nabla}}}{b^*}$ ] implies @xmath116 { \models_{_{\nabla}}}{c^*}$ ] for every sequence @xmath118 and thus , by definition [ def : ltlh - truth ] , iff for every sequence @xmath118 , @xmath263 { \models_{_{\nabla}}}{b^ * } { \supset}{c^*}$ ] . @xmath264 .
: : then @xmath265 . by definition [ def : ltlh - truth ] ,
@xmath115 { \models_{_{\nabla}}}{\mathsf{g}}{b^*}$ ] iff @xmath266 { \models_{_{\nabla}}}{b^*}$ ] iff ( by the induction hypothesis ) @xmath267 @xmath268 { \models_{_{\nabla}}}{b^*}$ ] for every sequence @xmath118 iff ( by definition [ def : ltlh - truth ] ) @xmath116 { \models_{_{\nabla}}}{\mathsf{g}}{b^*}$ ] , for every sequence @xmath118 . @xmath269 .
: : this case is very similar to the previous one and we omit it . @xmath270 . : : then @xmath271 . by definition [ def : ltlh - truth ] ,
we have @xmath272 @xmath273 @xmath93 @xmath94 { \models_{_{\nabla}}}{a^*}$ ] iff @xmath274 { \models_{_{\nabla}}}{c^*}$ ] or @xmath115 $ ] @xmath275 iff @xmath274 $ ] @xmath276 or @xmath277 $ ] @xmath278 iff @xmath274 $ ] @xmath276 or @xmath277 $ ] @xmath279 and @xmath280 { \models_{_{\nabla}}}{\nabla}{b^*}))$ ] iff @xmath274 { \models_{_{\nabla}}}{c^*}$ ] or @xmath281 { \models_{_{\nabla}}}{c^*}$ ] and @xmath282 implies @xmath192 @xmath283 { \models_{_{\nabla}}}{b^*}))$ ] iff ( by the induction hypothesis ) for every sequence @xmath118 , we have @xmath284 { \models_{_{\nabla}}}{c^*}$ ] or @xmath285 { \models_{_{\nabla}}}{c^*}$ ] and @xmath282 implies @xmath286 { \models_{_{\nabla}}}{b^*}))$ ] iff ( by definition [ def : ltlh - truth ] ) @xmath116 $ ] @xmath287 for every sequence @xmath118 . @xmath260 .
: : then @xmath261 . we have @xmath288 iff ( by definition [ def : ltl - truth ] ) @xmath289 implies @xmath290 iff ( by the induction hypothesis ) @xmath122 { \models_{_{\nabla}}}{b^*}$ ] implies @xmath122 { \models_{_{\nabla}}}{c^*}$ ] iff ( by definition [ def : ltlh - truth ] ) @xmath247 { \models_{_{\nabla}}}{b^ * } { \supset}{c^*}$ ] . @xmath264 . : : then @xmath265 .
we have @xmath291 iff ( by definition [ def : ltl - truth ] ) @xmath292 iff ( by the induction hypothesis ) @xmath293 { \models_{_{\nabla}}}{b^*}$ ] iff ( by lemma [ lem : last ] ) @xmath294 { \models_{_{\nabla}}}{b^*}$ ] iff ( by definition [ def : ltlh - truth ] ) @xmath247 { \models_{_{\nabla}}}{\mathsf{g}}{b^*}$ ] . @xmath269 .
: : this case is very similar to the previous one and we omit it . @xmath270 . : : then @xmath271 .
we have @xmath121 iff ( by definition [ def : ltl - truth ] ) @xmath295 and @xmath296 implies @xmath297 iff @xmath298 or @xmath299 and @xmath300 implies @xmath301 iff ( by the induction hypothesis ) @xmath247 { \models_{_{\nabla}}}{c^*}$ ] or @xmath302 { \models_{_{\nabla}}}{c^*}$ ] and @xmath300 implies @xmath303 { \models_{_{\nabla}}}{b^*})$ ] iff ( by lemma [ lem : last ] ) @xmath247 { \models_{_{\nabla}}}{c^*}$ ] or @xmath304 { \models_{_{\nabla}}}{c^*}$ ] and @xmath300 implies @xmath305 { \models_{_{\nabla}}}{b^*})$ ] iff @xmath247 { \models_{_{\nabla}}}{c^*}$ ] or @xmath306 { \models_{_{\nabla}}}{c^*}$ ] and @xmath307 implies @xmath305 { \models_{_{\nabla}}}{b^*})$ ] iff ( by definition [ def : ltlh - truth ] ) @xmath247{\models_{_{\nabla}}}{c^*}$ ] or @xmath308 @xmath309 { \models_{_{\nabla}}}{\mathsf{x}}{c^ * } \wedge { \nabla}{b^*})$ ] iff ( by definition [ def : ltlh - truth ] ) @xmath247{\models_{_{\nabla}}}{c^ * } \vee \ , { \mathsf{f}}({\mathsf{x}}{c^ * } \wedge { \nabla}{b^*}).$ ] the proofs of the statements @xmath189 and @xmath191 proceed in parallel and are by induction on the formula complexity .
the base case is when @xmath310 or @xmath311 and is trivial .
there is one inductive step case for each other formation case coming from the recursive definition of the grammar . along the proof ,
@xmath312 denote @xmath183-formulas while @xmath313 denote @xmath188-formulas .
: : by definition [ def : ltlh - truth ] , we have @xmath115 { \models_{_{\nabla}}}{{b}^l } { \supset}{{c}^l}\;$ ] iff @xmath315 @xmath273 @xmath316 { \models_{_{\nabla}}}{{b}^l } \,$ ] implies @xmath262 { \models_{_{\nabla}}}{{c}^l}.$ ] by the induction hypothesis , we see that this holds iff @xmath317 @xmath94 { \models_{_{\nabla}}}{{b}^l } \
, \mbox { implies } \ , { \mathcal{m } } , [ m_1 , \ldots,$ ] @xmath318 @xmath94 { \models_{_{\nabla}}}{{c}^l}\;$ ] for every sequence @xmath118 and thus , by definition [ def : ltlh - truth ] , iff for every sequence @xmath118 , @xmath319 @xmath320 @xmath94 $ ] @xmath321 . @xmath322 . : : @xmath115 { \models_{_{\nabla}}}{\mathsf{g}}{{b}^{l{\nabla}}}\;$ ] iff ( by definition [ def : ltlh - truth ] ) @xmath323 @xmath324 { \models_{_{\nabla}}}{{b}^{l{\nabla } } } \;$ ] iff ( by the induction hypothesis ) @xmath325 @xmath268 { \models_{_{\nabla}}}{{b}^{l{\nabla } } } \,$ ] for every sequence @xmath326 iff ( by definition [ def : ltlh - truth ] ) @xmath263 { \models_{_{\nabla}}}{\mathsf{g}}{{b}^{l{\nabla } } } \,$ ] for every sequence @xmath118 . @xmath327 . :
: this case is very similar to the previous one and we omit it . @xmath328 . : : @xmath115 { \models_{_{\nabla}}}{{b}^l } \;$ ] iff ( by the induction hypothesis ) @xmath319
@xmath329 @xmath330 { \models_{_{\nabla}}}{{b}^l}\;$ ] for every sequence @xmath331 and thus also @xmath319 @xmath332 @xmath333 { \models_{_{\nabla}}}{{b}^l}\;$ ] for every sequence @xmath118 . @xmath334 .
: : @xmath115 { \models_{_{\nabla}}}{{b}^{l{\nabla } } } { \supset}{{c}^{l{\nabla}}}\;$ ] iff ( by definition [ def : ltlh - truth ] ) @xmath335 @xmath316 { \models_{_{\nabla}}}{{b}^{l{\nabla } } } \,$ ] implies @xmath262 { \models_{_{\nabla}}}{{c}^{l{\nabla}}}.$ ] by the induction hypothesis , this holds iff @xmath336 { \models_{_{\nabla}}}{{b}^{l{\nabla } } } \ , \mbox { implies } \ , { \mathcal{m}},$ ] @xmath337 @xmath93 @xmath333 { \models_{_{\nabla}}}{{c}^{l{\nabla}}}\;$ ] for every sequence @xmath118 and thus , by definition [ def : ltlh - truth ] , iff for every sequence @xmath118 , @xmath338 $ ] @xmath339 . @xmath340 . : : @xmath115 { \models_{_{\nabla}}}{\nabla}{{b}^{l{\nabla}}}\;$ ] iff ( by definition [ def : ltlh - truth ] ) @xmath341 { \models_{_{\nabla}}}{{b}^{l{\nabla } } } \;$ ] iff ( by the induction hypothesis ) @xmath342 { \models_{_{\nabla}}}{{b}^{l{\nabla } } } \;$ ] for every sequence @xmath326 iff ( by definition [ def : ltlh - truth ] ) @xmath338 { \models_{_{\nabla}}}{\nabla}{{b}^{l{\nabla } } } \,$ ] for every sequence @xmath118 .
consider the case in which the last rule application is a @xmath141 , where @xmath343 and @xmath344 : @xmath345{\beta b_1:{\mathsf{g}}b } { \deduce{\beta b_1 b_2:b}{\deduce{\pi}{[b_1 { \leqslant}b_2 ] } } } \ ] ] where @xmath198 is a proof of @xmath346 from hypotheses in @xmath214 , with @xmath26 fresh and with @xmath347 . by the induction hypothesis , for all interpretations @xmath104 , if @xmath216 , then @xmath217 . we let @xmath104 be any interpretation such that @xmath218 , and show that @xmath348 .
let @xmath220 and @xmath222 $ ] . since @xmath26 is fresh , we can extend @xmath104 to an interpretation ( still called @xmath104 for simplicity ) such that @xmath349 for an arbitrary @xmath350 .
the induction hypothesis yields @xmath217 , i.e. , @xmath229{\models_{_{\nabla}}}b$ ] , and thus , since @xmath228 is arbitrary , we obtain @xmath351 { \models_{_{\nabla}}}{\mathsf{g}}b$ ] .
it follows @xmath352 .
now consider the case in which the last rule applied is @xmath40 and @xmath353 : @xmath354{\beta b_1 b_2 : a } { \deduce{\beta b_1 : { \mathsf{g}}a}{\pi } & b_1 { \leqslant}b_2 } \ ] ] where @xmath198 is a proof of @xmath355 from hypotheses in @xmath356 , with @xmath357 for some set @xmath356 of formulas . by applying the induction hypothesis on @xmath198
, we have : @xmath358 we proceed by considering a generic @xmath6-model @xmath71 and a generic interpretation @xmath104 on it such that @xmath218 and showing that this entails @xmath359 since @xmath360 , we deduce @xmath361 and , from the induction hypothesis , @xmath362 . furthermore @xmath218 entails @xmath363 .
then , by definition [ def : ltlh - truth ] , we obtain @xmath364 .
now consider the case in which the last rule applied is @xmath37 and @xmath353 : @xmath365{\beta b_1 b_2 : a } { \deduce{\beta b_1 b_3 : { \nabla}a}{\pi } & b_1 { \leqslant}b_2 & b_2 { \leqslant}b_3 } \ ] ] where @xmath198 is a proof of @xmath366 from hypotheses in @xmath356 , with @xmath367 for some set @xmath356 of formulas . by applying the induction hypothesis on @xmath198
, we have : @xmath368 we proceed by considering a generic @xmath6-model @xmath71 and a generic interpretation @xmath104 on it such that @xmath218 and showing that this entails @xmath359 since @xmath360 , we deduce @xmath361 and , from the induction hypothesis , @xmath369 .
furthermore @xmath218 entails @xmath363 and @xmath370 .
then , by definition [ def : ltlh - truth ] , we obtain @xmath364
. finally , consider the case in which the last rule applied is @xmath151 and @xmath231 : @xmath371{\beta b : a } { \deduce{\beta b_0 : a}{\pi ' } & b_0 { \leqslant}b & \deduce{\beta b_j : a}{\deduce{\pi}{[b_0 { \leqslant}b_i ] & [ b_i { \lhd}b_j ] & [ \beta b_i : a ] } } } \ ] ] where @xmath198 is a proof of @xmath372 from hypotheses in @xmath373 and @xmath374 is a proof of @xmath375 from hypotheses in @xmath356 , with @xmath376 and @xmath377 for some set @xmath356 of formulas .
the side - condition on @xmath151 ensures that @xmath152 and @xmath153 are fresh in @xmath198 .
hence , by applying the induction hypothesis on @xmath198 and @xmath374 , we have : @xmath378 we proceed by considering a generic @xmath6-model @xmath71 and a generic interpretation @xmath104 on it such that @xmath218 and showing that this entails @xmath379 first , we note that @xmath360 and therefore @xmath218 implies @xmath361 and , by the induction hypothesis on @xmath374 , @xmath380 .
now let @xmath381 for some natural number @xmath72 . from @xmath218 , we deduce @xmath382 and thus @xmath383 for some @xmath384 .
we show by induction on @xmath385 that @xmath234 . as a base case
, we have @xmath386 ; it follows that @xmath387 and thus trivially that @xmath380 entails @xmath234 .
let us consider now the induction step . given a label @xmath388 such that @xmath389 ,
we show that the induction hypothesis @xmath390 entails the thesis @xmath234 .
we can build an interpretation @xmath391 that differs from @xmath104 only in the points assigned to @xmath152 and @xmath153 , namely , @xmath392[b_j
\mapsto n+k]$ ] .
it is easy to verify that the interpretation @xmath391 is such that the following three conditions hold : furthermore , the side - condition on the rule @xmath151 ensures that @xmath104 and @xmath391 agree on all the labels occurring in @xmath356 , from which we can infer @xmath396 .
it follows @xmath397 and thus , by the induction hypothesis on @xmath198 , @xmath398 .
we conclude @xmath399 by observing that @xmath400 .
we present here the @xmath47-derivations of the remaining axioms of @xmath74 . note that , for simplicity , we use also some rules ( i.e. , @xmath401 , @xmath402 , @xmath403 , @xmath404 , @xmath405 and @xmath406 ) concerning derived operators .
they can be easily derived from the set of rules in figure [ fig : rules ] .
@xmath408 { b:{\mathsf{g } } ( a { \supset}b ) { \supset}({\mathsf{g}}a { \supset}{\mathsf{g}}b ) } { \infer[{{\supset}\!\!i}^2 ] { b : { \mathsf{g}}a { \supset}{\mathsf{g}}b } { \infer[{{\mathsf{g}}i}^3 ] { b : { \mathsf{g}}b } { \infer[{{\supset}\!\!e } ] { bc : b } { \infer[{{\mathsf{g}}e } ] { bc : a { \supset}b } { [ b:{\mathsf{g } } ( a { \supset}b)]^1 & [ b{\leqslant}c]^3 } & \infer[{{\mathsf{g}}e } ] { bc : a } { [ b : { \mathsf{g}}a]^2 & [ b { \leqslant}c]^3 } } } } } \ ] ] @xmath410 @xmath411 { b:{\mathsf{x}}\neg a { \supset}\neg { \mathsf{x}}a } { \infer[{\mathit{ser}_{\lhd}}^2 ] { b:\neg { \mathsf{x}}a } { \infer[{{\perp}\!\!e}^3 ] { b:\neg { \mathsf{x}}a } { \infer[{{\supset}\!\!e } ] { bc:{\perp } } { \infer[{{\mathsf{x}}e } ] { bc:\neg a } { [ b:{\mathsf{x}}\neg a]^1 & [ b { \lhd}c]^2 } & \infer[{{\mathsf{x}}e } ] { bc : a } { [ b:{\mathsf{x}}a]^3 & [ b { \lhd}c]^2 } } } } } \\ \ \\ \infer[{{\supset}\!\!i}^1 ] { b:\neg { \mathsf{x}}a { \supset}{\mathsf{x}}\neg a } { \infer[{{\mathsf{x}}i}^2 ] { b : { \mathsf{x}}\neg a } { \infer[{\supset}^3 ] { bc:\neg a } { \infer[{{\supset}\!\!e } ] { b:{\perp } } { [ b:\neg { \mathsf{x}}a]^1 & \infer[{{\mathsf{x}}i}^4 ] { b:{\mathsf{x}}a } { \infer[{\mathit{lin}_{\lhd } } ] { bd : a } { [ b { \lhd}c]^2 & [ b { \lhd}d]^4 & [ bc : a]^3 } } } } } } \end{array}\ ] ] @xmath414 { b:{\mathsf{g}}a { \supset}(a \wedge { \mathsf{x}}{\mathsf{g}}a ) } { \infer[\wedge i ] { b : a \wedge { \mathsf{x}}{\mathsf{g}}a } { \infer[{\mathit{refl}_{\leqslant}}^2 ] { b : a } { \infer[{\mathit{last } } ] { b : a } { \infer[{{\mathsf{g}}e } ] { b b : a } { [ b : { \mathsf{g}}a]^1 & [ b{\leqslant}b]^2 } } } & \infer[{{\mathsf{x}}i}^3 ] { b:{\mathsf{x}}{\mathsf{g}}a } { \infer[{{\mathsf{g}}i}^4 ] { bc:{\mathsf{g}}a } { \infer[{\mathit{last } } ] { bcd : a } { \infer[{\mathit{base}_{\leqslant}}^5 ] { bd : a } { [ b { \lhd}c]^3 & \infer[{\mathit{trans}_{\leqslant}}^6 ] { bd : a } { [ b { \leqslant}c]^5 & [ c { \leqslant}d]^4 & \infer[{{\mathsf{g}}e } ] { bd : a } { [ b : { \mathsf{g}}a]^1 & [ b { \leqslant}d]^6 } } } } } } } } } \ ] ] left - to - right direction : @xmath255 { b : { \mathsf{f}}({\mathsf{x}}b \wedge { \nabla}a ) { \supset}(a \wedge { \mathsf{x}}(b \vee { \mathsf{f}}({\mathsf{x}}b \wedge { \nabla}a)))}{\infer[{{\mathsf{f}}e}^2 ] { b : a \wedge { \mathsf{x}}(b \vee { \mathsf{f}}({\mathsf{x}}b \wedge { \nabla}a ) ) } { [ b:{\mathsf{f}}({\mathsf{x}}b \wedge { \nabla}a)]^1 & \infer[{\wedge i } ] { b : a \wedge { \mathsf{x}}(b \vee { \mathsf{f}}({\mathsf{x}}b \wedge { \nabla}a))}{\infer[{\mathit{last } } ] { b : a}{\infer[{\mathit{refl}_{\leqslant}}^3 ] { bb : a}{\infer[{{\nabla}e } ] { bb : a}{\infer[{\wedge e } ] { bc:{\nabla}a } { [ bc:{\mathsf{x}}b \wedge { \nabla}a]^2 } & [ b{\leqslant}b]^3 & [ b { \leqslant}c]^2 } } } & \deduce { b:{\mathsf{x}}(b \vee { \mathsf{f}}({\mathsf{x}}b \wedge { \nabla}a))}{\pi_1 } } } } } \ ] ] where @xmath418 is the following derivation : @xmath419 { b:{\mathsf{x}}(b \vee { \mathsf{f}}({\mathsf{x}}b \wedge { \nabla}a ) ) } { \infer[{\mathit{split}_{\leqslant}}^5 ] { bb':b\vee { \mathsf{f}}({\mathsf{x}}b \wedge { \nabla}a ) } { [ b{\leqslant}c]^2 & [ b{\lhd}b']^4 & \infer[{\vee i } ] { bb':b \vee { \mathsf{f}}({\mathsf{x}}b \wedge { \nabla}a)}{\infer[{\mathit{last } } ] { bb':b}{\infer[{{\mathsf{x}}e } ] { bcb':b}{\infer[{\wedge e } ] { bc:{\mathsf{x}}b } { [ bc:{\mathsf{x}}b \wedge { \nabla}a]^2 } & [ c { \leqslant}b']^5 } } } & \infer[{\mathit{lin}_{\lhd}}^6 ] { bb':b \vee { \mathsf{f}}({\mathsf{x}}b \wedge { \nabla}a)}{b{\lhd}b ' & b { \lhd}b '' & \deduce { bb'':b \vee
{ \mathsf{f}}({\mathsf{x}}b \wedge { \nabla}a)}{\pi_2 } & [ bb':b \vee { \mathsf{f}}({\mathsf{x}}b \wedge { \nabla}a)]^6 } } } } \ ] ] and @xmath420 is the following derivation : @xmath421 { bb'':b \vee { \mathsf{f}}({\mathsf{x}}b \wedge { \nabla}a)}{\infer[{{\mathsf{f}}i } ] { bb'':{\mathsf{f}}({\mathsf{x}}b \wedge { \nabla}a)}{\infer[{\wedge i } ] { bb''c:{\mathsf{x}}b \wedge { \nabla}a}{\infer[{{\mathsf{x}}i}^7 ] { bb''c:{\mathsf{x}}b}{\infer[{\mathit{last } } ] { bb''cc':b}{\infer[{{\mathsf{x}}e } ] { bcc':b}{\infer[{\wedge e } ] { bc:{\mathsf{x}}b } { [ bc:{\mathsf{x}}b \wedge { \nabla}a]^2 } & [ c { \lhd}c']^7 } } } & \infer[{{\nabla}i}^8 ] { bb''c:{\nabla}a}{\infer[{\mathit{last } } ] { bb''d : a}{\infer[{\mathit{base}_{\leqslant}}^9 ] { bd : a}{[b{\lhd}b'']^5 & \infer[{\mathit{trans}_{\leqslant}}^{10 } ] { bd : a}{[b{\leqslant}b'']^9 & [ b '' { \leqslant}d]^8 & \infer[{{\nabla}e } ] { bd : a}{\infer[{\wedge e } ] { bc:{\nabla}a } { [ bc:{\mathsf{x}}b \wedge { \nabla}a]^2 } & [ b{\leqslant}d]^{10 } & [ d{\leqslant}c]^8 } } } } } } & [ b''{\leqslant}c]^5 } } } \ ] ] @xmath255 { b : ( a \wedge { \mathsf{x}}(b \vee { \mathsf{f}}({\mathsf{x}}b \wedge { \nabla}a ) ) ) { \supset}{\mathsf{f}}({\mathsf{x}}b \wedge { \nabla}a)}{\infer[{\mathit{ser}_{\lhd}}^2 ] { b:{\mathsf{f}}({\mathsf{x}}b \wedge { \nabla}a)}{\infer[{\vee e}^3 ] { b:{\mathsf{f}}({\mathsf{x}}b \wedge { \nabla}a)}{\infer[{{\mathsf{x}}e } ] { b e : b \vee { \mathsf{f}}({\mathsf{x}}b \wedge { \nabla}a)}{\infer[{\wedge e } ] { b:{\mathsf{x}}(b \vee { \mathsf{f}}({\mathsf{x}}b \wedge { \nabla}a))}{[\varphi]^1 } & [ b { \lhd}e]^2 } & \deduce { b:{\mathsf{f}}({\mathsf{x}}b \wedge { \nabla}a)}{\deduce{\pi_1 } { [ b e : b]^3 } } & \deduce { b:{\mathsf{f}}({\mathsf{x}}b \wedge { \nabla}a)}{\deduce{\pi_2 } { [ b e:{\mathsf{f}}({\mathsf{x}}b \wedge { \nabla}a)]^3 } } } } } } \ ] ] where @xmath418 is the following derivation : @xmath423 { b:{\mathsf{f}}({\mathsf{x}}b \wedge { \nabla}a)}{\infer[{{\mathsf{f}}i } ] { b:{\mathsf{f}}({\mathsf{x}}b \wedge { \nabla}a)}{\infer[{\wedge i } ] { b b:{\mathsf{x}}b \wedge { \nabla}a}{\infer[{{\mathsf{x}}i}^5 ] { b b:{\mathsf{x}}b}{\infer[{\mathit{last } } ] { b b f : b}{\infer[{\mathit{lin}_{\lhd}}^6 ] { b f : b}{[b{\lhd}e]^2 & [ b{\lhd}f]^5 & [ b e : b]^3 & [ b f : b]^6 } } } & \infer[{{\nabla}i}^7 ] { b b:{\nabla}a}{\infer[{\mathit{last } } ] { b b':a}{\infer[{\mathit{eq}_{\leqslant } } ] { b':a}{[b{\leqslant}b']^7 & [ b'{\leqslant}b]^7 & \infer[{\wedge e } ] { b : a}{[\varphi]^1 } } } } } & [ b{\leqslant}b]^4 } } } \ ] ] @xmath420 is the following derivation : @xmath424 { b:{\mathsf{f}}({\mathsf{x}}b \wedge { \nabla}a ) } { [ b e:{\mathsf{f}}({\mathsf{x}}b \wedge { \nabla}a)]^3 & \infer[{\mathit{base}_{\leqslant}}^9 ] { b:{\mathsf{f}}({\mathsf{x}}b \wedge { \nabla}a ) } { [ b { \lhd}e]^2 & \infer[{\mathit{trans}_{\leqslant}}^{10 } ] { b:{\mathsf{f}}({\mathsf{x}}b \wedge { \nabla}a ) } { [ b{\leqslant}e]^9 & [ e{\leqslant}c]^8 & \infer[{{\mathsf{f}}i } ] { b:{\mathsf{f}}({\mathsf{x}}b \wedge { \nabla}a ) } { \infer[{\wedge i } ] { b c:{\mathsf{x}}b \wedge { \nabla}a } { \infer[{{\mathsf{x}}i}^{11 } ] { b c:{\mathsf{x}}b}{\infer[{\mathit{last } } ] { b c f : b}{\infer[{{\mathsf{x}}e } ] { b e c f : b}{\infer[{\wedge e } ] { b e c : b } { [ b e c:{\mathsf{x}}b
\wedge { \nabla}a]^8 } & [ c{\lhd}f]^{11 } } } } & \deduce { b c:{\nabla}a}{\pi_3 } } & [ b{\leqslant}c]^{10 } } } } } } \ ] ] and @xmath425 is the following derivation : @xmath426 { b c:{\nabla}a}{\infer[{\mathit{split}_{\leqslant}}^{13 } ] { b d : a}{[b{\leqslant}d]^{12 } & \infer[{\wedge e } ] { b : a}{[\varphi]^1 } & \infer[{\mathit{last } } ] { b d : a } { [ d : a]^{13 } } & \infer[{\mathit{last } } ] { b d : a}{\infer[{\mathit{lin}_{\lhd}}^{14 } ] { b e d : a}{[b{\lhd}f]^{13 } & [ b{\lhd}e]^2 & [ f{\leqslant}d]^{13 } & \infer[{{\nabla}e } ] { b e d : a}{\infer[{\wedge e } ] { bec:{\nabla}a } { [ bec:{\mathsf{x}}b \wedge { \nabla}a]^8 } & [ e { \leqslant}d]^{14 } & [ d { \leqslant}c]^{12 } } } } } } } \ ] ] proof of the axiom @xmath427 @xmath255 { b : b \vee ( { \mathsf{f}}({\mathsf{x}}b \wedge { \nabla}a ) ) { \supset}{\mathsf{f}}b } { \infer[{\vee e}^2 ] { b:{\mathsf{f}}b } { [ b : b \vee ( { \mathsf{f}}({\mathsf{x}}b \wedge { \nabla}a))]^1 & \infer[{\mathit{refl}_{\leqslant}}^3 ] { b:{\mathsf{f}}b}{\infer[{{\mathsf{f}}i } ] { b:{\mathsf{f}}b}{\infer[{\mathit{last } } ] { bb : b } { [ b : b]^2 } & [ b { \leqslant}b]^3 } } & \infer[{{\mathsf{f}}e}^4 ] { b:{\mathsf{f}}b } { [ b:{\mathsf{f}}({\mathsf{x}}b \wedge { \nabla}a)]^2 & \infer[{\mathit{ser}_{\lhd}}^5 ] { b:{\mathsf{f}}b}{\infer[{\mathit{base}_{\leqslant}}^6 ] { b:{\mathsf{f}}b}{[c { \lhd}d]^5 & \infer[{\mathit{trans}_{\leqslant}}^7 ] { b:{\mathsf{f}}b}{[b { \leqslant}c]^4 & [ c { \leqslant}d]^6 & \infer[{{\mathsf{f}}i } ] { b:{\mathsf{f}}b}{\infer[{\mathit{last } } ] { bd : b}{\infer[{{\mathsf{x}}e } ] { bcd : b}{\infer[{\wedge e } ] { bc:{\mathsf{x}}b } { [ bc:{\mathsf{x}}b \wedge { \nabla}a]^4 } & [ c { \lhd}d]^5 } } & [ b{\leqslant}d]^7 } } } } } } } } \ ] ] [ fig : axiom - a8 ] | until is a notoriously difficult temporal operator as it is both existential and universal at the same time : @xmath0 holds at the current time instant @xmath1 iff either @xmath2 holds at @xmath1 or there _ exists _ a time instant @xmath3 in the future at which @xmath2 holds and such that @xmath4 holds in _ all _ the time instants between the current one and @xmath3 .
this `` ambivalent '' nature poses a significant challenge when attempting to give deduction rules for until . in this paper , in contrast , we make explicit this duality of until by introducing a new temporal operator @xmath5 that allows us to formalize the `` history '' of until , i.e. , the `` internal '' universal quantification over the time instants between the current one and @xmath3 .
this approach provides the basis for formalizing deduction systems for temporal logics endowed with the until operator . for concreteness , we give here a labeled natural deduction system for a linear - time logic endowed with the new history operator and show that , via a proper translation , such a system is also sound and complete with respect to the linear temporal logic @xmath6 with until . |
coulombic effects are often paramount in soft matter systems , where the large dielectric constant of the solvent ( say water ) invites ionizable groups at the surface of macromolecules to dissociate @xcite .
while a realistic treatment requires considering three dimensional systems , interesting progress has been achieved for lower dimensional problems where the key mechanisms can be studied in greater analytical detail @xcite . in particular , a one dimensional model was introduced in the 1960s by lenard and prager independently , for which a complete thermodynamic solution was provided @xcite .
this model has been further studied in ref .
@xcite , but it turns out that some interesting features have been overlooked in relation with the like - charge attraction phenomenon @xcite . this striking non mean - field effect , relevant for strongly coupled charged matter @xcite is the thread in our study .
the paper is organized as follows .
the model is first defined in section [ sec : equal - charges ] .
it mimics the screening of charged colloids .
the coulomb potential in one dimension between two charges @xmath1 and @xmath2 located along a line with coordinates @xmath3 and @xmath4 is @xmath5 therefore , the electric field created by one particle is of constant magnitude .
this fact simplifies the study of the equilibrium statistical mechanics of such systems , and allows to obtain some of its properties by simple arguments .
furthermore , it also allows for an explicit computation of the partition function @xcite .
the system under scrutiny can be envisioned as a collection of parallel charged plates , able to move along a perpendicular axis .
the salient properties of this system can be obtained by simple arguments which we present in section [ sec : equal - charges ] , followed afterwards by a more technical analysis where the explicit calculation of the partition function is performed , first in the isobaric and then in the canonical ensemble .
after having presented the symmetric case , section [ sec : different - charges ] will generalize the investigation to the situations where the two screened charges are different .
noteworthy is that parity of the particle number considerations will play an important role in the remainder .
consider two charges @xmath1 along a line located at @xmath6 and @xmath7 . between the charges there are @xmath0 counterions of charge @xmath8 between them .
consider the equilibrium thermal properties of this system at a temperature @xmath9 , and as usual define @xmath10 with @xmath11 the boltzmann constant .
this simple model mimics the screening and effective interaction between two charged colloids in a counterion solution , without added salt . in one dimension , @xmath12 has dimensions of inverse length , therefore it is convenient to use rescaled units in which all distances are measured in units of @xmath13 : @xmath14 .
it is also convenient to work with a dimensionless pressure @xmath15 where @xmath16 is the pressure ( equal to the force , in one dimensional systems ) .
the potential energy ( dimensionless , measured in units of @xmath17 ) of the system is @xmath18 before presenting the technical analysis , we start by simple and more quantitative considerations . the possibility of attraction between the two @xmath19 charges at @xmath20 and @xmath21 is related to the parity of @xmath0 .
if @xmath0 is odd , @xmath22 , then @xmath23 counterions will form a double layer around each charge @xmath1 .
this will form two compound objects with charge @xmath24 each one , located around @xmath20 and @xmath21 .
there will be in addition one counterion between these two object , which is essentially free , as the electric field created by the charges located on each side around @xmath20 and @xmath21 cancel each other . when @xmath21 is large enough , consider figure [ fig:1dmodel - odd - attract ] .
the right side of the system composed of one charge @xmath1 and @xmath23 counterions has charge @xmath25 .
the left side which , for the sake of the argument , has the free counterion plus the compound charge , exhibits a total charge @xmath26 .
thus the force exerted by the left side on the right side is @xmath27 , an attractive force .
thus one expects that @xmath28 , for @xmath29 .
mobile ions ( counter - ions ) have charge @xmath30 and the confining objects have charge @xmath1 , so that the whole system is electro - neutral . here , @xmath22 is odd , so that a single ion ( referred to as the misfit since the net electric force acting on it vanishes ) `` floats '' in between the two screened boundaries which attract , each , @xmath23 ions in their vicinity ( see also fig .
[ fig:1dmodel - odd ] ) .
this single free counterion provides the binding mechanism responsible for long range attraction . in the canonical treatment ,
@xmath21 is held fixed , while in the isobaric situation , it is a fluctuating quantity .
[ fig:1dmodel - odd - attract],scaledwidth=70.0% ] on the other hand , if @xmath0 is even , there will not be a free counterion between the layers , which will be completely neutral , thus one expects that @xmath31 when @xmath29 , as shown in figure [ fig:1dmodel - even ] . ) . at large distance , the two double - layers ( made up of an ion @xmath1 and @xmath23 counter - ions ) decouple since they are neutral . no misfit ion is present to mediate attraction , and the pressure is repulsive at all distances .
[ fig:1dmodel - even],scaledwidth=70.0% ] the previous intuition , providing a large distance attraction for odd @xmath0 , can be substantiated by a simple calculation .
use will be made here of the contact theorem @xcite , an exact relation between the force exerted on the charge @xmath1 , and the ionic density at contact ( stemming from the mobile charges @xmath30 ) . such a relation is particularly useful for discussing the like - charge attraction phenomenon @xcite .
the argument allowing to get the contact density is two - fold , and goes as follows .
leftmost counterions in fig .
[ fig:1dmodel - odd - attract ] , one obtains an ion with charge @xmath32 .
this newly defined system has the same large distance pressure as that of fig .
[ fig:1dmodel - odd - attract ] .
[ fig : argument_trick],scaledwidth=70.0% ] first , we argue that at large @xmath21 , the @xmath23 counterions that are closest to each boundary remain in their vicinity , while the middle free counterion ( the misfit in figs .
[ fig:1dmodel - odd - attract ] and [ fig:1dmodel - odd ] ) , which does not feel any electric field by symmetry , tends to be unbounded and no longer contributes to the pressure ( discarding @xmath33 terms ) . in a second step ,
we thus compute the contact density in a system of an isolated charge @xmath19 , with a double - layer of @xmath23 ions in the vicinity ( the total charge of this composite object , shown on the right hand - side of fig .
[ fig:1dmodel - odd - attract ] ) is @xmath25 .
the solution to this problem is not immediate , but can be found by a convenient mapping onto a more convenient problem , shown in fig .
[ fig : argument_trick ] .
as illustrated in the figure , we regroup the @xmath34 leftmost counterions in a single ion , having charge @xmath35 . at large distances ,
this regroupment does not influence the distribution of counterions around the rightmost ion @xmath19 , and thus leaves the large @xmath21 pressure unaffected .
the next important argument is that the pressure can be equivalently computed from the contact density at the rightmost , or leftmost charge @xmath19 .
it is thus simpler to perform the calculation in the newly defined regrouped system ( left hand side of fig .
[ fig : argument_trick ] ) . the regrouped ion with charge
@xmath35 is in the electric field of the charge @xmath1 on its left , and of the composite system on its right having charge @xmath25 .
this amounts to a field @xmath36 .
hence , the electric potential energy reads @xmath37 .
the corresponding boltzmann weight gives the density of the regrouped ion @xmath38\ ] ] where due account was taken of normalization ( @xmath39 ) .
the contact density @xmath40 finally yields the pressure through the contact theorem @xmath41 .
we get here @xmath42 ( or equivalently @xmath43 ) , a result which by construction holds in the large @xmath21 limit .
the reason for a non vanishing pressure at large distance is that the @xmath23 counter - ions can not exactly screen the charge of an ion @xmath1 .
it is no longer the case when @xmath0 is even , in which case @xmath44 for @xmath29 .
the present results will be fully corroborated by direct partition function calculations .
returning to the case when @xmath22 is odd , we can also estimate the first correction to the pressure for large @xmath21 . consider that @xmath21 is fixed ( canonical ensemble ) and large .
since the system is somehow equivalent to two double layers with a free counterion in between , this counterion will contribute to the pressure ( denoted as @xmath45 in the canonical , fixed-@xmath21 ensemble ) with a correction @xmath33 .
this estimate can be made more quantitative .
the available space for the free counterion is not @xmath21 , but it is rather @xmath21 minus the space occupied by the diffuse counterion layers , given by @xmath46 the thermal average position of the @xmath23-th counterion if they have been ordered @xmath47 , in the limit @xmath29 .
thus @xmath48 this is illustrated in figure [ fig:1dmodel - odd ] . in the following section ,
we evaluate explicitly @xmath49 and find @xmath50 then , for large @xmath21 , we expect @xmath51 , scaledwidth=70.0% ] in the other limiting case @xmath52 , the result is @xcite @xmath53 , that can be understood as all the @xmath0 counterions are squeezed in a small distance @xmath21 .
thus we see that the pressure is positive ( repulsive force ) for small separations @xmath52 then changes to negative pressure ( attractive force ) for large @xmath21 .
we will show in the following section that the @xmath54 corrections in ( [ eq : plinfty ] ) are actually exponentially small , in the canonical ensemble , therefore equation ( [ eq : plinfty ] ) gives a fairly good approximation for the pressure for a large set of values of the separation @xmath21 . from this , one can estimate the distance @xmath55 , at which the effective force between the two charges becomes attractive @xmath56 figure [ fig : pvsl - canon ] shows the pressure @xmath45 as a function of @xmath21 , for @xmath57 and for @xmath58 particles . for @xmath57 ( odd ) the pressure changes its sign at @xmath59 , while for @xmath58 the pressure is always positive .
summarizing , in the case of odd @xmath0 , the possibility of having an effective attraction for large separations @xmath21 is due to the sharing of the `` free '' ion which leads to the creation of opposite charges objects ( ions @xmath1 plus their counterion clouds ) .
although the analytical results presented here are valid only for this one - dimensional model , the same physical mechanism has also been observed in three dimensional systems @xcite .
it can also be surmised that in situation of odd @xmath0 where the free counterion has a varying charge , attraction will be all the stronger as the charge will increase in absolute value . in addition , the very mechanism brought to the fore here indicates that at mean - field level , where the discrete nature of ions is discarded , attraction should be suppressed , which indeed is the case @xcite .
the ( canonical ) pressure @xmath45 as a function of the separation @xmath21 , for @xmath57 ( continuous bottom line ) and @xmath58 ( dashed upper line ) . for @xmath0 odd
the pressure becomes negative at large distances .
, scaledwidth=50.0% ] the equilibrium thermodynamics of the one - dimensional two - component coulomb gas was solved simultaneously but independently by lenard @xcite and prager @xcite . in the present model , only one type of identical particles ( the counterions )
are present .
it is convenient to order the particles as @xmath60 .
then , rearranging the terms in ( [ eq : pot ] ) , the potential energy of the system can be written as @xmath61 and @xmath62 notice that in the case @xmath22 , the particle with position @xmath63 does not appear in the potential energy .
it is the free counterion ( misfit ) discussed in the previous section , whose role is crucial for the possibility of like - charge attraction .
the canonical configuration integral is @xmath64 as mentioned by lenard in his seminal paper @xcite `` the ( configuration ) integral is elementary ( because ) the class of functions consisting of exponential of linear functions is closed under the operation of indefinite integral ( ... ) however the task of evaluating ( it ) is not trivial '' . for small
@xmath0 one can compute by hand @xmath65 , and for larger given values of @xmath0 it can be obtained numerically with the aid of a computer algebra system software program . by inspection of the integral ( [ eq : z ] ) , one can deduce that @xmath65 is a linear combination of products of exponentials of @xmath21 and linear functions of @xmath21 .
one can also deduce the argument of each exponential function of @xmath21 by keeping track of the factor that multiplies each @xmath66 in the integral ( [ eq : z ] ) .
these come from the explicit term in @xmath67 ( for instance , for @xmath68 it is @xmath69 in the case @xmath0 odd ) , but after each successive integration , the factor of @xmath70 will by added to the one of @xmath71 due to the upper limit of integration . taking that into account , one realizes that the exponentials of @xmath21 in @xmath65 are of the form @xmath72 in the case @xmath0 odd , and @xmath73 in the case @xmath0 even .
thus , the configuration canonical integral is expected to be of the form @xmath74 and @xmath75 the non trivial task is to evaluate explicitly the coefficients @xmath76 , @xmath77 , @xmath78 and @xmath79 .
this is done in section [ sec : canonical - exact ] . in @xcite ,
the present system was studied , but an exact analytical explicit evaluation of the partition function for an arbitrary number of particles was not achieved .
rather , an interesting reformulation of this model was proposed , by mapping it into a quantum mechanical problem , following a technique put forward by edwards and lenard @xcite .
it was shown in @xcite that the configuration integral is given by @xmath80 where @xmath81 is the solution of a set of @xmath0 coupled elementary linear differential equations @xmath82 with the initial condition @xmath83 . integrating this equation one has @xmath84 then , starting from the known @xmath85 one has to perform successively @xmath0 integrals ( [ eq : dean - system ] ) to obtain @xmath86 and the configuration integral .
this task is equivalent to performing directly the @xmath0 integrals of the configuration integral ( [ eq : z ] ) .
thus , unfortunately , the method proposed in @xcite does not provide any computational advantage over a direct numerical evaluation of the partition function . here
, our goal is to obtain an explicit analytical expression for the configuration integral for an arbitrary number of particles @xmath0 .
using lenard @xcite and prager @xcite method , we will first compute the partition function of the constant pressure ensemble @xmath87 which is the laplace transform of the canonical configuration integral @xmath65 .
this is a straightforward application of the technique of lenard and prager , and it is actually much simpler than the complete work presented in @xcite , since all particles are identical and we will not have to deal with the combinatorial problem of studying the different configurations of charges .
then , we shall invert the laplace transform to obtain the canonical , constant `` volume '' @xmath21 , configuration integral @xmath88 . since we are interested in finite systems , the results from the canonical ensemble and the constant pressure ensemble will differ , and it is of interest to compare them .
to introduce the technique used to compute the partition function , we undertake in this section a preliminary , simpler task , based on the same technique : the exact evaluation of the diffuse layer size @xmath49 .
this quantity appeared in the discussion of section [ sec : like - charge - attraction ] .
consider here that @xmath29 and @xmath22 . the double layer composed by the charge @xmath1 at @xmath21 and its corresponding @xmath23 counterions are thereby ` sent to infinity ' .
the remaining @xmath34 counterions , however , still feel the electric field created by this far charged double layer .
the potential energy part which depends on the position of the remaining counterions is @xmath90 we wish to evaluate @xmath91 let @xmath92 then @xmath93 .
following lenard @xcite and prager @xcite it is convenient to re - write the potential energy as @xmath94\ ] ] with the convention that @xmath95 .
let us define @xmath96 { x}/2 } \
, h({x})\ ] ] where @xmath97 is the heaviside step function .
then @xmath98 we notice that @xmath99 is the laplace transform ( evaluated at @xmath100 ) of the @xmath23-fold convolution product @xmath101 .
the laplace transform @xmath102 of @xmath103 is elementary @xmath104 then @xmath105 computing the derivative of @xmath106 we obtain @xmath107 thus proving ( [ eq : xp ] ) .
consider now the finite system with @xmath108 .
we will detail the calculations in the case @xmath22 odd , the case @xmath0 even can be obtained by a simple adaptation of the same technique . as it was done in the previous section ,
it is convenient to re - write the potential energy ( [ eq : uodd ] ) as @xmath109 where , by convention , we defined @xmath95 and @xmath110 . with @xmath103 defined in ( [ eq : fj ] ) , we notice again that the canonical partition function is a convolution product of @xmath111 functions @xmath103 @xmath112 the isobaric partition function @xmath113 is the laplace transform of @xmath65 , and we have @xmath114 ^ 2 } = \prod_{k=0}^{p } \frac{1}{\left[\left(k+\frac{1}{2}\right)^2+p\right]^2 } \label{eq : zp_odd } \end{aligned}\ ] ] where @xmath115 .
factoring @xmath116 , the above product can be expressed in terms of gamma functions @xmath117 the average length of the system is given by the usual thermodynamic relation @xmath118\end{aligned}\ ] ] where @xmath119 .
we can notice that this expression has a pole for @xmath43 , from which we obtain the behavior when @xmath120 , @xmath121 , in agreement with the general discussion of section [ sec : like - charge - attraction ] .
when @xmath0 is even this pole is absent ( see below ) . if @xmath122 is even , similar calculations lead to @xmath123 and @xmath124 notice an important difference in the analytic structure of the partition function in the case @xmath0 odd ( [ eq : zc_convol])([eq : zp_odd ] ) and @xmath0 even ( [ eq : zc_even])([eq : zp_even ] ) : for @xmath0 even , there is a single function @xmath125 in the convolution product , leading to a pole of order one for @xmath126 , in contrast to the case @xmath0 odd , where the functions @xmath127 appear twice in the convolution product and the pole for the smallest value of @xmath128 is of order two and it is for @xmath43 , rather than @xmath126 .
in the case @xmath0 even , the term @xmath129 corresponds to the coupling of the left diffuse layer with the free counterion and the coupling of this same free counterion with the right diffuse layer . on the other hand in the case @xmath0
odd , the term @xmath125 corresponds to the direct coupling of the left and right diffuse layers . the average length , for @xmath122 even , is @xmath130 we note that @xmath131 when @xmath31 , in contrast to what happens when @xmath0 is odd , where @xmath131 when @xmath132 .
we return to the case @xmath22 odd .
to compute the canonical partition function , we need to invert the laplace transform computed in the previous section @xmath133 ^ 2 } \right ) ( l).\ ] ] this rather technical part of the analysis is presented in appendix [ app : a ] , where it is shown that @xmath134 ^ 2 e^{-(j+\frac{1}{2})^2 l } \left [ l+\frac{2}{2j+1}\left ( \sum_{k = p - j+1}^{p+j+1 } \frac{1}{k } -\frac{1}{2j+1 } \right ) \right ] \,.\ ] ] from this expression , we obtain the canonical pressure @xmath135 , @xmath136 ^ 2 } \left [ l+\frac{2}{2j+1}\left ( { \displaystyle \sum_{k = p - j+1}^{p+j+1 } } \frac{1}{k } -\frac{3}{2j+1 } \right ) \right ] e^{-(j+\frac{1}{2})^2 l } } { { \displaystyle \sum_{j=0}^{p } } \left[\frac{2j+1}{(p - j)!(p+j+1)!}\right]^2 \left [ l+\frac{2}{2j+1}\left ( { \displaystyle \sum_{k = p - j+1}^{p+j+1 } } \frac{1}{k } -\frac{1}{2j+1 } \right ) \right ] e^{-(j+\frac{1}{2})^2 l } } \,.\ ] ] for @xmath122 even , the results are @xmath137 ^ 2 } \left [ l + \frac{1}{j}\left ( \sum_{k = p - j+1}^{p+j } \frac{1}{k } - \frac{1}{2j } \right ) \right ] \,,\ ] ] and @xmath138 ^ 2 } \left [ l + \frac{1}{j}\left ( { \displaystyle\sum_{k = p - j+1}^{p+j } } \frac{1}{k } -\frac{3}{2j } \right ) \right ] } { \frac{1}{(p!)^4}- { \displaystyle \sum_{j=1}^p } \frac{(2 j)^2 e^{-j^2 l}}{[(p+j)!(p - j)!]^2 } \left [ l + \frac{1}{j}\left ( { \displaystyle\sum_{k = p - j+1}^{p+j } } \frac{1}{k } -\frac{1}{2j }
\right ) \right ] } \,.\ ] ] with the exact expressions obtained above , we can prove rigorously the limiting behavior of the pressure when @xmath29 and @xmath52 discussed in section [ sec : like - charge - attraction ] .
let us consider first the case @xmath22 odd . in the canonical ensemble , the behavior of the pressure @xmath45 when @xmath139 , is obtained from the term @xmath140 of ( [ eq : zexactnimpar ] ) , confirming the prediction ( [ eq : plinfty ] ) of section [ sec : like - charge - attraction ] .
furthermore , we realize that the next to next to leading order correction is exponentially small @xmath141 in contrast , when @xmath122 , the pressure tends to 0 exponentially fast when @xmath29 @xmath142 the behavior of the pressure is different in the isobaric ensemble .
consider again first the case @xmath22 . from ( [ eq : lave ] ) , we already know that when @xmath43 , @xmath143 . denoting @xmath115 , one can expand ( [ eq : lave ] ) for small @xmath144 and invert the relation to obtain @xmath145 as a function of @xmath146 when @xmath120 . for instance , to order @xmath147 , eq .
( [ eq : lave ] ) is @xmath148 where @xmath149 ^ 2 } = 2{\cal h}^{(2)}_p - \frac{p(3p+4)}{(p+1)^2 } \,,\ ] ] with @xmath150 the harmonic numbers .
inverting that relation , up to order @xmath151 , gives @xmath152 notice a factor 2 of difference in the next to leading order correction ( the @xmath153 term ) in the pressure in the isobaric ensemble and the canonical ensemble .
furthermore , in the isobaric ensemble the next to next to leading order corrections are algebraic and not exponential as in the canonical ensemble . for @xmath122 , the behavior of the pressure , in the isobaric ensemble , when @xmath154 , is @xmath155 ^ 3 } + o\left({\langle l \rangle}^{-4}\right ) \,.\ ] ] notice again the different behavior with respect to the canonical ensemble .
here in the isobaric ensemble , the pressure vanishes as @xmath156 , whereas in the canonical ensemble it vanishes exponentially fast , as @xmath157 .
let us study the other limiting behavior of the pressure , for small separations @xmath21 .
let us focus on the case @xmath22 first .
it is not completely straightforward to obtain the behavior of the pressure in the canonical ensemble when @xmath52 directly from expression ( [ eq : pexactnimpar ] ) .
rather , it is better to return to ( [ eq : zc_convol ] ) , and notice that if @xmath52 , then the convolution product @xmath158 behaves as @xmath159 which is independent of @xmath160
. then , @xmath161 and @xmath162 we deduce that the pressure behaves as @xmath163 a result already noticed in @xcite .
( [ eq : pcanol0 ] ) also holds when @xmath122 . in the isobaric ensemble ,
when @xmath22 , if @xmath164 , then , necessarily , @xmath165 in ( [ eq : lave ] ) .
expanding that equation to order @xmath166 , one obtains @xmath167 this result also holds true for @xmath122 .
notice again the difference between the canonical ( [ eq : pcanol0 ] ) and isobaric ensemble ( [ eq : pisol0 ] ) , where the leading term changes from @xmath168 to @xmath169 . when @xmath22 is odd , the pressure changes of sign when @xmath21 varies .
it is positive for @xmath52 and negative for @xmath139 .
we already obtained an approximation of the value @xmath55 of @xmath21 when this occurs in the canonical ensemble , see ( [ eq : lstar_cano ] ) , up to exponentially small corrections . in the isobaric ensemble ,
one just has to put @xmath126 in ( [ eq : lave ] ) to obtain the exact value @xmath170 for this quantity , the predictions from the canonical ensemble ( [ eq : lstar_cano ] ) and the isobaric ensemble ( [ eq : lstar_isop ] ) are again different .
figure [ fig : pvsl - iso - cano ] shows the pressure as a function of the separation , for @xmath171 , in the isobaric ensemble and the canonical ensemble .
notice that the pressure from the canonical ensemble is smaller that the one in the isobaric ensemble for the same separation .
figure [ fig : lstarvsn ] shows the value of @xmath55 for which the pressure changes of sign as a function of @xmath0 , when @xmath0 is odd , in both ensembles .
notice again that in the canonical ensemble , the change of sign of the pressure occurs for smaller values @xmath55 of the separation than in the isobaric ensemble .
the pressure @xmath145 as a function of the separation @xmath21 , for @xmath171 .
the top continuous line represents the result from the isobaric ensemble , and the dotted bottom line those from the canonical ensemble .
, width=264 ] the value of the separation @xmath55 for which the pressure vanishes and changes sign as a function of @xmath0 for @xmath0 odd .
the filled squares represent the results from the isobaric ensemble , and the filled disks , their canonical counterpart.,width=264 ]
in this section we consider a generalization of the previous model , where the two charges located at @xmath172 and at @xmath173 are @xmath174 and @xmath175 , respectively , which can be eventually different .
the overall system should be neutral , therefore @xmath176 , @xmath177 being charge of one counterion .
it is convenient to introduce the notation @xmath178 and @xmath179 such that @xmath180 and @xmath181 .
the electroneutrality relation is @xmath182 .
the charge asymmetry can be characterized by the quantity @xmath183 , which allows to write @xmath184 and @xmath185 .
the potential energy of the system is now @xmath186 the overall effect of the charge asymmetry is to introduce a global electric field proportional to @xmath187 ( the term in @xmath188 ) .
adapting the ideas of section [ sec : isobaric - exact ] to the present case , we can obtain the isobaric partition function .
once again , the results differ depending on the parity of the number of counterions @xmath0 . for @xmath22 odd , @xmath189 \left[(k+\frac{1+a}{2})^2 + p\right]}\ ] ] while for @xmath122 even , @xmath190 \left[(k+\frac{a}{2})^2+p\right ] } \,.\ ] ] the above formulas highlight the difference between the two cases , depending on the parity of @xmath0 . however both formulas can be summarized in a single one as @xmath191 where we defined @xmath192 taking the derivative of ( [ eq : zp - a ] ) with respect to @xmath145 , we obtain the relation between the average length @xmath193 of the system and the pressure @xmath145 in the isobaric ensemble @xmath194 if @xmath195 is not an integer ( @xmath174 and @xmath175 are not integer multiples of @xmath196 ) , or @xmath197 ( @xmath174 and @xmath175 have opposite signs ) , then @xmath113 has simple poles .
but when @xmath198 is an integer and @xmath199 , the partition function @xmath113 turns out to have some double poles .
this corresponds to the case when @xmath200 and @xmath201 are both positive integers . in that case
it is best to reorder the products in ( [ eq : zp - a ] ) to make those double poles more apparent .
the result depends on the parity of @xmath200 and @xmath201 ( both have the same parity ) .
if @xmath200 and @xmath201 are odd , then @xmath178 and @xmath179 are half integers : @xmath202 and @xmath203 .
the notation @xmath204 denotes the floor function of @xmath205 ( largest integer less or equal than @xmath205 ) .
the isobaric partition function ( [ eq : zp - a ] ) becomes @xmath206 ^ 2 } \prod_{\ell={\left\lfloor q _ { < } \right\rfloor}+1}^{{\left\lfloor q _ { > } \right\rfloor}}\frac{1}{(\ell+\frac{1}{2})^2+p } \,,\ ] ] and the corresponding equation of state is @xmath207 when @xmath178 and @xmath179 are positive integers , these expressions become @xmath208 and @xmath209 to compute the canonical partition function , one has to perform the inverse laplace transform of the expressions obtained in the last section . from the above discussion , it is clear that the results will have a different analytical structure depending on whether the isobaric partition function has simple or double poles , that is , depending on whether @xmath187 is an integer or not . if @xmath187 is not an integer , or @xmath210 , all poles of @xmath113 are simple poles , and we obtain from ( [ eq : zp - a ] ) : @xmath211 this formula is valid whenever @xmath200 and @xmath201 are not integers , or if @xmath178 and @xmath179 have opposite signs ( @xmath212 and @xmath213 ) .
if @xmath187 is an integer , with @xmath199 , then using ( [ eq : zp - na - odd ] ) , we obtain , when @xmath178 and @xmath179 are half integers , @xmath214 \nonumber\\ & + \sum_{j={\left\lfloor q _ { < } \right\rfloor}+1}^{{\left\lfloor q _ { > } \right\rfloor } } \frac{e^{-(j+\frac{1}{2})^2{l}/4 } ( j-{\left\lfloor q _ { < } \right\rfloor}-1)!\,(2j+1)(-1)^{j-{\left\lfloor
\right\rfloor}-1 } } { ( { \left\lfloor q _ { < }
\right\rfloor}+1+j)!({\left\lfloor q _ { > } \right\rfloor}+1+j)!({\left\lfloor q _ { > } \right\rfloor}-j ) ! } \ , , \nonumber\\\end{aligned}\ ] ] and when @xmath178 and @xmath179 are integers , @xmath215 \nonumber\\ & + \sum_{j = q_{<}+1}^{q _ { > } } \frac{e^{-j^2{l } } ( j - q_{<}-1)!\,(2j)(-1)^{j - q _ { < } } } { ( q_{<}+j)!(q_{>}+j)!(q_{>}-j ) ! } + \frac{1}{\left(q_1!q_2!\right)^2 } \ , . \end{aligned}\ ] ] the two previous results ( [ eq : zc - na - odd ] ) and ( [ eq : zc - na - even ] ) show the different analytical structure of the two cases which depend on the parity of @xmath200 and @xmath201 , in particular the existence of a term independent of @xmath21 in the case @xmath200 and @xmath201 even , and the form of the argument of the exponentials @xmath216 ( for @xmath200 even ) , as opposed to @xmath217 ( for @xmath200 odd ) .
however , both results ( [ eq : zc - na - odd ] ) and ( [ eq : zc - na - even ] ) can be subsumed in a single formula as follows .
let us define @xmath218 \ , , \nonumber\end{aligned}\ ] ] for @xmath219 , and , when @xmath220 is an integer , define @xmath221 also , let @xmath222 then , both results ( [ eq : zc - na - odd ] ) and ( [ eq : zc - na - even ] ) are equivalent to @xmath223 for small separations @xmath21 , the results ( [ eq : pcanol0 ] ) , @xmath224 ( canonical ) and ( [ eq : pisol0 ] ) , @xmath225 ( isobaric ) , still hold independently of the charge asymmetry @xmath187 .
thus , the effective interaction is always repulsive at short distance , irrespective of the charges @xmath174 and @xmath175 , even in the case where these charges are opposite .
indeed , the pressure is dominated here by the entropy cost for confining the ions in a narrow domain . the behavior for large separations @xmath21 will depend on whether the charges @xmath174 and @xmath175 are multiples of @xmath177 or not , and their relative signs .
there are four cases to consider
. * opposite charges . *
first , suppose that @xmath226 , the charges at the edges have opposite signs .
this corresponds to the case @xmath197 , and the canonical partition function is obtained with eq .
( [ eq : zc - a ] ) . from that expression
, we deduce that for @xmath21 large , the leading order is given by the term @xmath140 of that sum .
therefore , the effective force is attractive and given by @xmath227 where here @xmath228 corresponds to the charge of the edge particle which has the same sign as the small ions .
this result can actually be obtained by simple arguments .
the small ions will be repelled by the particle with charge corresponding to @xmath220 and attracted to the other edge where there is a particle with charge @xmath229 with @xmath230 . by electroneutrality , the charge of the compound object formed by the small ions and @xmath229 will be @xmath231 . the effective force between this object and the other opposite charge @xmath232 is repulsive , equal to @xmath233 , thus recovering ( [ eq : pcq1q2neg ] ) .
application of the contact theorem of course yields the same result , since the density of counterions vanishes at contact with @xmath220 ( a similar effect was reported in @xcite ) . * like - charges that are not integer multiples of @xmath234 . * to discuss this situation
, we keep in mind that @xmath235 and @xmath236 are not integers .
the small ions of charge @xmath177 will be divided into two parts that will try to screen the charges @xmath174 and @xmath175 .
a number @xmath237 of counterions will partially screen @xmath174 and @xmath238 ions will partially screen the other charge @xmath175 .
each edge , with its screening cloud of counterions , will have a charge @xmath239 and @xmath240 respectively , where @xmath241 denotes the fractional part of @xmath205 .
however , since @xmath178 and @xmath179 are not integers , we have @xmath242 : there is still one counterion to take into consideration .
this counterion feels the electric field created by the charge difference @xmath243 , therefore it will be attracted to the edge which has the largest remaining charge ( in the sense of the largest between @xmath244 and @xmath245 ) . to fix
the ideas suppose @xmath246 .
the remaining ion will become part of the screening cloud of @xmath174 , and the charge of that compound object will be @xmath247 .
then the effective force between the two edges will be @xmath248 , the last equality coming from the fact that @xmath249 .
summarizing , in general we expect an attractive force at large separations given by @xmath250 this can be verified by identifying the largest argument of the exponentials in the canonical partition function ( [ eq : zc - a ] ) or , equivalently , the largest pole of the isobaric partition function ( [ eq : zp - a ] ) .
the poles of the isobaric partition function are @xmath251 , with @xmath252 going from 0 to @xmath0 .
then , one can notice that @xmath253 varies from @xmath254 up to @xmath213 by integer steps of 1 .
from this one - dimensional array of points , we are interested in the one that is the closest to 0 .
that is precisely @xmath255 , in agreement with ( [ eq : pcqnotint ] ) .
one can also notice from ( [ eq :
zc - a ] ) that in the canonical ensemble , the next to leading order correction to ( [ eq : pcqnotint ] ) is exponentially small of order @xmath256 . * like - charges that are half - integer multiples of @xmath234 . *
a degenerate case of the previous situation is when @xmath178 and @xmath179 are half - integers , that is @xmath257 . in this case
the canonical partition function is given by ( [ eq : zc - na - odd ] ) instead of ( [ eq : zc - a ] ) .
the leading order is still given by ( [ eq : pcqnotint ] ) , specifically @xmath258 . but the correction to leading order is not exponentially small , it can be read from the term @xmath140 of ( [ eq : zc - na - odd ] ) @xmath259 with @xmath260 we find here the generalization of the charge - symmetric case ( @xmath261 ) discussed in section [ sec : equal - charges ] . each charge @xmath174 and @xmath175
is screened by @xmath237 and @xmath238 ions .
the remaining counterion is free to roam in a region of size @xmath262 , and with zero electric field .
this ion contributes to the pressure with a term @xmath263 . here
@xmath264 is the size of the screening layer of @xmath237 counterions formed around @xmath174 and @xmath265 the size of the layer of @xmath238 counterions formed around @xmath175 ( compare ( [ eq : l1l2 ] ) to ( [ eq : xp ] ) , when @xmath266 ) .
* like - charges that are natural integer multiples of @xmath234 .
* in this case , the screening is not frustrated as in all the previous situations .
simply @xmath178 counterions will screen the charge @xmath174 forming a neutral object , and similarly around @xmath175 there will be a screening cloud of @xmath179 counterions . since both objects with their screening clouds are neutral , the effective force between them is expected to be @xmath267 .
this can be verified from the expression for the partition function applicable here , eq .
( [ eq : zc - na - even ] ) .
if @xmath29 , we have @xmath268\right ] + o(e^{-4l})\,.\ ] ] therefore , @xmath269+o(e^{-2l } ) \,.\ ] ] with the above results , we can obtain an explicit expression for the density profile of counterions @xmath270 notice that due to the fact that each particle only feels a constant electric field proportional to the difference between the number of charges which are at its left and right sides , the potential energy has the following property @xmath271 this can be interpreted as follows .
if the particle at position @xmath272 is fixed , the system decouples into two independent systems , one of size @xmath272 with @xmath273 particles , and the other one of size @xmath274 with @xmath275 particles , with the appropriate charges at each boundary ( obtained by summing the charges at the left side and right sides of @xmath276 of the original system ) .
then , the computation of the integrals in ( [ eq : density - def ] ) simply yields the product of the two partition functions of each subsystem , @xmath277 where each @xmath65 should be replaced by its appropriate corresponding expression from ( [ eq : zc - a ] ) or ( [ eq : zc - a - general ] ) .
from this expression we can verify the known relation between the contact density at @xmath172 ( or @xmath173 ) and the pressure @xcite .
indeed , notice that @xmath278 on the other hand , from eq .
( [ eq : zc - a ] ) we can verify that @xmath279 where this last relation was obtained by writing @xmath280 in ( [ eq : zc - a ] ) , and recalling that @xmath184 .
therefore , we find @xmath281 the last equality is obtained using the same argument on @xmath173 in @xmath282 .
let us consider the case @xmath283 , ie .
figure [ fig : density ] shows a plot of the density profile for @xmath57 and @xmath58 .
notice that in the case @xmath58 even , the density falls off quickly to zero far from the boundaries @xmath172 and @xmath173 . on the other hand ,
when @xmath57 is odd , the density does not fall to zero , but goes to a non - vanishing value shown by the horizontal line .
this corresponds to the density of the free counterion , responsible for the effective attraction between the two charges @xmath174 and @xmath175 as discussed earlier . the density profile for @xmath57 and @xmath58 counter - ions and @xmath285 .
notice that in the case where the number of counter - ions is odd , @xmath57 , the density far from the edges converges to a non zero value @xmath286 , here close to @xmath287.,width=415 ] to quantify this behavior , consider expression ( [ eq : density ] ) for the density in the case @xmath22 , and @xmath261 , @xmath288 in this sum , the partition function @xmath289 corresponds to a system with charges @xmath290 and @xmath291 at its boundaries .
if @xmath292 , these two charges carry opposite signs , therefore , @xmath293 is given by eq .
( [ eq : zc - a ] ) . then ,
if @xmath294 , @xmath295 . on the other hand , the second partition function , @xmath296 , corresponds to a system with charges @xmath297 and @xmath290 at its edges .
if @xmath292 , these two charges carry the same sign and are half integers multiples of @xmath177 , therefore @xmath298 should be obtained by using eq .
( [ eq : zc - na - even ] ) . in particular one can notice that if @xmath299 , then @xmath300 .
therefore , in the sum ( [ eq : densitya0 ] ) all terms with @xmath301 decay exponentially fast when @xmath205 is far from the boundaries : they are of order @xmath302 . the same argument could be applied to all the terms with @xmath303 , with the roles of @xmath293 and @xmath304 interchanged .
then , only one term in the sum ( [ eq : densitya0 ] ) survives , it corresponds to @xmath305 , which is precisely the index of the position of the free counterion . in this term , both @xmath306 and @xmath296 with @xmath305 , correspond to a system with charges @xmath290 and @xmath307 at its edges ( notice the opposite signs ) , and those partition functions should both be computed using ( [ eq : zc - a ] ) .
the leading order of these partition functions , when @xmath294 , is @xmath308 while the leading order of the denominator of ( [ eq : densitya0 ] ) is @xmath309 this gives @xmath310 this is the analytical confirmation of the intuitive analysis of section [ sec : like - charge - attraction ] where it was explained that when @xmath0 is odd , there is one free ion roaming between the two charges with an available space equal to @xmath311 , as shown in figure [ fig:1dmodel - odd ] . in the case where @xmath0 is
even , a similar analysis shows that all terms of the sum ( [ eq : density ] ) fall of exponentially fast when @xmath205 is far from the boundaries .
it is interesting to consider the limit @xmath312 . due to the electroneutrality condition @xmath313
, one needs to consider different situations : whether @xmath174 and @xmath175 are kept finite , then necessarily the charge of the counterions @xmath177 should vanish as @xmath314 .
then we notice that this is also a mean field regime .
the other possible limit is to consider that @xmath177 has a non vanishing finite value , then @xmath174 and/or @xmath175 should go to infinity as @xmath0 . momentarily , it is best to return to dimensional units @xmath316 and @xmath16 : the rescaling by @xmath317 is not appropriate here , because @xmath315 . consider the equation of state ( [ eq : lave - a ] ) derived in the isobaric ensemble , which now reads @xmath318 where @xmath319 and @xmath320 . since @xmath315 , one can recognize a riemann sum and replace it by an integral .
this finally leads to @xmath321 we recover here the implicit relation between @xmath322 and @xmath16 from the mean field theory as described by the poisson
boltzmann equation @xcite .
indeed , referring for instance to @xcite , where the mean field regime of the present problem was considered , eq .
( [ eq : lave - pb ] ) can be directly obtained from a simple linear combination of eqs .
( 16 ) and ( 17 ) of @xcite .
notice that the interesting effects , such as like - charge attraction , stemming from the discrete nature of the charges , are lost in this mean field limit .
like - charges will always have a repulsive effective interaction in the mean field regime @xcite .
a related comment is that the asymptotic negative pressure reported for odd @xmath0 in section [ sec : equal - charges ] , @xmath323 , vanishes in the limiting process addressed here .
it should be noted that the present limit is also the thermodynamic limit , since we have to remember that @xmath177 is of order @xmath314 , therefore in the left hand side of ( [ eq : lave - pb ] ) @xmath324 should be of order @xmath0 . to make this more apparent , introduce the average distance per ion @xmath325 ( inverse of the density ) , then ( [ eq : lave - pb ] ) becomes @xmath326 in this situation , the charges at the edges @xmath174 and @xmath175 should be of order @xmath0 , or at least one of them .
consider the case when both @xmath235 and @xmath236 are of order @xmath0 .
then , when @xmath312 , eq .
( [ eq : lave - a ] ) can be put in the following form by shifting the index of the summation by @xmath327 , @xmath328 notice that by shifting the index @xmath252 by one , we can replace @xmath329 by @xmath330 if necessary .
one can then write @xmath331 notice that in this analysis , the limit depends on how @xmath178 and @xmath179 are taken to infinity , and assumes that the fractional part of them is kept fixed as @xmath0 is increased . to cover the whole range of values for @xmath193 from 0 to @xmath332 ,
it is necessary that @xmath145 covers the range from @xmath333 to @xmath332 .
we recover the same phenomenology as in the case @xmath0 finite , when @xmath154 , @xmath334 .
so , the pressure can become attractive , except in the case where @xmath178 and @xmath179 are integers .
( [ eq : lave - inf ] ) can be made more explicit in two particular cases .
when @xmath178 and @xmath179 are integers , @xmath335 and when @xmath178 and @xmath179 are half integers , @xmath336 when @xmath178 and @xmath179 are not integers , the value of @xmath193 for which the pressure changes of sign is given by putting @xmath126 in ( [ eq : lave - inf ] ) @xmath337 when @xmath178 and @xmath179 are half - integers this reduces to @xmath338 .
we have studied a simple one - dimensional system as a model to understand the effective interaction between charged particles that are screened by counterions only .
this model evidences the possibility of attraction between two like - charges at large separation .
the physical phenomenon behind this attraction is a frustration of the screening process due to the discrete nature of the electric charges .
more specifically , if the two like - charges are not integers multiples of the charge of the counterions , a perfect screening of the charges is not possible , and there will be a `` misfit '' counterion , responsible for the over - screening of one of the like - charges , leading to an effective attractive force . a by - product is that in the mean - field limit where discreteness effects are washed out , no like - charge attraction is possible , a well - known phenomenon .
the present model is in addition interesting from a purely theoretical perspective , since it is exactly solvable : it is possible to compute explicitly its partition functions ( isobaric and canonical ) , the pressure ( effective force ) and the density profile of the counterions .
although the specific exact results and expression for the effective force are particular to this one - dimensional model , the physical mechanism responsible for the attraction between like - charges could also be applicable for three dimensional situations @xcite . in particular the case @xmath339 leads to an equation of state that is equivalent to that found under strong coupling for three dimensional planar interfaces , screened by point counter - ions interacting through the standard @xmath340 coulomb potential @xcite .
* acknowledgements .
* this work was supported by an ecos nord / colciencias - men - icetex action of colombian and french cooperation .
g. t. acknowledges support from fondo de investigaciones , facultad de ciencias , universidad de los andes , project `` apantallamiento y atraccin de cargas similares en sistemas de coulomb de una dimensin '' , 2015 - 2 .
the inverse laplace transform can be computed with integral inversion formula which can be evaluated using the residue theorem @xmath341 ^ 2 } \right ) ( l ) = \sum_{j=0}^{p } \mathop\text{res}_{p=-(j+\frac{1}{2})^2 } \frac { e^{pl}}{\prod_{k=0}^{p}\left[\left(k+\frac{1}{2}\right)^2+p\right]^2}\ ] ] each residue is straightforward to compute @xmath342 ^ 2 } = \frac{e^{-\left(j+\frac{1}{2}\right)^2l } } { \prod_{k=0,k\neq j}^{p}\left[\left(k+\frac{1}{2}\right)^2- \left(j+\frac{1}{2}\right)^2\right]^2 } \left ( l - \sum_{l=0 , l\neq j}^{p } \frac{2}{\left(l+\frac{1}{2}\right)^2-\left(j+\frac{1}{2}\right)^2 } \right ) \,.\ ] ] writing @xmath343 the above product and sum can be simplified @xmath344 } = \frac{(-1)^{j } ( 2j+1)}{(p - j)!(p+j+1 ) ! } \,,\ ] ] and @xmath345 gathering all results , the exact explicit result for the canonical partition function is found in the form of eq .
( [ eq : zexactnimpar ] ) . | the possibility that like - charges can attract each other under the mediation of mobile counterions is by now well documented experimentally , numerically , and analytically . yet , obtaining exact results is in general impossible , or restricted to some limiting cases .
we work out here in detail a one dimensional model that retains the essence of the phenomena present in higher dimensional systems .
the partition function is obtained explicitly , from which a wealth of relevant quantities follow , such as the effective force between the charges or the counterion profile in their vicinity .
isobaric and canonical ensembles are distinguished .
the case of two equal charges screened by an arbitrary number @xmath0 of counterions is first studied , before the more general asymmetric situation is addressed .
it is shown that the parity of @xmath0 plays a key role in the long range physics . |
the problem of susceptibility of chaotic systems to perturbations has attracted much attention in the last decade [ 1 - 9 ] .
this problem is fundamental , since it includes the determination of the response of a material system to an imposed external electromagnetic field , the setup that is typical for many experiments .
due to the sensitivity of classical phase trajectories or quantum energy spectra and stationary wavefunctions of chaotic systems to small changes of their parameters , the problem is challengingly difficult . a consistent and noncontroversial picture covering ( albeit qualitatively ) all the essential cases of the response hasnt been yet drawn at present . from the point of view of general theory ,
the problem is related to the applicability of the concept of quantum - classical correspondence to chaotic systems , that is a long - standing question in its own right @xcite .
we shall study a one - particle system with the hamiltonian of the form @xmath1 , where @xmath2 is the hamiltonian of the unperturbed system ; @xmath3 and @xmath4 are the operators of cartesian components of the momentum and of the position of the particle .
the classical system with the hamiltonian function @xmath5 will be assumed to be strongly chaotic , that is , nearly ergodic on the energy surfaces in a wide range of the energy values , system with @xmath6 degrees of freedom . in the perturbation operator @xmath7 the active variable @xmath8
is one of the cartesian coordinates of the particle , coupled to the external homogeneous force field . the amplitude @xmath9 in the following will be referred to as field . in the following
we shall deal with the quasiclassical case , when the planck constant is small in comparison of the action scale of the system @xmath10 . under the influence of the perturbation the energy value @xmath11 varies in a quasirandom way .
these variations frequently can be described as a process of the energy diffusion @xcite , when for the ensemble with the microcanonical initial energy distribution @xmath12 the dispersion of the energy increases linearly with time , @xmath13 , where @xmath14 is the energy diffusion coefficient .
if the external field @xmath9 is sufficiently small in comparison with the appropriately averaged values of the forces acting on a particle in the unperturbed system , then in the classical model the energy diffusion coefficient @xmath15 can be expressed through the characteristics of the unperturbed chaotic motion of the active coordinate , namely @xmath16 where @xmath17 is the power spectrum of the active coordinate ( the fourier transform of its autocorrelation function ) for the motion on the surface with the constant energy value @xmath18 [ 9 ] . the same expression ( 1 ) in the case of weak perturbation can be obtained in the classical limit from the quantum model .
the evolution of the quantum system can be treated as a sequence of one - photon transitions between stationary states of the unperturbed system @xmath19 , accompanied with the absorption or emission of the quanta @xmath20 . for small @xmath21
the energy spectrum of @xmath22 is quasicontinuous , thence the rates of transition are given by the fermi golden rule ( fgr ) @xmath23 where @xmath24 is the matrix element of the active coordinate , and @xmath25 is the density of states near the final state of the transition .
although the matrix elements @xmath24 in quantum chaotic systems fluctuate wildly with the variation of @xmath26 [ 10 , 11 ] , the averaged squared quantity @xmath27 in the limit @xmath28 is smooth ; it is proportional to the power spectrum @xmath29 of the coordinate @xcite , @xmath30 from eqs .
( 2 ) and ( 3 ) we have for the transition rate @xmath31 then for the energy dispersion for small @xmath32 we have @xmath33 , that returns us to the eq.(1 ) for the energy diffusion coefficient . it can be shown that the same expression for @xmath15 holds also for large @xmath32 @xcite .
the energy absorption in chaotic systems comes as an epiphenomenon of the energy diffusion [ 4 ] . with the account of the dependence on the energy of the power spectrum @xmath34 and the density of states
@xmath35 the diffusion becomes biased , and the energy absorption rate @xmath36 is given by the formula @xcite @xmath37 although for weak fields @xmath15 does not depend on the planck constant @xmath21 , the condition of the applicability of eq .
( 2 ) does .
the fgr is , after all , only a formula of the first order perturbation theory .
it is based on the assumption that the transition process has a resonant character - that the width @xmath38 of the energy distribution of states populated from the original one , given by the weisskopf - wigner formula @xmath39 @xcite , is small in comparison with the quanta energy @xmath20 . from eq .
( 4 ) it is evident that in the classical limit @xmath28 this condition will be violated . in the following
we shall use the border value of the field @xmath40 , defined by the condition @xmath41 , and refer to the domain @xmath42 as the range of the strong field . by analogy with other models , for strong fields one can expect a slow - down of the growth of the energy diffusion coefficient @xmath15 and of the energy absorption rate @xmath36 .
for example , for a two - level system with relaxation the quadratic dependence the absorption rate @xmath43 for small field turns into a field - independent value @xmath44 for strong one .
the border is determined by the condition @xmath45 , where @xmath46 is the rabi frequency and @xmath47 are longitudinal and transversal relaxation rates correspondingly @xcite .
the rate of transitions from discrete to continuous energy spectrum ( that are basically covariant with the energy absorption rate @xmath36 ) , studied in the context of the theory of photoionization , for sufficiently strong fields can even decrease with the increase of @xmath9 - the effect that is known as atom stabilization by the strong field @xcite . for our model we can describe the slow - down by a simple estimate . assuming the exponential decay of the initially populated state and using the weisskopf - wigner formula for the energy distribution of the final states [ 16 ]
, we can rewrite the expression for the rate of the transitions ( with the absorption of quanta ) in the form @xmath48 } \,\rho \left ( \omega \right)d\omega,\ ] ] where @xmath49 is a matrix element of perturbation taken as a function of the transition frequency and @xmath50 . in the weak field
the expression in square brackets turns into the @xmath51 - function , and eq .
( 6 ) returns the usual fgr . however , if we dare to use eq .
( 6 ) as an equation that is valid for any magnitude of the perturbation , then in the case of strong fields we will have @xmath52 the rate of transitions @xmath53 , and , consequently , the energy diffusion coefficient @xmath15 and the energy absorption rate @xmath36 , become proportional to the magnitude of perturbation @xmath9 and to the square root of the density of states .
this derivation of eq .
( 7 ) is too risky - but in the following sections this conclusion will be put on a more solid footing .
the slow - down of the energy diffusion in quantum chaotic systems in strong harmonic fields has been first demonstrated by cohen and kottos @xcite . however , their analytical estimates and data of numerical experiments are in quantitative disagreement with the results of the present paper .
lastly , it must be noted that the strong field regime is easily attainable to experiments .
for example , for the excitation of multiatomic molecules with the infrared laser radiation the border field corresponds to the intensity value @xmath54 , that has been reached in experiments long time ago @xcite .
for the system with the hamiltonian @xmath55 we take the wave function in the form of the expansion in the basis of stationary @xmath56 states of @xmath22 , @xmath57 then for the amplitudes @xmath58 we obtain the system of equations @xmath59 where the quantities @xmath60 are the rabi frequencies of transitions .
we shall use the initial conditions @xmath61 : at the initial moment only one of stationary states , @xmath62 , is populated .
following @xcite , we shall assume that @xmath24 are independent random gaussian variables with zero mean value and the dispersion , given by eq .
the system of equations ( 9 ) will be treated as a member of the corresponding statistical ensemble .
we are going to concentrate on the process of the energy diffusion .
then in the zeroth approximation we can restrict ourselves by consideration of the evolution of probability density in a narrow energy range around the initial state and use the power spectrum and the density of states values at this energy : @xmath63 and @xmath64 .
for the calculation of the absorption coefficient the global dependence on the energy must be restituted .
the power spectrum @xmath65 has the symmetry property @xmath66 . the dependence @xmath65 in the domain @xmath67 in typical strongly chaotic systems , such as nonlinear oscillators [ 20 ] and billiards [ 21 , 22 ] , has the form of an asymmetric peak .
we shall define the peak value of the rabi frequency simply as @xmath46 , the frequency of the maximum as @xmath68 and the characteristic width of the peak as @xmath69 .
typically the ratio @xmath70 is about few units .
right after the switching on the perturbation all amplitudes ( but that of the initially populated state ) grows in absolute value linearly in time . at this ballistic stage
the energy dispersion grows quadratically in time : @xmath71 where @xmath72 is a numerical constant .
this stage is limited by the depletion of the initial population and lasts until the depletion time @xmath73 . at this time
considerably populated levels are spread over the energy range @xmath74 that contains many levels ( since @xmath75 with @xmath6 ) .
we shall expect that at the next stage the ensemble averaged density of probability is a smooth function with a characteristic scale @xmath76 .
it is convenient to write indices in eq .
( 9 ) as arguments of functions .
we shall use the frequency distance from the initial level as a basic independent variable @xmath77 , @xmath78 will be denoted as @xmath79 , where @xmath80 .
dummy variables @xmath81 and @xmath82 will have the same meaning . by formal integration of eq . ( 9 ) and subsequent recurrent substitution we obtain the equation for the rate of change of the local probability density @xmath83 : @xmath84 summation in this formula goes over discrete values of @xmath81 and @xmath82 , and this equation is still exact . now we assume , that the amplitudes @xmath85 are random processes , that are not correlated for different states : @xmath86 . then for the averaged probability density we can retain in the eq . (
11 ) only diagonal terms : @xmath87 with the assumption that the averaged @xmath88 is a smooth function of @xmath89 , and slowly varying function of @xmath32 , we may cast the product of amplitudes in the form @xmath90 where @xmath91 is the normalized ( @xmath92 ) autocorrelation function of the probability amplitudes . by replacing the averaged square rabi frequency by its value from eq .
( 3 ) ( that depends only on the difference @xmath93 ) , substituting the summation over the states by the integration weighted with the density of states , and averaging over the time intervals that are much larger than the field period , we come to the equation @xmath94 from now we drop the angular brackets and deal only with ensemble averaged quantities
. if the rate of variations of @xmath95 is small in comparison with the decay of correlations of amplitudes given by @xmath91 , then we can treat eq .
( 13 ) as summation over the probability flow that comes from the different parts of the frequency range with the constant rate , @xmath96 this approximate expression to some extent replaces the fermi golden rule for strong perturbations . to construct the kinetic equation one must take into account both incoming and outgoing flows of probability . by taking into account the total probability flow , expanding @xmath97 in the taylor series
, we obtain a diffusion equation with the probability diffusion coefficient in the energy scale @xmath98 where @xmath99 now the problem reduces to the calculation of the integral @xmath100 . for large enough times the average probability density ,
that is governed by the diffusion equation , varies slowly , and we can treat the system eq .
( 9 ) as a set of equations in which all @xmath101 are non - correlated random processes with the same statistical properties .
then by averaging the equation for the squared time derivative of an amplitude , we obtain the expression for the mean squared frequency of these processes , that also gives the estimate for the square of the correlation decay rate @xmath102 : @xmath103 from eq . ( 18 ) for the decay correlation rate we have the estimate @xmath104 where @xmath105 is a numerical constant in the domain of strong field the autocorrelation function is the fastest component under the integration sign in eq . ( 17 ) .
then we have @xmath106 . by the substitution of this expression into eq .
( 16 ) we obtain the estimate of the energy diffusion coefficient : @xmath107 where @xmath108 is a numerical constant ; this expression agrees with the dependence that was derived in sec . 1 .
one must recall that this expression is valid only for the nearly resonant perturbation frequency ( @xmath109 ) .
we do not enter here into the studies of the dependence of @xmath15 on the perturbation frequency @xmath110 in the wider domain , postponing it for the further studies .
to check the analytical results of the preceding section , the system of equations ( 9 ) has been integrated numerically . the number of equations varied from @xmath111 to @xmath112 with the purpose to suppress the influence of the border .
the envelope of the rabi frequencies has been taken in the double lorentzian form @xmath113.\ ] ] all calculations have been carried for the `` resonant '' perturbation frequency @xmath114 and for the peak width @xmath115 . [ !
ht ] the dependence of the logarithm of the probability density on the dimensionless frequency @xmath116 for the time values @xmath117 ( a ) , @xmath118 ( b ) , and @xmath119 ( c ) .
the grassy lines - values of @xmath120 , averaged over 10 different sets of matrix elements , solid lines - fitted parabolas . to avoid the overlap of graphs , the plots for cases ( b ) and ( c )
are shifted upwards by 3 and 6 units respectively.,title="fig : " ] fig .
1 shows the distribution of probability as a function of dimensionless frequency @xmath121 for different moments of time .
it is clearly seen that even for relatively small time @xmath122 the distribution has very accurate gaussian form , with deviations noticeable only for @xmath123 .
thus we have quantitative support for our conclusion about the diffusive character of the energy evolution . [ !
ht ] dependence of the autocorrelation function of the probability amplitudes on the dimensionless time shift @xmath124 for three different sets of parameters with @xmath125 ( black squares ) , @xmath126 ( open circles ) and @xmath127 ( black triangles ) and the same value of @xmath128 .
the statistical errors are about the size of the data symbols . ,
title="fig : " ] fig 2 depicts the form of the normalized autocorrelation function of the probability amplitudes .
values @xmath91 have been calculated numerically for three sets of parameters with different values of @xmath129 but the same value of the product @xmath130 .
it is clearly seen that @xmath91 for these sets are nearly identical , as expected .
the decay rate @xmath102 taken from the equation @xmath131 is @xmath132 , that supports the estimate eq .
( 19 ) .
ht ] the dependence of the ratio @xmath133 of the energy diffusion constant @xmath15 to the doubled fermi transition rate @xmath134 on the logarithm of the ratio of the rabi frequency to its border value @xmath135 , @xmath136 .
the dashed line represents the curve @xmath137 that corresponds to the theoretical dependence eq .
( 20 ) ; it is fitted to the last three points.,title="fig : " ] figure 3 represents the dependence of the ratio @xmath138 of the energy diffusion constant @xmath15 to the doubled fermi transition rate @xmath139 on the logarithm of the ratio of the rabi frequency to its border value @xmath140 , @xmath141 .
it is seen that for the weak field this ratio comes close to the asymptotic limit ( unity ) , decreases in the vicinity of the border and decays as @xmath142 for sufficiently strong fields .
the agreement with the theoretical estimates is quite convincing .
from the comparison of the numerical data with the theoretical estimates we can conclude that the approach of the sec .
2 gives reasonably accurate description of the energy evolution process in strong fields , in spite of numerous simplifying approximations used in the calculation . to improve the accuracy and , firstly , to derive from the first principles the equation for the correlation function @xmath143
, the model of the random process must be improved .
we used the model of the stationary process , whereas from eq .
( 9 ) one can conclude that the model of periodic random process would be more appropriate .
the main conclusion from the results of our calculation is a qualitative one : by substituting eq .
( 3 ) in eq . ( 16 ) , we obtain @xmath144^{{1 \mathord{\left/ { \vphantom { 1 2 } } \right .
\kern-\nulldelimiterspace } 2 } } $ ] .
this quantity in the classical limit @xmath28 vanishes along with the energy absorption rate @xmath36 ( see eq . ( 5 ) ) .
that means the violation of the quantum - classical correspondence for the absorption - and , more generally , for the linear susceptibility of chaotic systems to harmonic external fields .
the author acknowledges the support by the `` russian scientific schools '' program ( grant # nsh - 1909.2003.2 ) .
n. brenner and sh .
fishman 1996 _ phys .
lett . _ * 77 * 3763 | the energy evolution of a quantum chaotic system under the perturbation that harmonically depends on time is studied for the case of large perturbation , in which the rate of transition calculated from the fermi golden rule exceeds the frequency of perturbation .
it is shown that the energy evolution retains its diffusive character , with the diffusion coefficient that is asymptotically proportional to the magnitude of perturbation and to the square root of the density of states .
the results are supported by numerical calculation .
they imply the absence of the quantum - classical correspondence for the energy diffusion and the energy absorption in the classical limit @xmath0 .
pacs numbers 05.45.-a |
for any @xmath2 , fix any @xmath3^s$ ] and consider probabilities @xmath4 associated with components of @xmath5 .
let the vector @xmath6 of probabilities itself be random and dirichlet - distributed with parameters @xmath7 .
let @xmath8 be a vector of @xmath9 independent samples drawn from the associated categorical distribution over components of @xmath5 .
note that the components of @xmath10 are conditionally independent , conditioned on @xmath6 , but are not unconditionally independent .
conditioned on @xmath6 , the mean of each @xmath11 is @xmath12 .
let @xmath13^n$ ] and @xmath14 for each @xmath15 .
then , the distribution of @xmath6 conditioned on @xmath16 is dirichlet with parameters @xmath17 .
let @xmath18 be distributed @xmath19 , ( \alpha^\top { \bf 1})^{-1}\right)$ ] .
let @xmath20 be a vector of @xmath21 independent samples distributed according to @xmath22 .
the distribution of @xmath18 conditioned on @xmath23 is @xmath24 , where @xmath25 in this paper , we establish that = 0mu = 0mu = 0mu @xmath26 , where @xmath27 denotes second - order stochastic dominance . in other words , conditioned on identical outcomes ,
the posterior mean of the categorical distribution second - order - stochastically dominates the posterior mean of the gaussian distribution .
this result extends earlier work relating variances of posterior means under gaussian and dirichlet models @xcite .
our result provides a dominance relation that applies to all moments .
our interest in this result stems from its significance in the area of _ reinforcement learning _
@xcite , where we have used it to establish a notion of stochastic optimism achieved by particular reinforcement algorithms that generate randomized value functions to explore in an efficient manner @xcite .
this paper presents the result and its proof in a form that will be cited by our work on reinforcement learning and that will be accessible to researchers more broadly .
in this section we will review several notions of partial orderings for real - valued random variables .
all random variables we define will be with respect to the probability space @xmath28 .
[ def : fsd ] let @xmath10 and @xmath29 be real - valued random variables .
we say that @xmath10 is ( first order ) stochastically dominant for @xmath29 if for all @xmath30 , @xmath31 we write @xmath32 for this relationship .
first order stochastic dominance defines a partial ordering between random variables but it also quite a blunt notion of dominance that will be insufficient for our purposes .
consider @xmath33 and @xmath34 with @xmath35 .
these random variables can not be related in terms of fsd . however , in the context of gambling we might imagine that the return from @xmath10 is in some sense preferable to @xmath29 , since they have the same mean but @xmath10 is somehow less risky . our next definition formalizes this notion .
[ def : ssd ] let @xmath10 and @xmath29 be real - valued random variables .
we say that @xmath10 is second order stochastically dominant for @xmath29 if for all @xmath36 convex and non - decreasing , @xmath37 \ge { \mathds{e}}[u(y)].\ ] ] we write @xmath38 for this relationship .
[ prop : ssd equiv ] let @xmath10 and @xmath29 be real - valued random variables with finite expectation .
the following are equivalent : 1 .
@xmath38 2 .
for any @xmath36 concave and increasing @xmath39 \ge { \mathds{e}}[u(y)]$ ] 3 . for any @xmath40 , @xmath41 .
4 . @xmath42 for @xmath43 and @xmath44 = 0 $ ] for all values @xmath45 .
this follows from a simple integration by parts @xcite .
second order stochastic dominance @xmath38 ensures that @xmath46 \ge { \mathds{e}}[y]$ ] .
it also establishes that for _ any _ convex loss @xmath47 that @xmath10 is less `` spread out '' than @xmath29 in the sense @xmath48 ) ] \le { \mathds{e}}[l(y-{\mathds{e}}[y])]$ ] .
motivated by this equivalence , we introduce another related dominance condition .
[ def : single crossing ] let @xmath10 and @xmath29 be real - valued random variables with cdfs @xmath49 and finite expectation .
we say that @xmath10 single - crossing dominates @xmath29 if and only if @xmath46 \ge { \mathds{e}}[y]$ ] and there a crossing point @xmath50 such that : @xmath51 we write @xmath52 for this relationship .
single crossing dominance is actually a stronger condition than ssd , as we show in proposition [ prop : relate scd ssd ] .
in general the reverse implication is not true , as we demonstrate in example [ ex : ssd not sc ] . [
prop : relate scd ssd ] let @xmath10 and @xmath29 be real - valued random variables with finite expectation then @xmath53 suppose @xmath52 with single crossing point @xmath54 .
let @xmath55 .
by @xmath52 we know @xmath56 for all @xmath57 and that @xmath58 is decreasing for all @xmath59 .
now we consider the limit @xmath60 - { \mathds{e}}[y ] \ge 0 $ ] .
hence @xmath61 for all @xmath40 , which shows that @xmath62 by proposition [ prop : ssd equiv ] .
[ ex : ssd not sc ] consider @xmath63 and let @xmath64 where @xmath65)$ ] and independent of @xmath29 . by proposition
[ prop : ssd equiv ] @xmath38 , however @xmath10 is not single crossing dominant for @xmath29 .
we display the cdfs of these variables in figure [ fig : double cross ] , they are not single crossing . in particular
the ordering of @xmath49 switches at least three points @xmath66 .
the main technical result in this paper comes in theorem [ thm : dir_norm ] , which we prove in section [ sec : proof thm ] .
[ thm : dir_norm ] let @xmath67 for @xmath3^s$ ] fixed and @xmath68 with @xmath69 and @xmath70 .
let @xmath71 with @xmath72 , then @xmath46 = { \mathds{e}}[y]$ ] and @xmath38
. at first glance , theorem [ thm : dir_norm ] may seem quite arcane , it provides an ordering between two paired families of gaussian and dirichlet distributions in terms of ssd .
the reason this result is so useful is that , given matched prior distributions , the resultant posteriors for the gaussian and dirichlet models will remain ordered in this way _ for any _ observation data .
the condition @xmath70 is technical but does not pose significant difficulties so long as at the posterior is updated with at least two observations .
we present this result as corollary [ cor : dir_norm ] .
[ cor : dir_norm ] let @xmath67 for @xmath3^s$ ] fixed and @xmath68 with @xmath69 .
let @xmath71 with @xmath72 .
let @xmath73 be the data from @xmath9 i.i.d .
samples from the categorical distribution @xmath6 and values @xmath5 .
let @xmath74 be the posterior distribution for @xmath75 and @xmath76 be the posterior distribution for @xmath77 but updating according to a mis - specified likelihood as if the observations were @xmath78 .
then , for all datasets @xmath73 such that @xmath79 we can guarantee that @xmath80 .
this result is a consequence of theorem [ thm : dir_norm ] together with algebraic relations for the conjugate updates of @xmath74 and @xmath76 given any data @xmath73 .
we write @xmath81 for the number of observations from each category @xmath5 in the dataset @xmath73 with @xmath82
. then we can write the posterior distribution @xmath83 for @xmath84 and @xmath85 . in a a similar way we can compute the posterior distribution of @xmath86 where we update with an misspecified likelihood as if the data were @xmath78 .
once again we can use a conjugate form for the update @xmath87 explicitly , @xmath88 @xmath89 we conclude by application of theorem [ thm : dir_norm ] on the updated posterior parameters @xmath90 .
the complete proof of theorem [ thm : dir_norm ] is long but the essential argument is simple .
we outline the main arguments below and fill in the details in sections [ app : beta dir ] and [ app : gauss beta ] .
first , we consider an auxilliary random variable @xmath91 with @xmath92 and @xmath93 . in lemma
[ lem : dir beta ] we show that @xmath94 .
next , we show that this auxilliary beta @xmath74 is single crossing dominant for the approximating gaussian posterior , @xmath95 . therefore , by proposition [ prop : relate scd ssd ] @xmath38 .
the main difficulty in this proof comes in establishing @xmath95 .
to do this we use a laborious calculus argument together with repeated applications of the mean value theorem .
our proof requires separate upper and lower bounds for different regions of @xmath90 and @xmath96 , but no real insight beyond that .
we believe that there should be a much more enlightened and elegant method to obtain these results .
we begin our proof of theorem [ thm : dir_norm ] with an intermediate comparison of the dirichlet distribution to a matched beta posterior .
we first state a more basic result that we will use on gamma distributions . [
le : gamma ] let @xmath97 and @xmath98 be independent random variables .
then the conditional expectations @xmath99 = \frac{k_1}{k_1 + k_2 } ( \gamma_1+\gamma_2)$ ] and @xmath100 = \frac{k_2}{k_1 + k_2 } ( \gamma_1+\gamma_2).$ ] [ lem : dir beta ] let @xmath101 for the random variable @xmath68 and constants @xmath102 and @xmath103 . without loss of generality ,
assume @xmath104 .
let @xmath105 and @xmath106 .
then , there exists a random variable @xmath107 such that , for @xmath108 , @xmath109 = { \mathds{e}}[x]$ ] and so @xmath94 .
let @xmath110 , with @xmath111 independent , and let @xmath112 , so that @xmath113 let @xmath114 and @xmath115 so that @xmath116 define independent random variables @xmath117 and @xmath118 so that @xmath119 take @xmath120 and @xmath121 to be independent , and couple these variables with @xmath122 so that @xmath123 note that @xmath124 and @xmath125 .
let @xmath126 and @xmath127 , so that @xmath128 and @xmath129 couple these variables so that @xmath130 and @xmath131 we can now say @xmath132 & = & { \mathds{e}}[(1- \tilde{p } ) v_1 + \tilde{p } v_d | x ] = { \mathds{e}}\left[\frac{v_1 \overline{\gamma}^0}{\overline{\gamma } } + \frac{v_d \overline{\gamma}^1}{\overline{\gamma } } \big| x\right ] \\ & = & { \mathds{e}}\left[{\mathds{e}}\left[\frac{v_1 \overline{\gamma}^0 + v_d \overline{\gamma}^1}{\overline{\gamma } } \big| \gamma , x\right ] \big| x \right ] = { \mathds{e}}\left[\frac{v_1 { \mathds{e}}[\overline{\gamma}^0 | \gamma ] + v_d { \mathds{e}}[\overline{\gamma}^1 | \gamma]}{\overline{\gamma } } \big| x \right ] \\ & = & { \mathds{e}}\left[\frac{v_1 \sum_{i=1}^d { \mathds{e}}[\gamma^0_i | \gamma_i ] + v_d \sum_{i=1}^dxp[\gamma^1_i | \gamma_i]}{\overline{\gamma } } \big| x \right ] \\ & \stackrel{\text{(a)}}{= } & { \mathds{e}}\left[\frac{v_1 \sum_{i=1}^d \gamma_i \alpha_i^0 / \alpha_i + v_d \sum_{i=1}^d \gamma_i \alpha_i^1/\alpha_i}{\overline{\gamma } } \big| x \right ] \\ & = & { \mathds{e}}\left[\frac{v_1 \sum_{i=1}^d \gamma_i ( v_i - v_1 ) + v_d \sum_{i=1}^d \gamma_i ( v_d - v_i)}{\overline{\gamma } ( v_d - v_1 ) } \big| x \right ] \\ & = & { \mathds{e}}\left[\frac{\sum_{i=1}^d \gamma_i v_i}{\overline{\gamma } } \big| x \right ] = { \mathds{e}}\left[\sum_{i=1}^d p_i v_i \big| x \right ] = x,\end{aligned}\ ] ] where ( a ) follows from lemma [ le : gamma ] .
therefore , @xmath74 is a mean - preserving spread of @xmath10 and so by proposition [ prop : ssd equiv ] , @xmath94 .
we complete the proof of theorem [ thm : dir_norm ] by showing that this auxilliary beta random variable @xmath74 defined in lemma [ lem : dir beta ] is second order stochastic dominant for the gaussian posterior @xmath29 .
[ lem : gauss beta ] let @xmath133 for any @xmath134 and @xmath135 .
then , @xmath95 ( and by proposition [ prop : relate scd ssd ] this implies @xmath136 ) whenever @xmath137 .
we want to prove that the cdfs cross at most once on @xmath138 . by the mean value theorem @xcite ,
it is sufficient to prove that the pdfs cross at most twice on the same interval .
we lament that the proof as it stands is so laborious , but our attempts at a more elegant solution has so far been unsuccessful .
the remainder of this appendix is devoted to proving this `` double - crossing '' property via manipulation of the pdfs for different values of @xmath139 .
we write @xmath140 for the density of the normal @xmath29 and @xmath141 for the density of the beta @xmath74 respectively . we know that at the boundary @xmath142 and @xmath143 where the @xmath144 represents the left and right limits respectively .
as these densities are positive over the interval , we can consider the log pdfs @xmath145 @xmath146 the function @xmath147 is injective and increasing ; if we can show that @xmath148 has at most two solutions on the interval we will be done .
instead we will attempt to prove an even stronger condition , that @xmath149 has at most one solution in the interval .
this sufficient condition may be easier to deal with since we can ignore the distributional normalizing constants .
@xmath150 finally we consider an even stronger condition , if @xmath151 has no solution then @xmath152 must be monotone over the region and so it can have at most one root .
@xmath153 with these definitions now let us define : @xmath154 our goal now is to show that @xmath155 does not have any solutions for @xmath156 $ ] .
once again , we will look at the derivatives and analyze them for different values of @xmath157 .
@xmath158 @xmath159 our proof will proceed by considering specific ranges for the values of @xmath157 and use different calculus arguments for each of these regions . by symmetry in the problem
, we only need to prove the result for @xmath160 . within this section of possible parameter values
we will need to subdivide the quadrant into three proof regions .
@xmath161 , @xmath162 and @xmath163 .
these regions completely cover all @xmath137 and hence suffice to complete the proof of lemma [ lem : gauss beta ] .
can be verified individually , in this case the pdfs do not intersect at any point . ] in this region we will show that @xmath165 has no solutions .
we write @xmath166 and @xmath167 as before .
@xmath168 @xmath169 @xmath170 we note that @xmath171 and so @xmath172 is a concave function . if we can show that the maximum of @xmath173 lies below @xmath174 then we know that there can be no roots .
we now attempt to solve @xmath175 : @xmath176 where here we write @xmath177 .
we ignore the case @xmath178 as a trivial special case .
we write @xmath179 and evaluate the function @xmath173 at its minimum @xmath180 . @xmath181
therefore the lemma holds for all @xmath182 in the case of @xmath184 we know that @xmath185 is a convex function on @xmath138 .
if we solve @xmath186 and @xmath187 then we prove our statement .
we will write @xmath188 for convenience .
first we solve @xmath189 in terms of @xmath190 , @xmath191 we can now evaluate the function @xmath192 at its minimum @xmath180 . @xmath193 as long as @xmath194 we have shown that the cdfs are single crossing .
we note that for all @xmath195 @xmath196 this completes the proof for @xmath197 .
our argument for this final region is no different than before , although it is slightly more involved .
the key additional difficulty is that it in this region is not enough to only look at the derivatives of the log likelihoods ; we need to use some bound on the normalizing constants to get our bounds . in @xmath199 , we know that @xmath200 so we will make use of an upper bound to the normalizing constant of the beta distribution , the beta function . @xmath201
the intuition is that , because in @xmath199 the value of @xmath202 is relatively small , this approximation will not be too bad .
therefore , we can explicitly bound the log likelihood of the beta distribution : @xmath203 we now repeat a familiar argument based upon explicit calculus .
we want to find two points @xmath204 for which @xmath205 .
since @xmath184 we know that @xmath192 is convex and so for all @xmath206 $ ] then @xmath207 .
we define the gap of the beta over the maximum of the normal log likelihood , @xmath208 if we can show the gap is positive then it must mean there are no crossings over the region @xmath209 $ ] .
this is because @xmath210 is concave and therefore totally above the maximum of @xmath211 over the whole region @xmath209 $ ] .
consider any @xmath212 ; we know from the ordering of the tails of the cdf that if there is more than one root in this segment then there must be at least three crossings .
if there are three crossings , then the second derivative of their difference @xmath192 must have at least one root on this region .
however we know that @xmath192 is convex , so if we can show that @xmath213 this can not be possible .
we use a similar argument for @xmath214 $ ] and complete this proof via laborious calculus .
we remind the reader of the definition in , @xmath215 .
for ease of notation we will write @xmath188 .
we note that : @xmath216 @xmath217 and we solve for @xmath218 .
this means that @xmath219 and clearly @xmath220 .
now , if we can show that , for all possible values of @xmath221 in this region @xmath222 , our proof will be complete . to make the dependence on @xmath223 more clear we write @xmath224 below = 1mu = 1mu = 1mu @xmath225 @xmath226
we will demonstrate that @xmath227 for all of the values in our region @xmath228 .
@xmath229 similarly , @xmath230 therefore , for any @xmath231 this means that @xmath232 .
therefore this expression @xmath233 is maximized over @xmath234 for @xmath235 .
we can evaluate this expression explicitly : @xmath236 this provides a monotonicity result which states that both @xmath237 are minimized at at the largest possible @xmath238 for any given @xmath234 over our region .
we will now write @xmath239 . if we can show that @xmath240 for all @xmath241 and @xmath242 we will be done with our proof .
we will perform a similar argument to show that @xmath243 is monotone increasing for all @xmath241 .
@xmath244 note that the function @xmath245 is increasing in @xmath234 for @xmath241 .
we can conservatively bound @xmath173 from below noting @xmath246 in our region .
@xmath247 we can use calculus to say that : @xmath248 this expression is monotone decreasing in @xmath234 and with a limit @xmath249 .
therefore @xmath250 for all @xmath234
. we can explicitly evaluate this numerically and @xmath251 so we are done . the final piece of this proof involves a similar argument for @xmath252 .
= 1mu = 1mu = 1mu @xmath253 once again we can see that @xmath254 is monotone increasing @xmath255 we complete the argument by noting @xmath256 .
this concludes our proof of the pdf double crossing in region @xmath199 .
the results of sections [ sec : r1 ] , [ sec : r2 ] and [ sec : r3 ] together prove lemma [ lem : gauss beta ] . by proposition [ prop : ssd equiv ] ,
lemmas [ lem : dir beta ] and [ lem : gauss beta ] together complete the proof of theorem [ thm : dir_norm ] .
this work was generously supported by a research grant from boeing , a marketing research award from adobe , and stanford graduate fellowships , courtesy of paccar . | we consider the problem of sequential learning from categorical observations bounded in @xmath0 $ ] .
we establish an ordering between the dirichlet posterior over categorical outcomes and a gaussian posterior under observations with @xmath1 noise .
we establish that , conditioned upon identical data with at least two observations , the posterior mean of the categorical distribution will always second - order stochastically dominate the posterior mean of the gaussian distribution .
these results provide a useful tool for the analysis of sequential learning under categorical outcomes . |
discovery of superconductivity in a hydrated sodium cobaltate na@xmath0coo@xmath1h@xmath2o @xcite has attracted much attention recently .
this material is of special interest since it is a @xmath8-electron system having a quasi - two - dimensional lattice structure , which resembles the high @xmath7 cuprates . on the other hand ,
it differs from the cuprates in that the co ions form a triangular lattice , and also there are several bands intersecting the fermi level , all of which are away from half - filling .
another interesting point concerning this material is that some experiments suggest possible occurrence of an unconventional superconductivity.@xcite in particular , there has been a debate as to whether the pairing occurs in the spin singlet channel or in the triplet channel .
an unchanged knight shift across @xmath7 found in some experiments suggests spin - triplet pairing,@xcite while the knight shift does decrease below @xmath7 in other experiments at low magnetic field.@xcite as for the occurrence of broken time reversal symmetry , several @xmath9 experiments show that such a possibility is rather small.@xcite there has also been a singlet - triplet debate theoretically .
some theories propose singlet pairing , @xcite , others triplet pairing mechanisms.@xcite naively , presence of possible ferromagnetic spin fluctuations reported in some experiments @xcite may support spin - fluctuation - mediated triplet pairing scenario .
couple of years ago , however , two of the present authors showed that @xmath7 of spin triplet superconductivity due to ferromagnetic spin fluctuations , if any , should in general be very low.@xcite a similar conclusion has been reached by monthoux and lonzarich.@xcite this is because the triplet pairing interaction mediated by spin fluctuations is proportional to @xmath10 ( in contrast to @xmath11 for the singlet case ) , where @xmath12 is the spin susceptibility , while the effective interaction that enters the normal self energy is proportional to @xmath11 , so that the suppression of pairing due to the normal self - energy overpowers the pairing effect .
we then showed that this difficulty for spin - fluctuation - mediated triplet pairing may be eased in systems having ` disconnected fermi surfaces ' , where the nodal lines of the gap can run _ in between _ the fermi surfaces to open up a full gap _ on the fermi surfaces_.@xcite since pairing with high angular momentum is degraded by the presence of gap nodes on the fermi surfaces , opening a full gap may result in an enhanced pairing.@xcite as an example , we considered the hubbard model on a triangular lattice , where the fermi surface becomes disconnected into two pieces centered around the k point and the k@xmath13 point ( see fig.[fig1](a ) ) when there are a small number of holes ( or electrons , depending on the sign of the hopping integral ) .
using fluctuation exchange ( flex ) approximation@xcite and solving the linearized liashberg equation , we have shown that a finite @xmath7 can be obtained for spin - triplet @xmath4-wave pairing , where the nodes of the gap run between the fermi surfaces.@xcite we have also confirmed this conclusion@xcite using dynamical cluster approximation,@xcite which is a non - perturbative approach . in a recent study,@xcite we have looked at the band calculation results of na@xmath14coo@xmath2@xcite from this point of view , where we find that the pocket - like fermi surfaces ( hole pockets ) near the k and k@xmath13 points ( as in the bottom figure of fig.[fig1](b ) ) , originating from the band having @xmath3 character to some extent in the notation of ref .
( denoted by the thick line in fig.[fig2](a ) ) , are disconnected in a similar sense as in the triangular lattice , namely the nodes of the @xmath4-wave gap do not intersect the fermi surfaces ( although they do intersect the large fermi surface around the @xmath15 point originating from the bands having @xmath16 character ) .
in fact , a strong motivation to focus on the @xmath3 hole pockets is the presence of van hove singularity ( vhs ) near the k point .
namely , as can be seen from the band structure shown in fig.[fig2 ] , there exist saddle points at points denoted as sp , where the density of states ( dos ) takes a large value .
since this large dos lies close to the fermi level@xcite , it is likely that the band structure around the k and k@xmath13 points strongly affects the low energy properties of this material .
in particular , ferromagnetic spin fluctuations may arise due to this high dos near the fermi level . in ref .
, we have adopted a model where we separate out the portion of the bands which has @xmath3 character to some extent , and in particular focus only on the upper @xmath3 band , which contributes to the formation of the pocket fermi surfaces and to the large dos at the vhs , while neglecting the lower @xmath3 band , which has only small dos near the fermi level due to the linear dispersion at the band top .
although this portion of the band , shown by the thick curves in fig.[fig2](a ) , is disconnected between m and @xmath15 points due to @xmath16-@xmath3 hybridization , it originally comes from a single band having @xmath3 character , as can be understood from the inset of fig.[fig2](a ) , where a tight binding band dispersion is given for a case when large @xmath16-@xmath3 level offset is introduced .
we have found that the thick portion of the band in fig.[fig2](a ) ( apart from the missing part between m and @xmath15 points ) can be roughly reproduced by a single band tightbinding model on a triangular lattice with hopping integrals up to fourth nearest neighbors .
namely , the dispersion of the tight binding model on an isotropic triangular lattice with only nearest neighbor hoppings takes its maximum at the k point ( fig.[fig1](a ) ) , while the band top moves towards the @xmath15 point when the fourth neighbor hopping is introduced ( fig.[fig1](b ) ) , resulting in a pocket like fermi surface that lies between the k and @xmath15 points when small amount of holes are present .
moreover , the band structure around the k point resembles that of the actual material ( compare fig.[fig1](b ) and the thick curve in fig.[fig2 ] ) , so the vhs due to the saddle points sp in fig.[fig2 ] is also reproduced by this effective model . as for electron - electron interactions
, we considered a model with only the on - site repulsion , where we found using flex that spin triplet @xmath4-wave pairing indeed dominates and has a finite @xmath7.@xcite in the present study , we continue to study the single band model having hoppings up to fourth nearest neighbors , but this time with repulsive interactions up to second nearest neighbors to investigate the effect of the off - site repulsions .
we find that ( i ) spin triplet @xmath4-wave pairing is robust to some extent even in the presence of off - site repulsions , but ( ii ) for sufficiently large off - site repulsions , singlet pairings can dominate over triplet pairings .
in standard notations , the hamiltonian considered in the present study is given in the form , @xmath17 where @xmath18 , @xmath19 , @xmath20 , and @xmath21 are the nearest , second nearest , third nearest and fourth nearest neighbor hopping integrals , respectively,@xcite and @xmath22 , @xmath23 , @xmath24 , represent on - site , nearest neighbor , second nearest neighbor electron - electron repulsions , respectively .
the summations @xmath25 and @xmath26 are taken over pairs of nearest and next nearest neighbors , respectively .
we define the hole density as @xmath27 , where @xmath28 is the electron band filling ( number of electrons / number of sites ) .
@xmath29 is restricted to a small number in the present study because the @xmath3 hole pockets are small .
we take @xmath30 as the unit of the energy throughout the study . to make the calculation feasible at low temperatures even in the presence of off - site repulsions up to second nearest neighbors ,
here we adopt a random phase approximation ( rpa ) approach .
the singlet ( @xmath31 ) and triplet ( @xmath32 ) pairing interactions are given within rpa as follows .
@xmath33 @xmath34 here , @xmath35 is the fourier transform of the off - site repulsions . @xmath36 and @xmath37 are the spin and charge susceptibilities , respectively , which are given as @xmath38 here @xmath39 is the bare susceptibility given by @xmath40 with @xmath41 being the energy dispersion measured from the chemical potential and @xmath42 is the fermi distribution function .
we take 128@xmath43128 @xmath44-point meshes in the actual numerical calculations . to obtain the onset of the superconducting state
, we solve the gap equation within the weak - coupling theory , @xmath45 the transition temperature @xmath46 is determined by the condition , @xmath47 . in the present approach , @xmath48 dependence as well as the self energy correction is neglected .
although this approximation is quantitatively insufficient , we believe that the present approach suffices for the purpose of discussing the _ competition _ ( relative tendency toward pairing ) among various pairing symmetries . pairing symmetries
are classified according to the irreducible representations of the point group d@xmath49 , as shown in fig.[fig3 ] .
for the cases we considered , the @xmath4-wave pairing having @xmath50 symmetry turned out to have small values of @xmath51 , so in the following sections we consider @xmath6 , @xmath52 , @xmath53 , @xmath4 , @xmath5-wave pairings having @xmath54 , @xmath55 , @xmath56 , @xmath57 , @xmath58 symmetry , respectively .
we first show results for a case without off - site repulsions , where we take @xmath59 , @xmath60 , @xmath61 , @xmath62 and @xmath63 . as seen in fig.[fig4 ] , the @xmath4-wave pairing strongly dominates over other pairing symmetries , which is consistent with our previous study.@xcite the reason for this is : ( i ) the spin fluctuations are nearly ferromagnetic ( see fig.[fig8](a ) ) , and ( ii ) the nodes of the @xmath4-wave gap do not intersect the fermi surfaces as seen in fig.[fig5 ] .
we now turn on the off - site repulsions . as a case where a moderate nearest neighbor repulsion exists , we take @xmath64 and @xmath65 , while other parameter values are taken to be the same as in section [ nooffsite ] .
we see in fig.[fig6 ] that the @xmath4-wave pairing still dominates over the others , but it is suppressed compared to the case without off - site repulsions . on the other hand , @xmath51 of singlet @xmath5,@xmath53 , and @xmath6-wave pairings , along with the singlet gap functions shown in fig.[fig7 ] , are barely affected by the off - site repulsions . to understand the reason why @xmath4-wave is suppressed while singlet pairings are unaffected by off - site repulsions , we show in fig.[fig8 ] the pairing interaction in momentum space . here
we define the `` spin part '' and the `` charge part '' of the pairing interaction as @xmath66 and @xmath67 , respectively .
the spin part peaks near the @xmath15 point , while the charge part takes a broad negative minimum value around the m point .
thus , the effect of the pair scattering between two pockets that is mediated mainly by the charge part cancels out when there are nodes on the pockets ( fig.[fig9](a ) ) , while the sign change of the @xmath4-wave gap is unfavorable in the presence of the charge part , which has a negative sign ( fig.[fig9](b ) ) . for a larger nearest neighbor repulsion , where we take the parameters values to be the same as in section [ moderate ] except that @xmath68 , we find that @xmath4-wave pairing gives way to singlet pairings , as shown in fig.[fig10 ] . among the singlet pairings , which are found to closely compete in a wide range of parameter values ,
@xmath5-wave pairing surprisingly has the largest @xmath7 for the present set of parameter values .
note that @xmath4-wave is no longer full - gaped on the fermi surfaces due to the additional nodes as seen in fig.[fig11 ] .
the reason why triplet pairings are suppressed by off - site repulsions while singlet pairings are barely affected can intuitively be understood in a real space picture . as seen in fig.[fig5 ] and fig.[fig7 ] , the triplet pairing gap functions have relatively small number of nodes , which shows that these pairing takes place at short distances in real space .
in particular , @xmath4-wave gap in the absence of off - site repulsions ( fig.[fig5 ] ) essentially has the form @xmath69 , which can be rewritten , apart from a constant , as @xmath70 where @xmath71 , @xmath72 , @xmath73 , and @xmath74 . since @xmath75 , @xmath76 , and @xmath77
are the coordinates ( in units of the lattice constant ) of the six nearest neighbors , the @xmath4-wave pairing in the absence of off - site repulsion is essentially a nearest neighbor odd - parity pairing .
by contrast , the singlet pairings have many additional nodal structures compared to the basic nodal structure shown in fig.[fig3 ] .
this means that the pairs are mainly formed at large distances in real space .
in fact , the result that @xmath51 is unaffected by @xmath23 and @xmath24 is consistent with the picture that the singlet pairs are formed at distances farther than second nearest neighbors .
the reason why singlet pairs are formed at large distances while triplet pairs are formed at short distances can intuitively be understood from a nearly ferromagnetic spin alignment in real space as shown in fig.[fig12 ] .
namely , electrons at close distances have nearly parallel spins , while the electrons can have antiparallel spins only when they are separated far away . in order to reinforce this real space intuitive picture
, we look into a case where the spin susceptibility peaks at positions more closer to the @xmath15 point , namely when the spin structure is more purely ferromagnetic than in the cases considered in sections [ moderate ] and [ large ] .
we take @xmath78 and @xmath79 , which moves the vhs and the fermi level towards the band edge , thus making the spin structure more ferromagnetic as seen in fig.[fig13](a ) . as shown in fig.[fig14](a ) , the triplet pairings dominate strongly over singlet pairings , and among the triplet pairings , @xmath52-wave has the largest @xmath7 .
this result can again be understood from the above real space picture .
namely , since the spin susceptibility peaks around the @xmath15 point , the wave length of the spin alignment is large , so that the singlet pairing distance is large from fig.[fig12 ] . in fact
, the number of nodes in the singlet pairing gaps are found to increase compared to those in section[large ] , as seen in fig.[fig15 ] .
this makes the singlet pairing less competitive against triplet pairings .
moreover , since the wave length of the spin alignment is large , not only electrons residing at nearest neighbors but also those separated to some extent have parallel spins , so that spin triplet pairing can take place at relatively large distances , thereby circumventing the effect of off - site repulsions .
it can be seen from fig.[fig15 ] that the @xmath4-wave gap has additional nodes close to the fermi surfaces suggesting that the pairing takes place not only at nearest neighbors but also at larger distances .
pairing at large distances , and thus the presence of nodes near the fermi surfaces , degrades @xmath4-wave pairing , so that @xmath52-wave pairing dominates .
even if we decrease @xmath23 down to @xmath64 , @xmath52-wave still dominates over @xmath4-wave as seen fig.[fig14](b ) , in contrast to the case discussed in sections [ moderate ] .
this is because in the presence of purely ferromagnetic spin fluctuations , @xmath4-wave pairing takes place at large distances even for relatively small off - site repulsions , so that the additional nodes are located close to the fermi surfaces .
this can in fact be seen by comparing fig.[fig15 ] and fig.[fig16 ] , where the additional nodes in the @xmath4-wave gap do not move away so much from the fermi surfaces even if we reduce @xmath23 from 1.0 to 0.7 .
this is in contrast with the case of fig.[fig7 ] and fig.[fig12 ] , where the additional nodes in the @xmath4-wave gap clearly moves away from the fermi surfaces ( i.e. , the nearest neighbor pairing component becomes large ) as we reduce @xmath23 .
the reason why @xmath5-wave pairing , which has a very large angular momentum , may dominate in the presence of off - site repulsions can be found in the disconnectivity of the fermi surface _
peculiar _ to the present system .
namely , the fermi surfaces in the present case are disconnected in a way that they do not intersect the @xmath15-m lines nor the hexagonal brillouin zone edge , both of which are the nodal lines of the @xmath5-wave gap .
thus , the number of nodal lines that intersect the fermi surfaces is similar among the three singlet pairing symmetries , extended @xmath6 , @xmath53 , and @xmath5-waves , as seen in fig.[fig11 ] , thereby making the competition subtle .
in fact , we have looked into various sets of parameter values and found that the three singlet symmetries tend to have close values of @xmath51 . to show more clearly that the subtle competition between @xmath5-wave and other singlet pairings is due to the disconnectivity of the fermi surfaces peculiar to the present system , we compare the above results with those for cases with only nearest neighbor hopping , namely with a simple triangular lattice . in a simple triangular lattice with @xmath80 ,
the fermi surfaces are disconnected in a manner that they do not intersect the @xmath15-m lines , but they do intersect the brillouin zone edge .
the absence of @xmath21 moves the van hove singularity away from the band top , so in order to have a moderate @xmath7 as in the cases above , we either have to take larger values of @xmath29 , or increase the values of the electron - electron interactions . here
we take @xmath81 and the interactions the same as in section[large ] in fig.[fig17](a ) , while in fig.[fig17](b ) we take the same @xmath29 but take the interactions 1.5 times larger than in section[large ] .
it can be seen that in both cases , @xmath51 of the @xmath5-wave pairing is much smaller than that of the @xmath53- and the @xmath6-wave pairings .
this is because the fermi surfaces intersect the brillouin zone edge , i.e. , the nodes of the @xmath5-wave gap , as seen in fig.[fig18 ] and fig.[fig19 ] . in the present study
, we have looked into a model where nearly ferromagnetic spin fluctuations and charge fluctuations due to off - site repulsions coexist .
it is interesting to compare the present situation with the case where _ antiferromagnetic _ spin fluctuations and charge fluctuations coexist .
we have previously proposed that spin - triplet pairing is possible in a quarter - filled , quasi-1d organic material ( tmtsf)@xmath2x,@xcite where superconductivity lies right next to spin density wave ( sdw ) in the pressure - temperature phase diagram , so that the spin fluctuations should be antiferromagnetic .
naively , antiferromagnetic spin fluctuations would favor spin singlet @xmath53-wave pairing , while experiments suggest triplet pairing.@xcite our proposal is that if @xmath82 ( @xmath83 because the band is quarter - filled ) charge fluctuations coexist with @xmath82 spin fluctuations , as suggested from an experimental fact that cdw actually coexist with sdw@xcite , spin triplet ` @xmath4-wave ' like pairing , which is essentially a fourth neighbor pairing having a gap form of @xmath84 ( this is called @xmath4-wave in the sense that the gap changes sign as @xmath85 along the fermi surface ) , may become competitive against singlet ` @xmath53-wave ' like pairing , which is a second neighbor pairing with a gap form @xmath86 . in momentum space , this close competition arises because ( i ) the number of nodes of the @xmath4-wave gap on the fermi surface is the same with that of the @xmath53-wave because of the disconnectivity of the fermi surfaces due to quasi - one - dimensionality , and ( ii ) the pairing interactions due to spin and charge fluctuations ( apart from the first order terms ) have the form @xmath87 for singlet pairing and @xmath88 for triplet pairing , where @xmath89 and @xmath90 are contributions from spin and charge fluctuations , respectively , so that the absolute values of @xmath91 and @xmath92 becomes comparable when @xmath93 .
a similar theory has been proposed by fuseya _
et al._@xcite recently , we have confirmed this scenario microscopically by applying rpa to an extended hubbard model with off - site repulsions up to third nearest neighbors.@xcite here , a key parameter is the second nearest neighbor repulsion , due to which the @xmath82 charge fluctuations arises . in real space , this @xmath4-@xmath53 competition can be understood as follows .
when there are purely antiferromagnetic spin fluctuations , second neighbor singlet pairing is likely to occur , as seen from fig.[fig20 ] .
if we now turn on the second nearest neighbor repulsion , the second neighbor singlet pairing is degraded , and a fourth neighbor triplet pairing becomes competitive .
since this intuitive picture in real space is rather general , the tendency that spin singlet pairing , which is favored by antiferromagnetic spin fluctuations , gives way to triplet pairings in the presence of sufficiently large off - site repulsions , should hold in various situations .
in fact , such a singlet - triplet crossover due to off - site repulsions is seen also in a half - filled quasi-1d system,@xcite a model for sr@xmath2ruo@xmath94 on a 2d square lattice,@xcite and a model for na@xmath0coo@xmath2 on a 2d triangular lattice .
@xcite now , it is interesting to point out that the above tendency is exactly the opposite of what has happened in the present case with nearly ferromagnetic spin fluctuations : triplet pairings formed at relatively short distances give way to singlet pairings formed at farther distances when sufficiently large off - site repulsions are introduced .
thus , the effect of the off - site repulsion totally differs depending on whether the spin fluctuations are ferromagnetic or antiferromagnetic .
it is also important to note that in both cases , the disconnectivity of the fermi surfaces can assist the pairings at large distances ( singlet in the ferromagnetic case and triplet in the antiferromagnetic case ) , which have larger number of nodes in the gap function . in the present study ,
we have focused only on the @xmath3 fermi pockets .
although it is necessary to consider a multiband model@xcite in order to strictly take into account the effect of the fermi pockets , our single band approach is supported by a recent multiband flex study,@xcite where a similar @xmath4-wave pairing is found to dominate when spin fluctuations are not too purely ferromagnetic .
experimentally , the fermi pockets are not observed in arpes experiments @xcite up to date .
however , since these experiments are done for materials with relatively large na content , ( i.e. , large number of electrons in coo layers ) it is likely that the fermi level lies above the @xmath3 bands .
@xcite in fact , an experimental result that maximum @xmath7 is reached only when the content of na decreases sufficiently@xcite can be considered as an indirect support for scenario in which the @xmath3 band plays an important role for superconductivity .
let us finally discuss the pairing symmetry in the cobaltate in light of the present results along with the existing experiments .
we have seen that `` full - gapped '' @xmath4-wave proposed in ref .
is robust to some extent even in the presence of off - site repulsions .
existence of nodes in the gap function is suggested in a number of experiments,@xcite which may seem to contradict with the @xmath4-wave scenario . in the actual na@xmath0coo@xmath2 , however , there is also the large fermi surface around the @xmath15 point , which is not taken into account in the present study .
then we can propose a possible scenario where the ` active ' fermi surfaces for superconductivity are the pockets , while the large fermi surface is a passive one , in which superconductivity is induced by the ` active ' band via interband interactions . since the nodes of the @xmath4-wave gap intersects the large fermi surface , this can account for the experimental results suggesting the existence of gap nodes . in fact
, it has been shown that a @xmath95sr data can be explained by taking this view.@xcite on the other hand , singlet pairings are found to dominate over triplet pairings in the present model when spin fluctuations are not purely ferromagnetic and the off - site repulsions are sufficiently large . among the singlet symmetries , @xmath5-wave is another good candidate that competes with @xmath4-wave since it does not break time reversal symmetry below @xmath7 .
we have shown that @xmath5-wave may dominate over other symmetries , but the parameter range where this symmetry dominates is not so wide as that for the @xmath4-wave .
although extended @xmath6-wave is also consistent with almost all the existing experiments , we could not find a parameter regime where this symmetry dominates .
however , considering the fact that the competition within the singlet pairings is very subtle , and that the present model only roughly takes into account the electronic structure of the actual material , we believe the possibility of extended @xmath6-wave pairing can not be ruled out . since @xmath53-wave and @xmath52-wave pairings are likely to result in a broken time reversal symmetry and/or a full gapped state ( unless an accidental situation occurs ) below @xmath7 due to the two - fold degeneracy , we believe that the possibility of these pairing states are less likely .
we may conversely say from the present results that the spin fluctuation in the actual cobaltate is not _
ferromagnetic and/or the off - site repulsions are not too large , so that @xmath52-wave pairing does not dominate . at present , it is not clear experimentally whether the pairing occurs in the singlet or in the triplet channel because knight shift experiments are done on polycrystal samples .
in order to settle this singlet - triplet debate , knight shift experiment for high quality samples , or alternatively , tunneling spectroscopy studies based on a newly developed theory for triplet superconductors proposed by one of the present authors@xcite is required .
thus , our viewpoint is that possibility of neither singlet nor triplet pairings can not be ruled out experimentally at the present stage .
then , the bottom line obtained from the present study , combined with the existing experimental facts , is that triplet @xmath4-wave pairing is most likely , although there remains a possibility of singlet @xmath5-wave pairing , and also a possibility of extended @xmath6-wave pairing can not be ruled out .
this ambiguity is a direct consequence of the very point of the present study .
namely , the present system is highly peculiar in that anisotropic pairings having large angular momentum , such as @xmath4- and @xmath5- waves , are good candidates due to the disconnectivity of the fermi surfaces . the fact that these `` unusual '' pairing symmetries are good candidates makes the pairing competition subtle .
to summarize , we have studied an effective single band model for the hydrated sodium cobaltate focusing on the @xmath3 hole pockets , where we take into account off - site repulsions up to second nearest neighbors .
we find that @xmath4-wave pairing proposed in our previous study is robust to some extent in the presence of off - site repulsions . for large off - site repulsions
, @xmath4-wave gives way to singlet pairings that does not break time reversal symmetry below @xmath7 .
calculation has been performed at the facilities of the supercomputer center , institute for solid state physics , university of tokyo 99 k. takada , , h. sakurai , e. takayama - muromachi , f. izumi , r.a .
dilanian , and t. sasaki , nature * 422 * , 53 ( 2003 ) .
h. sakurai , , k. takada , s. yoshi , t. sasaki , k. kindo , and e. takayama - muromachi , phys .
b * 68 * , 132507 ( 2003 ) . t. waki , , c. michioka , m. kato , k. yoshimura , k. takada , h. sakurai , e. takayama - muromachi , cond - mat/0306036 .
t. fujimoto , , g .- q .
zheng , y. kitaoka , r.l .
meng , j. cmaidalka , c.w .
chu , phys .
* 92 * , 047004 ( 2004 ) .
k. ishida , , y. ihara , y. maeno , c. michioka , m. kato , k. yoshimura , k. takada , t. sasaki , h. sakurai , and e. takayama - muromachi , j. phys .
. jpn . * 72 * , 3041 ( 2003 ) y. kobayashi , , m. yokoi , and m. sato , j. phys .
. jpn . * 72 * , 2161 ( 2003 ) ; _ ibid _ * 72 * , 2453 ( 2003 ) . c. michioka , m. kato , and k. yoshimura , cond - mat/0403293 .
w.higemoto , , k.ohishi , a.koda , r. kadono , k. ishida , k. takada , k. sakurai , e. takayama - muromachi , t. sasaki , cond - mat/0310324 .
y.j . uemura , p.l .
russo , a.t .
savici , c.r .
wiebe , g.j.macdougall , g.m .
luke , m. mochizuki , y. yanase , m. ogata , m.l .
foo , r.j .
cava , cond - mat/0403031 .
w. koshibae and s. maekawa , phys .
lett . * 91 * , 253901 ( 2003 ) .
g. baskaran , phys .
lett . * 91 * , 097003 ( 2003 ) .
b. kumar , b.s .
shastry , phys .
b * 68 * , 104508 ( 2003 ) . q
wang , d .- h .
lee , and p.a .
lee , phys .
b * 69 * , 092504 ( 2004 ) .
m. ogata , j. phys .
. jpn . * 72 * , 1839 ( 2003 ) .
a. tanaka and x.hu , phys .
lett . * 91 * , 257006 ( 2003 ) .
h. ikeda , y. nishikawa , and k. yamada , j.phys.soc.jpn .
* 73 * , 17 ( 2004 ) .
y. tanaka , y. yanase , and m. ogata , j.phys.soc.jpn.*73 * , 319 ( 2004 ) .
k. kuroki , y. tanaka , and r. arita , cond - mat/0311619 , to appear in phys .
m.mochizuki , y. yanase , and m. ogata , cond - mat/0407094 .
y. yanase , m.mochizuki , and m. ogata , cond - mat/0407563 .
r. arita , , k. kuroki , and h. aoki , phys . rev .
b * 60 * , 14585 ( 1999 ) ; j. phys .
. jpn . * 69 * , 1181 ( 2000 ) .
p. monthoux and g. lonzarich , phys .
b * 59 * , 14598 ( 1999 ) .
k. kuroki and r. arita , phys .
b * 63 * , 174507 ( 2001 ) .
k. kuroki and r. arita , phys .
b * 64 * , 024501 ( 2001 ) .
t. kimura , h. tamura , k. kuroki , k. shiraishi , h. takayanagi , and r. arita , phys .
b * 66 * , 149237 ( 2002 ) .
k. kuroki , t. kimura , and r. arita , phys .
b * 66 * , 184508 ( 2002 ) .
t. kimura , zenitani , k.kuroki , r.arita , and h.aoki , phys .
b * 66 * , 212505 ( 2002 ) .
s. onari , r. arita , k. kuroki , and h. aoki , phys .
b * 68 * , 24525 ( 2003 ) .
bickers , d.j .
scalapino , and s.r .
white , phys .
lett . * 62 * , 961 ( 1989 ) .
t. dahm and l. tewordt , phys .
b * 52 * , 1297 ( 1995 ) .
s. grabowski , j.schmalian , m. langer , and k.h .
bennemann , solid state commun . * 98 * , 611 ( 1996 ) .
r.arita k. kuroki , and h. aoki , j.phys.soc.jpn . *
73 * , 553 ( 2004 ) .
hettler , a.n .
tahvildar - zadeh , m. jarrell , t. pruschke , and h.r .
krishnamurthy , phys .
b * 58 * , 7475 ( 1998 ) .
d.j.singh , phys .
b * 61 * , 13397 ( 2000 ) ; m.d .
johannes and d.j.singh , cond - mat/0401646 .
@xmath19 , @xmath20 , @xmath21 should be considered as fitting parameters determined so as to reproduce the upper portion of the @xmath3 bands in momentum space , rather than the amplitude of hopping from one @xmath8 orbital to another .
k. kuroki , r. arita , and h. aoki , phys .
b * 63 * , 094509 ( 2001 )
. k. kuroki , y. tanaka , t. kimura , and r. arita , phys .
b * 69 * , 214511 ( 2004 ) .
y. fuseya , y. onishi , h. kohno , and k. miyake , j. phys .
* 14 * , l655 ( 2002 ) .
lee , s.e .
brown , w.g .
clark , m.j .
strouse , m.j .
naughton , w. kang , and p.m. chaikin , phys .
lett . * 88 * , 017004 ( 2002 ) ; phys .
b * 68 * , 092510 ( 2003 ) .
lee , m.j .
naughton , g.m .
danner , and p.m. chaikin , phys .
lett . * 78 * , 3555 ( 1997 ) ; j. p. pouget and s. ravy , j. phys .
i * 6 * 1501 ( 1996 ) .
s. kagoshima , y. saso , m. maesato , r. kondo , and t. hasegawa , solid state comm . * 110 * , 479 ( 1999 ) .
y. tanaka and k. kuroki , cond - mat/0402672 .
hasan , y .- d .
chuang , a.p .
kuprin , y.kong , d.qian , y.w .
mesler , z. hussain , a.v .
fedorov , r. kimmerling , e. rotenberg , k. rossnagle , h. koh , m. rogado , m.l .
foo , and r.j .
cava , phys .
lett . * 92 * , 246402 ( 2004 ) .
yang , s .- c .
wang , a.k.p.sekharan , h. matsui , s.souma , t. sato , t. takahashi , t. takeuchi , j.c .
campuzano , r. jin , b.c.sales , d. mandrus , z. wang , and h. ding , phys .
lett . * 92 * , 246043 ( 2004 ) .
this view may further be supported by a recent study by karppinen _
( cond - mat/0402484 ) , which shows that the actual co valence can be smaller ( the fermi level can be higher ) than expected from the composition na@xmath0coo@xmath2 .
schaak , t. klimczuk , m.l .
foo , and r.j .
cava , nature * 424 * , 527 ( 2003 ) s.onari , r.arita , k.kuroki , and h. aoki , cond - mat/0312314 .
r. arita , s. onari , k. kuroki , and h. aoki , phys .
lett . * 92 * , 247006 ( 2004 ) .
y. tanaka and s. kashiwaya , to appear in phys .
b * 70 * ( 2004 ) . | we discuss the pairing symmetry of a cobaltate superconductor na@xmath0coo@xmath1 h@xmath2o by adopting an effective single band model that takes into account the @xmath3 hole pockets , as discussed in our previous paper [ to appear in phys .
rev .
lett . ] here we consider the off - site repulsions in addition to the on - site repulsion considered in our previous study .
we show that the spin - triplet @xmath4-wave pairing proposed in our previous study is robust to some extent even in the presence of off - site repulsions
. however , @xmath4-wave pairing gives way to singlet pairings for sufficiently large values of off - site repulsions . among the singlet pairings , @xmath5-wave and extended @xmath6-wave pairings are good candidates which do not break time reversal symmetry below @xmath7 in agreement with the experiments |
the usual task in quantum tomography is to determine an unknown quantum state from measurement outcome statistics .
there are two obvious ways to vary this setting .
first , our task need not be the determination of any possible input state but only some states belonging to a restricted subset of all states .
second , we typically have some prior information , or premise , which tells us that the input state belongs to some subset of states .
it is clear that with this additional information and restricted task , this problem should be easier than the problem of determing an unknown quantum state without any prior information . as an example , consider the usual optical homodyne tomography of a single mode electromagnetic field @xcite .
if the state is completely unknown , then , in principle , one needs to measure infinitely many rotated field quadratures @xcite . however , as soon as one knows that the state can be represented as a finite matrix in the photon number basis , then already finitely many quadratures are enough , the exact number depending on the size of the matrix @xcite .
it should be emphasized that the premise is not merely a mathematical assumption but carries also physical meaning .
indeed , it simply means that the probability of detecting energies above a certain bound is zero . since one might expect that also in general a given task and premise leads to the requirement of less or worse resources ,
an immediate question is the characterization of these resources .
the task and the premise can be described as subsets of the set of all states , hence this modified setting is specified by two subsets @xmath0 ( task ) and @xmath1 ( premise ) of all states .
clearly , we must have @xmath2 to make the formulation meaningful .
smaller @xmath0 means less demanding determination task and smaller @xmath1 means better prior knowledge . in this work
we study the previously explained question from the point of view of quantum observables , mathematically described as positive operator valued measures ( povms ) . a quantum observable
is called _ informationally complete _ if the measurement outcome probabilities uniquely determine each state @xcite , and this clearly relates to the usual task in quantum tomography . the previously described generalized setting leads to the concept of _ @xmath3-informational completeness_. we present a general formulation of this property , and then concentrate on some interesting special cases .
our main results are related to situations when the premise tells that the rank of the input state is bounded by some number @xmath4 , and the task is then to determine all states with rank @xmath5 or less .
we show , in particular , that if there is no premise and the task is to determine all states with rank less than or equal to @xmath6 , where @xmath7 is the dimension of the hilbert space of the quantum system , then we actually need an informationally complete observable .
perhaps the most important informationally complete observables are covariant phase space observables .
these are widely used in both finite and infinite dimensional quantum mechanics .
however , not all covariant phase space observables are informationally complete , and for instance noise can easily destroy this desired property .
we will show that even if a covariant phase space observable fails to be informationally complete , it can be @xmath3-informationally complete for some meaningful sets @xmath0 and @xmath1 .
_ notation .
_ we denote by @xmath8 the set of natural numbers ( containing @xmath9 ) and @xmath10 .
we use the conventions @xmath11 for all nonzero @xmath12 . for every @xmath13 ,
we denote by @xmath14 the largest integer not greater than @xmath15 , and we define @xmath16 . if not specified , @xmath17 is a finite dimensional or separable infinite dimensional complex hilbert space .
we denote @xmath18 .
we denote by @xmath19 the complex banach space of bounded linear operators on @xmath17 endowed with the uniform norm , and by @xmath20 the real banach subspace of selfadjoint operators .
if @xmath21 is a complex linear space such that @xmath22 whenever @xmath23 , we denote by @xmath24 the selfadjoint part or @xmath25 , and regard it as a real linear space . then @xmath26 ; in particular , @xmath27 .
we write @xmath28 for the complex banach space of the trace class operators on @xmath17 endowed with the trace class norm , and @xmath29 . clearly , if @xmath30 , then @xmath31 as linear spaces .
we denote by @xmath32 } = 1\}$ ] the set of all states ( i.e. , density operators ) on @xmath17 , and by @xmath33 the set of all pure states ( i.e. , one - dimensional projections ) .
in this section we generalize the linear algebra framework for quantum tomography as introduced in @xcite to a wider setting , also covering infinite dimensional hilbert spaces and arbitrary measurable spaces .
let @xmath34 be a measurable space .
an _ observable _ on @xmath35 is a map @xmath36 such that 1 . each @xmath37 is a positive operator ; 2 . for all finite or denumerably infinite partitions @xmath38 of @xmath35 into disjoint measurable sets @xmath39
, we have @xmath40 , the sum converging in the weak operator topology .
if @xmath41 is an observable on @xmath35 and @xmath42 , we can define the _ associate measurement outcome probability distribution _
@xmath43 on the measurable space @xmath34 , given by @xmath44}$ ] for all @xmath45 .
when @xmath35 is a finite or denumerable set , we will take @xmath46 , the set of all subsets of @xmath35 , and denote @xmath47 and @xmath48 for all @xmath49 for short .
a _ weak*-closed real operator system _ on @xmath17 is a weak*-closed real linear subspace @xmath50 such that @xmath51 .
( note that @xmath52 is a real operator system if and only if @xmath53 is an operator system in the standard sense of operator theory , and then we have @xmath54 @xcite ) .
if @xmath52 is a weak*-closed real operator system on @xmath17 , then its _ annihilator _ is the following closed subspace of @xmath55 @xmath56}=0 \
\forall a\in{\mathcal{r}}\ } \ , .\ ] ] since @xmath51 , we have @xmath57 } = 0 $ ] for all @xmath58 . any observable @xmath36 generates a weak*-closed real operator system on @xmath17 as the weak*-closure of the real linear span of its range ; we denote @xmath59 note that @xmath60}=0 \ \forall x\in{\mathcal{a}}\ } \ , .\end{aligned}\ ] ] conversely , we have the following facts .
[ prop : exm ] a. suppose @xmath52 is a weak*-closed real operator system on @xmath17
. then there exists a finite or denumerable set @xmath35 satisfying @xmath61 and an observable @xmath41 on @xmath35 such that @xmath62 .
b. suppose @xmath63 is a closed subspace such that @xmath57 } = 0 $ ] for all @xmath64 .
then @xmath65 for some observable @xmath41 .
a. if @xmath66 , then this is proved in ( * ? ? ?
* prop . 1 ) ( note that the proof is not affected if @xmath67 ) . for @xmath68 , we use the following slight modification of the proof of ( * ? ? ?
* theorem 2.2 ) .
we define the set @xmath69 then , @xmath70 is weak*-compact and metrizable , being a weak*-closed subset of the unit ball of @xmath19 . in particular , it is separable .
let @xmath71 be a weak*-dense subset of @xmath70 , and define an observable @xmath72 by @xmath73 the series converges in norm , thus also in the weak*-topology .
since @xmath70 is weak*-closed , we have @xmath74 .
each @xmath75 can be written in the form @xmath76 since @xmath77 , we conclude that @xmath78 . by this fact and weak*-density of the set @xmath71 in @xmath70
, it follows that @xmath62 .
b. let @xmath79 } = 0 \ \forall t\in\mathcal{x } \}$ ] .
it is straightforward to check that @xmath52 is a weak*-closed real operator system , hence @xmath62 for some observable @xmath41 by item ( a ) .
moreover , @xmath80 by the bipolar theorem ; see e.g. ( * ? ? ?
* v.1.8 ) .
we recall that an observable @xmath41 is called _ informationally complete _ if for any two different states @xmath81 , the measurement outcome distributions @xmath82 and @xmath83 are different @xcite . in other words
, each state leads to a unique measurement outcome distribution . in the following
we generalize this concept of informational completeness .
suppose that two nonempty subsets @xmath84 are given .
( they may also be equal . )
the larger subset @xmath1 corresponds to a certified premise , so we know that the initial state belongs to @xmath1 with certainty .
the smaller subset @xmath0 specifies the given state determination task ; we are required to determine the state whenever it belongs to @xmath0 .
therefore , we can fulfill the given task if we are able to differentiate every state in @xmath0 from every state in @xmath1 .
this additional aspect leads to the following generalization of informational completeness .
let @xmath85 .
an observable @xmath41 is _ @xmath3-informationally complete _ if @xmath86 for any two different states @xmath87 and @xmath88 . clearly , an observable is informationally complete in the usual sense if and only if it is @xmath89-informationally complete .
this obviously corresponds to the most demanding state determination task without having any prior information .
we note that the smaller the set @xmath1 is , the more informative the premise is .
likewise , the smaller the set @xmath0 is , the less demanding the task is .
for all nonempty subsets @xmath84 , we denote @xmath90 and @xmath91 every @xmath92 satisfies @xmath57}=0 $ ] .
[ prop : prem ] let @xmath93 . for an observable @xmath41 ,
the following conditions are equivalent : a. @xmath41 is @xmath3-informationally complete . b. @xmath94 . c. @xmath95 .
let @xmath87 and @xmath88 . then , using , @xmath96 } = { { \rm tr}\left[\varrho_2 { \mathsf{m}}(x)\right ] } \\ & \leftrightarrow \quad \forall x\in{\mathcal{a } } : { { \rm tr}\left[(\varrho_1-\varrho_2 ) { \mathsf{m}}(x)\right ] } = 0 \\ & \leftrightarrow \quad \varrho_1-\varrho_2 \in { \mathcal{r}}({\mathsf{m}})^\perp \ , .\end{aligned}\ ] ] thus
, @xmath41 is @xmath3-informationally complete if and only if @xmath94 .
since @xmath97 is a linear space , the equivalence of ( ii ) and ( iii ) follows .
a well known mathematical characterization of informationally complete observables is that @xmath98 @xcite . as an application of prop .
[ prop : prem ] we give a short derivation of this fact . [ prop : well - known ] an observable @xmath41 is informationally complete if and only if @xmath99 .
each nonzero element @xmath100 decomposes as @xmath101 , where @xmath102 are the positive and negative parts of @xmath103 @xcite . since @xmath57 }
= 0 $ ] , we have @xmath104 } = { { \rm tr}\left[t_-\right ] } \equiv c$ ] , and @xmath105 as otherwise @xmath106 . setting @xmath107
, we have @xmath108 .
thus , @xmath109 , and the claim follows by prop .
[ prop : prem ] .
the rank of an operator @xmath110 is the dimension of its range : @xmath111 . for each @xmath112
such that @xmath113 , we denote by @xmath114 the set of all states @xmath115 satisfying @xmath116 . clearly , @xmath117 and @xmath118 .
moreover , if @xmath119 , we denote by @xmath120 the set of states with finite rank . in this section
we investigate observables that are ( @xmath0,@xmath1)-informationally complete when @xmath121 and @xmath122 for some @xmath123 , @xmath124 with @xmath125 , or @xmath126 and @xmath127 , or @xmath128 . by the spectral theorem each @xmath129 has a spectral decomposition ; there exists an orthonormal basis @xmath130 of @xmath17 such that @xmath131 where @xmath132 and @xmath133}$ ] .
( in the infinite dimensional case the sum in the right hand side is infinite , in which case it converges in the trace class norm and is independent on the order of the terms . ) each @xmath134 is an eigenvalue of @xmath103 , and every eigenvalue of @xmath103 appears in this decomposition as many times as is its multiplicity .
clearly , @xmath135 is the number of nonzero eigenvalues of @xmath103 counted with their multiplicities .
in addition to @xmath135 , we will also need the following other characteristic numbers of @xmath103 : * @xmath136 = number of strictly positive eigenvalues of @xmath103 counted with their multiplicities ; * @xmath137 = number of strictly negative eigenvalues of @xmath103 counted with their multiplicities ; * @xmath138 = @xmath139 ; * @xmath140 = @xmath141 .
we clearly have @xmath142 note also that @xmath143 while @xmath144 and @xmath145 .
we will be interested in subspaces @xmath146 that are annihilators of certain weak*-closed real operator systems . by prop .
[ prop : exm ] these subspaces consist of operators @xmath129 with @xmath57}=0 $ ] .
the following lemma will be used later several times .
[ prop : pos - neg ] let @xmath129 be a nonzero operator with @xmath57}=0 $ ] .
we then have the following facts . a. the inequalities @xmath147 hold .
b. there are @xmath148 and @xmath149 such that @xmath150 and @xmath151 c. if @xmath152 and @xmath149 are such that @xmath153 then @xmath154 a. since @xmath155 , it must have a nonzero eigenvalue .
further , since @xmath57}=0 $ ] , it must have both strictly positive and strictly negative eigenvalues , i.e. , @xmath156 .
the third inequality in now follows from , and the remaining inequalities are clear .
b. let @xmath157 be the spectral decomposition of @xmath103 .
we denote by @xmath158 and @xmath159 the positive and negative parts of @xmath103 respectively .
we define @xmath160 and @xmath161 .
since @xmath57}=0 $ ] , we have @xmath162 and @xmath163 thus satisfy the required conditions .
c. we recall the following consequence of fan s theorem ( * ? ? ?
* theorem 1.7 ) : if two operators @xmath164 satisfy @xmath165 , then @xmath166 for every @xmath167 , where @xmath168 , @xmath169 are the eigenvalues of @xmath170 and @xmath171 , respectively , ordered in the decreasing order and repeated according to their multiplicities . from it
follows that @xmath172 and @xmath173 . by fan
s theorem , @xmath103 can not have more strictly positive eigenvalues than @xmath174 and thus @xmath175 . similarly , @xmath176 can not have more strictly positive eigenvalues than @xmath177 and thus @xmath178 .
since @xmath179 , we obtain .
[ prop : lemma2 ] we have the following facts .
a. @xmath180}=0,\ , { { \rm rank}\,}_\downarrow(t ) \leq t \mbox { and } { { \rm rank}\,}_\uparrow(t ) \leq p\}$ ] for all @xmath181 with @xmath182 .
b. @xmath183}=0 \mbox { and } { { \rm rank}\,}_\uparrow(t ) < \infty\}$ ] . c. @xmath184}=0 \mbox { and } { { \rm rank}\,}_\downarrow(t ) < \infty\}$ ] .
a. if @xmath185 , then @xmath186 } = 0,\ , { { \rm rank}\,}_+ ( t ) \leq m \mbox { and } { { \rm rank}\,}_- ( t ) \leq n\}\ ] ] as an immediate consequence of lemma [ prop : pos - neg]b , c . since @xmath187 the claim follows .
b. we have @xmath188 , hence the claim follows from ( a ) . c. similarly , @xmath189 , which by ( a ) and triviality of the condition @xmath190 implies the claim .
the following theorem characterizes @xmath191-informational completeness and @xmath192-informational completeness for all values of @xmath193 and @xmath4 in both cases @xmath194 and @xmath119 .
[ prop : structure ] let @xmath41 be an observable and @xmath123 with @xmath195
. a. the following conditions are equivalent : a. @xmath41 is @xmath191-informationally complete .
b. every nonzero @xmath100 has @xmath196 . b. if @xmath197 with @xmath198 , then the following conditions are equivalent : a. @xmath41 is @xmath192-informationally complete .
b. every nonzero @xmath100 has @xmath196 or @xmath199 .
these are all immediate consequences of prop .
[ prop : prem ] and lemma [ prop : lemma2]a . for ( a ) , note that , if @xmath200 , then the condition @xmath201 in lemma [ prop : lemma2]a is trivial . if @xmath119 , then we can also consider @xmath202-informational completeness and @xmath203-informational completeness .
the following theorem characterizes these two properties .
[ prop : structure - fin ] let @xmath41 be an observable and @xmath119 . a. the following conditions are equivalent : a. @xmath41 is @xmath203-informationally complete .
b. every nonzero @xmath100 has @xmath204 .
b. the following conditions are equivalent : a. @xmath41 is @xmath202-informationally complete .
b. every nonzero @xmath100 has @xmath205 .
these are all immediate consequences of prop . [
prop : prem ] and lemma [ prop : lemma2]b , c . with certain choices of @xmath193 and
@xmath4 the conditions in theorem [ prop : structure ] become simpler . in the following we list some special cases . since every nonzero @xmath129 with @xmath57}=0
$ ] satisfies @xmath206 ( the second inequality following from lemma [ prop : pos - neg]a ) we recover the following two consequences of theorem [ prop : structure ] , first presented in @xcite .
[ prop : s1s1 ] let @xmath41 be an observable .
the following conditions are equivalent : a. @xmath41 is @xmath207-informationally complete .
b. every nonzero @xmath100 satisfies @xmath208 .
[ prop : spsp ] let @xmath41 be an observable and @xmath123 such that @xmath195 .
the following conditions are equivalent : a. @xmath41 is @xmath209-informationally complete
. b. every nonzero @xmath100 has @xmath210 .
a necessary condition for these equivalent conditions is that every nonzero @xmath100 satisfies @xmath211 , and
a sufficient condition is that every nonzero @xmath100 satisfies @xmath212 . obviously , the set @xmath213 contains @xmath214 whenever @xmath215 .
it follows that a @xmath216-informationally complete observable is also @xmath217-informationally complete for all @xmath218 and @xmath219 .
physically speaking , smaller @xmath193 means easier task while smaller @xmath4 means stronger premise , hence the relation between the above properties is easy to understand .
moreover , if @xmath119 , then @xmath202-informational completeness implies @xmath203-informational completeness , and for the different kinds of informational completeness we have the following implications : @xmath220 for @xmath221 with @xmath125 . for a fixed dimension @xmath194 , there are @xmath222 pairs @xmath223 consisting of integers with @xmath125 .
we would thus expect to have @xmath222 different properties of @xmath192-informational completeness .
but as we will see , for some values of @xmath224 and @xmath225 the corresponding properties are equivalent . we will derive a complete classification of the inequivalent properties and it turns out that there are only @xmath226 inequivalent forms of @xmath192-informational completeness . .
each @xmath223 represents the property of @xmath192-informational completeness .
one property implies another one if the latter can be reached from the first by moving down and right .
equivalent properties are in the same box .
the box with thick boundary is the set of all properties that are equivalent to informational completeness . ]
[ prop : equivalence - p ] ( equivalence of different premises . )
let @xmath227 .
let @xmath123 be such that @xmath228 .
the following properties are equivalent : a. @xmath229-informational completeness .
b. @xmath191-informational completeness .
it is clear that ( ii)@xmath230(i ) . to show that ( i)@xmath230(ii ) , assume that @xmath41 is a @xmath229-informationally complete observable . by theorem [ prop : structure]b this means that every nonzero @xmath100 has @xmath196 or @xmath231 .
but the second condition can not hold since @xmath232 by lemma [ prop : pos - neg]a .
therefore , every nonzero @xmath100 has @xmath196 . then , theorem [ prop : structure]a shows that @xmath41 is @xmath191-informationally complete . [ prop : equivalence - t ] ( equivalence of different tasks . )
let @xmath227 .
let @xmath221 be such that @xmath233 .
the following properties are equivalent : a. @xmath234-informational completeness . b. @xmath192-informational completeness . c. @xmath235-informational completeness .
it is clear that ( iii)@xmath230(ii)@xmath230(i ) . to show that ( i)@xmath230(iii ) , assume that @xmath41 is a @xmath234-informationally complete observable . by theorem [ prop : structure]b this means that every nonzero @xmath100 has @xmath236 or @xmath199 .
but the first condition can not hold since @xmath237 for every nonzero selfadjoint operator @xmath103 .
therefore , every nonzero @xmath100 has @xmath199 .
then , by cor .
[ prop : spsp ] @xmath41 is @xmath235-informationally complete .
( _ dimension @xmath238_.)[ex:2 ] let @xmath239 . by prop .
[ prop : equivalence - p ] and prop .
[ prop : equivalence - t ] the property of @xmath192-informational completeness is equivalent to informational completeness for the all three possible pairs @xmath223 : @xmath240 , @xmath241 , @xmath242 . [
prop : inequivalence - p ] ( inequivalence of different premises . )
let @xmath243 .
let @xmath244 , @xmath245 be such that @xmath246 .
the following properties are not equivalent : a. @xmath247-informational completeness
. b. @xmath248-informational completeness .
fix an orthonormal basis @xmath130 and define an operator @xmath103 by @xmath249 it follows from @xmath250 that @xmath251 , hence , as we are also assuming @xmath252 , this definition makes sense . since @xmath253 and @xmath57}=0 $ ] , we conclude from prop .
[ prop : exm ] that there exists an observable @xmath41 such that @xmath254 . as @xmath255 and @xmath256 for every @xmath257
, it follows from theorem [ prop : structure]b that @xmath41 is @xmath247-informationally complete , but not @xmath248-informationally complete by theorem [ prop : structure]b ( if @xmath258 ) or theorem [ prop : structure]a ( if @xmath259 ) .
( _ dimension @xmath260_.)[ex:3 ] let @xmath261 . using prop .
[ prop : equivalence - p ] and prop .
[ prop : equivalence - t ] we see that the property of @xmath192-informational completeness is equivalent to informational completeness for five choices of @xmath223 : @xmath262 , @xmath263 , @xmath264 , @xmath240 and @xmath241 .
the remaining property , namely @xmath265-informational completeness , is not equivalent to @xmath266-informational completeness ( and hence not to any other ) by prop .
[ prop : inequivalence - p ] .
[ prop : inequivalence - t ] ( inequivalence of different tasks . )
let @xmath267 .
let @xmath268 and @xmath124 such that @xmath269 and @xmath270 .
the following properties are not equivalent : a. @xmath271-informational completeness .
b. @xmath272-informational completeness .
fix an orthonormal basis @xmath130 and define an operator @xmath103 by @xmath273 it follows from @xmath274 that @xmath275 , hence @xmath276 and , as we are also assuming @xmath277 , this definition makes sense . since @xmath253 and
@xmath57}=0 $ ] , we conclude from prop . [
prop : exm ] that there exists an observable @xmath41 such that @xmath254 .
we have @xmath278 for every @xmath257 . by theorem [ prop :
structure]b ( if @xmath197 ) or theorem [ prop : structure]a ( if @xmath279 ) , @xmath41 is @xmath271-informationally complete but not @xmath272-informationally complete .
( _ dimension @xmath280_.)[ex:4 ] let @xmath280 .
using prop .
[ prop : equivalence - p ] and prop .
[ prop : equivalence - t ] we see that the property of @xmath192-informational completeness is equivalent to informational completeness for five choices of @xmath223 : @xmath281 , @xmath282 , @xmath283 , @xmath262 and @xmath263 .
we also see that the properties corresponding to @xmath284 and @xmath264 are equivalent but inequivalent to informational completeness . by prop .
[ prop : inequivalence - p ] and prop .
[ prop : inequivalence - t ] , the remaining properties corresponding to @xmath240 , @xmath241 and @xmath242 are not equivalent to any other choices of @xmath223 .
a moment s thought shows that props .
[ prop : equivalence - p ] - [ prop : inequivalence - t ] give a complete classification of the @xmath192-informational completeness properties into equivalent and inequivalent collections in all finite dimensions . the classification for @xmath285
is summarized in fig .
[ fig : equivalence ] .
one of the most interesting consequences of this classification is the following .
( equivalence to informational completeness . )
let @xmath227 . for integers @xmath286 , @xmath192-informational completeness is equivalent to informational completeness if and only if @xmath287 and @xmath288 .
this result is implying , in particular , that if there is no premise and the task is to determine all states with rank less or equal to @xmath289 , then we actually need an informationally complete observable . when @xmath119 , the infinite dimensional state space has also proper subsets that do not exist in the finite dimensional case , hence leading to new kinds of tasks and premises .
for example , we can have a premise that the system has finite rank but we do not know any upper bound for its rank , i.e. , @xmath290 . similarly , we could be interested in the task of determining all states with finite rank , i.e. , @xmath126 , when @xmath127 .
we already characterized @xmath203- and @xmath202-informational completeness in theorem [ prop : structure - fin ] .
the next proposition shows that these two properties are not the same .
[ prop : sf ] let @xmath291 .
the following properties are all inequivalent : a. @xmath203-informational completeness . b. @xmath202-informational completeness . c. informational completeness . clearly , ( iii ) @xmath230 ( ii ) @xmath230 ( i ) .
hence we need to show that ( i ) @xmath292 ( ii ) and ( ii ) @xmath292 ( iii ) .
there clearly exist @xmath293 such that @xmath294}=0 $ ] , @xmath295 and @xmath296 . by prop .
[ prop : exm ] , there exists two observables @xmath297 and @xmath298 such that @xmath299 . as in the proofs of props .
[ prop : inequivalence - p ] and [ prop : inequivalence - t ] , we have @xmath300 and @xmath301 for all nonzero @xmath302 , and @xmath303 for all nonzero @xmath304 .
thus , by theorem [ prop : structure - fin ] the observable @xmath297 is @xmath203-informationally complete but not @xmath202-informationally complete .
similarly , by theorem [ prop : structure - fin ] and cor . [ prop : well - known ] the observable @xmath298 is @xmath202-informationally complete but not informationally complete .
the content of prop .
[ prop : sf ] is , essentially , that knowing that the unknown state has finite rank is useful information for state determination .
in this section we assume that @xmath194 and @xmath305 . by a _
minimal @xmath3-informationally complete observable _ we mean a @xmath3-informationally complete observable with minimal number of outcomes .
more precisely , an observable @xmath41 with an outcome space @xmath35 is minimal @xmath3-informationally complete if any other @xmath3-informationally complete observable @xmath306 with an outcome space @xmath307 satisfies @xmath308 . since @xmath17 is finite dimensional , the real vector spaces @xmath309 and @xmath55 are the same and @xmath310 we then see that a @xmath3-informationally complete observable @xmath41 with @xmath311 outcomes exists if and only if there is a @xmath312-dimensional subspace @xmath313 satisfying 1 .
@xmath57 } = 0 $ ] for all @xmath314 ; 2 . @xmath315 . indeed , in this case by prop .
[ prop : exm ] we can find an observable @xmath41 with @xmath316 and having @xmath317 outcomes . such an observable is @xmath3-informationally complete by prop .
[ prop : prem ] .
we thus conclude that seeking a minimal @xmath3-informationally complete observable is equivalent to looking for a real subspace @xmath313 satisfying ( 1 ) and ( 2 ) , and having maximal dimension among all subspaces of @xmath55 with the properties ( 1 ) and ( 2 ) .
once such a maximal subspace @xmath25 is found , then the minimal number of outcomes for a @xmath3-informationally complete observable is @xmath318 . if @xmath239 , then all @xmath192-informational completeness properties are equivalent ( see example [ ex:2 ] ) , hence the minimal number of outcomes is @xmath319 in all of them . if @xmath261 , then only @xmath265-informationally completeness is inequivalent to informational completeness ( see example [ ex:3 ] ) . in the latter case
the minimal number is @xmath320 , while in the first case a simple argument shows that the minimal number of outcomes is @xmath321 ; see prop . 5 in @xcite .
let us then assume @xmath322 and recall some bounds for the minimal number of @xmath209- and @xmath191-informationally complete observables . in these cases we need to find subspaces @xmath25 such that every nonzero @xmath314 satisfies @xmath57 } = 0 $ ] and @xmath210 ( cor .
[ prop : spsp ] ) or @xmath196 ( theorem [ prop : structure]a ) , respectively . to find a good upper bound for the minimal number of outcomes , we need to find as large @xmath25 as possible .
a useful method for constructing these kind of subspaces was presented in @xcite . using this method the following upper bounds ( a ) and ( b ) were proved in @xcite and @xcite , respectively .
[ prop : upper ] let @xmath323
. there exists a. @xmath209-informationally complete observable with @xmath324 outcomes .
b. @xmath191-informationally complete observable with @xmath325 outcomes . in the case of @xmath209-informationally complete observables
, it is possible to obtain lower bounds from the known non - embedding results for grassmannian manifolds @xcite . in some cases the obtained lower bounds agree or are very close with the upper bounds written in prop .
[ prop : upper]a .
in particular , it was proved in @xcite that in the case of @xmath265-informational completeness , the minimal number of outcomes is not a linear function of @xmath7 but differs from the upper bound @xmath326 at most @xmath327 . also a slightly better upper bound was derived , and these results give the exact answer for many @xmath7 .
for instance , for the dimensions between @xmath238 and @xmath328 , the results of @xcite give the exact minimal number in @xmath329 cases .
in the case of @xmath330-informational completeness , the upper bound for the minimal number of outcomes is @xmath331 @xcite . obviously , the known lower bound for minimal @xmath265-informational completeness is also a lower bound for minimal @xmath330-informational completeness .
we are not aware of any better lower bound . in the following subsection
we prove that the minimal number of outcomes for @xmath280 is @xmath332 .
this means that @xmath331 is generally just an upper bound for the minimal number of outcomes , not the exact answer .
our result for @xmath280 also implies that , as in the case of @xmath265-informational completeness , the minimal number is not a linear function of @xmath7 .
in this subsection we concentrate on minimal observables in dimension @xmath333 . a minimal informationally complete observable
has @xmath334 outcomes .
further , it was shown in @xcite that a minimal @xmath207-informationally complete observable has @xmath335 outcomes . in prop . [
prop : d=4/2 ] below we give the minimal numbers for the remaining three inequivalent properties ( see example [ ex:4 ] ) .
these results are summarized in fig .
[ fig : equivalence-4 ] . before deriving the minimal numbers we characterize these properties in convenient forms . .
each @xmath223 represents the property of @xmath192-informational completeness , and equivalent properties are in the same box .
as explained in example [ ex:4 ] , there are five inequivalent properties .
the big numbers give the minimal number of outcomes that an @xmath192-informationally complete observable must have . ]
[ prop : d=4/1 ] let @xmath280 .
an observable @xmath41 is a. @xmath336-informationally complete if and only if every nonzero @xmath337 satisfies @xmath338}>0 $ ] .
b. @xmath339-informationally complete if and only if every nonzero @xmath337 satisfies @xmath338}<0 $ ] . c. @xmath340-informationally complete if and only if every nonzero @xmath337 satisfies @xmath338}\neq 0 $ ] .
if an observable is @xmath340-informationally complete , then it is either @xmath336-informationally complete or @xmath339-informationally complete . by lemma [ prop : pos - neg]a
every nonzero @xmath100 has @xmath341 . since @xmath338}$ ] is the product of eigenvalues ,
we conclude that every nonzero @xmath337 satisfies a. @xmath338}>0 $ ] if and only if every nonzero @xmath337 satisfies @xmath342
. b. @xmath338}<0 $ ] if and only if every nonzero @xmath337 satisfies @xmath343 . c. @xmath338}\neq 0 $ ] if and only if every nonzero @xmath337 satisfies @xmath342 or @xmath343 . the claims ( a ) , ( b ) and ( c ) in prop .
[ prop : d=4/1 ] now follow from theorem [ prop : structure]a , cor .
[ prop : spsp ] and theorem [ prop : structure]b , respectively . to prove the last claim ,
suppose @xmath344 is a subspace such that every nonzero @xmath345 satisfies @xmath346}\neq 0 $ ] .
we need to prove that the sign of @xmath346}$ ] is constant for all nonzero @xmath347 . if @xmath348 , then this is clearly true .
so assume that @xmath349 .
we make a counter assumption : @xmath350 are two linearly independent matrices with @xmath346}<0 $ ] and @xmath351}>0 $ ] .
then @xmath352 for every @xmath353 , and @xmath354}=0 $ ] for some @xmath355 by the intermediate value theorem .
[ prop : d=4/2 ] let @xmath280 .
a. a minimal @xmath336-informationally complete observable has @xmath332 outcomes .
b. a minimal @xmath339-informationally complete observable has @xmath356 outcomes . c. a minimal @xmath340-informationally complete observable has @xmath332 outcomes .
\(a ) for all @xmath357 , denote by @xmath358 the complex linear space of @xmath359 complex matrices , and by @xmath360 the real space of selfadjoint elements in @xmath358 . by prop .
[ prop : d=4/1 ] we need to look for subspaces @xmath361 such that 1 .
@xmath362 } = 0 $ ] for every @xmath345 ; 2 .
@xmath346 } > 0 $ ] for every nonzero @xmath345 . indeed , if @xmath25 has maximal dimension among all subspaces of @xmath363 satisfying these two conditions , then any observable @xmath41 with @xmath316 and @xmath318 outcomes ( which exists by prop .
[ prop : exm ] and ) is minimal @xmath336-informationally complete .
it was shown in @xcite that the maximal dimension of a real subspace @xmath361 satisfying ( 2 ) is @xmath364 .
we show that if the additional requirement ( 1 ) is added , this maximal dimension remains the same , and thus the minimal number of outcomes is @xmath365 .
to do this , we introduce four @xmath366 matrices @xmath367 and define the following linear map @xmath368 @xmath369 note that @xmath370 next , we define five @xmath371 selfadjoint matrices @xmath372 and the following linear map @xmath373 @xmath374 clearly , @xmath375 } = 0 $ ] for all @xmath376 . moreover , it is easy to verify that the matrices @xmath377 are linearly independent .
it follows that the map @xmath378 is injective , hence the image @xmath379 of @xmath378 is a @xmath364-dimensional subspace of @xmath363 . finally , using the formula for the determinant of square block matrices ( * ? ? ?
* theorem 3 ) we obtain @xmath380 } = { { \rm det}\left[-a_0 ^ 2{\mathbbm{1}}- n(\vec{a})n(\vec{a})^*\right ] } = ( a_0 ^ 2 + { \left\|\vec{a}\right\|}^2)^2 \geq 0\ ] ] and @xmath381}=0 $ ] if and only if @xmath382 and @xmath383 .
thus , @xmath384 has the required properties .
\(b ) suppose @xmath385 is a real subspace of @xmath55 such that @xmath338}<0 $ ] for all nonzero @xmath314 .
we claim that @xmath386 .
to prove this , let us first make a counter assumption that @xmath349 .
we fix two linearly independent @xmath350 , and then the map @xmath387}\end{aligned}\ ] ] is continuous by the continuity of each map @xmath388 and @xmath389 . since the set @xmath390 is connected , this would imply that @xmath391 is constant , hence @xmath392 .
it follows that @xmath393 for all @xmath394 , but this is impossible as @xmath338}<0 $ ] .
thus , the counter assumption is false .
we still need to prove that there exists a @xmath395-dimensional subspace of @xmath55 such that @xmath338}<0 $ ] and @xmath57 } = 0 $ ] for all nonzero @xmath314 .
fix an orthonormal basis @xmath396 of @xmath17 , and set @xmath397 then , @xmath398 is a @xmath395-dimensional subspace of @xmath55 such that @xmath338}<0 $ ] and @xmath57 } = 0 $ ] for all nonzero @xmath314 .
thus , there exists an observable @xmath41 with @xmath399 outcomes such that @xmath400 , and such observable is minimal @xmath339-informationally complete by props .
[ prop : exm ] and [ prop : d=4/1 ] .
\(c ) this follows from prop .
[ prop : d=4/1 ] combined with items ( a ) and ( b ) above .
we now turn our attention to covariant phase space observables . after introducing these observables in the general case of a phase space defined by an abelian group
, we treat the finite and infinite dimensional cases separately . as an application we study the effect of noise on the observable s ability to perform the required state determination tasks .
let @xmath401 be a locally compact and second countable abelian group with the dual group @xmath402 .
the composition laws of @xmath401 and @xmath402 will be denoted by addition , and the canonical pairing of @xmath403 and @xmath404 will be denoted by @xmath405 .
we fix haar measures @xmath406 and @xmath407 on @xmath401 and @xmath402 , respectively . if @xmath408 is any bounded measure on @xmath409 , the _ symplectic fourier transform _ of @xmath408 is the bounded continuous function @xmath410 on @xmath409 given by @xmath411 this definition clearly extends to any integrable function : if @xmath412 , we define @xmath413 for the rest of this section , we will assume that the haar measures @xmath406 and @xmath407 are normalized so that @xmath414 whenever also @xmath415 .
let @xmath416 .
we define the following two unitary representations @xmath417 and @xmath418 of @xmath401 and @xmath402 on @xmath17 @xmath419 ( y ) = \psi(y - x ) \ , , \qquad [ v(\xi ) \psi ] ( y ) = { \left\langle\,\xi , y\,\right\rangle } \psi(y ) \ , .\ ] ] note that @xmath420 so that the following _ weyl map
_ @xmath421 is a projective square integrable representation of the direct product group @xmath409 on @xmath17 ( for square integrability of @xmath422 , see e.g. @xcite in the case @xmath423 , and ( * ? ? ?
* theorem 6.2.1 ) and @xcite for the general case ) .
the weyl map has the useful properties @xmath424 and @xmath425 for any set @xmath426 , we denote @xmath427 if @xmath428 is a symmetric set ,
i.e. , @xmath429 then from it follows that @xmath430 , and we can thus consider the selfadjoint part @xmath431 of @xmath432 .
if in addition @xmath433 , then @xmath431 is a weak * closed real operator system on @xmath17 .
let @xmath434 be the borel @xmath435-algebra of the locally compact and second countable space @xmath436 .
for any state @xmath437 , a _
covariant phase space observable _ with the _ fiducial state _
@xmath438 is the following observable @xmath439 on @xmath409 @xmath440 ( see @xcite or @xcite for the case of @xmath423 , and ( * ? ? ?
2.1 , p. 166 ) and ( * ? ? ?
* theorem 3.4.2 ) for the general form of covariant phase space observables ) .
the integral in the definition of @xmath439 is understood in the weak*-sense ,
i.e. , for all @xmath441 , @xmath442 } = \int_x { { \rm tr}\left[sw(x,\xi)\tau w(x,\xi)^*\right ] } { \,{\rm d}}x { \,{\rm d}}\xi \qquad \forall x\in{\mathcal{b}({\mathcal{g}}\times{\widehat{\mathcal{g } } } ) } \ , .\ ] ] more generally , the map @xmath443}$ ] is continuous and integrable for all @xmath444 , with @xmath445 } { \,{\rm d}}x { \,{\rm d}}\xi = { { \rm tr}\left[s\right]}{{\rm tr}\left[t\right]}\ ] ] ( see ( * ? ? ?
* lemma 3.1 ) for the case @xmath423 , and @xcite for the general case ) . for any @xmath446 ,
the _ inverse weyl transform _ of @xmath103 is the continuous function @xmath447 on @xmath409 given by @xmath448 } \ , .\ ] ] the zero - level set of @xmath103 is the closed set @xmath449 as usual , @xmath450 stands for the support of @xmath447 , that is , @xmath451 by injectivity of the inverse weyl transform @xcite , we have @xmath452 if and only if @xmath453 or , equivalently , @xmath454 . since @xmath455 , if @xmath129 , then the sets @xmath456 and @xmath450 are symmetric .
moreover , if @xmath103 is such that @xmath57}\neq 0 $ ] , then @xmath457 , since @xmath458}$ ] .
[ prop : covariant_nassela ] for any covariant phase space observable @xmath439 we have @xmath459 and @xmath460 note that @xmath461 if and only if @xmath462}=0 $ ] for all @xmath463 . in this case , using and , we obtain that the symplectic fourier transform @xmath464 } { \,{\rm d}}y { \,{\rm d}}\zeta = \widehat{s}(x,\xi)\overline{\widehat{\tau}(x,\xi ) } \equiv 0\ ] ] so that @xmath465 , but since @xmath466 is closed , this implies that @xmath467 . on the contrary , if @xmath468 is such that @xmath467 , then by the injectivity of the symplectic fourier transform we have @xmath462}=0 $ ] for all @xmath463 so that @xmath469 .
this shows the first equality in . for the second , note that , if @xmath470 are such that @xmath471 , then @xmath472 therefore , @xmath471 @xmath473 @xmath474 , hence the second equality in holds . in order to complete the proof
, we note that @xmath475 } = 0 \ \forall ( x,\xi ) \in { \rm supp}\ , \widehat{\tau}\ } \\ & = \{s\in{\mathcal{t}_s({\mathcal{h}})}\mid { \rm supp}\ , \widehat{\tau } \subseteq z(s)\ } \ , .\end{aligned}\ ] ] comparing this with , follows by the bipolar theorem .
it follows from prop .
[ prop : covariant_nassela ] that , vaguely speaking , the larger the support of the inverse weyl transform of the fiducial state is , the better the corresponding observable is from the state determination point of view .
in particular , the extreme case @xmath476 , if any , must yield an informationally complete observable .
the next proposition shows that this is indeed a necessary and sufficient condition for informational completeness .
the proof ( taken from @xcite ) is a straightforward generalization of the corresponding result for the case @xmath477 proved in @xcite .
[ prop : covariant_infocomplete ] the following conditions are equivalent : a. @xmath439 is informationally complete b. @xmath476 . if ( ii ) holds , then by prop .
[ prop : covariant_nassela ] and injectivity of the inverse weyl transform we have @xmath478 , so that @xmath439 is informationally complete by cor .
[ prop : well - known ] .
suppose then that ( ii ) does not hold .
in order to complete the proof , by we need to show that there exists a nonzero @xmath468 such that @xmath467 .
the set @xmath479 is nonempty , symmetric , open , and does not contain the identity @xmath480 .
let @xmath481 .
then we can find a symmetric open neighbourhood @xmath418 of @xmath480 such that @xmath482 and the measure @xmath483 of @xmath418 is finite . the
function @xmath484 ( convolution in @xmath409 ) is then of positive type ( * ? ? ?
* cor . 3.16 ) , so by bochner s theorem ( * ? ? ?
* theorem 4.18 ) there exists a positive bounded measure @xmath485 such that @xmath486 .
let @xmath487 , @xmath488 , be nonzero and define @xmath489 so that @xmath490 and @xmath491 .
now , @xmath492 for all @xmath493 , since @xmath494 implies @xmath495 .
therefore , @xmath496 for all @xmath493 so that @xmath497 for all @xmath493 , but @xmath498 since @xmath499 } \neq 0 $ ] .
finally , define @xmath500 and @xmath501 , so that at least one of @xmath502 , @xmath503 is nonzero and @xmath504 for all @xmath505 .
hence , @xmath506 and , as @xmath507 , the proof is complete . for any nonzero @xmath508 ,
we denote by @xmath509 the cyclic group with @xmath7 elements , and let @xmath510 . then , @xmath511 , the pairing of @xmath512 and @xmath513 being @xmath514
. moreover , the haar measures of @xmath509 and @xmath515 are just @xmath516 the respective counting measures .
let @xmath17 be a @xmath7-dimensional hilbert space , and choose an orthonormal basis @xmath517 of @xmath17 .
the weyl map @xmath518 is then given by @xmath519 since @xmath520 } = d \delta_{x , y } \delta_{\xi,\zeta}$ ] for all @xmath521 , the set @xmath522 is an orthonormal basis of the linear space @xmath19 endowed with the hilbert - schmidt inner product @xmath523}$ ] .
[ prop : covariant_nassela ] now reduces to @xmath524 so that the real operator system is completely characterized by the zero set @xmath525 .
therefore the essential question is the characterization of possible zero sets .
this is done in the next proposition .
[ prop : finite_zero_set ] for any state @xmath437 , we have @xmath526 and @xmath527 .
conversely , if @xmath528 is such that @xmath529 and @xmath530 , then there exists a state @xmath438 such that @xmath531 .
we have already observed that @xmath532 and @xmath533}>0 $ ] implies the first part of the proposition .
for the second part , suppose first that @xmath534
. then one may choose @xmath535 with @xmath536 , where @xmath537 and @xmath538 is a normalization constant , which gives @xmath539 and hence @xmath540 .
suppose now that @xmath541 .
for any @xmath542 define the function @xmath543 by taking the symplectic fourier transform we have for @xmath544 @xmath545 and @xmath546 and @xmath547 otherwise .
now define @xmath548 which then satisfies @xmath549 let @xmath550 be such that @xmath551 for all @xmath552 , and define @xmath553 then @xmath437 since @xmath554 and @xmath555 , and moreover @xmath556 if and only @xmath557 . that is ,
@xmath558 . since in the finite dimensional setting no topological considerations are needed , prop .
[ prop : covariant_infocomplete ] takes the following simple and well - known form .
[ prop : eqz0finite ] the following conditions are equivalent : a. the observable @xmath439 is informationally complete .
b. @xmath540 .
the next result shows that for covariant phase space observables in dimensions @xmath238 and @xmath260 all of the notions of informational completeness are equivalent . for @xmath239 ,
indeed this is true for arbitrary observables ( example [ ex:2 ] ) ; but the fact that in dimension @xmath260 all the notions of informational completeness are equivalent is specific to covariant phase space observables ( compare with example [ ex:3 ] ) .
let @xmath239 or @xmath261 .
then the following conditions are equivalent .
a. the observable @xmath439 is @xmath192-informationally complete for some @xmath221 such that @xmath559 .
b. the observable @xmath439 is informationally complete .
we already remarked that , in the case @xmath239 , all the properties of @xmath560-informational completeness are equivalent by example [ ex:2 ] . if @xmath261 , then by example [ ex:3 ] we still have to show that , for the observable @xmath439 , @xmath207-informational completeness implies informational completeness .
now , the observable @xmath439 is @xmath207-informationally complete if and only if either @xmath561 , in which case we are done , or @xmath562 for some invertible @xmath129 with @xmath57}=0 $ ] by ( * ? ? ?
5 ) . in particular , @xmath563 . in order to complete the proof we only need to show that this is not possible
by and linear independence of the set @xmath564 , we have @xmath565 .
but since @xmath525 is symmetric , @xmath527 and the dimension @xmath261 is odd , @xmath525 must contain an even number of points , hence @xmath566 is even .
as we have noted before , by increasing the size of the zero set the observable becomes less capable of performing state determination tasks .
the next proposition shows that already in the simplest case of an informationally _ incomplete _ observable , namely , one having a zero set consisting of a single point , certain tasks become impossible .
[ prop : z=1 ] suppose @xmath567 .
the condition @xmath568 can hold for some fiducial state @xmath438 only if @xmath7 is even .
if @xmath438 is a fiducial state with @xmath568 , then the observable @xmath439 is a. @xmath191-informationally complete for all @xmath569 .
b. not @xmath209-informationally complete for any @xmath570 .
let @xmath571 with @xmath544 . since @xmath525 is symmetric , we have @xmath572 , and this implies that @xmath7 is even and @xmath573 or @xmath574 . in particular , @xmath575 .
we fix a square root of @xmath405 and denote it by @xmath435
. then the operator @xmath576 is selfadjoint ( by ) and generates @xmath577 ( by ) . since @xmath578 and @xmath57 } = 0 $ ]
, we have @xmath579 .
the rest of the claim then follows from theorem [ prop : structure ] . for the next possible case ,
i.e. , a zero set consisting of two points , we can give the following characterization , analogous to prop .
[ prop : z=1 ] , in odd prime dimensions .
suppose @xmath7 is an odd prime number and @xmath438 is a fiducial state with @xmath580 .
the observable @xmath439 is a. @xmath191-informationally complete for all @xmath581 .
b. not @xmath209-informationally complete for any @xmath582 .
we have @xmath583 for some nonzero @xmath584 . as @xmath7 is odd
, @xmath238 has a multiplicative inverse in the ring @xmath509 and we can define the following projective representation @xmath585 of @xmath586 on @xmath17 @xmath587 note that , as @xmath588 , the operators @xmath589 are selfadjoint for all @xmath590 , and @xmath591 by . for all @xmath592 ,
the map @xmath593 is actually a unitary representation of @xmath509 , which is equivalent to the representation @xmath418 @xcite .
so , there is a hilbert basis @xmath594 such that @xmath595 for all @xmath596 . in particular
, for @xmath597 the eigenvalues of @xmath598 are @xmath599 for fixed @xmath600 and @xmath601 .
we thus see that the following three possibilities occur : 1 .
@xmath602 ; 2 .
@xmath603 , @xmath604 ; 3 .
@xmath605 , @xmath606 . in all three cases
we see that @xmath607 , and thus @xmath439 is @xmath191-informationally complete for @xmath608 by theorem [ prop : structure ] .
moreover , choosing @xmath598 as in item ( 1 ) , by the same theorem we see that @xmath439 is not @xmath209-informationally complete for @xmath609 .
we remark that if the dimension @xmath7 is not an odd prime , then in the case @xmath610 the observable @xmath439 need not be @xmath207-informationally complete . indeed ,
fix @xmath280 and let @xmath611 .
then @xmath612 and the elements of @xmath577 are thus @xmath613 now , for instance any @xmath614 , @xmath615 , will give @xmath616 which implies that @xmath439 is not @xmath207-informationally complete by cor .
[ prop : s1s1 ] . as a final result concerning the finite dimensional phase space ,
we show that there is an upper bound on the size of the zero set after which the observable fails to be even @xmath207-informationally complete .
let @xmath567 and denote by @xmath617 the number of @xmath395 s in the binary expansion of @xmath618 .
if a. @xmath619 , or b. @xmath620 , @xmath7 is odd and @xmath621 , or c. @xmath622 , @xmath7 is odd and @xmath623 , then @xmath439 is not @xmath207-informationally complete .
if @xmath624 is as in the statement , then by we have @xmath625 and @xmath439 is not @xmath207-informationally complete by ( * ? ?
* theorem 6 ) .
let @xmath626 , with dual group @xmath627 , pairing @xmath628 and haar measures on @xmath401 and @xmath402 coinciding with @xmath629 the lebesgue measure .. then , the weyl map acts on the hilbert space @xmath630 and is given by @xmath631 for all @xmath632 , where @xmath633 and @xmath634 are the usual @xmath311-dimensional position and momentum operators
. in this case , by it is the support of @xmath635 that is relevant for @xmath636 .
the characterization of the possible supports is a difficult task , but for the possible zero sets @xmath525 this can be done . for any state @xmath437
, @xmath525 is a closed set such that @xmath526 and @xmath527 .
conversely , if @xmath637 is a closed set such that @xmath529 and @xmath530 , then there exists a state @xmath437 such that @xmath531 .
we already remarked that @xmath525 is closed and symmetric , and @xmath533 } > 0 $ ] implies @xmath527 . in order to prove the converse statement ,
suppose first that @xmath534 .
choosing @xmath638 and defining @xmath639 , it is easy to check that @xmath540 . now let @xmath640 be a closed nonempty set such that @xmath529 and @xmath530 .
by @xcite , there exists a probability measure @xmath641 $ ] such that @xmath642 let @xmath550 be such that @xmath643 for all @xmath644 , and define @xmath645 then @xmath438 is positive , nonzero , and satisfies @xmath646 if and only if @xmath647 , i.e. , @xmath531 . in the finite dimensional setting we saw that the cardinality @xmath648 provides a natural way to characterize the state determination properties of the corresponding observables .
in particular , observables having a small zero set are able to perform more tasks than those having larger zero sets .
we see from the next proposition that in the infinite dimensional case a similar natural property is the compactness of the set .
[ prop : infinite_dimension ] a. @xmath649 if and only if @xmath439 is informationally complete . b. if @xmath525 is compact , then @xmath439 is @xmath203-informationally complete .
if @xmath650 is compact , then @xmath439 is not @xmath207-informationally complete .
d. if neither @xmath525 nor @xmath650 is compact , then @xmath439 may or may not be @xmath207-informationally complete .
part ( a ) is just a restatement of prop .
[ prop : covariant_infocomplete ] .
suppose then that @xmath525 is compact . by (
* theorem 2.2 ) , any @xmath468 such that @xmath651 is compactly supported , is necessarily of infinite rank . in particular , by and
compactness of @xmath525 every @xmath461 has @xmath652 .
item ( b ) then follows from theorem [ prop : structure - fin]a .
assume next that @xmath650 is compact and let @xmath653 be such that @xmath654 .
let @xmath655 be such that @xmath656 , and denote @xmath657 .
now define the unit vectors @xmath658 where @xmath659 denotes the characteristic function of the set @xmath428 and @xmath660 is the normalization constant . if @xmath661 , then @xmath662 and we have @xmath663 therefore , if @xmath664 , then for @xmath661 we have @xmath665 hence @xmath666 by , @xmath667 , hence ( c ) holds by prop .
[ prop : prem ] . for part ( d )
we will give two examples in the case @xmath668 .
for the first one , let @xmath653 and choose @xmath669\cup[r,6r ] \right)\times { \mathbb{r}}$ ] .
then a similar argument as before shows that @xmath667 , hence @xmath439 is not @xmath207-informationally complete by prop .
[ prop : prem ] . for the second example
, we refer to ( * ? ? ?
* prop . 9 ) where the authors constructed a state @xmath438 such that @xmath525 is nowhere dense but of infinite lebesgue measure . in other words
, @xmath439 is informationally complete but neither @xmath525 nor @xmath650 is compact . in any realistic measurement
one needs to take into account the effect of noise originating from various imperfections in the measurement setup .
this typically results in a smearing of the measurement outcome distribution which appears in the form of a convolution : if @xmath670 is the probability distribution corresponding to the ideal measurement of @xmath671 , the actually measured distribution is @xmath672 for some probability measure @xmath408 modelling the noise .
the convolution does not affect the covariance properties of the observable and hence the general structure of the observable remains the same .
that is , the actually measured observable is a covariant phase space observable @xmath439 with the smeared fiducial state @xmath673 the inverse weyl transform of @xmath438 now gives @xmath674 . in particular , we have @xmath675 where we have defined analogously @xmath676 .
consider next the special case where @xmath677 so that @xmath671 is informationally complete .
for instance , one may think of the measurement of the husimi @xmath633-function of a state , in which case @xmath678 and @xmath679 is the vacuum , i.e. , the ground state of the harmonic oscillator @xmath680 .
now the overall observable s ability to perform any state determination task is completely determined by the support of @xmath410 . in the specific example with the @xmath633-function we immediately see , e.g. , that any gaussian noise has no effect on the success of the task at hand . however , from prop .
[ prop : infinite_dimension ] we know that any @xmath408 with @xmath681 but with compact @xmath682 results in an observable which is not informationally complete but still allows one to determine any finite rank state under the premise that the rank is bounded by some arbitarily high finite number @xmath4 . finally , if @xmath683 is compact , then even the simplest task of determining pure states among pure states fails .
t.h . acknowledges financial support from the academy of finland ( grant no .
j.s . and a.t .
acknowledge financial support of the italian ministry of education , university and research ( firb project rbfr10coaq ) .
d. t. smithey , m. beck , m. g. raymer , and a. faridina .
measurement of the wigner distribution and the density matrix of a light mode using optical homodyne tomography : application to squeezed states and the vacuum . 70:12441247 , 1993 .
h. rubin and t. m. sellke .
zeroes of infinitely differentiable characteristic functions . in _
a festschrift for herman rubin _ , volume 45 of _ ims lecture notes monogr .
_ , pages 164170 .
statist . ,
beachwood , oh , 2004 . | the purpose of quantum tomography is to determine an unknown quantum state from measurement outcome statistics .
there are two obvious ways to generalize this setting .
first , our task need not be the determination of any possible input state but only some input states , for instance pure states .
second , we may have some prior information , or premise , which guarantees that the input state belongs to some subset of states , for instance the set of states with rank less than half of the dimension of the hilbert space .
we investigate state determination under these two supplemental features , concentrating on the cases where the task and the premise are statements about the rank of the unknown state .
we characterize the structure of quantum observables ( povms ) that are capable of fulfilling these type of determination tasks . after the general treatment we focus on the class of covariant phase space observables , thus providing physically relevant examples of observables both capable and incapable of performing these tasks . in this context , the effect of noise is discussed . |
the purpose of this letter is to discuss the relative motion of a spinning test particle in a gravitational field with respect to a nearby geodesic observer . to compare the theory with observation
, it proves useful to express such relative motion in a quasi - inertial fermi normal coordinate system that can be set up along the worldline of the reference observer @xcite .
this approach has been recently employed for the general motion of a test particle ; in particular , astrophysically significant results have been obtained for ultrarelativistic relative motion , i.e. motion with relative speed above a critical speed given by @xmath0 @xcite .
the present work aims to extend previous results to the case of a spinning test mass .
we choose units such that @xmath1 in the rest of this letter .
the motion of an extended body in a gravitational field can be described using the mathisson - papapetrou - dixon equations @xcite .
if we neglect the quadrupole and higher moments of the particle , then the mathisson - papapetrou ( `` pole - dipole '' ) equations suffice ; these are given by @xmath2 here @xmath3 is the proper time , @xmath4 is the four - velocity , @xmath5 is the four - momentum of the particle and @xmath6 is its spin tensor .
suppose that the particle has mass @xmath7 and spin @xmath8 and is moving in the gravitational field of an astronomical mass @xmath9 .
in general , the dipole interaction must be much smaller than the monopole interaction in our approximation scheme ; thus , the spin - curvature force in equation ( [ eq:1 ] ) , @xmath10 , must be much smaller than the newtonian force @xmath11 .
this means that the mller radius @xcite of the particle , @xmath12 , should be much smaller than the distance @xmath13 between @xmath7 and @xmath14
. equations ( [ eq:1 ] ) and ( [ eq:2 ] ) must be supplemented with additional constraints on the spin tensor . for extended bodies
, the appropriate condition turns out to be @xcite @xmath15 in section [ sec:2 ] , we develop an iterative scheme for the solution of equations ( [ eq:1])-([eq:3 ] ) .
section [ sec:3 ] is devoted to a discussion of the spin - curvature force .
the equations of relative motion are developed in section [ sec:4 ] and specialized to the exterior kerr spacetime in section [ sec:5 ] .
the effects of spin - curvature coupling along the rotation axis of a kerr black hole are of particular relevance to current theoretical speculation on the nature of gamma ray bursts .
we conclude with a brief discussion of our results in section [ sec:6 ] .
it is necessary to develop an iterative approximation method for the solution of equations ( [ eq:1])-([eq:3 ] ) .
we note that these equations imply that @xmath16 and @xmath17 where @xmath8 is the constant magnitude of the spin of the particle . differentiating ( [ eq:3 ] ) with respect to proper time and using ( [ eq:2 ] ) results in @xmath18 next ,
we multiply ( [ eq:5 ] ) by @xmath19 and use the antisymmetry of the spin tensor to get @xmath20 thus either @xmath21 or @xmath22 , where @xmath7 has the interpretation of the constant mass of the spinning particle .
the signature of the spacetime metric is assumed to be @xmath23 throughout this work .
the multipole approximation method discussed in the previous section implies that @xmath5 should be approximately parallel to @xmath24 except for small corrections due to the spin of the particle ; in fact , multiplication of ( [ eq:2 ] ) with @xmath25 results in @xmath26 we therefore assume that @xmath27 and write @xmath28 where the extra term @xmath29 is the small contribution of the spin to the canonical momentum . from ( [ eq:8 ] )
, we find @xmath30 ; on the other hand , multiplying ( [ eq:5 ] ) with @xmath31 results in @xmath32 it thus follows from equations ( [ eq:8 ] ) and ( [ eq:9 ] ) that @xmath33 , where @xmath34 is @xmath35 . differentiating ( [ eq:8 ] ) with respect to @xmath3 and using ( [ eq:1 ] ) we get @xmath36 to determine @xmath29 ,
we substitute ( [ eq:8 ] ) in ( [ eq:5 ] ) to get @xmath37 it is clear from equations ( [ eq:10])-([eq:12 ] ) that @xmath29 is generally small and of order @xmath38 ; moreover , @xmath39 is a force that is of second order in the spin of the particle . in this way , it is possible to develop an iterative scheme based on the small parameter @xmath40 . restricting our treatment to effects that are of first order in the spin , we drop @xmath41 in ( [ eq:11 ] ) and focus on the mathisson - papapetrou spin - curvature force @xmath42 it is useful to define the spin vector @xmath43 of the extended test particle in general as @xmath44 where @xmath45 is the levi - civita tensor and @xmath46 is the alternating symbol with @xmath47 . equation ( [ eq:14 ] )
implies that @xmath48 ; moreover , @xmath49 it follows from the results of this section that to linear order in spin , one can replace @xmath5 by @xmath50 in equations ( [ eq:14 ] ) and ( [ eq:15 ] ) .
consider the linear post - newtonian gravitational field of a source with mass @xmath14 and angular momentum @xmath51 given by @xmath52 where @xmath53 is the newtonian gravitoelectric potential , @xmath54 is the gravitomagnetic vector potential and @xmath55 .
the gravitomagnetic field is given by @xmath56 in analogy with electrodynamics .
if an ideal torque - free test gyroscope is placed at @xmath57 , its precession frequency is given by @xmath58,\label{eq:17}\ ] ] so that the gravitomagnetic field has the interpretation of a precession frequency in conformity with the gravitational larmor theorem .
expressing the metric tensor in ( [ eq:16 ] ) as @xmath59 , we note that @xmath60 , @xmath61 and @xmath62 .
the corresponding riemann curvature tensor is given by @xmath63 imagine a free test particle @xmath64 initially at rest in the stationary gravitational field of the source .
the particle carries an orthonormal tetrad frame @xmath65 , where @xmath66 is its local temporal axis and @xmath67 , @xmath68 , are its local spatial axes .
the curvature tensor as measured by @xmath64 is given by the projection of ( [ eq:18 ] ) on its local tetrad frame .
to calculate this curvature in the linear approximation in the gravitational potentials , one may set @xmath69 ; then , the measured components of the riemann tensor can be expressed as a @xmath70 matrix @xmath71\label{eq:19}\ ] ] with indices that range over @xmath72 . here
@xmath73 and @xmath74 are the electric and magnetic components of the curvature , respectively , and are @xmath75 symmetric and traceless matrices given by @xmath76,\label{eq:21}\end{aligned}\ ] ] where @xmath77 and @xmath78 .
the spin - curvature force experienced by the extended test particle @xmath79 is given by @xmath80 , so that using equation ( [ eq:13 ] ) , @xmath81 where @xmath82 .
more explicitly , this force can be written as @xmath83 which is the gravitational analog of the stern - gerlach force .
as an illustration , let us suppose that @xmath79 is initially at rest on the @xmath84 axis with its spin along the @xmath84 direction , which we take to be the rotation axis of the source ; then , @xmath85 where @xmath86 .
this mathisson - papapetrou force is repulsive ( attractive ) if @xmath79 spins in the same ( opposite ) sense as the rotation of the central source .
moreover , it is interesting to note that this force is always smaller than the newtonian force of attraction ( @xmath87 ) .
indeed , in realistic astrophysical situations , the mathisson - papapetrou dipole force would be very small in comparison with the newtonian monopole force .
this holds in general relativistic situations as well , as demonstrated in section [ sec:5 ] .
if instead of being at rest , @xmath79 is initially boosted with speed @xmath88 along the @xmath84 axis , then the spin - curvature force @xmath89 experienced by the boosted particle can be obtained via a lorentz transformation . in fact , the spin tensor remains invariant under a boost along the @xmath84 axis , so that @xmath90 are the only nonzero components of the spin tensor .
as discussed in the next section , the curvature tensor remains invariant as well . in any case
, the nonzero components of @xmath89 can be simply calculated to be @xmath91 where @xmath92 and @xmath93 is given by equation ( [ eq:24 ] ) .
it should be emphasized that equations ( [ eq:22])([eq:25 ] ) are valid only to first order in the spin of the extended test particle .
similar results have been derived before using the other principal interpretation of the mathisson - papapetrou equations , namely , within the context of a classical point particle with `` intrinsic '' spin ; that is , a point gyro that satisfies instead of ( [ eq:3 ] ) the pirani supplementary conditions @xmath94 @xcite .
imagine a free test `` observer '' @xmath95 following a timelike geodesic in the exterior gravitational field of an astronomical source .
let @xmath96 be an orthonormal tetrad frame that is parallel transported along the worldline of @xmath95 .
a fermi normal coordinate system @xmath97 can be set up in the neighborhood of this worldline based on @xmath98 as the local axes such that @xmath95 remains at the spatial origin of this coordinate system .
thus @xmath95 has fermi coordinates @xmath99 , where @xmath100 is the proper time of @xmath95 .
we need to express the motion of a spinning test particle @xmath79 @xmath101 with respect to @xmath95 .
the equation of relative motion has been derived in general in @xcite and is given by @xmath102 where the four - velocity @xmath24 in fermi coordinates is @xmath103 with @xmath104 .
the requirement that the worldline of @xmath79 be timelike can be expressed as @xmath105 the metric in fermi coordinates is given by @xmath106 where @xmath107 the fermi coordinates are admissible in a cylindrical region with @xmath108 along the reference worldline such that @xmath109 is a certain minimum radius of curvature of spacetime .
in the next section , equation ( [ eq:27 ] ) is studied for the case of motion along the rotation axis of a kerr source .
the motion of spinning test particles in the kerr field has been the subject of previous investigations @xcite ; however , in the present work we consider the deviation of such motion relative to a reference geodesic in the context of a fermi normal coordinate system that is constructed along the reference geodesic .
this approach turns out to be of direct observational relevance for the motion of clumps in astrophysical jets as discussed in section [ sec:6 ] .
the purpose of this section is to study the motion of @xmath79 relative to @xmath95 along the rotation axis of a kerr source .
the kerr metric is given by @xmath110 in boyer - lindquist coordinates , where @xmath111 and @xmath112 . here
@xmath14 and @xmath113 are respectively the mass and the specific angular momentum ( @xmath114 ) of the source .
we assume that the free reference particle @xmath95 moves along the rotation axis on an escape trajectory .
the geodesic equations of motion of @xmath95 reduce to @xmath115 here @xmath116 is a constant of integration such that for @xmath117 is the lorentz factor of the particle as measured by the static inertial observers at spatial infinity .
we integrate system ( [ eq:32 ] ) with the initial conditions that at @xmath118 , @xmath119 and @xmath120 .
moreover , the spinning test particle @xmath79 also starts from this same event with speed @xmath121 relative to @xmath95 and moves along the rotation axis . to describe this relative motion , we establish a fermi coordinate system along the worldline of @xmath95 . in @xmath122 coordinates , the orthonormal tetrad frame @xmath123 is such that @xmath124 where @xmath125 and @xmath126 are given by ( [ eq:32 ] ) .
the axial symmetry about the rotation axis implies that there is a simple rotational degeneracy in the choice of @xmath127 and @xmath128 .
the projection of the curvature tensor on the tetrad frame of @xmath95 , given by equation ( [ eq:30 ] ) , turns out to be independent of the explicit choice for @xmath127 and @xmath128 and may be expressed as before in terms of @xmath73 and @xmath74 such that @xmath129 and @xmath130 , where @xmath131 thus for a given @xmath13 , the curvature measured by @xmath95 is completely independent of @xmath132 .
this remarkable fact is a consequence of the degenerate nature of the kerr solution , which implies that the rotation axis of the exterior kerr spacetime corresponds to two special tidal directions for ingoing and outgoing trajectories ; in fact , @xmath73 and @xmath74 are invariant under a boost along the rotation axis @xcite .
the equation of motion of @xmath79 relative to @xmath95 is given by equations ( [ eq:26 ] ) and ( [ eq:27 ] ) , where we need to evaluate the force term involving @xmath133 and @xmath134 in the fermi system .
to simplify matters , we assume that @xmath135 , which can be determined by equation ( [ eq:34 ] ) , and @xmath136 . as in the previous section ,
the spin of the test mass @xmath79 is assumed to be along the axis of rotation , so that the only nonzero components of the spin tensor are @xmath137 .
it follows that @xmath138 and @xmath139 are the only nonzero components of @xmath140 , where @xmath141 and @xmath142 is the mller radius of @xmath79 . with these simplifications and taking only the terms given in ( [ eq:29 ] ) into account , equation ( [ eq:27 ] ) reduces to @xmath143 this equation takes a dimensionless form if all lengths are expressed in units of @xmath144 and @xmath145 is replaced by @xmath146
. then , @xmath147 by assumption .
the term @xmath148 in equation ( [ eq:35 ] ) is due to the fact that we express the equation of motion of @xmath79 with respect to the fermi time @xmath100 rather than @xmath3 , the proper time of @xmath79 . for @xmath149 , equation ( [ eq:35 ] )
reduces to a jacobi - type equation for the relative motion of @xmath79 with respect to @xmath95 @xcite . to characterize the relative motion invariantly ,
let us imagine a set of static observers in the fermi system situated along the @xmath150 axis and let @xmath151 be the lorentz factor of @xmath79 as measured by these static fermi observers that are in general accelerated .
then @xmath152 , where @xmath153 in our approximation scheme ; therefore , @xmath154 , since @xmath155 .
equation ( [ eq:35 ] ) contains the `` electric '' and `` magnetic '' curvatures @xmath156 and @xmath157 , respectively , that can be expressed as functions of the fermi coordinate time @xmath100 .
that is , the equation for @xmath158 in ( [ eq:32 ] ) can be integrated such that with @xmath159 , we find @xmath160 and hence @xmath161 and @xmath162 that must then be substituted in ( [ eq:35 ] ) .
this system has been numerically integrated with initial conditions that at @xmath163 , @xmath164 and @xmath165 , and the results are presented in figures 1 and 2 , where @xmath151 is plotted versus @xmath166 for ultrarelativistic relative motion along the rotation axis of a maximal @xmath167 kerr source .
we note that with our initial conditions both @xmath168 and @xmath169 at @xmath170 are given by @xmath171 , which is the lorentz factor corresponding to @xmath172 . for motion along the @xmath150 axis , @xmath154 ; however , it turns out that a similar plot for @xmath168 would be indistinguishable from figure [ fig:1 ] .
we choose @xmath173 in figures 1 and 2 , since the results turn out to be independent of the choice of @xmath132 so long as @xmath95 is slowly outgoing . for highly ultrarelativistic motion as in figure [ fig:1 ] , the spin of the clump has a negligible influence on its deceleration .
in fact , with the same conditions as in figure [ fig:1 ] but for @xmath174 and 10 , the resulting figures turn out to be indistinguishable from figure [ fig:1 ] .
let us note that @xmath175 is not allowed here , since the back reaction of such a particle on the gravitational field of the source can not be neglected and hence our test particle approximation scheme would break down .
the case where the spin of @xmath79 is antiparallel to the @xmath150 axis can be treated by formally letting @xmath176 .
the corresponding results are presented in the bottom panel of figure [ fig:2 ] .
the graphs in the top panel are slightly above the corresponding ones in the bottom panel , though the trends are essentially the same .
let us suppose that the clump is homogeneous and cylindrical with radius @xmath177 , and the axis of the cylinder coincides with the @xmath150 axis .
thus @xmath178 , where @xmath179 is the frequency of rotation of the clump such that @xmath180 .
moreover , @xmath181 implies that @xmath182 . in figures [ fig:1 ] and [ fig:2 ] we have @xmath183
; therefore , we have chosen @xmath184 to represent in effect the maximum spin of the clump .
this choice is consistent with various scenarios that have been proposed for jet creation near the poles of a rapidly rotating kerr black hole ( see @xcite , chapters 9 and 11 , and the references cited therein ) .
if the initial lorentz factor @xmath185 of @xmath79 is close to the critical value @xmath186 , then the spin - curvature force is rather small but may not be negligible very close to the central black hole .
however , for @xmath187 , the contribution of the spin - curvature force turns out to be essentially negligible and the spinning particle @xmath79 decelerates toward the critical speed .
this confirms our previous work @xcite on the speed of jets in microquasars @xcite . in this paper
, we have shown that the current speculation @xcite that the spin - curvature coupling could be a prominent driver of astrophysical jets from black holes is overly optimistic .
we show quite generally that the standard tidal effects @xcite will dominate the spin - curvature effects for plasma clumps in astrophysical jets .
synge , relativity : the general theory ( north - holland , amsterdam , 1960 ) . c. chicone , b. mashhoon and b. punsly , int . j. mod
d 13 ( 2004 ) 945 ; + c. chicone and b. mashhoon , class .
quantum grav . 19
( 2002 ) 4231 ; + c. chicone and b. mashhoon , ann .
( leipzig ) 14 ( 2005 ) 290 .
m. mathisson , acta phys . pol . 6
( 1937 ) 167 ; + a. papapetrou , proc .
209 ( 1951 ) 248 ; + w.g .
dixon , phil .
lond . a 277 ( 1974 ) 59 . c. mller , commun .
dublin inst .
adv . stud .
a ( 1949 ) no .
dixon , gen .
. grav . 4 ( 1973 ) 199 .
b. mashhoon , f. gronwald and h.i.m .
lichtenegger , lect .
notes phys .
562 ( 2001 ) 83 .
c. chicone and b. mashhoon , class .
quantum grav . 22
( 2005 ) 195 .
b. mashhoon , j. math .
phys . 12 ( 1971 ) 1075 .
tod , f. de felice and m. calvani , nuovo cimento b 34 ( 1976 ) 365 .
s. b. faruque , phys .
lett . a 327 ( 2004 ) 95 .
d. bini , f. de felice and a. geralico , class quantum grav . 21
( 2004 ) 5441 .
r. plyatsko , phys .
d 58 ( 1988 ) 084031 ; + r. plyatsko and o. bilaniuk , class . quantum grav .
18 ( 2001 ) 5187 ; + r. plyatsko , class . quantum grav . 22 ( 2005 ) 1545 .
o. semerk , mon . not .
r. astron .
( 1999 ) 863 .
beem and p.e .
parker , j. math .
( 1990 ) 819 ; + j.k .
beem and s.g .
harris , gen .
grav . 25 ( 1993 )
939 ; + j.k .
beem and s.g .
harris , gen .
grav . 25 ( 1993 ) 963 ; + g.s . hall and a.d .
hossack , j. math .
34 ( 1993 ) 5897 .
b. mashhoon , phys .
lett . a 163 ( 1992 ) 7 ; + b. mashhoon and j.c .
mcclune , mon . not .
r. astron .
( 1993 ) 881 .
b. punsly , black hole gravitohydromagnetics ( springer - verlag , new york , 2001 ) .
r. fender , lect .
notes phys . 589 ( 2002 )
van putten , astro - ph/0409062 . | the relative motion of a classical relativistic spinning test particle is studied with respect to a nearby free test particle in the gravitational field of a rotating source .
the effects of the spin - curvature coupling force are elucidated and the implications of the results for the motion of rotating plasma clumps in astrophysical jets are discussed . |
compact and regular two dimensional arrangements of oxycyanine dye bound to hydrophobic carbon tails show remarkably narrow absorption and fluorescence peaks [ 1 - 5 ] . in these so called j - aggregates
it is believed that the excitation of a dye molecule becomes delocalized ( exciton ) due to the strong dipole - dipole interactions between delocalized electrons in adjacent dye molecules .
it was also shown [ 3 - 5 ] that these structures can support practically loss - free energy transport over large distances .
molecules of a suitable acceptor ( a molecule similar to that used in j - aggregate , but with thiocyanine instead of oxycyanine ) were able to absorb almost all the energy , even at concentrations as low as 1:10 000 .
it is , however , astonishing , that energy transport is more efficient when the coupling to the acceptor is weak [ 5 ] .
quantum mechanical mechanism of this effect was developed in [ 7,8 ] in terms of exciton trapping by shallow bound state of large spatial extend .
however , the theory assumed ideal excitons , which would imply extremely small absorption width in j - aggregate . clearly a dissipation mechanism is needed to explain the observed experimentally absorption width , which , moreover , should not destroy the delicate exciton trapping needed to explain long range energy transport .
investigation of possible dissipation mechanisms is the aim of this paper .
it will be shown that the usually assumed [ 6 ] exciton - phonon coupling is not up to the task and that only a mechanism destroying excitons will do .
we propose , therefore , a new phonon - exciton interaction , which has the required properties and , moreover , leads to a solvable model .
then we demonstrate that our interaction permits a long range energy transfer in two dimensional structures .
we start with the one dimensional hamiltonian for the excitons [ 6,7 ] .
we have @xmath0,\ ] ] where @xmath1 , @xmath2 are the creation and annihilation operators , respectively , of an excitation at site n , @xmath3 is the excitation energy of a single dye molecule and j is the resonant interaction energy between adjacent sites .
we use one dimensional formalism for simplicity s sake , two dimensional generalization is immediate .
hamiltonian ( 1 ) is diagonalized in momentum space .
introducing bloch waves with quasi momentum k we have @xmath4 our hamiltonian ( 1 ) is now @xmath5 with the dispersion relation @xmath6 again for simplicity we use scalar photons with the coupling to the dye within rotating wave approximation [ 8 ] , so the hamiltonian for the exciton - photon system is @xmath7 where the second integral is the energy of the photons with the usual dispersion relation @xmath8 and the last term is the photon - exciton interaction . note that the sum of exciton and phonon numbers is conserved ( by the virtue of rotating wave approximation ) and that we have delta of momentum conservation along the chain . at the momenta of interest
the molecular formfactor @xmath9 is practically a constant , which we denote @xmath10 , and its momentum dependence is of importance in loop integrals only : it gives the natural line width of an atomic state , much smaller than the experimentally observed width [ 8 ] .
it is important to realize that since the wavelength of visible light is of the order 1000 , this corresponds to momenta of the order @xmath11 @xmath12 , so to a good approximation light produces excitons with negligible momentum , i.e. from the bottom of the band . in this form
the model is analytically solvable . working within one photon - one exciton sector
we use for the wave function the ansatz @xmath13 where @xmath14 is the fock vacuum ( which is an exact eigenstate of the theory ) .
this leads to a system of equations for the unknown functions @xmath15 and @xmath16 @xmath17 where the photon wavefunction @xmath16 is a sum of plane and scattered wave @xmath18 where @xmath19 is the momentum of the incoming light . system ( 9 ) is easily solved and we get for the experimentally observed photon wavefunction @xmath20 as we see , the scattered wave has a cylindrical outgoing wave ( note the one dimensional delta function ) multiplied by scattering amplitude .
the absorption intensity is proportional therefore to @xmath21 where we have introduced the notation @xmath22 for small @xmath23 ( our case ) the real part of @xmath24 is negligible , while the imaginary part is @xmath25 the absorption probability has therefore the familiar lorentzian shape @xmath26 due to the smallness of @xmath10 [ 8 ] , the absorption peak has a finite , albeit very small with . in order to get a more realistic value , we need a dissipation mechanism , i.e. interaction with phonons .
like everybody else we assume that excitons interact with phonons , since the j - aggregate structure is not very stiff .
we take the phonon hamiltonian in the familiar form @xmath27 where @xmath28 , @xmath29 are the ( bosonic ) creation and annihilation operators for phonons with the dispersion relation @xmath30 one usually assumes a trilinear coupling exciton - phonon [ 6 ] @xmath31 however , this coupling does not change the exciton number : this interaction can only push excitons to the bottom of the band , but never destroy them ! since excitons are created with negligible momentum anyway , this mechanism , although certainly present , can not contribute significantly to the observed light absorption width .
we need therefore an interaction which will explicitly destroy excitons , regardless their energy .
the simplest possibility is the annihilation of an exciton into 2 phonons .
after all , the excited dye molecule contains a delocalized electron , which can not only reradiate light , but also give up its energy to the vibrations of the whole structure .
we of course realize that it is difficult for two phonons to carry away 2 - 3 ev of energy , so we call them effective. we therefore propose the interaction in the form @xmath32 where @xmath33 is the formfactor , which , for the sake of simplicity , we assume is constant .
our interaction has the great advantage that it leads to a solvable model .
our excitons coupled to photons and phonons are described by a free part @xmath34 and the interactions @xmath35 one exciton sector is solved by the ansatz @xmath36 which leads to a coupled set of equations @xmath37 where the energy is @xmath38 the scattered photon wave function can be written in the form @xmath39 where we have introduced the function @xmath40 absorption cross section is therefore proportional to @xmath41 remembering that excitons are created with zero momentum we need the function ( 26 ) at k=0 , we have @xmath42 assuming linear dispersion for the phonons ( cf . eq .
16 ) we can perform the integration explicitly @xmath43 the logarithm does not play , numerically speaking , any significant role , therefore we have obtained a familiar lorentzian shape .
its width is given by @xmath44 and depends on @xmath45 and @xmath46 .
since experimentally it is of the order of 0.026 ev , we can estimate @xmath45 if we assume @xmath46 , which we arbitrally have taken to be 4.0 ev .
the @xmath47 parameter gives the position of the absorption peak , taking it to be 3.104 ev we get an excellent agreement with the experiment .
presence of an acceptor molecule , which we place at site @xmath48 , introduces an additional term in the exciton hamiltonian , which now reads @xmath0-(\omega-\omega_{0})a_{0}^{+}a_{0 } , \ ] ] where @xmath49 is the acceptor excitation energy [ 7 ] . in momentum space
this is equivalent to the introduction of a separable potential with constant formfactor @xmath50 @xmath51 and @xmath52 as shown in [ 7,8 ] for small values of lambda , leading to a shallow bound state , we get large absorption , while paradoxically large lambda does induce strong absorption .
the question now is : will the interaction with phonons destroy this delicate mechanism ?
a satisfying feature of our interaction is that the model is still analytically solvable .
the full interaction is now @xmath53 which gives now a set of equations in one exciton - one photon - two phonons sectors ( which are coupled ) @xmath54 again the system can be solved and we get the scattering part of the photon wave function @xmath55 @xmath56 @xmath57 showing the effects of acceptor and phonon interactions in all its gory details .
the last line is the effect of the acceptor .
we will obtain a significant enhancement if this term is small .
as assumed above we can take @xmath58 where @xmath59 is the experimental width .
the whole behavior is therefore controlled by the integral @xmath60 for suitable values of coupling lambda real part of @xmath61 can be canceled , leading to a large enhancement , provided that the imaginary part is reasonably small .
inspection of fig . 1 , where the y axis is at the threshold for exciton production , shows indeed that large values of @xmath62 ( i.e. lambda small ! ) are favored . absorption spectra with @xmath62 small ( fig . 2 ) and @xmath62 ( fig .
3 ) show an enhancement of the order of @xmath63 .
the effect is there , but disappointingly small .
the reason lies in the large value of the imaginary part of sigma ( cf .
fig . 1 ) .
we must notice however that this is a one dimensional model . in two dimensions
we would get formally the same expressions , but all integrations are 2 dimensional .
in particular the function sigma , after angular integration would be @xmath64 and has integration measure @xmath65 , which suppresses small @xmath23 - just what is needed .
the imaginary part at threshold is now strongly reduced ( fig . 4 ) .
the absorption enhancement at small lambda now is of the order of @xmath66 ( fig .
it is seen that the fact that j aggregate is a two dimensional structure is of paramount importance .
we have shown that our postulated mechanism : exciton annihilation into 2 phonons leads to a solvable model .
it describes correctly the width of j - aggregate absorption spectra . in the case of introduced acceptors , the large absorption on shallow bound states is confirmed , but two dimensional geometry seems essential .
of course full two dimensional calculation is needed and will be presented shortly . in order to describe fluorescence spectra
, we have to include the usual energy loss by phonon radiation mechanism [ 6 ] . in conclusion
let us remark that absorption energy on shallow bound states , postulated already in [ 7 ] not only survives the introduction of realistic dissipation mechanisms , but acquires new features allowing for a better agreement with experiment .
g. scheibe , angew .
1937 , vol .
50 , p. 212
e.e jelley , nature 1937 , vol .
139 , p. 631
h. kuhn , j. photochem .
1979 , vol .
10 , p. 111
d. mbius , phys .
1978 , vol .
82 , p. 848
d. mbius and hans kuhn , j. appl .
1988 , vol .
64 , p. 5138
davydov , teorija molekuljarnych eksitonov , mockva 1968 e.a .
bartnik and k.j .
blinowska , phys .
a * 143 * , 448 ( 1982 ) .
bartnik and j.a .
tuszyski , phys .
e * 48 * , 1516 ( 1993 ) . | this paper presents a theoretical model intended to address the question of energy transfer in two - dimensional molecular assemblies such as scheibe aggregates .
a new phonon - exciton interaction is introduced to explain the exciton width in j aggregates .
it is shown that the long range energy transport can occur for weakly interacting acceptors . |
due to the difficulty of obtaining experimental evidence , theoretical and numerical analyses are useful .
four separate scenarios for the pressure temperature ( @xmath10 ) phase diagram have been proposed : \(i ) the _ stability limit _ ( sl )
scenario @xcite hypothesizes that the superheated liquid - gas spinodal at negative pressure re - enters the positive @xmath11 region below @xmath12 . in this view , the liquid state is delimited by a single continuous locus , @xmath13 , bounding the superheated , stretched and supercooled states .
there is no reference to the phase into which the liquid transforms when @xmath14 . as the spinodal is approached , @xmath4 , @xmath5 , and @xmath15 .
a thermodynamic consequence of the sl scenario is that the intersection of the retracing spinodal with the liquid
vapor coexistence line must be a critical point @xcite .
the presence of a such a critical point in the liquid vapor transition , although possible , is not confirmed by any experiment .
this fact poses a serious challenge to the sl scenario .
\(ii ) the _ liquid liquid critical point _
( llcp ) scenario @xcite hypothesizes a first order phase transition line between two liquids a low density liquid ( ldl ) , and a high density liquid ( hdl ) which terminates at a liquid liquid critical point @xmath16 .
hdl is a dense liquid with a highly disordered structure , whereas ldl has a lower density and locally tetrahedral order .
the experimentally observable high density amorphous ( hda ) and low density amorphous ( lda ) solids correspond , in this scenario , to a structurally arrested state of hdl and ldl respectively @xcite .
starting from @xmath16 , the locus of maxima of the correlation length @xmath17 ( the widom line ) projects into the one phase region @xcite .
asymptotically close to the critical point , response functions can be expressed in terms of @xmath17 , hence , these too will show maxima , e.g. , as a function of @xmath18 upon isobaric cooling . these maxima will diverge upon approaching @xmath16 . furthermore , for @xmath19 , the pressure of @xmath16 , the response functions will diverge by approaching the spinodal converging to @xmath16 .
specific models suggest @xcite that @xmath20 , but the possibility @xmath21 has also been proposed @xcite .
\(iii ) the _ singularity free _ ( sf ) scenario @xcite hypothesizes that the low-@xmath18 anticorrelation between volume and entropy is sufficient to cause the response functions to increase upon cooling and display maxima at non zero @xmath18 , without reference to any singular behavior .
specifically , sastry _ et al . _
@xcite consider the temperature of maximum density ( tmd ) line , where the density has a maximum as a function of temperature , and prove a general thermodynamic theorem establishing the proportionality between the slope of the tmd , @xmath22 , and the temperature derivative of @xmath4 .
thus , since the tmd has negative slope in water , it follows that @xmath23 , and therefore @xmath4 must increase upon cooling , whether there exists a singularity or not .
\(iv ) the _ critical point free _ ( cpf ) scenario @xcite hypothesizes an order disorder transition , with possibly a weak first order transition character , separating two liquid phases and extending to @xmath24 down to the superheated limit of stability of liquid water .
this scenario effectively predicts a continuous locus of stability limit spanning the superheated , stretched and supercooled state , because the spinodal associated with the first
order transition will intersect the liquid
gas spinodal at negative pressure .
no critical point is present in this scenario .
these four scenarios predict fundamentally different behavior , though each has been rationalized as a consequence of the same microscopic interaction : the h bond .
a question that naturally arises is whether the macroscopic thermodynamic descriptions are in fact connected in some way .
previous works have attempted to uncover relations between several of the scenarios , for example between ( i ) and ( ii ) @xcite or ( ii ) and ( iii ) @xcite . here
we offer a relation linking all four scenarios showing that ( a ) all four can be included in one general scheme , ( b ) the balance between the energies of two components of the h bond interaction determines which scenario is valid .
morevover , we argue that current values for these energies support the llcp scenario .
we analyze a microscopic model @xcite of water in which the fluid is divided into @xmath25 cells with nearest neighbor ( n.n . ) interactions .
the division is such that each cell is in contact with four n.n .
, mimicking the first shell of liquid water .
the case of a shared h bond , due to more than four molecules in the first shell , is assimilated with the case in which a h bond is broken , since the interaction energy of a shared bond is less than half the energy of a single h bond @xcite .
the goal of the model is to represent , microscopically , the essential features of the interaction among water molecules , while being able to qualitatively understand the importance of each of these features . to this end
the interaction among cells is separated into four distinct components .
the first component of the interaction is due to the short range repulsion of the electron clouds .
this is incorporated into the model by assigning to each cell @xmath26 $ ] ( a ) a volume @xmath27 , where @xmath28 is the exclusion volume per molecule , and ( b ) a maximum of one molecule .
the second component includes all the isotropic long range attractive interactions , such as the instantaneous induced dipole - dipole ( london ) interactions between the electron clouds of different molecules or the isotropic part of the hydrogen bond @xcite .
we refer to this component as the van der waals attractive interaction , keeping in mind , however , that this component includes not only the ( weak ) london dispersion interaction , but also the ( stronger ) isotropic interaction of the hydrogen bond .
the overall sum of the isotropic attractive and repulsive interactions can be represented in different ways .
the one we adopt in a mean field ( mf ) treatment is @xmath29 where if @xmath30 we set the index @xmath31 , and if @xmath32 we set @xmath33 , hence @xmath31 if the density in the cell is gas - like , and @xmath33 if the density in the cell is liquid - like ; @xmath34 is the characteristic energy of the attraction and the sum is over all n.n .
pairs @xmath35 .
the characteristic feature of h@xmath36o is its ability to form h bonds between neighboring molecules .
this interaction has a strong directional component due to the dipole - dipole interaction between the highly concentrated positive charge on each h and each of the two excess negative charges concentrated on the o of another water molecule .
accordingly , the third component incorporated here is this orientational
dependent interaction , which includes the covalent component of the bond @xcite . to account for the orientational degrees of freedom of each water molecule ,
we assign to each cell @xmath37 four bond variables @xmath38 ( one for each n.n .
cell @xmath39 ) , representing the orientation of molecule @xmath37 with respect to molecule @xmath39 .
we choose the parameter @xmath40 by selecting 30@xmath41 as the maximum deviation from a linear bond , i.e. @xmath42 , hence every molecule has @xmath43 possible orientations .
( the effect of choosing a different value for @xmath40 has been analyzed in @xcite . )
we say that a bond is formed between cells @xmath37 and @xmath39 if @xmath44 .
experiments show that formation of the h bonds leads to an open locally tetrahedral structure that induces an increase of volume per molecule @xcite .
this effect is incorporated in the model by considering the total volume to be given as @xmath45 where @xmath46 is the total number of h bonds , @xmath47 if @xmath48 , @xmath49 if @xmath50 , and @xmath51 is the volume increase per h bond @xcite .
bond formation also leads to a decrease in the local potential energy , hence we add to the hamiltonian in eq .
( [ vdw ] ) the term @xmath52 where @xmath53 is the characteristic energy of the directional component of the h bond .
another key experimental fact is that at low @xmath18 the o o o angle distribution in water becomes sharper around the tetrahedral value @xcite , suggesting an interaction that induces a cooperative behavior among bonds . for water , four body and higher order interactions seem to be negligible with respect to the three body term @xcite . hence , the fourth component to the interaction potential is the many
body effect due to h bonds @xcite , which minimizes the energy when the h bonds of nearby molecules assume a tetrahedral orientation .
this is accomplished by further adding to the hamiltonian in eqs .
( [ vdw ] ) and ( [ hb ] ) the term @xmath54 where @xmath55 is the characteristic energy of the cooperative component of the h bond , and @xmath56 indicates one of the six different pairs of the four bond variables of molecule @xmath37 .
this interaction introduces a cooperative behavior among bonds , which may be fine tuned by changing @xmath57 .
choosing @xmath58 leads to h bonds which form independent of neighboring bonds @xcite , while @xmath59 leads to fully dependent bonds @xcite .
the total hamiltonian is now given by @xmath60 this model is studied using both mf analysis and monte carlo ( mc ) simulations @xcite .
details of the mf and mc techniques are available elsewhere @xcite . in the following we adopt @xmath61 , @xmath62 and @xmath63 .
three qualitatively different phase diagrams are found , dependent on the strengths of the h bond energy parameters , @xmath64 and @xmath65 ( fig . [ 1 ] ) . when @xmath66 the model coincides with that proposed in @xcite , which gives rise to the sf scenario ( fig .
[ 1]a ) . for @xmath67
the model displays a liquid liquid transition ending in a llcp at @xmath68 ( fig .
[ 1]b ) @xcite . for @xmath69 , where @xmath70 and @xmath71 are fitting parameters ,
a llcp occurs at @xmath21 ( fig .
[ 1]c ) . for @xmath72 ,
a liquid - liquid transition with no critical point is found , consistent with the cpf scenario ( fig . [ 1]d ) . in fig . [ 2 ] we summarize these results in the @xmath64 vs. @xmath65 parameter space . in the following
we discuss how , by tuning @xmath64 and @xmath65 , we can pass from one scenario to another in a continuous way .
\(i ) by beginning with the llcp scenario , and studying the limit @xmath73 , we find @xmath74 .
moreover , we find that @xmath4 and @xmath6 diverge as @xmath75 for any value of @xmath57 , including @xmath76 and @xmath74 .
further , we find for the entropy @xmath77 that , for any value of @xmath57 , @xmath78 . hence @xmath79 diverges as @xmath75 when @xmath80 .
when @xmath81 ( @xmath66 ) , @xmath5 is constant , as in the sf scenario @xcite .
therefore , the sf scenario coincides with the llcp scenario in the limiting case of @xmath74 , for @xmath73 ( fig . [ 1]a ) .
\(ii ) again , beginning with the llcp scenario , and increasing @xmath65 while keeping other parameters constant , we observe that @xmath16 moves to larger @xmath18 and lower @xmath11 , with @xmath21 for @xmath82 ( fig.[1]c ) .
\(iii ) with further increase of @xmath65 , @xmath16 approaches , and eventually reaches , the liquid gas spinodal . for larger values of @xmath65 only the liquid
liquid transition remains , which is precisely the cpf scenario @xcite ( fig .
hence the cpf scenario differs from the llcp scenario only in that @xmath16 is now inaccessible , lying beyond the region of liquid states .
the same result may be obtained by decreasing @xmath64 , while fixing @xmath65 and other parameters . here a decrease of @xmath83 moves @xmath16 to lower pressure , i.e. towards the liquid spinodal , while the entire liquid - liquid phenomena moves to succesively lower temperature . in all cases ,
the location of @xmath16 varies continuously with variation of @xmath64 and @xmath65 .
\(iv ) in the case of the cpf scenario , we find that the superheated liquid - gas spinodal merges with the supercooled liquid - liquid spinodal , as in ref .
this gives rise to a liquid spinodal which retraces in the @xmath11@xmath18 plane .
this feature resembles the main characteristic of the sl scenario , where the high-@xmath18 liquid has a limit of stability at @xmath24 that retraces toward @xmath84 at low @xmath18 . here
this retracing locus is formed by two spinodal lines , with different signs of the slope , that merge at @xmath24 .
therefore , in the framework of the present model , the cpf scenario and the sl scenario coincide , corresponding to the case in which the cooperative behavior is very strong .
for the cell model , we can derive @xmath85 and @xmath86 here @xmath87 and @xmath88 are constants and , in the mf context , @xmath89 .
symbols @xmath90 , where @xmath91 is @xmath92 or @xmath93 , represent terms of order @xmath94 or higher , that are negligible when @xmath95 .
our mf results confirm the relations in eq .
( [ tc ] ) and ( [ pc ] ) , with @xmath96 and @xmath97 , with negligible @xmath90 terms . therefore , we can rewrite the above relations as @xmath98 , when @xmath63 . as a consequence , for the case @xmath99 , we find @xmath100 , which is exactly what we find numerically in fig . [ 2 ] along the line separating the llcp scenario with @xmath20 ( valid for @xmath101 ) and the llcp scenario with @xmath21 ( valid for @xmath102 ) .
it is possible to show that eq .
( [ pc ] ) can be generalized to @xmath103 , where @xmath104 and @xmath105 are the @xmath18 and @xmath11 along the liquid liquid transition line .
our mf results are in good agreement with this prediction .
we can estimate the equation of the line separating the llcp scenario with @xmath21 and the cpf / sl scenario in the @xmath64@xmath92 plane , by using the eq .
( [ pc ] ) , together with the equation for the liquid gas spinodal .
in particular , we adopt a parametric fit , in terms of the parameter @xmath64 , of the spinodal pressure with respect to the spinodal temperature , and we evaluate the line separating the llcp and cpf / sl scenarios for @xmath106 when @xmath16 is on the spinodal . from this approximate approach , we derive that @xmath107 , with @xmath108 , of the same order of magnitude of the fitting parameter @xmath109 in fig .
yet , @xmath110 , the value of @xmath111 in fig . [ 2 ] , as a consequence of the strong approximations made .
to test the validity of our mf calculations , we perform mc simulations in the @xmath112 ensemble @xcite . to this end ,
\(i ) we consider that the total volume is @xmath113 , where @xmath114 is a dynamical continuous variable ; \(ii ) we assume that the system is homogeneous with all the variables @xmath115 set to 1 ; with this assumption the gas state occurs when @xmath116 ; ( iii ) we replace the isotropic repulsive and attractive terms of the hamiltonian in eq .
( [ ham ] ) with a lennard jones potential , more suitable for continuous distances @xmath117 between particles , with attractive energy @xmath34 plus a hard core repulsion at distance @xmath118 @xmath119 & \text{if $ r > r_0$}.\\ \end{cases } \label{lj}\ ] ] here @xmath120 and @xmath121 is the system dimension @xcite ( the hard core repulsion reduces the computational cost and does not change the phase diagram ) ; the distance between two n.n .
molecules is @xmath122 , and the distance @xmath117 between two generic molecules is the cartesian distance between the centers of the cells in which they are enclosed .
\(iv ) we consider the system in @xmath123 dimensions . while the mf results are valid for any dimension so long as the number of n.n .
molecules is four , the mc results hold for a system with coordination number four and two dimensions . since the results in the two cases are qualitatively comparable , we do not expect a strong dependence of the phase diagram on dimension .
we simulate this system for @xmath124 molecules arranged on a square lattice , adopting wolff s algorithm to equilibrate at low @xmath18 @xcite , for different values of @xmath65 , keeping constant @xmath125 , and @xmath126 ( fig .
[ mc ] ) . for large values of @xmath65 ( @xmath127 ) , we find a hdl ldl first
order phase transition that merges with the superheated liquid spinodal as in the cpf scenario ( fig .
[ mc]a ) . at lower @xmath92 ( @xmath128 ) ,
a hdl ldl critical point appears at @xmath24 , from which emanates the locus of @xmath5 maxima ( used here as an approximation of the liquid liquid widom line ) , which intersects the superheated liquid spinodal ( fig . [ mc]b ) . by further decreasing @xmath65 ( @xmath129 ) , the hdl ldl critical point occurs at @xmath84 , with the line of @xmath5 maxima intersecting the @xmath130 axis ( fig .
[ mc]c ) . for @xmath131 ,
approaching zero , we find that the temperature of the hdl ldl critical point approaches zero and the critical pressure increases toward the value @xmath132 independent of @xmath92 . in this case
, we can show that eq .
( [ pc ] ) still holds , but with @xmath133 .
the line of @xmath5 maxima approaches the @xmath134 axis for @xmath135 .
these results confirm the qualitative behavior found with the mf calculations .
to show that our analysis offers a general framework within which to analyze the supercooled water phase diagram in terms of the interplay between the strengths of the directional contribution to the h bond interaction and its cooperative part , we compare our results with those from other thermodynamic models that can reproduce more than one scenario by tuning appropriate parameters @xcite .
one free energy model with cooperative interactions is the one introduced by tanaka @xcite .
he shows that , as in the sf scenario , water s anomalies are the effect of the excitation of locally favored structures upon cooling , which have lower energy and larger volume than normal - liquid structures . as in our model , in tanaka s model increasing the cooperativity among excitations of locally favored structures leads to the llcp scenario .
moreover , tanaka s model llcp is regulated by relations such as our eq.([tc ] ) and ( [ pc ] ) .
therefore , by increasing the strength of the cooperative interaction , the llcp will eventually reach the limit of stability of the liquid , as in the cpf / sl scenario .
we next consider the free energy model introduced by poole et al .
@xcite , in which a van der waals free energy is augmented to include the effect of h bond formation .
the h bond interaction is characterized by two free parameters : the strength of the h bond , and a geometrical constraint on h bond formation .
the fraction of molecules that form h bonds with decreased energy and entropy is determined by a distribution over molar volumes , the width of which is @xmath136 .
poole et al .
show that , by keeping @xmath136 fixed , their model displays a sl scenario for weak h bond energy , and a llcp at positive pressure for strong h bond energy .
this corresponds in our model to increase the h bond coupling @xmath64 from @xmath137 to @xmath138 , while keeping @xmath139 fixed .
next we study the effect of varying the other h bond parameter in the poole et al .
model , the width @xmath136 . keeping the h bond energy fixed
, we produce the llcp phase behavior at large @xmath136 and the sl phase behavior at small @xmath136 .
hence a decrease of @xmath136 has the same effect on the phase diagram as an increase in the h bond cooperativity in our model .
this result is consistent with that of borick et al .
@xcite for their hamiltonian model that incorporates the cooperativity of h bonds trough the same mechanism used by poole et al . , i.e. by adopting a distribution with width @xmath136 that makes the h bond strength density dependent . by decreasing @xmath136 , borick et al .
find that the llcp moves to lower @xmath11 and higher @xmath18 .
this behavior makes sense physically , as a more all - or - nothing distribution of h bonds ( small @xmath136 ) implies a more cooperative process of bond formation .
it also implies that the models of poole et al . and borick et al
. give rise to the sf scenario only in the limiting case of infinite @xmath136 .
we conclude that all four models give a consistent physical picture .
this suggests that our result , expressed in terms of strength of the directional and cooperative components of the h bond , as summarized in fig .
[ 2 ] , is general .
in the framework of the scheme presented here , in which directionality and cooperativity are the two relvant physical parameters , we propose that the way to understand which scenario best describes water is to probe the energy of the covalent part of the h bond interaction @xcite and the energy of the cooperative component of the h bond interaction @xcite . experiments measure h bonds in ice ih to be approximately 3 kj / mol stronger than in liquid water @xcite . attributing this increase to a cooperative interaction among h bonds @xcite , we can estimate the value of @xmath57 in the cell model to be @xmath140 1.0 kj / mol .
an estimate of the van der waals attraction , based on isoelectronic molecules at optimal separation , yields @xmath141 5.5 kj / mol @xcite .
the optimal h bond energy , @xmath142 , has been measured to be @xmath14023.3 kj / mol @xcite . by considering tetrahedral clusters of h bonded molecules , with h bond and van der waals interactions between n.n .
molecules ( and appropriately reduced van der waals interactions between second and third n.n .
molecules ) , we derive the value for the directional component of the h bond , @xmath143 12.0 kj / mol .
other experimental estimates suggest that breaking the directional component of the h bond requires @xmath140 6.3 kj / mol @xcite .
both estimates from experiments fall within the range of @xmath144 , with @xmath145 , i. e. with @xmath146 . therefore , within our model , these values lead to the llcp scenario with @xmath20 . in particular , mf calculations with @xmath147 , @xmath148 and @xmath149 , predict a llcp at @xmath150 and @xmath151 .
we have shown that a microscopic cell model of water , by taking into account the cooperativity among h bonds , is able to produce phase behaviors consistent with any of the proposed scenarios for water s phase diagram .
it is the amount of cooperativity in relation to the strength of the directional component of the h bond that establishes which scenario holds . for no amount of cooperativity
, the sf scenario is recovered . by increasing the amount of cooperativity in relation to the h bond directional strength , a liquid
liquid transition grows out from the @xmath134 axis , ending in a llcp . with sufficiently strong cooperativity
, this llcp lies beyond the region of stable liquid states , leaving only the liquid
liquid transition , consistent with the cpf scenario . in this case
the spinodal associated with the transition acts as the line predicted in the sl scenario .
comparison with previous models gives consistent results .
hence we argue that each of the four scenarios proposed for the phase diagram of liquid water may be viewed as a special case of our general scheme .
this scheme is based on the assumption that water - water interaction is characterized by an isotropic component , a directional component and a cooperative component , and that h bond formation leads to an open local structure .
alternative mechanisms , based only on isotropic interactions @xcite or only on directional interactions @xcite have been considered and their relevance for the water case is an open question . finally , estimates for the three components of the h bond interaction , based on experimental data , lead to the conclusion that the llcp scenario with a positive critical pressure holds for water .
we thank p. poole , s. sastry , f. sciortino , and f. starr for helpful discussions , and nsf grant che0616489 for support .
thanks the spanish ministerio de ciencia e innovacin grant fis2009 - 10210 ( co - financed feder ) .
xu l , kumar p , buldyrev sv , chen s - h , poole ph , sciortino f , stanley he ( 2005 ) relation between the widom line and the dynamic crossover in systems with a liquid
liquid phase transition _ proc natl acad sci usa _ 102:16558 - 16562 .
borick ss , debenedetti pg , sastry s ( 1995 ) a lattice model of network - forming fluids with orientation - dependent bonding - equilibrium , stability , and implications for the phase - behavior of supercooled water .
_ j. phys .
_ 11:3781 - 3792 .
pedulla jm , vila f , and jordan kd ( 1996 ) binding energy of the ring form of ( h@xmath36o)@xmath152 : comparison of the predictions of conventional and localized - orbital mp2 calculations .
_ j chem phys _ 105:11091 - 11099 .
ohno k , okimura m , akai n , katsumoto y ( 2005 ) the effect of cooperative hydrogen bonding on the oh stretching - band shift for water clusters studied by matrix - isolation infrared spectroscopy and density functional theory .
_ phys chem chem phys _ 7:3005 - 3014 .
xu l , kumar p , buldyrev sv , chen s - h , poole ph , sciortino f , stanley he ( 2005 ) relation between the widom line and the dynamic crossover in systems with a liquid - liquid phase transition .
_ proc natl acad sci usa _ 102:16558 - 16562 . ) , fixed h bond volume increase ( @xmath126 ) , and different values of the h bond cooperativity strength @xmath92 .
( a ) _ singularity - free _ scenario ( @xmath153 ) . at high @xmath18 ,
liquid ( l ) and gas ( g ) phases are separated by a first order transition line ( thick line ) ending at a critical point @xmath154 , from which a l
g widom line ( double dot dashed line ) emanates . in the liquid phase , maxima of @xmath4 and @xmath6
converge to a locus ( dot dashed line ) . at @xmath16 both @xmath4 and @xmath6 and have diverging maxima .
the locus of the maxima is related to the l - l widom line for @xmath155 ( see text ) .
( b ) _ liquid liquid critical point _ scenario with positive critical pressure ( for @xmath156 ) . at low @xmath18 and high @xmath11 ,
a high density liquid ( hdl ) and a low density liquid ( ldl ) are separated by a first order transition line ( thick line with hdl / ldl labeled ) ending in a critical point @xmath16 , from which the l - l widom line ( dot dashed line ) emanates .
other symbols are as in the previous panel .
( c ) _ liquid liquid critical point _
scenario with negative critical pressure ( for @xmath157 ) . here
the l - l widom line ( dot dashed line ) is shown intersecting the l - g spinodal ( dotted line ) .
other symbols are as in the previous panel .
( d ) _ critical point free _ scenario ( @xmath158 ) . the hdl
ldl coexistence line extends to the superheated liquid region at @xmath24 , reaching with the liquid spinodal ( dotted line ) .
the stability limit ( sl ) of water at ambient conditions ( hdl ) is delimited by the superheated liquid to gas spinodal and the supercooled hdl to ldl spinodal ( dashed line ) , giving a re - entrant behavior as hypothesized in the sl scenario .
other symbols are as in the previous panels . in all panels ,
@xmath159 is the boltzmann constant .
, title="fig : " ] ) , fixed h bond volume increase ( @xmath126 ) , and different values of the h bond cooperativity strength @xmath92 .
( a ) _ singularity - free _ scenario ( @xmath153 ) . at high
@xmath18 , liquid ( l ) and gas ( g ) phases are separated by a first order transition line ( thick line ) ending at a critical point @xmath154 , from which a l
g widom line ( double dot dashed line ) emanates . in the liquid phase , maxima of @xmath4 and @xmath6
converge to a locus ( dot dashed line ) . at @xmath16 both @xmath4 and @xmath6 and have diverging maxima .
the locus of the maxima is related to the l - l widom line for @xmath155 ( see text ) .
( b ) _ liquid liquid critical point _
scenario with positive critical pressure ( for @xmath156 ) . at low @xmath18 and high @xmath11 , a high density liquid ( hdl ) and a low density liquid ( ldl )
are separated by a first order transition line ( thick line with hdl / ldl labeled ) ending in a critical point @xmath16 , from which the l - l widom line ( dot dashed line ) emanates .
other symbols are as in the previous panel .
( c ) _ liquid liquid critical point _
scenario with negative critical pressure ( for @xmath157 ) . here
the l - l widom line ( dot dashed line ) is shown intersecting the l - g spinodal ( dotted line ) .
other symbols are as in the previous panel .
( d ) _ critical point free _ scenario ( @xmath158 ) . the hdl
ldl coexistence line extends to the superheated liquid region at @xmath24 , reaching with the liquid spinodal ( dotted line ) .
the stability limit ( sl ) of water at ambient conditions ( hdl ) is delimited by the superheated liquid to gas spinodal and the supercooled hdl to ldl spinodal ( dashed line ) , giving a re - entrant behavior as hypothesized in the sl scenario .
other symbols are as in the previous panels . in all panels
, @xmath159 is the boltzmann constant .
, title="fig : " ] ) , fixed h bond volume increase ( @xmath126 ) , and different values of the h bond cooperativity strength @xmath92 .
( a ) _ singularity - free _ scenario ( @xmath153 ) . at high
@xmath18 , liquid ( l ) and gas ( g ) phases are separated by a first order transition line ( thick line ) ending at a critical point @xmath154 , from which a l
g widom line ( double dot dashed line ) emanates . in the liquid phase , maxima of @xmath4 and @xmath6 converge to a locus ( dot dashed line ) . at @xmath16 both @xmath4 and @xmath6 and have diverging maxima .
the locus of the maxima is related to the l - l widom line for @xmath155 ( see text ) .
( b ) _ liquid liquid critical point _
scenario with positive critical pressure ( for @xmath156 ) . at low @xmath18 and high @xmath11 , a high density liquid ( hdl ) and a low density liquid ( ldl )
are separated by a first order transition line ( thick line with hdl / ldl labeled ) ending in a critical point @xmath16 , from which the l - l widom line ( dot dashed line ) emanates .
other symbols are as in the previous panel .
( c ) _ liquid liquid critical point _
scenario with negative critical pressure ( for @xmath157 ) . here
the l - l widom line ( dot dashed line ) is shown intersecting the l - g spinodal ( dotted line ) .
other symbols are as in the previous panel .
( d ) _ critical point free _ scenario ( @xmath158 ) . the hdl
ldl coexistence line extends to the superheated liquid region at @xmath24 , reaching with the liquid spinodal ( dotted line ) .
the stability limit ( sl ) of water at ambient conditions ( hdl ) is delimited by the superheated liquid to gas spinodal and the supercooled hdl to ldl spinodal ( dashed line ) , giving a re - entrant behavior as hypothesized in the sl scenario .
other symbols are as in the previous panels . in all panels , @xmath159 is the boltzmann constant .
, title="fig : " ] ) , fixed h bond volume increase ( @xmath126 ) , and different values of the h bond cooperativity strength @xmath92 .
( a ) _ singularity - free _ scenario ( @xmath153 ) . at high @xmath18 ,
liquid ( l ) and gas ( g ) phases are separated by a first order transition line ( thick line ) ending at a critical point @xmath154 , from which a l
g widom line ( double dot dashed line ) emanates . in the liquid phase , maxima of @xmath4 and @xmath6
converge to a locus ( dot dashed line ) . at @xmath16 both @xmath4 and @xmath6 and have diverging maxima .
the locus of the maxima is related to the l - l widom line for @xmath155 ( see text ) .
( b ) _ liquid liquid critical point _
scenario with positive critical pressure ( for @xmath156 ) . at low @xmath18 and high @xmath11 , a high density liquid ( hdl ) and a low density liquid ( ldl )
are separated by a first order transition line ( thick line with hdl / ldl labeled ) ending in a critical point @xmath16 , from which the l - l widom line ( dot dashed line ) emanates .
other symbols are as in the previous panel .
( c ) _ liquid liquid critical point _
scenario with negative critical pressure ( for @xmath157 ) . here
the l - l widom line ( dot dashed line ) is shown intersecting the l - g spinodal ( dotted line ) .
other symbols are as in the previous panel .
( d ) _ critical point free _ scenario ( @xmath158 ) . the hdl
ldl coexistence line extends to the superheated liquid region at @xmath24 , reaching with the liquid spinodal ( dotted line ) .
the stability limit ( sl ) of water at ambient conditions ( hdl ) is delimited by the superheated liquid to gas spinodal and the supercooled hdl to ldl spinodal ( dashed line ) , giving a re - entrant behavior as hypothesized in the sl scenario .
other symbols are as in the previous panels . in all panels , @xmath159 is the boltzmann constant .
, title="fig : " ] , ( directional component ) and @xmath92 ( cooperative component ) , both in units of the van der waals energy @xmath160 , obtained from mf calculations .
the ratio @xmath161 is kept constant .
( i ) if @xmath153 ( red line along x axis ) , the singularity free ( sf ) scenario is realized , independent of @xmath64 .
( ii ) for large enough @xmath92 , water would possess a first order liquid
liquid phase transition line terminating at the liquid
gas spinodal the critical point free ( cpf ) scenario ; the liquid spinodal would retrace at negative pressure , as in the stability limit ( sl ) scenario ( yellow region in top left ) .
( iii ) for other combinations of @xmath64 and @xmath92 , water would be described by the liquid liquid critical point ( llcp ) scenario . for larger @xmath92 ,
the llcp is at negative pressure ( ochre region between dashed lines ) . for smaller @xmath92 ,
the llcp is at positive pressure ( orange region in bottom right ) .
dashed lines separating the three different regions correspond to mean field results of the microscopic cell model .
equations for the lines are @xmath100 and @xmath162 , with @xmath163 and @xmath164 .
the @xmath10 phase diagram evolves continuously as @xmath64 and @xmath92 change . ] , @xmath126 and vary the parameter @xmath92 .
( a ) for @xmath165 ( i.e. @xmath166 where @xmath167 and @xmath111 are given in the text and both are approximately 1/3 ) , we find a liquid
liquid phase transition ( thick line with circles ) ending on the superheated liquid - to - gas spinodal ( dotted line with diamonds ) as in the cpf scenario .
( b ) for @xmath168 ( i.e. @xmath169 ) , the liquid
liquid phase transition ends in a liquid liquid critical point ( llcp ) at negative pressure .
( c ) for @xmath156 ( i.e. @xmath170 ) , the llcp ends at positive pressure and the line of specific heat maxima ( crosses ) , emanating from the llcp , is shown only for positive pressure .
errors are of the order of the symbol sizes .
lines are guides for the eyes .
other model parameters are as for mf calculations ( see text).,title="fig : " ] , @xmath126 and vary the parameter @xmath92 .
( a ) for @xmath165 ( i.e. @xmath166 where @xmath167 and @xmath111 are given in the text and both are approximately 1/3 ) , we find a liquid
liquid phase transition ( thick line with circles ) ending on the superheated liquid - to - gas spinodal ( dotted line with diamonds ) as in the cpf scenario .
( b ) for @xmath168 ( i.e. @xmath169 ) , the liquid
liquid phase transition ends in a liquid liquid critical point ( llcp ) at negative pressure .
( c ) for @xmath156 ( i.e. @xmath170 ) , the llcp ends at positive pressure and the line of specific heat maxima ( crosses ) , emanating from the llcp , is shown only for positive pressure .
errors are of the order of the symbol sizes .
lines are guides for the eyes .
other model parameters are as for mf calculations ( see text).,title="fig : " ] , @xmath126 and vary the parameter @xmath92 .
( a ) for @xmath165 ( i.e. @xmath166 where @xmath167 and @xmath111 are given in the text and both are approximately 1/3 ) , we find a liquid
liquid phase transition ( thick line with circles ) ending on the superheated liquid - to - gas spinodal ( dotted line with diamonds ) as in the cpf scenario .
( b ) for @xmath168 ( i.e. @xmath169 ) , the liquid
liquid phase transition ends in a liquid liquid critical point ( llcp ) at negative pressure .
( c ) for @xmath156 ( i.e. @xmath170 ) , the llcp ends at positive pressure and the line of specific heat maxima ( crosses ) , emanating from the llcp , is shown only for positive pressure .
errors are of the order of the symbol sizes .
lines are guides for the eyes .
other model parameters are as for mf calculations ( see text).,title="fig : " ] | four scenarios have been proposed for the low temperature phase behavior of liquid water , each predicting different thermodynamics . the physical mechanism which leads to each is debated .
moreover , it is still unclear which of the scenarios best describes water , as there is no definitive experimental test . here
we address both open issues within the framework of a microscopic cell model by performing a study combining mean field calculations and monte carlo simulations .
we show that a common physical mechanism underlies each of the four scenarios , and that two key physical quantities determine which of the four scenarios describes water : ( i ) the strength of the directional component of the hydrogen bond and ( ii ) the strength of the cooperative component of the hydrogen bond .
the four scenarios may be mapped in the space of these two quantities .
we argue that our conclusions are model - independent . using estimates from experimental data for h bond properties
the model predicts that the low - temperature phase diagram of water exhibits a liquid liquid critical point at positive pressure .
water s phase diagram is rich and complex : more than sixteen crystalline phases @xcite , and two or more glasses @xcite have been reported .
the liquid state also displays interesting behavior , such as the density maximum for 1 atm at @xmath0c .
the volume fluctuations @xmath1 , entropy fluctuations @xmath2 , and cross - fluctuations between volume and entropy @xmath3 , proportional to the magnitude of isothermal compressibility @xmath4 , isobaric specific heat @xmath5 , and isobaric thermal expansivity @xmath6 , respectively , show anomalous increases in magnitude upon cooling @xcite .
further , these quantities display an apparent divergence for 1 atm at @xmath7c @xcite , hinting at interesting phase behavior in the supercooled region .
microscopically , water s anomalous liquid behavior is understood as resulting from the tendency of neighboring molecules to form hydrogen ( h ) bonds upon cooling , with a decrease of local potential energy , decrease of local entropy , and increase of local volume due to the formation of local open structures of bonded molecules .
different models include these h - bond features , but depending on the assumptions and approximations of each model , different conclusions are obtained for the low t phase behavior .
the relevant region of the bulk liquid state can not be probed experimentally , and none of the theories tested , because crystallization of bulk water is unavoidable below the homogeneous nucleation temperature @xmath8 ( @xmath9c at 1 atm ) . |
this paper is a continuation of the ongoing study of cobordism of grassmann manifolds .
let @xmath7 denote one of the division rings @xmath8 of reals , @xmath9 of complex numbers , or @xmath1 of quaternions .
let @xmath10@xmath7 .
then the grassmannian manifold @xmath2 is defined to be the set of all @xmath11-dimensional ( left ) subspaces of @xmath12 .
@xmath2 is a closed manifold of real dimension @xmath13 . using the orthogonal complement of a subspace
one identifies @xmath2 with @xmath14 in @xcite , sankaran has proved that , for @xmath0 or @xmath15 the grassmannian manifold @xmath2 bounds if and only if @xmath16 where , given a positive integer @xmath17 , @xmath18 denotes the largest integer such that @xmath19 divides @xmath20 given a positive integer @xmath4 , let @xmath21@xmath22 denote the set of bordism classes of all non - bounding grassmannian manifolds @xmath2 having real dimension @xmath4 such that @xmath23 the restriction @xmath24 is imposed because @xmath25 and , for @xmath26 @xmath2 bounds .
thus , @xmath27 \in { \frak{n } } _ * \mid nkt = d , k < n,\ { \rm and}\ \nu(n+k ) \le \nu(k ) \ } \subset { \frak{n}}_d.$ ] the purpose of this paper is to prove the following : [ t : gd ] @xmath21@xmath22 is a linearly independent set in the @xmath28-vector space @xmath5
. similar results for dold and milnor manifolds can be found in @xcite and @xcite respectively .
the real grassmannian manifold @xmath29 is an @xmath30-dimensional closed manifold of @xmath11-planes in @xmath31 .
it is well - known ( see @xcite ) that the mod-2 cohomology of @xmath29 is given by @xmath32/ \{w.\bar{w}=1\},\ ] ] where @xmath33 and @xmath34 are the total stiefel
whitney classes of the universal @xmath11-plane bundle @xmath35 and the corresponding complementary bundle @xmath36 both over @xmath37 respectively . for computational convenience in this cohomology
one uses the flag manifold @xmath38 consisting of all ordered @xmath39-tuples @xmath40 of mutually orthogonal one - dimensional subspaces of @xmath31 with respect to the ` standard ' inner product on @xmath31 .
it is standard ( see @xcite ) that the mod-2 cohomology of @xmath38 is given by @xmath41 \bigg/ \left\ { \prod_{i=1}^{n+k } ( 1+e_i ) = 1 \right\},\ ] ] where @xmath42 are one - dimensional classes .
in fact each @xmath43 is the first stiefel whitney class of the line bundle @xmath44 over @xmath38 whose total space consists of pairs , a flag @xmath45 and a vector in @xmath46 .
there is a map @xmath47 which assigns to @xmath48 , the @xmath11-dimensional subspace @xmath49 . in the cohomology ,
@xmath50 @xmath51 is injective and is described by @xmath52 in @xcite , stong has observed , among others , the following facts : [ f : val ] the value of the class @xmath53 on the fundamental class of @xmath29 is the same as the value of @xmath54 on the fundamental class of @xmath38 . [
f : for ] in @xmath55 one has @xmath56 if @xmath57 and the set @xmath58 in particular @xmath59 for each @xmath60 , @xmath61 [ f : top ] in the top dimensional cohomology of @xmath62 a monomial @xmath63 represents the non - zero class if and only if the set @xmath64 the tangent bundle @xmath65 over @xmath29 is given ( see @xcite ) by @xmath66 in particular , the total stifel
whitney class @xmath67 of the tangent bundle over @xmath29 maps under @xmath68 to @xmath69 choosing a positive integer @xmath70 such that @xmath71 we have , using fact [ f : for ] , @xmath72 @xmath73 thus , the @xmath17th stiefel whitney class @xmath74 maps under @xmath68 to the @xmath17th elementary symmetric polynomial in @xmath43 , @xmath75 each with multiplicity @xmath76 , and @xmath77 , @xmath78 each with multiplicity@xmath79 therefore , if @xmath80 denotes the expression of the power sum @xmath81 as a polynomial in elementary symmetric polynomials @xmath82 s in @xmath83 ` unknowns ' @xmath84 , @xmath85 , we have ( see @xcite ) @xmath86 thus we have a polynomial @xmath87 of stiefel
whitney classes of @xmath29 such that @xmath88
it is shown in @xcite that @xmath89 = [ g_k({\bbb{r}}^{n+k})]^4 \quad \text{in}\ { \frak{n}}_{4nk}.\ ] ] from this , we have , in particular , @xmath90=[g_k({\bbb{r}}^{n+k})]^t \quad \text{in}\ { \frak{n}}_{nkt}.\ ] ] for this one has to simply observe that the mod-2 cohomology of the @xmath91-grassmannian is isomorphic as ring to that of the corresponding real grassmannian by an obvious isomorphism that multiplies the degree by @xmath92 .
on the other hand , since @xmath93 is a polynomial ring over the field @xmath6 , we have the following : [ r : ind]a set @xmath94,[m_2 ] , \dots , [ m_m]\}$ ] is linearly independent in @xmath5 if and only if the set @xmath94^{2^\beta},[m_2]^{2^\beta } , \dots , [ m_m]^{2^\beta}\}$ ] is linearly independent in @xmath95 , @xmath96 therefore , noting that @xmath97 , or @xmath98 , it is enough to prove theorem [ t : gd ] for real grassmannians only .
thus , from now onwards , we shall take @xmath99\ \mid \
nk = d,\ k < n,\ \text{and } \ ; { \nu}(n+k ) \leq { \nu}(k ) \}.\ ] ] if @xmath29 is an odd - dimensional real grassmannian manifold then both @xmath100 and @xmath11 must be odd , and so @xmath101 this means that @xmath29 bounds and so it follows that @xmath102 if @xmath4 is odd .
therefore we assume that @xmath4 is even .
[ l : flag ] in @xmath55 one has , for @xmath103 @xmath104 & = e_1^{k-1}e_2^{k-2 } \dots e_{k - j}^j \cdot e_{k-(j-1)}^{n+k - j}\cdot e_{k-(j-2)}^{n+k-(j-1 ) } \dots e_k^{n+k-1}.\end{aligned}\ ] ] note that 1 .
if @xmath105 then the exponent of @xmath43 in the product @xmath106 is greater than or equal to @xmath107 , and 2 . @xmath108 therefore , invoking fact [ f : for ] , the lemma follows.@xmath109 [ p : odd]@xmath73 let @xmath110 \in { \cal{g}}(d)\ \mid \ n+k$ ] is odd @xmath111 .
then @xmath112 is linearly independent in @xmath5 .
arrange the members of @xmath112 in descending order of the values of @xmath76 , so that @xmath113,\ [ g_{k_2}({\bbb{r}}^{n_2+k_2})],\ \dots , \ [ g_{k_s}({\bbb{r}}^{n_s+k_s})]\},\ ] ] where @xmath114 . note that @xmath115 and @xmath116 . for a @xmath4-dimensional grassmannian manifold @xmath29 ,
consider the polynomials @xmath117 @xmath73 of stiefel
whitney classes of @xmath37 where @xmath118 then , for each @xmath119 @xmath120 , we have , using , @xmath121 & = \left(\underset{1 \leq j \leq k_{\ell}}{\prod } \left(\underset{1 \leq i \leq k_{\ell}}{\sum } e_i^{n_{\ell}+k_{\ell}-(2j-1 ) } \right)\right)e_1^{k_{\ell}-1}e_2^{k_{\ell}-2 } \dots e_{k_{\ell}-1}e_{k_{\ell}+1}^{n_{\ell}-1}e_{k_{\ell}+2}^{n_{\ell}-2}\\[.2pc ] & \quad\ \dots e_{n_{\ell}+k_{\ell}-1 } \\[.2pc ] & = e_1^{n_{\ell}}e_2^{n_{\ell}+1 } \dots e_{k_{\ell}}^{n_{\ell}+k_{\ell}-1}e_{k_{\ell}+1}^{n_{\ell}-1 } e_{k_{\ell}+2}^{n_{\ell}-2 } \dots e_{n_{\ell}+k_{\ell}-1 } , \end{aligned}\ ] ] applying lemma [ l : flag ] repeatedly for successive values of @xmath107 . thus , in view of facts [ f : val ] and [ f : top ] , the stiefel
whitney number @xmath122 \rangle \ ; \ne \ ; 0\ ] ] for each @xmath123 , @xmath124 on the other hand , using , it is clear that @xmath125 \rangle \ ; = \ ; 0\ ] ] for each @xmath126 , since @xmath127 therefore , it follows that the @xmath128 matrix @xmath129 \rangle ] _ { 1 \le { \ell}\le s,\ ; 1 \le h \le s}\ ] ] is non - singular ; being lower triangular with 1 s in the diagonal .
this completes theproof.@xmath109 now we shall complete the proof of theorem [ t : gd ] using induction on @xmath4 . first note that @xmath130\ } = \{[{\bbb{r}}p^2]\},\\[.2pc ] { \cal g}(4 ) & = \{[g_1({\bbb{r}}^{4 + 1})]\ } = \{[{\bbb{r}}p^4]\},\end{aligned}\ ] ] and so both are linearly independent in @xmath131 @xmath132 respectively .
assume that the theorem holds for all dimensions less than @xmath4 .
we have @xmath133 , where @xmath134 \in { \cal g}(d ) | n+k\ \ \text{is even}\}\ ] ] and @xmath135 \in { \cal g}(d ) | n+k\ \ \text{is odd}\}.\ ] ] observe that if @xmath136 \in { \cal e}(d)$ ] then both @xmath100 and @xmath11 are even with @xmath137 on the other hand , @xmath138 \
\in \ { \cal e}(d)$ ] if @xmath139 thus , @xmath140 if and only if @xmath141 in view of proposition [ p : odd ] , we may assume without any loss that @xmath142 then , by the above observation and by theorem 2.2 of @xcite every member of @xmath143 is of the form @xmath144 ^ 4 $ ] , where @xmath144 $ ] @xmath145 @xmath146 by induction hypothesis , @xmath147 is linearly independent in @xmath148 .
so , by remark [ r : ind ] , @xmath149 again note that if @xmath136 \ ; \in \ ; { \cal e}(d),$ ] then , by ( [ e : sym ] ) , the polynomial @xmath150 @xmath151 @xmath152 .
so , for each of the polynomials @xmath153 @xmath154 considered in proposition [ p : odd ] , we have @xmath155 \rangle \ ; = \ ; 0.\ ] ] therefore , writing @xmath156 , [ g_{k_{s+2}}({\bbb{r}}^{n_{s+2}+k_{s+2}})],\dots , [ g_{k_{s+q}}({\bbb{r}}^{n_{s+q}+k_{s+q}})]\},\end{aligned}\ ] ] where @xmath157 we see that the @xmath158 matrix @xmath129 \rangle ] _ { 1 \le { \ell}\le s,\ ; 1 \le h \le s+q}\ ] ] is of the form @xmath159}.\ ] ] thus , no non - trivial linear combination of members of @xmath160 can be expressed as a linear combination of the members of @xmath161 this , together with ( [ e : ind ] ) and proposition [ p : odd ] , proves that the set @xmath133 is linearly independent in @xmath5 .
hence , by induction , theorem [ t : gd ] is completely proved . using the decomposition of the members of @xmath143 , and the polynomials @xmath162 , in the lower dimensions together with the _ doubling homomorphism _ defined by milnor @xcite , one can obtain a set of polynomials of stiefel
whitney classes which yield , as in proposition [ p : odd ] , a lower triangular matrix for @xmath143 with @xmath163 s in the diagonal . thus using ( [ m : od ] )
we have a lower triangular matrix , with 1 s in the diagonal , for the whole set @xmath164
part of this work was done under a dst project | this note proves that , for @xmath0 or @xmath1 , the bordism classes of all non - bounding grassmannian manifolds @xmath2 , with @xmath3 and having real dimension @xmath4 , constitute a linearly independent set in the unoriented bordism group @xmath5 regarded as a @xmath6-vector space .
= msam10 at 10pt [ theore]*theorem * [ theore]proposition [ theore]lemma [ theore]definition [ theore]corollary [ theore]remark [ theore]example [ theore]fact |
comparative genomics is nowadays a classical field in computational biology , and one of its typical problems is to cluster sets of orthologous genes that have virtually the same function in several genomes .
a very strong paradigm is that groups of genes which remain `` close '' during evolution work together ( see for instance @xcite , @xcite , @xcite ) .
thus , a widely used approach to obtain interesting clusters is to try to cluster genes or other biological units ( for instance unique contigs of protein domains ) according to their common proximity on several genomes . for this goal ,
many different cluster models have been proposed , like common intervals in @xcite , conserved intervals in @xcite , @xmath8-patterns in @xcite , gene teams in @xcite , domain teams in @xcite , approximate common intervals in @xcite and so on , considering different chromosome models ( permutations , signed permutations , sequences , graphs , etc . ) or different distance models ( accepting gaps , distance modeled as weighted graphs , etc . ) among all those models , the first proposed , and still one of the most used in practice , is the concept of _ common interval _ on genomes represented by permutations .
a set of genes form a common interval of @xmath0 genomes if it appears as a segment on each of the @xmath0 unsigned permutations that represent the genomes .
the orders inside the segments might be totally different .
the model of conserved interval is close to the model of common interval but considers signed permutations .
recently , nested common intervals ( easily extensible to nested conserved intervals ) were introduced in @xcite based on real data observation by @xcite .
a common interval @xmath2 of size @xmath3 is nested if @xmath9 or if it contains at least one nested common interval of size @xmath10 .
@xcite pointed that the nestedness assumption can strengthen the significance of detected clusters since it reduces the probability of observing them randomly .
an @xmath11 time algorithm to compute all nested common intervals between two permutations has been presented in @xcite while between @xmath0 permutations a recent @xmath6 algorithm is proposed in @xcite , where _ nocc _ is the number of solutions . in this paper , we exhibit two simple algorithms to easily compute nested common and conserved intervals of @xmath0 permutations from their natural tree representations . also , with the same simplicity , we propose to deal with a generalization of nested common intervals , called @xmath1-nested common intervals , and with its variant for conserved intervals ( which are a signed version of common intervals ) , called @xmath1-nested conserved intervals .
these new classes allow - as @xmath1 grows - for a less constraint containment between the intervals in the family .
indeed , a nested interval @xmath2 must contain a nested interval of size @xmath10 or be a unit interval .
a @xmath1-nested common ( resp .
conserved ) interval must contain a @xmath1-nested common ( resp . conserved ) interval of size at least @xmath4 or be a unit interval .
nested intervals are indead @xmath12-nested . from a biological point of view , this is equivalent to modeling clusters with a larger variability in gene content and gene order , thus allowing algorithms to deal with annotation errors .
however , the study and validation of this new interval model is deferred to further applied studies . in this paper
we focus on the algorithmic aspects . given a set @xmath13 of @xmath0 permutations on @xmath14 elements representing genomes with no duplicates , our simple algorithms for finding all @xmath1-nested common or conserved intervals of @xmath13 run in @xmath15-time and need @xmath16 additional space , where _ nocc _ is the number of solutions . in this way
, our algorithm for common intervals performs as well as the algorithm in @xcite for the case of @xmath0 permutations , and proposes an efficient approach for the new classes of @xmath1-nested intervals .
moreover , a slight modification of our approach allows us to count the number of @xmath1-nested common or conserved intervals of @xmath0 permutations in @xmath7 time .
efficiently counting the number of @xmath1-nested common intervals without enumerating them is very usefull when one needs to compute similarity functions between genomes that are expressed in terms of number of intervals .
see for instance @xcite .
the paper is organized as follows . in section [
sect : generalities ] we present the main definitions for common and conserved intervals , and precisely state the problem to solve . in section [ sect : pq ] we focus on @xmath1-nested common intervals , recalling the data structure called a @xmath17-tree , giving a characterization of @xmath1-nested common intervals and showing how @xmath17-trees can be used to find all @xmath1-nested common intervals . in section [
sect : conservedtree ] we adopt a similar approach for conserved intervals , with the difference that another tree structure must be used in this case . in section [
sect : conclusion ] we eventually conclude .
a _ permutation _ @xmath18 on @xmath14 elements is a complete linear order on the set of integers @xmath19 .
we denote @xmath20 the identity permutation @xmath21 . an _ interval _ of a permutation @xmath22 is a set of consecutive elements of the permutation @xmath18 .
an interval of a permutation will be denoted by giving its first and last positions , such as @xmath23 $ ] .
such an interval is also said _ delimited _ by @xmath24 ( left ) and @xmath25 ( right ) .
an interval @xmath23 = \{i , i+1 , \ldots , j\}$ ] of the identity permutation will be simply denoted by @xmath26 .
[ def : common - intervals ] let @xmath27 be a set of @xmath0 permutations on @xmath14 elements .
a _ common interval _ of @xmath28 is a set of integers that is an interval in each permutation of @xmath28 .
the set @xmath19 and all singletons ( also called unit intervals ) are common intervals of any non - empty set @xmath13 of permutations .
moreover , one can always assume that one of the permutations , say @xmath29 , is the identity permutation @xmath20 .
for this , it is sufficient to renumber the elements of @xmath29 so as to obtain @xmath20 , and then to renumber all the other permutations accordingly .
then the common intervals of @xmath13 are of the form @xmath30 with @xmath31 .
define now a _ signed _ permutation as a permutation @xmath18 whose elements have an associated sign among @xmath32 and @xmath33 , making each element to be respectively _ positive _ or _
negative_. negative elements are denoted @xmath34 while positive elements are simply denoted @xmath24 , or @xmath35 for emphasizing positivity .
a permutation is then a signed permutation containing only positive elements .
let @xmath36 @xmath37 @xmath38 be a set of signed permutations over @xmath19 , with first element @xmath39 and last element @xmath40 , for each @xmath41 .
assume @xmath42 .
a _ conserved interval _ of @xmath13 is either a unit interval or a common interval @xmath43 of @xmath13 ( ignoring the signs ) which is delimited , in each @xmath44 , either by @xmath45 ( left ) and @xmath46 ( right ) , or by @xmath47 ( left ) and @xmath48 ( right ) .
[ def : conserved ] in the subsequent , we assume that @xmath36 @xmath37 @xmath38 with @xmath49 .
moreover , when we deal with conserved intervals , the permutations are assumed to satisfy the hypothesis in definition [ def : conserved ] .
now , we are ready to introduce the new classes of intervals .
[ def : bnested - intervals ] let @xmath50 be a set of @xmath0 permutations on @xmath14 elements and let @xmath1 be a positive integer .
a common ( respectively conserved ) interval of @xmath13 is _ @xmath1-nested _ if either @xmath9 or @xmath2 strictly contains a common ( resp . conserved ) interval of size at least @xmath4 .
we are interested in efficient algorithms for finding and counting all @xmath1-nested common ( resp
. conserved ) intervals of @xmath50 , without redundancy .
obviously , unit intervals are , by definition , @xmath1-nested common ( resp
. conserved ) intervals .
as a consequence , from now on and without any subsequent specification , we focus on finding @xmath1-nested common ( resp
. conserved ) intervals of size at least 2 .
the following notions will be very useful in the subsequent .
let @xmath27 be a set of @xmath0 permutations on @xmath14 elements and let @xmath1 be a positive integer .
a common ( resp . conserved ) interval of @xmath13 is _
@xmath1-small _ if its size does not exceed @xmath1 .
otherwise , the interval is _
@xmath1-large_. notice that all @xmath1-small intervals are @xmath1-nested , by definition and since unit intervals are @xmath1-nested .
this section is divided into three parts .
the first one recalls a tree structure that we associate to common intervals of permutations , the @xmath17-trees .
the second one discusses the properties of @xmath1-nested common intervals . finally , we give the algorithms for efficiently computing and counting the @xmath1-nested common intervals .
let @xmath51 be a family of intervals from @xmath20 containing the interval @xmath52 .
pq - tree representing the family @xmath53 _ is a tree @xmath54 satisfying : * its nodes are in bijection with a subset @xmath55 of intervals from @xmath53 , the root corresponding to @xmath52 * its arcs represent all the direct ( not obtained by transitivity ) inclusions between intervals in @xmath55 * each node is labeled @xmath18 ou @xmath56 , and an order is defined for the children of each @xmath56-node * an interval @xmath2 of @xmath20 belongs to @xmath53 iff either it corresponds to a node , or there exists a unique @xmath56-node @xmath57 such that @xmath2 is the union of intervals corresponding to successive children of @xmath57 , according to the order defined for @xmath57 .
[ def : pq ] note that the size of the tree is in @xmath58 , thus allowing to drastically reduce the memory space needed to store all the intervals in @xmath53 .
when labels @xmath18 and @xmath56 are forgotten , the tree @xmath54 is called the _ inclusion tree _ of @xmath55 .
given the @xmath17-tree representing a family @xmath53 , we denote by @xmath59 the interval from @xmath55 corresponding to a node @xmath60 .
we also denote , for each interval @xmath2 from @xmath53 , by @xmath61 the _ domain _ of @xmath2 defined as follows .
if @xmath62 , then @xmath61 is the set of its children . if @xmath63 , then by condition 4 . in definition [ def :
pq ] , let @xmath64 be the children of the @xmath56-node @xmath57 such that @xmath65 . then @xmath66 .
fundamental results on @xmath17-trees involve _ closed _ families of intervals . a _ closed _ family @xmath67 of intervals of the permutation @xmath20 is a family that contains all singletons as well as the interval @xmath52 , and that in addition has the following property : if @xmath68 and @xmath69 , with @xmath70 , belong to @xmath67 , then @xmath71 , @xmath72 , @xmath73 and @xmath74 belong to @xmath51 .
the construction of a @xmath17-tree for a closed family of intervals relies on strong intervals : let @xmath53 be a family of closed intervals from @xmath20 .
an interval @xmath30 is said to _ overlap _ another interval @xmath75 if they intersect without inclusion , _
i.e. _ @xmath76 or @xmath77 .
an interval @xmath2 of @xmath67 is _ strong _ if it does not overlap any other interval of @xmath53 , and is _ weak _ otherwise .
[ def : strong ] notice that @xmath52 and the unit intervals are always strong . also , the family of strong intervals of @xmath53 is _ laminar _ ( that is , every two distinct intervals are either disjoint or included in each other ) and , as @xmath52 belongs to the family , it is possible to define for them an inclusion tree
. then it can be shown that : [ th1 ] given a closed family @xmath78 of intervals of @xmath20 , let @xmath55 be the set of strong intervals from @xmath53 and let @xmath54 be the inclusion tree of @xmath55 .
then the @xmath17-tree obtained by the following rules represents the family @xmath53 : * label with @xmath18 each node @xmath60 of @xmath54 such that @xmath79 for all @xmath80 with @xmath81 . * label with @xmath56 each node @xmath82 of @xmath54 not labeled @xmath18 , and define the order @xmath83 of its children such that @xmath84 for all @xmath85 .
common intervals of permutations ( including those of size 1 ) are obviously a closed family of intervals from @xmath20 , thus theorem [ th1 ] applies .
moreover , the @xmath17-tree for common intervals ( hereafter simply denoted @xmath86 ) may be computed in linear time : @xcite the construction of the @xmath17-tree @xmath86 of common intervals of a set @xmath13 of @xmath0 permutations on @xmath14 elements may be done in @xmath7 time .
it is easy to note here that the leaves of @xmath86 are the singletons .
also , the intervals @xmath87 , @xmath88 , associated to the children of a @xmath56-node are contiguous , i.e. @xmath89 .
this is due to condition 4 in the definition of a @xmath17-tree and to the assumption , that the reader must keep in mind , that @xmath49 ( thus all the common intervals are of the form @xmath30 ) .
an example is given in fig .
[ figtotal2b ] .
p4cmlp4 cm & & @xmath90 @xmath91 the @xmath17-tree for the set of common intervals of @xmath28 is shown on the left .
+ let @xmath92 be a set of permutations on @xmath14 elements such that @xmath49 , and let @xmath86 be the @xmath17-tree representing the common intervals of @xmath13 .
say that a common interval @xmath2 is a _
@xmath18-interval _ if it is strong and there is a @xmath18-node @xmath60 with @xmath93 .
otherwise , @xmath2 is a _
@xmath56-interval_. with the aim of identifying the particular structure of @xmath1-nested common intervals among all common intervals , we first prove that : [ lemma : nonnested ] let @xmath2 be a @xmath1-nested common interval with @xmath94 , @xmath95
. then each of the intervals @xmath96 , @xmath97 , is either a @xmath1-small or a @xmath1-nested common interval .
assume a contrario that some @xmath96 is of size @xmath98 and is not @xmath1-nested .
let @xmath99 be a @xmath1-nested common interval with the property that @xmath100 and @xmath101 is minimal with this property .
now , since @xmath102 is @xmath1-nested , we have that @xmath102 strictly contains @xmath96 and thus @xmath103
. then @xmath102 must contain some @xmath1-nested common interval @xmath104 with @xmath105 .
furthermore , @xmath104 and @xmath96 are disjoint since @xmath96 is strong and by the minimality of @xmath102 we have that @xmath104 can not contain @xmath96 .
but then @xmath106 , a contradiction .
it is easy to see that : [ rem : inclusion ] _ let @xmath107 be common intervals such that @xmath108 and @xmath104 is @xmath1-nested with @xmath109 .
then @xmath110 is @xmath1-nested , since @xmath111 .
_ now , the characterization of @xmath1-nested intervals corresponding to a @xmath18-node is obtained as follows .
[ lemma : round ] let @xmath2 be a @xmath18-interval with @xmath112 .
then @xmath2 is a @xmath1-nested common interval if and only if there is some @xmath113 , such that @xmath96 is a @xmath1-nested common interval of size at least @xmath4 .
since @xmath2 is a @xmath18-interval , its maximal common subintervals are @xmath96 , @xmath97 .
the @xmath114 part follows directly from the definition . for the
@xmath115 part , assume by contradiction that the affirmation does not hold
. then none of the intervals @xmath96 , @xmath97 , is @xmath1-nested of size at least @xmath4 , but since @xmath86 is @xmath1-nested we deduce that some interval @xmath96 exists containing a @xmath1-nested common interval @xmath104 of size at least @xmath4 . but
this is impossible according to remark [ rem : inclusion ] .
the structure of @xmath1-nested common intervals given by consecutive children of a @xmath56-node is more complex . in the next lemmas
we show that at most one of the intervals @xmath96 composing such an interval may be @xmath1-large ( see also fig .
[ bstruct ] ) .
[ lemma : firstlast ] let @xmath2 be a @xmath56-interval with @xmath112 . then @xmath2 is a @xmath1-nested common interval if and only if @xmath116 is @xmath1-small and @xmath117 is a @xmath1-nested common interval , or @xmath118 is @xmath1-small and @xmath119 is a @xmath1-nested common interval . recall that for a @xmath56-interval @xmath2 , the order @xmath120 implies that the equation @xmath121 holds for all @xmath122 with @xmath123 .
@xmath115 : since @xmath2 is @xmath1-nested , it contains some @xmath1-nested interval @xmath104 such that @xmath124 . choose @xmath104 as large as possible .
now , @xmath104 can not be strictly included in some non @xmath1-nested @xmath96 by remark [ rem : inclusion ] , thus @xmath125 with @xmath126 , and @xmath127 or @xmath128 .
assume w.l.o.g .
that @xmath129 .
then @xmath116 is @xmath1-small ( since @xmath130 ) and @xmath117 is @xmath1-nested by remark [ rem : inclusion ] since it contains @xmath104 or is equal to @xmath104 .
@xmath114 : let @xmath131 or @xmath132 according to which proposition holds .
we have that @xmath133 since @xmath134
. then @xmath2 is @xmath1-nested .
[ lemma : not2large ] let @xmath2 be a @xmath56-interval with @xmath112 which is a @xmath1-nested common interval .
then at most one of the intervals @xmath96 , @xmath97 , is @xmath1-large , and in this case this interval is a @xmath1-nested common interval . by contradiction ,
assume there exist @xmath1-nested common @xmath56-intervals that contain at least two @xmath1-large intervals of type @xmath96 , and let @xmath2 be a smallest such interval w.r.t .
let @xmath135 ( resp .
@xmath136 ) , with @xmath137 , be such that @xmath138 ( resp .
@xmath139 ) is @xmath1-large and @xmath140 ( resp .
@xmath141 ) is minimum ( resp .
maximum ) with this property . then @xmath142 and @xmath143 , otherwise by lemma [ lemma : firstlast ] the minimality of @xmath2 is contradicted .
but now lemma [ lemma : firstlast ] is contradicted , since @xmath116 and @xmath118 are both @xmath1-large .
then , at most one of the intervals @xmath96 , @xmath97 , is @xmath1-large .
to finish the proof , assume that @xmath96 ( for some fixed @xmath122 ) , is the unique @xmath1-large interval and apply lemma [ lemma : nonnested ] to @xmath96 to deduce that @xmath96 is @xmath1-nested .
-nested @xmath56-interval .
, width=188 ] we are able now to prove the theorem characterizing @xmath1-nested common intervals .
[ th : all ] let @xmath2 be a common interval of @xmath13 .
@xmath2 is @xmath1-nested if and only if : 1 . either @xmath2 is a @xmath18-interval and there exists @xmath144 such that @xmath145 is a @xmath1-nested common interval of size at least @xmath4 .
2 . or @xmath2 is a @xmath56-interval with the property that all intervals @xmath96 with @xmath146 are @xmath1-small , with one possible exception which is a @xmath1-large @xmath1-nested interval .
lemma [ lemma : round ] proves the theorem in the case where @xmath2 is a @xmath18-interval .
when @xmath2 is a @xmath56-interval , lemma [ lemma : not2large ] proves affirmation @xmath147 .
consider algorithm [ algo : bnested1 ] , which computes all @xmath1-nested common intervals .
for a node @xmath148 , the notations @xmath149 and @xmath150 respectively indicate the minimum and the maximum value in @xmath151 .
figure [ fig : algo ] illustrates our algorithm . [ algo : bnested1 ] perform a post - order traversal of @xmath86 algorithm [ algo : bnested1 ] correctly computes all the @xmath1-nested common intervals , assuming the @xmath17-tree is already built , in @xmath152 time , where @xmath153 is the number of @xmath1-nested common intervals in @xmath13 . to show the algorithm correctness , note first that all the leaves are output in step 4 , and they are @xmath1-nested common intervals .
moreover , all @xmath1-nested common intervals corresponding to @xmath18-nodes are correctly output in step 9 according to theorem [ th : all]@xmath154 .
next , @xmath56-intervals corresponding to a @xmath56-node @xmath60 are generated in steps 12 - 22 by starting with each child @xmath148 of @xmath60 , and successively adding right children @xmath155 as long as condition @xmath147 in theorem [ th : all ] is satisfied ( step 15 ) .
let us analyze now the running time .
the @xmath17-tree has size @xmath16 , and the traversal considers every node @xmath60 exactly once .
working once on the children of each node takes @xmath16 .
the test in line 8 considers every child of a @xmath18-node one more time , so that the @xmath16 time is ensured when the @xmath56-interval generation is left apart
. now , during the generation of the @xmath56-intervals , a node @xmath155 that belongs to no @xmath1-nested common interval is uselessly included in some interval candidate at most once by left initial positions for the scan ( beginning line 12 ) , which is in total bounded by @xmath14 since there exists a linear number of initial positions in the @xmath17-tree . at each iteration of the loop line 15 ,
a unique distinct @xmath1-nested interval is output , or @xmath156 ( that happens once for each node since @xmath157 is incremented at each iteration ) , or @xmath158 ( that also happens once for each node since it ends the loop ) .
the total number of iterations is thus @xmath159 , each iteration taking @xmath160 .
the overall running time is thus in @xmath161 where _ nocc _ is the total number of @xmath1-nested common intervals .
-nested intervals of a @xmath56-interval , where @xmath162 ( resp .
@xmath163 ) means @xmath1-small ( resp .
@xmath1-large ) .
the algorithm considers all positions from left to right and expands the @xmath1-nested interval while it is possible.,width=340 ] the previous approach can be modified to count the @xmath1-nested common intervals instead of enumerating them , by simply analyzing more precisely the structure of the @xmath56-nodes .
the goal is to count the @xmath1-nested common intervals in a time proportional to the number of children instead of the number of @xmath1-nested common intervals .
to perform the count , we assume that the traversal is post - order , that it marks each vertex as @xmath1-nested or not , and that it computes the cardinality of each @xmath96 . * @xmath18-nodes . * obviously , a @xmath18-node ( including the leaves ) counts for 1 if the associated interval is @xmath1-nested , and for 0 otherwise
. * @xmath56-nodes . * to make the computation for a @xmath56-node with children @xmath164 @xmath165
, the algorithm looks for the @xmath1-large @xmath1-nested intervals @xmath166 and counts + 1 ) the @xmath1-nested common intervals containing each @xmath166 , and + 2 ) the @xmath1-nested common intervals generated by maximal sets of consecutive @xmath1-small common intervals @xmath96 . to this end ,
the vertices @xmath167 are considered from left to right in order to identify both the @xmath1-large @xmath1-nested common intervals @xmath166 , and the maximal sets of consecutive @xmath1-small common intervals @xmath96 . then : to solve 1 ) , each encountered @xmath1-large @xmath1-nested interval @xmath166 has the following treatment .
we count the number of consecutive @xmath1-small nodes @xmath167 on its right ( resp .
left ) , denoted by @xmath168 ( resp .
@xmath169 ) .
then we compute the number of @xmath1-nested common intervals which contain child @xmath155 as @xmath170 to solve 2 ) , for each maximal set of consecutive @xmath1-small common intervals @xmath96 , assuming it has size @xmath171 , we count @xmath172 one may easily decide whether the interval corresponding to the @xmath56-node itself is @xmath1-nested or not , and compute its size .
all these operations may be performed in @xmath173 time , where @xmath174 is the number of sons of the @xmath56-node .
* example .
* on the example in fig .
[ fig : algo ] , we count ( from left to right ) : ( a ) for the first @xmath1-large strong child to the left : @xmath175 @xmath1-nested common intervals ; ( b ) for the @xmath1-large second strong child : @xmath176 @xmath1-nested common intervals ; ( c ) for the last @xmath1-large strong child : @xmath177 @xmath1-nested common intervals .
we sum up to obtain @xmath178 @xmath1-nested common intervals containing one @xmath1-large @xmath1-nested interval @xmath96 .
now we add the nested intervals generated between the @xmath1-large intervals : @xmath179 .
altogether , the @xmath56-node of figure [ fig : algo ] generates @xmath180 @xmath1-nested intervals .
[ [ complexity ] ] complexity + + + + + + + + + + the time complexity of the counting procedure is obviously @xmath181 the size of the underlying @xmath17-tree .
the time needed to get the pq - tree itself given @xmath0 permutations is however @xmath7 , as indicated before .
as stated before , we assume the set @xmath13 of permutations has the properties required in definition [ def : conserved ] . as conserved intervals are common intervals , one may be tempted to follow the same approach using @xmath17-trees .
unfortunately , the inclusion tree of strong conserved intervals does not define a @xmath17-tree representing the family of conserved intervals , due to the fact that a conserved interval can not be written as a disjoint union of strong conserved intervals .
the resulting _ ordered tree _ has been used in the literature @xcite , but its underlying properties have not been clearly stated .
we do it here , before using these properties .
we start by an easy property about intersection of conserved intervals : [ lemstru2 ] let @xmath182 and @xmath183 be two conserved intervals of @xmath13 with @xmath184 .
then the intervals @xmath185 , @xmath186 , @xmath187 and @xmath188 are conserved intervals . as an element @xmath60 from @xmath2 fulfills @xmath189 and
an element @xmath82 from @xmath104 fulfills @xmath190 then an element @xmath57 from @xmath191 fulfills @xmath192 .
these elements are exactly those between @xmath140 and @xmath46 and @xmath185 is thus a conserved interval .
similarly , @xmath193.the elements from @xmath194 are all the elements not lower than @xmath141 and not larger than @xmath46 , so @xmath186 is a conserved interval . finally , the elements from @xmath195 are all elements greater than @xmath140 and smaller than @xmath157 , so @xmath188 is a conserved interval .
the notions of strong / weak intervals and of a frontier are essential in our study .
a conserved interval @xmath2 of @xmath13 is strong if it has cardinality at least two , and does not overlap other conserved intervals .
otherwise , it is _
weak_. notice that unit conserved intervals are not strong .
let @xmath196 be a conserved interval .
a set @xmath197 of elements satisfying @xmath198 is a _ set of frontiers _ of @xmath2 if @xmath199 is a conserved interval , for all @xmath200 with @xmath201 .
an element of @xmath2 is a _ frontier _ of @xmath2 if it occurs in at least one set of frontiers of @xmath2 .
the two following properties are easy ones : [ lemstru1 ] let @xmath196 be a conserved interval and @xmath202 be a set of frontiers of @xmath2 .
the elements of @xmath202 are either all positive or all negative . by the definition of a conserved interval , its two endpoints
have the same sign . by the definition of a set on frontiers , any two frontiers are the extremities of some conserved interval . [ unionfrontier ]
let @xmath196 be a conserved interval , and let @xmath203 and @xmath204 be two sets of frontiers of @xmath2 .
then @xmath205 is also a set of frontiers of @xmath2 .
we show that any interval between two elements of @xmath206 is conserved .
let @xmath207 and @xmath208 , and suppose that @xmath209 . if @xmath210 , then we have @xmath211 and we are done . if @xmath212 , then @xmath213 and we are also done . if @xmath214 and @xmath215 , then lemma [ lemstru2 ] allows to conclude .
the same proof holds if @xmath216 .
let now @xmath86 be the inclusion tree of strong intervals from @xmath13 , in which every node @xmath60 corresponds to a strong interval denoted @xmath59 , and node @xmath60 is the parent of node @xmath82 iff @xmath59 is the smallest strong conserved interval strictly containing @xmath217 .
then @xmath86 contains two types of nodes : those corresponding to strong conserved intervals with no internal frontier , and those corresponding to strong conserved intervals with at least one internal frontier .
we will show that weak conserved intervals are the conserved strict subintervals of the latter ones , defined by two frontiers .
overall , we have a structure working pretty much as a @xmath17-tree , but which can not be mapped to a @xmath17-tree .
this is proved in the next theorem .
given a conserved interval , denote by @xmath218 the smallest strong conserved interval such that @xmath219 .
[ theostruc2 ] each conserved interval @xmath2 of @xmath13 admits a unique maximal ( w.r.t . inclusion ) set of frontiers denoted @xmath220 .
moreover , each conserved interval @xmath2 of @xmath13 satisfies one of the following properties : * @xmath2 is strong * @xmath2 is weak and there exists a unique strong conserved interval @xmath104 of @xmath13 , and two frontiers @xmath221 with @xmath222 , such that @xmath223
. moreover @xmath224 and @xmath225 let us first prove the uniqueness of the maximal set of frontiers
. by contradiction , assume two distinct maximal sets of frontiers @xmath226 and @xmath227 exist for a conserved interval @xmath2 .
using lemma [ unionfrontier ] , we deduce that @xmath206 is a larger set of frontiers of @xmath2 , a contradiction .
let us now prove that if @xmath2 is not strong , then there exists a unique strong conserved interval @xmath104 , and two frontiers @xmath221 with @xmath222 , such that @xmath223 . _ existence of @xmath104 .
_ let @xmath228 be a weak conserved interval of @xmath13 .
then there is another interval @xmath229 overlapping it , either on its left ( i.e. @xmath230 with @xmath231 ) or on its right ( i.e. @xmath232 with @xmath233 ) . according to lemma [ lemstru2 ] , @xmath234 is a conserved interval , and @xmath235 is a set of frontiers for @xmath236 ( not necessarily all distinct ) . if @xmath236 is not strong then it is overlapped by another interval @xmath237 .
we build an increasing sequence of intervals @xmath238 , with @xmath239 overlapped by @xmath240 , and @xmath241 , until we find a strong conserved interval @xmath242 ( the process ends since @xmath52 is strong ) .
each time lemma [ lemstru2 ] ensures that @xmath243 is a conserved interval , and @xmath244 is a set of frontiers for @xmath243 ( where @xmath245 are the endpoins of @xmath240 ) .
but then @xmath246 is a subset of the maximal set of frontiers @xmath247 of @xmath242 and the two frontiers of @xmath242 defining @xmath248 are @xmath249 , since @xmath250 .
_ uniqueness of @xmath104 .
_ assume a contrario that two strong intervals @xmath251 and @xmath236 exist with @xmath252 , @xmath253 and such that @xmath254 with @xmath255 and @xmath256 .
then , clearly , @xmath257 and this is the left endpoint of @xmath2 , whereas @xmath258 and this is the right endpoint of @xmath2 . since @xmath251 and @xmath236 are both strong , they can not overlap .
assume then w.l.o.g .
that @xmath251 is strictly included in @xmath236 , and more precisely that @xmath259 ( then @xmath260 due to inclusion ) .
now , @xmath261 overlaps @xmath251 ( which is forbidden ) unless @xmath262 , in which case we have @xmath263 and this is the left endpoint of @xmath2 .
now , since @xmath2 is strictly included in @xmath104 , we deduce @xmath264 and @xmath265 overlaps @xmath251 .
this contradiction proves the uniqueness of the strong interval @xmath104 satisfying condition 2 .
let us prove that @xmath224 .
suppose by contradiction that @xmath220 contains a frontier @xmath266 not in @xmath267 .
recall that @xmath268 and @xmath269 , @xmath270 belong to @xmath267 .
now , for each @xmath271 with @xmath272 , we have that @xmath273 is a conserved interval either by lemma [ lemstru2 ] applied to @xmath274 and @xmath275 ( when @xmath276 ) or since it is the union of the two conserved intervals @xmath277 and @xmath278 ( when @xmath279 ) .
symmetrically , @xmath280 is also a conserved interval when @xmath271 , @xmath281 .
but then @xmath282 is a set of frontiers of @xmath104 larger than @xmath267 , which contradicts the maximality of @xmath283 eventually , let us prove that @xmath225 assume that @xmath284 @xmath285 , that @xmath223 and that , by contradiction , there is a smallest strong conserved interval @xmath286 which contains @xmath2 .
then @xmath287 since otherwise @xmath104 and @xmath288 would overlap . w.l.o.g .
assume that @xmath289 .
then @xmath290 overlaps @xmath288 , and this contradicts the assumption that @xmath288 is strong .
we only need three more results before dealing with @xmath1-nested conserved intervals .
[ memecontainer ] let @xmath2 and @xmath104 be two conserved intervals .
if there exists a frontier @xmath291 , then @xmath218 @xmath292 @xmath293 since both @xmath218 and @xmath294 are strong , and since they share a common element , one contains the other . now , suppose w.l.o.g . that @xmath295 .
by contradiction , assume that @xmath296 and let @xmath297 .
since @xmath298 and that by theorem [ theostruc2 ] we have @xmath299 , we deduce that @xmath300 and thus @xmath278 and @xmath301 are conserved intervals . now , one of them necessarily overlaps @xmath218 , since at least one of @xmath302 is not an endpoint of @xmath218 .
but this is impossible . therefore @xmath218
@xmath292 @xmath293 [ coro1 ] let @xmath2 be a ( weak or strong ) conserved interval with frontier set @xmath303 and let @xmath304 be a conserved interval .
then exactly one of the following cases occurs for the interval @xmath43 : 1 .
either there exists @xmath305 such that @xmath306 , and then there exist @xmath122 and @xmath307 such that @xmath308 .
2 . or @xmath43 contains no frontier of @xmath220 , and then there exists @xmath122 such that @xmath309 . obviously , the two cases can not hold simultaneously . moreover , in case 2 .
, the deduction is obvious .
let us focus now on the case 1 .
let @xmath43 contain some frontier @xmath310 .
consider now the two intervals @xmath275 and @xmath311 then either _
@xmath312 , or _
( ii ) _ @xmath313 , or _ ( iii ) _
@xmath314 or @xmath315 , or ( iv ) @xmath43 overlaps @xmath275 or @xmath316 .
the two first cases are trivial , let us consider the two last cases . in case _
assume w.l.o.g .
that @xmath314 ( the case @xmath315 is symmetric ) .
then @xmath317 since @xmath304 . since @xmath43 contains the elements of @xmath2 that are smaller or equal to @xmath46 , then @xmath318 is a conserved interval .
thus @xmath319 is a frontier set of @xmath2 .
if @xmath320 then according to lemma [ unionfrontier ] we have that @xmath321 is a frontier - set of @xmath2 contradicting the maximality of @xmath220 .
we deduce that there exists @xmath307 such that @xmath322 .
now we are done , since @xmath323 . in case _
( iv ) _ , @xmath43 is necessarily weak since it overlaps another conserved interval
. w.l.o.g .
we assume that @xmath275 overlaps @xmath324 let us first prove that @xmath325 is also a frontier of @xmath43 .
indeed , assume _ a contrario _ that @xmath326 , and denote @xmath327 . with @xmath328 , we have either @xmath329 and by lemma [ lemstru2 ] for @xmath275 and @xmath330 we deduce that @xmath331 is conserved , or @xmath332 and then lemma [ lemstru2 ] for @xmath333 and @xmath334 we deduce that @xmath335 is conserved .
then @xmath336 is a set of frontiers of @xmath43 , which contradicts the maximality of @xmath337 .
now , by lemma [ memecontainer ] for @xmath338 ( whose frontier set contains @xmath325 by theorem [ theostruc2 ] ) and @xmath218 we deduce that @xmath339 . thus using theorem [ theostruc2 ]
, we conclude that @xmath340 for some @xmath341 .
but since @xmath304 and @xmath342 , we are done .
the following theorem ensures that in the inclusion tree @xmath86 of the strong conserved intervals of @xmath13 , weak intervals are exactly the intervals extending between two frontiers of a strong interval .
moreover , each weak interval is uniquely represented in such a way .
in addition , the computation of the tree and of all the frontier sets is linear .
let @xmath86 be the inclusion tree @xmath86 of strong conserved intervals of a set @xmath13 of permutations .
then : 1 .
a conserved interval @xmath2 of @xmath13 is weak if and only if there exists a strong interval @xmath104 of @xmath13 and two frontiers @xmath221 such that @xmath223 . moreover , in this case @xmath104 is unique . 2 .
for each strong conserved interval @xmath2 of @xmath86 with parent @xmath104 in @xmath86 , there is a unique conserved interval @xmath343 defined by successive frontiers in @xmath267 such that @xmath344 .
3 . the tree @xmath86 , the maximal set of frontiers @xmath267 of each strong conserved interval @xmath104 and the interval @xmath345 of each strong conserved interval @xmath346 may be computed in global @xmath16 time and @xmath16 space .
[ th : arbret ] concerning affirmation 1 , the
part is deduced directly from theorem [ theostruc2 ] , whereas the @xmath114 part is ensured by the definition of a set of frontiers .
again by theorem [ theostruc2 ] , we deduce the uniqueness of @xmath104 .
affirmation 2 results from lemma [ coro1 ] .
according to affirmation 1 in this lemma , @xmath2 can not contain a frontier of @xmath104 , since otherwise @xmath2 would be of the form @xmath199 , with @xmath347 , and thus would not be strong .
thus , by affirmation 2 in lemma [ coro1 ] , we deduce the existence of @xmath345 , which is necessarily unique by the definition of the frontiers .
we focus now on affirmation 3 . in @xcite , a conserved interval
is called _ irreducible _ if it can not be written as the union of smaller conserved intervals .
it is easy to notice that the set of irreducible intervals of size at least two is exactly composed of the intervals @xmath348 , where @xmath349 are two consecutive frontiers of a strong conserved interval of @xmath13 .
indeed , affirmation 1 shows that the only irreducible weak conserved intervals @xmath223 are those for which @xmath350 , and obviously the only irreducible strong intervals @xmath2 are those with @xmath351 , which are of the form @xmath352 , with @xmath353 . to show affirmation 3
, we notice that the number of irreducible intervals is in @xmath16 [ @xcite ] , and that they may be computed in @xmath16 time and space for an arbitrary number @xmath0 of permutations using generators from @xcite .
knowing irreducible conserved intervals , the computation of strong intervals , of their set of frontiers , as well as that of the tree is quite easy .
first , one must plot on the identity permutation the @xmath16 irreducible intervals of size at least two , by marking the left and right endpoint of each such irreducible interval .
notice that each element @xmath354 of the permutation @xmath20 has at most two marks , the equality occurring only when @xmath354 is an internal ( that is , different from an endpoint ) frontier of a strong interval .
we assume the right mark of @xmath354 ( when it exists ) always precedes the element @xmath354 whereas the left mark ( when it exists ) always follows the element @xmath354 , so that a left - to - right traversal of @xmath20 allows to close the interval with right endpoint @xmath354 before opening the interval with left endpoint @xmath354 . replacing left and right marks with respectively ( square ) left and right brackets indexed by their corresponding element @xmath354 on @xmath20
, we obtain an expression @xmath355 which has correctly nested brackets , since irreducible intervals may only overlap on one element . moreover , if @xmath2 and @xmath104 are strong intervals such that @xmath104 is the parent of @xmath2 , then @xmath356 and thus these intervals are closed exactly in this order during a left - to - right traversal of the expression @xmath355 . the expression
@xmath355 then allows , during a left - to - right traversal , to discover the strong intervals according to a post - order traversal of @xmath86 , which is built on the fly . a strong conserved interval @xmath104 is obtained by chaining as long as possible neighboring irreducible intervals , i.e. such that the right bracket of an interval is followed by the endpoint @xmath354 of the interval and by the left bracket of the next interval .
its frontiers are given by the endpoints of the chained irreducible intervals . also , since @xmath357 for all strong intervals @xmath2 , it is easy to identify @xmath345 since it is the interval which closes immediately after @xmath2 during the traversal .
* example .
* let @xmath358 , where @xmath359 . the strong intervals are @xmath360 ( with frontier set @xmath361 ) , @xmath362 ( with frontier set @xmath363 ) and @xmath364 ( with frontier set @xmath365 ) . to build @xmath86 , we first plot the irreducible intervals , i.e. @xmath366 and @xmath367 on @xmath29 , and we obtain the expression : @xmath368_2\,\ , -2\,\ , ] _
4\,\ 4\,\ , [ _ 4 \,\,\,]_5\,\ , 5\,\ , [ _ 5\,\ , -8\,\ , [ _ 8 \,\,\,]_7\,\ , -7\,\ , [ _ 7 \,\,\,]_6\,\ , -6\,\ , ] _ 9\,\ , 9\,.\ ] ] a left - to - right traversal of @xmath355 allows to find first the interval @xmath362 which is included in @xmath369 , and thus the node @xmath362 of @xmath86 is built , and a node @xmath60 starting with @xmath369 is created and defined as the parent of @xmath362 .
next , the interval @xmath369 in @xmath60 is continued with @xmath370 ( just change the interval inside the already existing node ) , and continued with another interval @xmath371 , where @xmath372 is not yet known .
still , @xmath367 is discovered as a subinterval of @xmath371 , and it may be continued with @xmath373 , thus creating together the interval @xmath364 which is another child of @xmath60 .
once this is done , we read @xmath374 which indicates that @xmath375 .
thus @xmath86 has three nodes , the root @xmath360 ( with frontier set @xmath361 discovered during the traversal ) , and its two children @xmath362 ( with frontier set @xmath363 ) and @xmath364 ( with frontier set @xmath365 discovered during the traversal ) .
recall that by definition all @xmath1-small conserved intervals are @xmath1-nested conserved intervals .
the characterization below of @xmath1-nested conserved intervals has some similarities with that of @xmath1-nested common intervals represented in the @xmath17-tree as @xmath56-intervals ( see affirmation @xmath147 in theorem [ th : all ] ) .
let @xmath104 be a conserved interval and let @xmath376 be its maximal set of frontiers .
we say that @xmath104 _ contains a @xmath1-_gap _ at position @xmath305 _ if @xmath377 is @xmath378-large .
furthermore , we say that a conserved interval @xmath43 _ falls in the gap _ between @xmath325 and @xmath379 of @xmath104 if @xmath380 and @xmath338 has the parent @xmath294 .
in other words , @xmath104 is represented in @xmath86 by the node @xmath294 and @xmath43 is represented by the node @xmath338 , in such a way that the former node is the parent of the latter one .
a @xmath1-gap at position @xmath305 is said _ good _ if it contains at least one @xmath1-nested strong conserved interval @xmath2 with @xmath381 , or equivalently @xmath382 .
then a good @xmath1-gap is a @xmath1-nested conserved interval .
[ thgap ] a conserved interval @xmath104 of @xmath13 is @xmath1-nested if and only if it contains no @xmath1-gap , or if it contains exactly one good @xmath1-gap @xmath383 .
we first prove the @xmath114 part .
let @xmath104 be a conserved interval with maximal set of frontiers @xmath384 .
if @xmath104 contains exactly one @xmath1-gap , let it be at position @xmath305 .
otherwise , let us fix arbitrarily @xmath314 . in both cases ,
@xmath377 is a conserved @xmath1-nested interval : either because it is @xmath1-small , or because it is a good @xmath1-gap .
now , for all @xmath385 in increasing order , we deduce by induction that @xmath386 is a conserved interval ( by definition of the set of frontiers ) containing @xmath387 , and thus it is @xmath1-nested .
thus @xmath316 is a conserved @xmath1-nested interval .
similarly , we deduce by induction that @xmath388 is a conserved @xmath1-nested interval , for all decreasing values of @xmath389 .
thus @xmath390 is @xmath1-nested .
now let us prove the @xmath115
part , by proving ( a ) that an conserved interval with two @xmath1-gaps or more is not @xmath1-nested , and ( b ) that if a @xmath1-nested conserved interval contains one @xmath1-gap then this @xmath1-gap is good .
proof of ( a ) .
assume by contradiction that @xmath104 with @xmath391 is conserved @xmath1-nested and contains two @xmath1-gaps at initial positions @xmath305 and @xmath392 ( @xmath393 ) .
the @xmath1-nestedness of @xmath104 implies the existence of a @xmath1-nested conserved interval @xmath394 such that @xmath395 .
now , @xmath396 , otherwise @xmath288 misses at least @xmath397 elements from @xmath377 ( which is @xmath378-large ) and thus @xmath398 , a contradiction .
similarly , @xmath399 .
we deduce that @xmath288 contains both @xmath379 and @xmath400 , as well as at least one additionnal element on the left of @xmath379 and one additionnal element on the right of @xmath400 . by lemma [ coro1 ] with @xmath401 and @xmath402
we have that @xmath403 with @xmath404 and @xmath405 .
thus the existence of the @xmath1-nested interval @xmath104 containing the two @xmath1-gaps implies the existence of a smaller @xmath1-nested interval @xmath288 still containing the two @xmath1-gaps
. now assume a maximal size series @xmath406 of @xmath1-nested conserved intervals has been built similarly , each interval being strictly included in the previous one , all containing the two @xmath1-gaps .
such a series ends with @xmath407 , since otherwise ( if the last interval is larger ) it is possible to construct a smaller @xmath1-nested interval included in the last interval and containing @xmath408 thus @xmath407 is @xmath1-nested .
but this is not possible , as it can not strictly contain another @xmath1-nested common interval of size at least @xmath409 .
as before , such an interval needs to contain the two gaps , and it is therefore not strictly included in @xmath408 proof of ( b ) .
assume now that @xmath104 with @xmath391 is conserved @xmath1-nested and has a unique @xmath1-gap situated at position @xmath305 .
as before , @xmath104 must contain a @xmath1-nested conserved interval @xmath288 with @xmath410 and then @xmath396 implying by lemma [ coro1 ] that @xmath288 must contain @xmath377 .
the smallest interval obtained following the same reasoning is then @xmath377 itself , which must be @xmath1-nested . as this interval has no internal frontiers
( otherwise @xmath411 would not be maximal ) , any @xmath1-nested common interval @xmath196 of size at least @xmath412 it contains satisfies @xmath413 .
if @xmath2 is strong then we are done , otherwise @xmath218 is strong and has all the required properties .
theorem [ th : arbret ] allows to count and to enumerate efficiently all @xmath1-nested conserved intervals .
the computation may be performed , as was the case for common intervals , in a single post - order traversal of the inclusion tree @xmath86 , focusing on each strong conserved interval @xmath2 of @xmath13 .
first compute the @xmath1-gaps of @xmath2 .
mark @xmath2 as @xmath1-nested if it contains no @xmath1-gap or one @xmath1-gap @xmath377 , with @xmath414 , that is a good one .
the latter verification assumes that during the treatment of each child @xmath102 of @xmath2 , if @xmath102 is detected as @xmath1-nested , then @xmath415 ( which is an interval @xmath416 with @xmath417 ) is marked as good if and only if @xmath415 is a @xmath1-gap and @xmath418 ( otherwise , @xmath415 is not marked at all ) .
now , if @xmath2 is @xmath1-nested and @xmath346 , the same type of mark is performed on @xmath345 if the conditions are fulfilled . then , applying algorithm [ alg2 ] on each strong conserved interval @xmath2 allows to enumerate all the @xmath1-nested conserved intervals generated by the frontiers of @xmath2 , according to theorem [ th : arbret ] .
affirmation 1 in theorem [ th : arbret ] ensures that each interval is output exactly once .
the running time of algorithm [ alg2 ] is clearly linear in the number of intervals output plus the numbers of children of @xmath2 , yielding a global @xmath152 time . to simply count the number of @xmath1-nested conserved intervals in @xmath2
, we must follow the same approach as for common intervals .
good and not good @xmath1-gaps are identified during a search among the intervals @xmath377 , where @xmath419 @xmath285 .
then , for each good @xmath1-gap we compute the number @xmath305 ( respectively @xmath392 ) of successive @xmath378-small intervals @xmath348 on its left ( respectively right ) .
we count the number of @xmath1-nested conserved intervals containing the good @xmath1-gap as @xmath420 next , for each maximal set of successive @xmath378-small intervals @xmath348 we add @xmath421 @xmath1-nested common intervals , where @xmath171 is the number of @xmath378-small intervals in the set .
all these operations obvisously take @xmath16 time and space .
in this paper we introduced the family of @xmath1-nested common intervals of @xmath0 permutations , and showed that it may be computed in time proportional to its cardinality .
this approach extends to any closed family of intervals that is represented by a @xmath17-tree .
we also show that our approach can be applied to conserved intervals , whose structure and properties are close but still different from those of common intervals .
the interest of our generalization of common / conserved intervals for finding conserved clusters of genes should be attested by further experiments .
also , other applications may be devised , such as helping the identification of orthologs / paralogs or defining distances between genomes in an evolutionary approach .
these are the close perspectives of our work .
a. bergeron , m. blanchette , a. chateau , and c. chauve .
reconstructing ancestral gene orders using conserved intervals .
in _ proceedings of wabi _ , volume 3240 of _ lecture notes in computer science _ , pages 1425 , 2004 .
a. bergeron , c. chauve , f. de montgolfier , and m. raffinot .
computing common intervals of k permutations , with applications to modular decomposition of graphs .
_ siam journal of discrete mathematics _ , 220 ( 3):0 10221039 , 2008 .
g. fertin and i. rusu . computing genomic distances : an algorithmic viewpoint . in _ algorithms in computational molecular biology : techniques , approaches and applications , m. elloumi , a. y. zomaya _ eds . _ _ , pages 773797 .
wiley series in bioinformatics , 2011 .
s. pasek , a. bergeron , j. l. risler , a. louis , e. ollivier , and m. raffinot .
dentification of genomic features using microsyntenies of domains : domain teams .
_ genome research _ , 150 ( 6):0 867874 , june 2005 . | in this paper we explain how to easily compute gene clusters , formalized by classical or generalized nested common or conserved intervals , between a set of @xmath0 genomes represented as @xmath0 permutations . a @xmath1-nested common ( resp . conserved ) interval @xmath2 of size @xmath3 is either an interval of size 1 or a common ( resp .
conserved ) interval that contains another @xmath1-nested common ( resp . conserved ) interval of size at least @xmath4 . when @xmath5 , this corresponds to the classical notion of nested interval .
we exhibit two simple algorithms to output all @xmath1-nested common or conserved intervals between @xmath0 permutations in @xmath6 time , where _ nocc _ is the total number of such intervals .
we also explain how to count all @xmath1-nested intervals in @xmath7 time .
new properties of the family of conserved intervals are proposed to do so . |
this supplemental material describes a quantum monte carlo ( qmc ) algorithm for estimating the rnyi entropy , @xmath92 , \label{renyi}\ ] ] using an extended - ensemble version of the so - called `` ratio method '' first applied in this context by humeniuk and roscilde @xcite . in the present work
, the algorithm will be specialized to the case of the second rnyi entropy ( @xmath54 ) , although extensions to other @xmath93 are relatively straightforward .
similarly , although we concentrate on applications to the spin-1/2 xy model , generalizations to hamiltonians with other symmetries are possible .
qmc estimators for the second rnyi entropy @xmath61 rely on the fact that the trace over powers of the reduced density matrix can be related to a ratio of partition functions : @xmath94}{z[a{=}0,2,t ] } , \label{zratio}\ ] ] where the numerator is the partition function of a multi - sheeted riemann surface @xcite , ( a `` replicated '' qmc simulation cell ) and the denominator is the square of the regular partition function , @xmath56 = z[t]^2 $ ] . in the replicated case , the geometry of the entangled region @xmath2 dictates the `` boundaries '' of the @xmath95 dimensional qmc simulation cell ; world - lines in region @xmath2 are periodic in imaginary time with period @xmath96 ( or @xmath97 ) , while in region @xmath3 , 2 ( or @xmath93 ) independent replicas exist with periodicity @xmath98 . since the logarithm in eq .
( [ renyi ] ) reduces the calculation of @xmath61 to the difference in free energies between systems described by the two partition functions , thermodynamic integration or wang - landau techniques can be used to devise qmc estimators .
in contrast , _ ratio methods _ ( not to be confused with the related `` ratio trick '' coined in ref .
@xcite ) give an estimator for the ratio of partition functions , and act as valuable alternatives to explicit calculation of free energies .
the most distinct advantage of ratio methods is their ability to calculate @xmath99 directly at a given fixed temperature
. they can however be inefficient for large entropies ( large subregions @xmath2 ) , and may require several separate simulations of different subregion geometries ( that can be combined with the ratio trick ) to combat this .
we begin by generalizing eq .
( [ zratio ] ) to the problem of calculation the ratio of replicated partition functions defined with two arbitrary regions @xmath2 and @xmath100 ; namely @xmath101/z[a,2,t]$ ] . as with all ratio methods ,
the qmc estimator in this case is based on a simple identity @xcite : @xmath102 where we have simplified our previous notation @xmath103 $ ] .
the expectation value is taken in the @xmath104 ensemble and @xmath105 and @xmath106 are the weights of a configuration @xmath17 in the @xmath107 and @xmath104 ensembles respectfully .
the difficulty of applying this identity in a straightforward manner is that it is valid only when the configuration space of @xmath107 , @xmath108 , is contained within the configuration space of @xmath109 , @xmath110 that is @xmath111 for any @xmath17 .
unfortunately , this condition is not automatically satisfied when the standard labelling variables are used in conventional qmc techniques .
however , for the xy model , by introducing a new label , it is possible to recast the configuration spaces of both ensembles such that @xmath112 . in other words
, a common label @xmath17 can be found that enumerates all possible configurations within both ensembles . here
we will consider a stochastic series expansion ( sse ) representation of the partition function @xcite .
a term in the sse expansion of @xmath113 can be represented by a spin state and an operator list , where the operators are terms in the hamiltonian .
alternatively , we can represent the same configuration by a linked list of vertices @xcite ; an example of a set of vertices is displayed in fig .
[ fig : vertices ] . we can therefore label the configuration by the list of links , @xmath114 , and the set of vertex types @xmath115 for a given boundary condition defined by @xmath2 as shown in fig .
[ fig : cells ] .
here we have defined @xmath114 to include all `` internal '' links between vertices and we link the `` exterior '' vertices to the boundaries of the operator string ; we specifically do not close the links at the boundaries , as one would do in a regular @xmath95 dimensional qmc simulation cell @xcite these two labels are sufficient to identify the configuration space of many models .
thus , a partition function can be expressed as the following double sum : @xmath116 where @xmath117 is the weight of a configuration labelled by @xmath114 and @xmath115 .
this general expression can be simplified for the xy model .
the model s sse vertices are displayed in fig .
[ fig : vertices ] . with an appropriate choice of adjustable sse constants ,
the weight of each vertex becomes equal to @xmath118 @xcite .
hence , the weight of a vertex configuration does not depend on a particular combination of vertices in the list and , therefore , is completely defined by the length of the corresponding operator list alone .
we get that @xmath119 and eq .
( [ eq : gen_part_func ] ) is simplified to : @xmath120 the second sum counts the degeneracy of vertex configurations compatible with the boundary conditions between replicas . motivated by the search for a variable independent of those boundary conditions , we introduce a new label , @xmath121 , that enumerates all possible partitions of a vertex configuration labelled by @xmath122 into a set of non - overlapping segments .
the following algorithm is used to construct a single instance of those segments : 1 .
pick an unmarked leg located on a boundary slice .
mark it as visited .
2 . by following the linked list , switch to a leg connected to it .
3 . the new leg belongs to a vertex . *
if this vertex is unmarked , pick with an equal probability one of two possible non - bounce moves for this vertex and switch to the corresponding leg .
mark this vertex as visited and store the move type .
* if the vertex is marked , switch to the next leg by performing a move of the same type that was done before .
repeat steps 2 - 4 until a leg on a boundary slice is reached . by repeating this algorithm for all legs located on the boundary slices ,
all open segments are traced out .
however , it is possible that some of the legs located on the inner slices have been left unmarked after this procedure . in order to partition those remaining legs too , the closed segments ( loops ) need to be traced .
this is achieved by adjusting two steps of the algorithm .
now in the first step , the choice of legs to be picked is extended to all interior legs .
once the initial leg is picked , the algorithm proceeds in the same way until it reaches the same leg again .
hence , the condition to terminate the execution of the forth step has to be modified appropriately . by construction
, any two segments built in such a manner can never pass through the same leg and , therefore , are non - intersecting . the loops tracing continues until all legs are marked . by the end of this procedure every leg belongs to one single segment ( closed or open ) .
this constitutes a single instance of the partitioning of a vertex configuration into a set of non - overlapping segments .
an example of such partition is shown in the first cell of fig .
[ fig : cells ] .
note that at each vertex , there are two choices how to proceed with the construction of a segment .
each of them leads to a different partition .
therefore , a simulation cell that contains @xmath123 vertices can be partitioned in @xmath124 distinct ways .
this is the range of the newly introduced label @xmath125 .
since @xmath123 just counts the number of vertices without discerning their types , the number of partitions for a particular @xmath115 is determined by the @xmath114 label only .
this fact allows to rewrite eq .
( [ eq : xy_part_func ] ) in a new form : @xmath126 this expression is obtained from eq .
( [ eq : xy_part_func ] ) by introducing a third sum via the substitution : @xmath127 where the new sum is performed over all different partitions of @xmath115 .
it can be also shown that for any vertex configuration in @xmath2 , @xmath115 , partitioned as @xmath121 , there exists a vertex configuration in @xmath100 , @xmath128 , with exactly the same partitioning , that is @xmath129 .
the proof is by construction .
if @xmath115 and @xmath128 were the same , the task is trivial .
otherwise , @xmath115 has to be modified in order to satisfy the boundary conditions of @xmath100 .
an example of such process is displayed in fig .
[ fig : cells ] . here
, the first and thirds simulation cells represent @xmath115 and @xmath128 correspondingly .
the second cell depicts an intermediate step of the correctional procedure . here ,
the open segments are connected into a loop along which the boundary spins mismatches are fixed one - by - one .
further details of the algorithm are given below .
proceeding column by column , consider each pair of boundary legs , @xmath130 , to be matched with respect to the new boundary conditions @xmath100 .
if the legs align , @xmath131 , proceed to the next pair . otherwise , randomly choose one of the two legs in the pair .
say it is @xmath132 .
since this leg is located on a boundary slice , it belongs to an open segment .
flip all legs belonging to this segment .
now , the original pair of legs is properly aligned , however there might be another mismatch at the other end of the segment .
call the new pair @xmath133 where @xmath134 is the leg that has just been flipped as part of the open segment . by the same logic as before , @xmath135 must belong to an open segment whose other end is identified as another boundary leg @xmath136 . if @xmath137 , flip this segment in order to align the @xmath138 pair and move on to the next pair @xmath139 .
otherwise , proceed to the same pair without flipping the segment . in this way ,
one - by - one pairs of boundary legs are aligned with respect to @xmath100 boundary condition along a loop of open segments .
an important subtlety occurs at the last step of this algorithm when the last pair @xmath140 is considered .
unlike previously , @xmath141 can not be flipped if those legs do not align .
an attempt to do so would entail another iteration of corrections with the same result , thus , initiating the algorithm in an infinite loop .
however , this does not occur in the xy - model due to the special properties of its vertices .
notice that the only vertex move that connects two anti - aligned legs is the `` switch - and - reverse '' move ( fig.[fig : vertices ] ) ; this is the only move that reverses the vertical directionality of propagation of the segment s head .
consequently , once a segment tracing is initiated with the choice of a leg and its state , the spin state of the leg at the segment s head is determined by the vertical direction that the segment passes through the leg . on the last boundary connection , the direction of motion along the segment must be the same as the initial direction , and therefore the initial spin state at the head of the segment under construction is always the same its final state .
we see then that for any segment partition of @xmath115 , it is always possible to construct a @xmath128 with the same segment partition . now that we have shown the equivalence between any two configuration spaces constraint by boundary conditions @xmath2 and @xmath100 in terms of the segment partitions , we have achieved our initial goal to find a label @xmath125 that can be used to apply the eq .
[ eq : simple_id ] .
however , we still need to compute the weights of @xmath142 and @xmath143 . upon a close inspection of the inner double sum in eq . [ eq : xy_part_func_2 ]
, we realize that the same segment partitions can be generated from different vertex configurations .
this degeneracy can be exploited in order to replace the double sum with a single sum : @xmath144 where the inner sum iterates over all unique segment partitions for a given linked list and @xmath145 is the degeneracy of the segment partition labelled as @xmath146 .
as will be shown , the degeneracy depends on the boundary conditions between replicas and , thus , the superscript must be included . in order to calculate the degeneracy of a segment partition @xmath146 ,
connect open segments using boundary conditions @xmath2 to form @xmath147 loops that cross boundary slices .
in addition to those loops , there are also @xmath148 inner loops ( closed segments ) .
since all those loops do not intersect with each other by construction , the spins within them can be flipped independently .
each combination of the loops flips leads to another vertex configuration .
equivalently , all of those vertex configurations generate the same segment partition @xmath149 .
there are in total @xmath150 loops with each one being in one of two states : flipped or not - flipped .
each combination of those states corresponds to a different vertex configuration . in total
, there are @xmath151 of such combinations which constitutes the degeneracy @xmath145 . with the last piece of the puzzle in our hands ,
we infer the segment partition weight from eq .
( [ eq : xy_part_func_3 ] ) : @xmath152 since the inner - loops are unaffected by the boundary conditions , upon the substitution of this weight in the eq .
( [ eq : simple_id ] ) their number drops out and an elegant expression is obtained : @xmath153 in practice , this estimator can be implemented in fewer steps that were required to prove its validity .
it requires two routines .
one routine traces out a random single segment partition @xmath146 for a vertex configuration @xmath115 as was outlined before . in order to speed up the execution ,
it is not necessary to identify the closed segments .
the end product of this routine is to associate the pairs of boundary spins that are connected via open segments .
once this step is done , the second routine takes the set of those pairs together with replicas boundary conditions as its inputs .
its task is to count @xmath154 .
this routine is executed for both @xmath2 and @xmath100 with the same open segments . in the end , @xmath155 and @xmath156
are known and the estimator can be evaluated according to eq .
( [ eq : estimator ] ) .
to illustrate the efficiency of the new estimator , its raw measurements are compared to the original ratio method @xcite in fig .
[ fig : methcomp ] , on the 2d xy model of interest in the main text . here
, the deterioration of both estimators statistics is seen as the difference between the partition functions regions a , @xmath157 , grows large .
as the reference values , we employ the results obtained from the ratio trick @xcite , which constitute a compilation of the extended ensemble ( ee ) ratio method results from many different monte carlo simulations , each executed with @xmath158 .
note that the original ratio method results are based on five times more monte carlo sweeps that were involved to produce the ee ratio method results .
even with such advantage , the ratio method statistics becomes increasingly poor towards @xmath159 . after this threshold
, the estimator seems no longer capable to capture any meaningful statistics within the running time of its simulations .
the performance of the ee is strikingly better .
even when @xmath160 , the largest possible increment in the system , it produces an accurate result with a precision comparable to the ratio method precision at @xmath161 . | in the face of mounting numerical evidence , metlitski and grover [ arxiv:1112.5166 ] have given compelling analytical arguments that systems with spontaneous broken continuous symmetry contain a sub - leading contribution to the entanglement entropy that diverges logarithmically with system size .
they predict that the coefficient of this log is a universal quantity that depends on the number of goldstone modes . in this paper
, we confirm the presence of this log term through quantum monte carlo calculations of the second rnyi entropy on the spin 1/2 xy model .
devising an algorithm to facilitate convergence of entropy data at extremely low temperatures , we demonstrate that the single goldstone mode in the ground state can be identified through the coefficient of the log term .
furthermore , our simulation accuracy allows us to obtain an additional _ geometric _ constant additive to the rnyi entropy , that matches a predicted fully - universal form obtained from a free bosonic field theory with no adjustable parameters .
_ introduction _ in condensed matter , the entanglement entropy of a bipartition contains an incredible amount of information about the correlations in a system . in spatial dimensions @xmath0 , quantum spins or bosons
display an entanglement entropy that , to leading order , scales as the boundary of the bipartition @xcite .
subleading to this `` area - law '' are various constants and particularly in gapless phases functions that depend non - trivially on length and energy scales .
some of these subleading terms are known to act as informatic `` order parameters '' which can detect non - trivial correlations , such as the topological entanglement entropy in a gapped spin liquid phase @xcite . at a quantum critical point , subleading terms contain novel quantities that identify the universality class , and potentially can provide constraints on renormalization group flows to other nearby fixed points @xcite . in systems with a continuous broken symmetry ,
evidence is mounting that the entanglement entropy between two subsystems with a smooth spatial bipartition contains a term , subleading to the area law , that diverges logarithmically with the subsystem size .
first observed in spin wave @xcite and finite - size lattice numerics @xcite , the apparently anomalous logarithm had no rigorous explanation until 2011 , when metlitski and grover developed a comprehensive theory @xcite .
they argued that , for a finite - size subsystem with length scale @xmath1 , the term is a manifestation of the two long - wavelength energy scales corresponding to the spin wave gap , and the `` tower of states '' arising from the restoration of symmetry in a finite volume @xcite .
remarkably , their theory not only explains the subleading logarithm , but predicts that the coefficient is directly proportional to the number of goldstone modes in the groundstate .
furthermore , describing a goldstone mode with a free scalar field theory allows them to predict the value for an additional additive _ geometric _ constant , which is fully universal and should therefore be the same across a wide range of continuum theories and lattice models . in this paper
, we confirm these predictions in a striking way , through large - scale quantum monte carlo ( qmc ) simulations on a spin-1/2 xy model on the square lattice . by employing an extended - ensemble generalization of a _ ratio method _ @xcite
, we are able to carefully converge the mutual information of the second rnyi entropy to its low - temperature value . there
, finite - size scaling reveals the coefficient of the subleading logarithmic term to precisely match the prediction of metlitski and grover , identifying the lone goldstone mode in the theory .
remarkably , our simulations are accurate enough to also measure the universal additive geometric constant , which is consistent with the predicted relationship to the constant which appears in a lattice - regularized free scalar field theory @xcite .
_ entanglement entropy in the tower of states
_ to obtain a qualitative understanding of the origin of the logarithmic correction in metlitski and grover s theory , it is simplest to first envision decoupling the two spatial subsystems , @xmath2 and @xmath3 , that define the entangled bipartition .
the low - energy degrees of freedom in each subsystem can be described by an @xmath4 rotor ( @xmath5 for the xy model ) , representing the direction of the order parameter .
here we are only allowing global fluctuations of the order parameter within each subsystem , such that we may approximate the state of @xmath2 and @xmath3 each as a single independent quantum rotor .
the effective hamiltonian of each subsystem is @xmath6 , where @xmath7 is the total angular momentum operator with eigenvalues @xmath8 and @xmath9 is the effective moment of inertia which is extensive , proportional to the magnetic susceptibility @xmath10 : @xmath11 in @xmath12 spatial dimensions @xcite .
thus the energy scale of the tower @xmath13 vanishes with the system volume , faster than any other energy scale .
the eigenstates of @xmath7 result in the famous `` tower of states '' observed routinely in computational studies of systems with continuous symmetry breaking in a finite volume @xcite .
the interaction between @xmath2 and @xmath3 which aligns the subsystem order parameters may be introduced via a goldstone mode hamiltonian @xmath14 which couples the two rotors .
the energy scale of the @xmath14 is the goldstone mode gap @xmath15 which is the scale of the lowest energy spin waves . since @xmath16 where @xmath17 is the spin - wave velocity , @xmath18 in the thermodynamic limit for @xmath19 . in the limit @xmath20 , there are no _ relative _ fluctuations in the order parameter between subsystems , and @xmath2 and @xmath3 act as a single rigid rotor . for finite @xmath15 , there will be relative fluctuations between the subsystems order parameters due to the zero point fluctuations of @xmath14 . to estimate the entanglement entropy contribution from the tower of states
, we can count the number of accessible " states of subsystem @xmath2 , @xmath21 , when the total system is in the ground state , and use @xmath22 . in the limit @xmath23 and the rotors are rigidly coupled , the ground state is the ground state of the total system tower of states with zero total angular momentum : @xmath24 . in this case
all states in the @xmath2 subsystem tower are accessible to the ground state , as each state in @xmath2 can be paired with an appropriate state in @xmath3 to form a state with nonzero overlap with the @xmath24 state . however , as discussed above , by including a _ finite _ @xmath25 and thus allowing _ relative _ fluctuations of the subsystem order parameter between @xmath2 and @xmath3 , the fluctuations in the subsystem angular momentum are finite and determined by the ratio of the energy scales : @xmath26 @xcite .
in fact , the reduced matrix of the subsystem takes the form of a thermal density matrix with an effective `` entanglement hamiltonian '' given by @xmath27 and the `` entanglement temperature '' given by @xmath15 @xcite ; the resulting tower of states structure in the entanglement spectrum has been seen in numerics @xcite .
thus the inclusion of goldstone modes cuts off the accessible states of the subsystem to those with an energy below the spin wave gap , as illustrated in fig .
[ fig : tower ] . as an example of this mechanism ,
consider the case of @xmath5 ( valid for our xy model simulations below ) . here
the rotors have a single component of angular momentum @xmath28 and the orientation of the rotors is described by a single angle @xmath29 .
for @xmath30 the ground state has @xmath31 , which has nonzero overlap with states of equal and opposite @xmath28 in each subsystem : @xmath32 ; consequently all @xmath33 state are accessible in this limit .
we may include the effect of the lowest goldstone mode by treating the dynamics of the relative angle between subsystems @xmath34 as a single harmonic oscillator with frequency @xmath15 and moment inertia @xmath35 , with an effective hamiltonian @xmath36 here , the fluctuations in the relative angular momentum @xmath37 are given by the ground state fluctuations of a harmonic oscillator : @xmath38 .
the key point here is that because the order parameter is canonically conjugate to the rotor angular momentum , _ increasing _ the relative fluctuations in the order parameter _
reduces _ the fluctuations in @xmath39 .
thus , allowing relative fluctuations of the order parameter between subsystems effectively cuts off subsystem rotor states that are accessed in the ground state at order @xmath40 a relationship that holds for all @xmath41 @xcite .
we may therefore estimate @xmath21 by counting the number of states ( in @xmath2 s tower of states ) that lie below @xmath15 .
for systems with @xmath4 symmetry , the tower of states is described by a rotor living on an @xmath42 dimensional sphere , where @xmath43 is the number of goldstone modes .
the dengeneracy of each energy level is of order @xmath44 .
we then may estimate the total number of states below @xmath15 by integrating the degeneracy up to the cutoff @xmath45 : @xmath46 using the relation @xmath47 from hydrodynamic spin - wave theory where @xmath48 the stiffness @xcite , the entanglement entropy correction due to the tower of states becomes @xmath49 we see that the logarithmic correction to the area law arises due to the quasi - degeneracy of accessible bulk subsystem states , that scales as a power law in @xmath1 for systems with spontaneously broken continuous symmetries .
this contrasts with the leading - order area law , arising from the exponential scaling of the number of local boundary states with the boundary area . clearly , the prefactor of the logarithmic correction is a universal number that simply counts the number of goldstone modes .
_ quantum monte carlo procedure
_ in order to examine the effects of broken continuous symmetry on entanglement , we implement a highly - efficient stochastic series expansion ( sse ) qmc algorithm @xcite for the 2d spin-1/2 xy model , @xmath50 .
this model is known to realize a ground state where the @xmath51 symmetry is spontaneously broken , resulting in one goldstone mode .
the finite - temperature sse algorithm uses a version of the directed - loop updates specialized to the xy model @xcite . to measure the entanglement entropy
, we employ a replicated simulation cell @xmath52 $ ] , which gives access to second rnyi entropy @xmath53 /(1-n)$ ] , with @xmath54 .
this is done through the ratio @xmath55/z[a{=}0,2,t]$ ] , where @xmath56 = z[t]^2 $ ] ; the square of the unmodified partition function @xcite .
one therefore needs to evaluate a difference in free energies or , alternatively , a ratio of partition functions , between the replicated and unmodified simulation cells with qmc .
several options are available , such as thermodynamic integration @xcite and wang - landau sampling @xcite .
alternatively , methods such as the _ ratio method _ @xcite allow one to compute the partition function ratio directly .
as described in the next section , for this model , careful convergence to low temperature is required .
hence , in the supplemental material , we develop a highly - efficient variant of the ratio method , dubbed the _ extended ensemble ratio method _ , for the xy model
. _ results _ we perform large - scale calculations to obtain the second rnyi entropy of the xy model on @xmath57 square lattices with periodic boundary conditions .
the resulting torus is partitioned into two cylindrical regions of linear dimension @xmath58 and @xmath59 . for system sizes
@xmath60 , separate finite - temperature tests are performed to explore the convergence of @xmath61 . in agreement with the expected scaling of the tower - of - states gap
, those results confirm that the convergence temperature scales approximately as @xmath62 .
we use this to estimate the convergence temperatures for @xmath63 .
since sse qmc simulations necessarily run at finite temperature , very small thermal contributions to @xmath61 are expected , which we observe to significantly affect our finite - size scaling analysis below .
fortunately , this thermal contribution can be essentially eliminated by employing the mutual information , @xmath64 @xcite .
assuming that bulk ( volume - law ) contributions from the two subsystems @xmath2 and @xmath3 approximately cancel the bulk term @xmath65 in @xmath66 , the scaling form described by metlitski and grover becomes : @xmath67 here , @xmath68 is a non - universal constant , @xmath48 and @xmath17 are the spin - wave stiffness and velocity , and @xmath69 is the geometric constant that depends on the aspect ratios of the cylinders @xmath2 and @xmath3 @xcite . note , since non - universal ( cutoff )
dependences are all contained within @xmath48 and @xmath17 , this geometric constant remains fully universal .
for the spin-1/2 xy model on the square lattice , @xmath70 and @xmath71 were obtained from ref .
@xcite .
figure [ fig : conver ] illustrates a representative convergence test for different system sizes . the mutual information peaks at temperatures above the kosterlitz - thouless transition of @xmath72 ( which can be detected by the crossing of the finite - size curves ;
see ref .
@xcite ) . for @xmath73 ,
the mutual information reaches a minimum ( at @xmath74 in fig .
[ fig : conver ] ) before undergoing a slow rise .
this rise continues until the approximate ground state is reached , for temperature below the finite - size scaling gap , which for system sizes larger than @xmath75 occurs for @xmath76 .
thus , although the method of thermodynamic integration is useful to produce the general shape of the @xmath66 curve for a wide range of temperatures , it is difficult to control the systematic error introduced by numerical integration at low temperatures for @xmath77 . therefore , data used in the below fits was converged at very low temperatures using the extended ensemble ratio method , described in the supplemental material .
figure [ fig : eevslv3 ] illustrates the resulting temperature - converged mutual information for a variety of system sizes , as a function of the `` width '' of the cylindrical region , @xmath78 . since , for a subsystem @xmath2 and its complement @xmath3 , @xmath79 only at @xmath80 , the symmetry of the entanglement entropy about @xmath81 provides a sensitive test of temperature convergence .
employment of the `` bare '' rnyi entropy results in a very slight asymmetry in the curve ; use of @xmath66 restores this symmetry producing high - quality data that can be fitted using eq .
( [ mi ] ) .
the results of this analysis are illustrated in fig [ fig : fits2 ] .
here , @xmath66 is calculated at various aspect ratios ( the vertical cuts in fig .
[ fig : eevslv3 ] ) and fit to the functional form eq .
( [ mi ] ) . specifically , to extract the coefficient of the subleading logarithm , the mutual information was fit to @xmath82 , where @xmath83 and @xmath12 are adjustable fit parameters . as illustrated in fig .
[ fig : eevslv3 ] ( a ) , there is definative evidence for the existence of a logarithm ; furthermore , independent fits for the four aspect ratios studied each give @xmath84 to within error bars .
even more striking , we are able to extract the universal shape - dependence of the geometric constant @xmath69 . to do so , fits were performed to the functional form @xmath85 , where @xmath86 is fixed at unity in order to remove one parameter from the analysis .
thus calculated , @xmath69 for @xmath5 in two dimensions can be compared via a zero - parameter fit to the subleading constant term @xmath87 calculated in a free scalar field theory @xcite through the relation @xmath88 , valid for the second rnyi entropy .
the free field result @xmath87 , which depends on the aspect ratio @xmath89 , can be calculated numerically for free bosons on the lattice using the correlation matrix technique ( as in ref .
@xcite ) . as illustrated in fig .
[ fig : eevslv3 ] ( b ) , the resulting theoretical curve is in excellent agreement with our qmc results for @xmath69 .
_ discussion _ in this paper , we have employed large - scale quantum monte carlo ( qmc ) simulations on the spin-1/2 xy model to demonstrate the presence of a logarithmic correction to the rnyi entropy due to spontaneous breaking of continuous symmetry .
this term arises from the presence of two infra - red energy scales in the problem ; the spin - wave gap , and the `` tower of states '' .
the coefficient of this logarithm is predicted in ref .
@xcite to be @xmath90 , where @xmath86 is the number of goldstone modes and @xmath12 the spatial dimension .
we confirm this prediction in a striking manner , through finite - size scaling studies on the square lattice xy model , recovering to high precision the expected @xmath91 . in order to do so
, a qmc algorithm was developed to sample the second rnyi entropy at a fixed temperature with high efficiency ( described in the supplemental material ) .
remarkably , in addition to confirming @xmath91 , our simulation technique is able to converge the value of an additional additive geometric constant @xmath69 , which is fully universal since all short - distance physics is confined to the ( known ) spin wave stiffness and velocity , contained within the argument of the logarithm .
the resulting @xmath69 has a functional dependence on the geometric aspect ratio of the entangled bipartition .
this function matches , to within error bars , that calculated using a free scalar field theory regularized on a toroidal square lattice , with no adjustable parameters . this is a rare and striking example of complete quantitative agreement between a universal quantity calculated in continuum field theory and finite - size lattice simulations .
the ability of the rnyi entanglement entropy to mediate between continuum theory and lattice simulation illustrates the utility of such geometric quantities as an alternative paradigm to study correlations in condensed - matter systems . in this case , with the full understanding of the universal structure of the entanglement entropy in the presence of a spontaneously broken continuous symmetry , the door is now open to the direct search and detection of goldstone modes in a large variety of systems , through the widely accessible rnyi entropies .
we are thankful for enlightening discussions with a. luchli , d. poilblanc , k resch , and particularly , m. metlitski and t. grover , without whom this work would not have been possible .
support was provided by nserc , the canada research chair program , and the perimeter institute ( pi ) for theoretical physics .
research at pi is supported by the government of canada through industry canada and by the province of ontario through the ministry of economic development & innovation .
s.i . acknowledges support from the fp7/erc starting grant no .
306897 .
the simulations were performed on the computing facilities of sharcnet . * supplementary material for `` detecting goldstone modes with entanglement entropy '' * |
symmetries and killing vectors play an important role in exact solutions , especially in the electro vacuum ones e.g. @xcite .
a general plane symmetric metrics in cylindrical coordinates is taken as a seed solution @xmath0 where @xmath1 , @xmath2 and @xmath3 are arbitrary functions of coordinates @xmath4 and @xmath5 . the generating conjecture , formulated in @xcite , @xcite , @xcite , is applied to azimuthal @xmath6 killing vector belonging to rotation in planes of symmetry .
components of this killing vector are @xmath7 expected form of `` charged '' metrics is again the general plane symmetric metrics @xmath8 where @xmath9 , @xmath10 and @xmath11 are functions different from ( [ seed ] ) and they will be evaluated out of the einstein - maxwell equations . according to the conjecture we can form an electro - magnetic potential from killing vector @xmath12
if we use the electro - magnetic potential ( [ pot ] ) for purely electro - magnetic energy - momentum tensor , two einstein equations @xmath13 and @xmath14 for the space - time ( [ newm ] ) are @xmath15 these lead to the restrictions on function @xmath2 in seed metric ( [ sedm ] ) @xmath16 then the maxwell equations for ( [ newm ] ) are automatically satisfied , the nonzero covariant and contravariant components of the maxwell electro - magnetic tensor are @xmath17 there are other four non - trivial einstein equations @xmath18 , @xmath19 , @xmath20 and @xmath21 with unknown functions @xmath9 , @xmath10 and @xmath11 to be solved @xmath22 where subscript denotes partial derivation .
now we are trying to solve these equations with the assumption that @xmath10 is a function of coordinate @xmath5 only . consequently equations ( [ einst2 ] ) and ( [ einst4 ] )
lead to independence of functions @xmath9 and @xmath11 on coordinate @xmath4 . therefore all unknown functions @xmath23 , @xmath24 and @xmath25 now depend on coordinate @xmath5 only .
we should solve two einstein equations @xmath18 ( [ einst1 ] ) and @xmath21 ( [ einst4 ] ) @xmath26 from ( [ einst02 ] ) we evaluate @xmath27 which is subsequently put into ( [ einst01 ] ) then we get only one equation for two unknown functions @xmath28 and @xmath29
@xmath30 we can choose one arbitrary function @xmath31 , let us choose it in the form @xmath32 we obtain the functions @xmath28 and @xmath29 @xmath33 where @xmath34 and @xmath35 are integration constants .
next , the new metrics components for the solution ( [ newm ] ) can be computed @xmath36 where @xmath31 is an arbitrary function and @xmath34 and @xmath35 are arbitrary constants .
the metric tensor must satisfy hilbert conditions which lead to inequalities @xmath37 now we are looking at simplifying resulting metrics tensor using constants substitutions and coordinate transformations @xmath38 where striped coordinates denote the new ones .
the transformation ( [ transf ] ) is formally equivalent to the choice of function @xmath31 as @xmath39 thus , from the metric tensor we have removed the arbitrary function @xmath31 which can be interpreted as an arbitrary transformation in coordinate @xmath5 .
the resulting form of the metric tensor is @xmath40 where we have abandoned the strips and the hilbert conditions give @xmath41 moreover , we can look for solutions beyond hilbert conditions .
they can be gained from einstein equations ( [ einst1 ] ) , ( [ einst2 ] ) , ( [ einst3 ] ) and ( [ einst4 ] ) with the assumption that @xmath10 is function only of coordinate @xmath4
. then equations ( [ einst1 ] ) and ( [ einst2 ] ) lead to independence of functions @xmath9 and @xmath11 on coordinate @xmath5 . so all unknown functions @xmath42 , @xmath43 and @xmath44 depend only on coordinate @xmath4 now .
the resulting metric tensor has the form @xmath45 corresponding hilbert conditions for this solution are @xmath46 the solution ( [ solution2 ] ) can be obtained directly from the solution ( [ solution ] ) exchanging coordinates @xmath4 and @xmath5 .
let us explore some properties of the solutions ( [ solution ] ) and ( [ solution2 ] ) now .
metrics ( [ solution ] ) is singular at @xmath47 .
singularities at @xmath48 are only coordinate ones , there exist horizons . on the other hand
, @xmath49 is a real physical singularity , and scalar invariants of the riemann tensor diverge here as shown @xmath50 the electro - magnetic potential in solution ( [ solution ] ) is @xmath51 the components of maxwell electro - magnetic tensor are @xmath52 we can write down three usual dimensional components of the maxwell tensor as @xmath53 , \\ \vec{b}&=&\left[0,0,-k\left(\frac{k+z}{z}\right)^2\right].\end{aligned}\ ] ] in this solution there exists only a magnetic field . the magnetic field @xmath54 is shown in fig .
[ fig1 ] ( where @xmath55 ) .
it is also possible to evaluate geodesic trajectories of motion when testing particles are fixed in @xmath4,@xmath5 plane .
the null geodesic lines are given explicitly by @xmath56 where @xmath57 is an integration constant .
such geodesic lines are shown in fig .
[ fig2 ] .
the motion of the charged test particles in the @xmath4,@xmath5 plane is given by @xmath58 the dots denote differentiation with respect to an affinne parameter .
it is clear that the central singular plane acts as a source of repulsive power .
let us consider a static plane - symmetric solution as appears in @xcite ( with null cosmological constant ) @xmath59 where @xmath60 is an arbitrary function of coordinate @xmath5 , a prime denotes differentiation with respect to @xmath5 , @xmath61 and @xmath62 are constants and @xmath63 is the charge per unit area .
when we use the transformations @xmath64 or analogous ones @xmath65 our solution ( [ gsol1 ] ) - ( [ gsol4 ] ) turns to be identical with the mcvittie ( [ solutiong ] ) with @xmath66 .
the mcvittie solution @xcite is thus generated by the killing @xmath67 vector from the seed metric ( [ sedm ] ) .
let us connect both ( [ solution ] ) and ( [ solution2 ] ) solutions together through the horizons @xmath68 .
we will use retarded and advanced null geodesic coordinates @xmath69 and @xmath70 like this @xmath71 after transformation of the metric tensor from the cartesian coordinate couple @xmath5 , @xmath4 to the null one @xmath69 , @xmath70 we get @xmath72 exponential transformation is used here ( which doubles the regions as in kruskal analytical extension of schwarzchild solution @xcite and leaves the coordinates null ) @xmath73 now again , we diagonalize the metric by using the new cartesian coordinate couple @xmath74 , @xmath75
@xmath76 this leads to the metrics tensor @xmath77 where @xmath5 , which is @xmath5 in ( [ solution ] ) for @xmath78 and @xmath4 in ( [ solution2 ] ) for @xmath79 , is given implicitly by the equation @xmath80 at first sight there are no sign changes between @xmath81 and @xmath82 , the hilbert conditions are always satisfied for all values of @xmath74 , @xmath75 .
the resulting space - time contains every original region i : @xmath78 , ii : @xmath83 , and iii : @xmath84 twice as shown in fig .
furthermore , in ( [ kruskal ] ) there are still singularities at @xmath85 but no more at @xmath86 .
lines with constant @xmath5 form over @xmath75 , @xmath74 plane a hyperbolic paraboloid and lines with constant @xmath4 form lines @xmath87 as shown in fig .
singular points @xmath88 are represented by values @xmath89 of @xmath90 , the apparent singularity @xmath86 is moved to infinity in these coordinates .
null geodesic lines in @xmath91 are straight lines @xmath92 as displayed by dashed lines in fig .
[ fig5 ] .
the transformation to the penrose conformal diagram is done by these transformations ( again leave null coordinates ) @xmath93 new null coordinates @xmath94 are limited by @xmath95 so the infinity has been conformaly moved to finite distance .
now it is possible to draw the penrose diagram ( fig .
[ fig6 ] ) . here
are marked geodesic future and past @xmath96 , null future and past @xmath97 and the time - like infinity @xmath98 .
for the physical singularity @xmath99 is @xmath100 and both are time - like , so it is possible for an observer to avoid falling in it .
we have found two electro - vacuum plane symmetric exact solutions of the einstein - maxwell equations using the generating conjecture and the killing vector @xmath67 of the plane symmetric vacuum seed metric .
we have shown that one of them is equivalent to the mcvittie solution of a charged infinite thin plane .
we also carried out analytical extension of coordinates , which can cover both found solutions with only one coordinate chart .
the corresponding penrose conformal diagram has been obtained as well .
99 bik j. , pravdov a. , symmetries of asymptotically flat electrovacuum spacetimes and radiation , j. mathem .
39 , 6011 ( 1998 ) horsk j. , mitskievich n. v. : czech j. phys .
b 39 ( 1989 ) , page 957 cataldo m. , horsk j. , mitskievich n. v. : killing vectors and einstein - maxwell field , differential geometry and its applications , proc . conf . brno ( 1989 ) novotn j. , horsk j. : generation method for solutions of einstein - maxwell equations and its applications , proc .
opava ( 1992 ) amundsen p. a. , grn .
: physical review d , vol .
27 , num . 8
( 1983 ) , page 1731 mcvittie g. c.:proc .
london 124 , 366 ( 1929 ) hawking s. w. , ellis g. f. r. : the large scale structure of space - time , cambridge university press ( 1973 ) | for the plane symmetry we have found the electro - vacuum exact solutions of the einstein - maxwell equations and we have shown that one of them is equivalent to the mcvittie solution of a charged infinite thin plane .
the analytical extension has been accomplished and the penrose conformal diagram has been obtained as well . |
recent cosmological observations @xcite have demonstrated that our universe is undergoing the phase of an accelerating expansion .
although general relativity ( gr ) developed in the last century has been successful in many ways of explaining various experimental results in gravity , the nature of the accelerating universe now rises as a small cloud shrouding it .
we thereby look for a more general theory that comprises gr yet able to solve the accelerating problem referred to as dark energy @xcite . by virtue of the local gauge principle ,
one leads to incorporate poincar group as gauge group of a principal bundle , such that the local lorentz symmetry of the spacetime is preserved @xcite .
an attempt is to release torsion from a connection , rather than the levi - civita connection in the standard gr , which also acts as a dynamical field like metric tensor .
such spacetime is usually called riemann - cartan manifold . the gauge theory based on this manifold ,
is known as poincar gauge theory ( pgt ) @xcite
. it has been investigated in @xcite that there are six modes by the decomposition of the connection according to torsion tensor in the linearlized theory , classified as @xmath2 , @xmath3 and @xmath4 in terms of spins and parities . among them , the @xmath5 mode @xcite , also called the _ scalar - torsion _ mode , does not directly interact with any known fundamental source @xcite .
along with the property induced by the nonlinear equation set , this @xmath5 mode is our main concern . in @xcite ,
shie , nester and yo ( sny ) have examined models with the spin @xmath5 mode in pgt to achieve the late time accelerating expansion of the universe . in other words , the geometric effect of torsion
is treated as an effective dark energy . in particular , they have presented two cases with the solutions of a constant curvature and positive kinetic energy , respectively .
the first case is an extremely simple solution existing inside the system of differential equations formed by the spin @xmath5 mode , which provides a modeling for the late time de sitter universe for dark energy .
note that this simple solution violates the positivity argument @xcite .
the second one is referred to as the normal case , which conforms with the regular positivity condition , but it gives rise to no obvious analytic solution .
torsion cosmology related to the scalar - torsion mode has been also explored in @xcite . in this work ,
we concentrate on these two cases and present numerical solutions of the late - time acceleration behavior corresponding to the equation of state ( eos ) , defined by @xmath6 , where @xmath7 and @xmath8 are the energy density and pressure of the relevant component of the universe , respectively .
this paper is organized as follows : in sec .
ii , we review the scalar - torsion of the spin @xmath5 mode in pgt and give equations of motion for cosmology . in sec .
iii , we show our numerical results on the cosmological evolutions for the scalar - torsion mode .
we present our conclusions in sec . iv .
pgt of gravity starts with a lagrangian 4-form on @xmath9-spacetime : @xmath10 where @xmath11 is a set of the orthonormal dual basis , @xmath12 is the connection 1-form with respect to @xmath11 , @xmath13 is the matter lagrangian , and @xmath14 is the gravitational lagrangian that can be made up by certain combinations . in @xcite , sny studied the spin @xmath5 mode , given by @xcite @xmath15 where @xmath16 is the hodge dual map , @xmath17 is the volume 4-form of the space - time and @xmath18 with @xmath19 are the irreducible pieces of the torsion 2-form @xmath20 , defined by @xcite @xmath21 the coefficients of @xmath14 in ( [ e : sny ] ) are constrained by the positivity argument @xcite such that @xmath22 the independent variation of ( [ e : lagrangian ] ) with respect to @xmath23 yields @xcite @xmath24 where the gauge field momenta are given by @xmath25 and the source terms are defined by @xmath26 corresponding to the symmetric energy - momentum tensor - valued 4-form , asymmetric vector - valued 3-form usually called canonical energy - momentum tensor , and tensor - valued 3-form known as canonical spin angular momentum tensor , respectively
. one can also write the decompositions into the basis of @xmath27 @xcite : @xmath28 where @xmath29 .
the equations of motion are given symbolically as @xmath30 we shall describe our universe with the flrw cosmology , which is homogeneous and isotropic with the metric @xmath31 where @xmath32 is the constant curvature . for simplicity , we shall only consider the flat universe with @xmath33 . for ( [ e : sny ] ) of the sny model in the frlw cosmology with no spin source @xmath34 ,
the main field equation ( [ e : eom ] ) leads to @xcite @xmath35 where @xmath36 , @xmath37 , and @xmath38 , which is the time component of the torsion trace , defined by @xmath39 . here ,
@xmath40 in ( [ e : main eq1])-([e : main eq3 ] ) denotes the affine curvature with respect to the curvature 2-form @xmath41 , given by @xmath42 hence , one obtains the relation @xmath43 where @xmath44 represents the curvature of the levi - civita connection induced by ( [ e : frw ] ) .
the energy - momentum tensor @xmath45 is defined as ( [ e : canonical em tensor ] ) and @xmath46 stands for the trace @xmath47 .
explicitly , one has @xmath48 with the subscript @xmath49 representing the ordinary matter including both dust and radiation . to see the geometric effect of torsion , we can write down the friedmann equations as @xmath50 with @xmath51 in gr , where @xmath52 and @xmath53 denote the critical energy density and total pressure of the universe , while @xmath54 and @xmath55 correspond to the energy density and pressure of some effective field , respectively . by comparing the equation of motion of the scalar - torsion mode in pgt ( [ e : rho_t ] ) to the friedmann equations ( [ e : friedmanneq ] ) ,
one obtains @xmath56 which will be regarded as the effective torsion dark energy density and pressure , respectively . from ( [ e : friedmanneq ] ) , we get the continuity equation , @xmath57 which can also be derived by applying the identity @xmath58 where @xmath59 is the covariant derivative with respect to the levi - civita connection and @xmath60 is the effective energy - momentum tensor of torsion dark energy . in addition , from ( [ e : main eq1 ] ) - ( [ e : main eq3 ] ) , one can check that the continuity equation for the torsion field is also valid , @xmath61 @xmath62 consequently , we obtain the continuity equation for the ordinary matter to be @xmath63 by assuming no coupling between radiation and dust , the matter densities of radiation ( @xmath0 ) and dust ( @xmath64 ) in scalar - torsion cosmology share the same evolution behaviors as gr , @xmath61 @xmath65 and @xmath66 , respectively . in order to investigate the cosmological evolution ,
it is natural to define the total eos by @xcite @xmath67 which leads to @xmath68 where @xmath69 and @xmath70 with @xmath71 , representing the energy density ratios and eoss of matter and torsion , respectively .
note that the eos in ( [ e : w_t ] ) , which is commonly used in the literature , e.g. @xcite , can be examined by cosmological observations in @xcite .
in particular , it can be used to distinguish the modified gravity theories from @xmath72cdm @xcite .
consequently , the evolution of the torsion dark energy can be described in terms of solely @xmath73 by @xmath74 in the following section , we shall focus on this important quantity .
the evolution of torsion cosmology is determined by ( [ e : main eq1 ] ) - ( [ e : main eq3 ] ) . in general
, one needs to solve the dynamics of @xmath40 , @xmath75 and @xmath76 by the system of ordinary differential equations .
however , one easily sees that in ( [ e : main eq3 ] ) there exists a special case , in which the constant scalar affine curvature @xmath77 is a possible solution @xcite . recall that in order to conform with the positive kinetic energy argument , the condition ( [ e : condition ] ) is needed .
however , since the special case yields the negative curvature @xmath78 with a negative matter density @xmath79 , the condition of @xmath80 is required @xcite . in this section ,
we concentrate on the eos of the scalar - torsion mode in both special and normal cases .
we also present the cosmological evolution of the density ratio , defined by @xmath81 , from a high redshift to the current stage . in this special case
, we take the assumptions of @xmath83 and @xmath84 in @xcite .
the evolution equations ( [ e : main eq1 ] ) - ( [ e : main eq3 ] ) reduce to @xmath85 for the numerical calculation , we rescale the parameters as follows : @xmath86 where @xmath87 is the matter density at @xmath88 and the scalar affine curvature is a constant @xmath89 . from ( [ eq : sc01 ] ) , ( [ eq : sc02 ] ) and ( [ eq : sc03 ] ) , we obtain the following dimensionless equations , @xmath90 where the prime @xmath91 stands for @xmath92 and @xmath93 . using ( [ e : conservation_rho_t ] ) , ( [ e : w_t ] ) and ( [ eq : scrhot ] ) , we find that @xmath94 [ cols= " < , < " , ] in fig .
[ fg:2]a , we show the evolution of the density ratio , @xmath95 , as a function of the redshift @xmath96 .
the figure demonstrates that the torsion density @xmath54 dominates the universe in the high redshift regime ( @xmath97 ) with the general parameter and initial condition selection , while the matter - dominated regime is reached only within a very short time interval . in fig .
[ fg:2]b , we show that @xmath73 has an asymptotic behavior at the high redshift regime , @xmath61 @xmath98 . moreover ,
in the low redshift regime , it may even have a phantom crossing behavior , @xmath61 , the torsion eos could cross the phantom divide line of @xmath99 . as a result ,
the scalar - torsion mode is able to account for the late - time accelerating universe . finally , we remark that the studies in refs .
@xcite only indicated an oscillating behavior without the asymptotic one above for the torsion density .
we have studied the torsion eos of the two cases of the scalar - torsion mode in pgt of gravity , which are suitable for explaining the late - time accelerating universe but each of them possesses a quite different cosmological behavior in the high redshift regime . for the first case , which violates the positive kinetic energy and
has a constant affine curvature @xmath40 , the torsion eos has asymptotic behaviors : @xmath100 in the radiation - dominated stage , @xmath101 in the matter - dominated stage , and finally a late - time de - sitter solution corresponding to @xmath99 .
the torsion density ratio of @xmath102 in the high redshift regime is a `` negative '' constant . for the second one , which has the positive kinetic energy , under the general selection of parameters and initial conditions
, the torsion eos still shows an asymptotic behavior , @xmath103 , in the high redshift regime , while it could cross the phantom divide line in the low redshift regime . the most confusing phenomenon in this spin
@xmath5 scalar - torsion cosmology is that the universe is dominated by torsion in the high redshift regime even though there exists a narrow window for the matter - dominated epoch .
we are grateful to professor h. j. yo and professor j. m. nester for the inspiring and helpful discussions .
this work was partially supported by national center of theoretical science , national tsing hua university and national science council ( nsc-98 - 2112-m-007 - 008-my3 ) of r.o.c .
a. g. riess _ et al .
_ [ supernova search team collaboration ] , astron .
j. * 116 * , 1009 ( 1998 ) . s. perlmutter _ et al .
_ [ supernova cosmology project collaboration ] , astrophys .
j. * 517 * , 565 ( 1999 ) .
d. n. spergel _ et al . _
[ wmap collaboration ] , astrophys .
j. suppl .
* 148 * , 175 ( 2003 ) .
m. tegmark _ et al . _
[ sdss collaboration ] , phys .
* d69 * , 103501 ( 2004 ) .
s. w. allen , r. w. schmidt , h. ebeling , a. c. fabian , l. van speybroeck , mon . not .
soc . * 353 * , 457 ( 2004 ) .
e. j. copeland , m. sami and s. tsujikawa , int .
j. mod .
d * 15 * , 1753 ( 2006 ) ; m. li , x. -d .
li , s. wang and y. wang , commun .
* 56 * , 525 ( 2011 ) .
f. w. hehl , `` four lectures on poincare gauge field theory , '' in proc . of the 6th course of the school of cosmology and gravitation on spin , torsion , rotation , and supergravity , held at erice , italy , 1979 ( p. g. bergmann and v. de sabbata , eds . ) , p. 5 .
plenum , 1980 .
k. hayashi and t. shirafuji , prog .
* 64 * , 866 ( 1980 ) ; * 64 * , 883 ( 1980 ) ; * 64 * , 1435 ( 1980 ) ; * 64 * , 2222 ( 1980 ) ; * 65 * , 525 ( 1981 ) ; * 66 * , 318 ( 1981 ) ; * 66 * , 2258 ( 1981 ) . | we investigate the scalar - torsion mode in a cosmological model of the poincar gauge theory of gravity . we treat the geometric effect of torsion as an effective quantity , which behaves like dark energy , and study the effective equation of state ( eos ) of the model .
we concentrate on two cases with the constant curvature solution and positive kinetic energy , respectively . in the former , we find that the torsion eos has different values in the various stages of the universe
. in particular , it behaves like the radiation ( matter ) eos of @xmath0 ( @xmath1 ) in the radiation ( matter ) dominant epoch , while in the late time the torsion density is supportive for the accelerating universe . in the latter
, our numerical analysis shows that in general the eos has an asymptotic behavior in the high redshift regime , while it could cross the phantom divide line in the low redshift regime . |
lower dimensional models of shells are preferred in numerical computations to three - dimensional models when the thickness of the shells is ` very small ' .
a lot of work has been done on the lower dimensional approximation of boundary value and eigenvalue problem for elastic plates and shells ( cf .
recently some work has been done on the lower dimensional approximation of boundary value problem for piezoelectric shells ( cf .
@xcite ) . in this paper
, we would like to study the limiting behaviour of the eigenvalue problems for thin piezoelectric shallow shells .
we begin with a brief description of the problem and describe the results obtained .
let @xmath0 with @xmath1 and the mapping @xmath2 is given by @xmath3 for all @xmath4 , where @xmath5 is an injective mapping of class @xmath6 and @xmath7 is a unit normal vector to the middle surface @xmath8 of the shell .
let @xmath9 with meas(@xmath10 and meas(@xmath11 .
let @xmath12 and let @xmath13 .
the shell is clamped along the portion @xmath14 of the lateral surface .
then the variational form of the eigenvalue problem consists of finding the displacement vector @xmath15 , the electric potential @xmath16 and @xmath17 satisfying eq .
( [ eq : a9 ] ) .
we then show that the component of the eigenvector involving the electric potential @xmath18 can be uniquely determined in terms of the displacement vector @xmath19 and the problem thus reduces to finding @xmath20 satisfying equations ( [ eq : a31 ] ) and ( [ eq : a32 ] ) . after making appropriate scalings on the data and the unknowns , we transfer the problem to a domain @xmath21 which is independent of @xmath22 . then we show that the scaled eigensolutions converge to the solutions of a two - dimensional eigenvalue problem ( [ eq : e55 ] ) .
throughout this paper , latin indices vary over the set @xmath23 and greek indices over the set @xmath24 for the components of vectors and tensors . the summation over repeated indices will be used .
let @xmath25 be a bounded domain with a lipschitz continuous boundary @xmath26 and let @xmath27 lie locally on one side of @xmath26 .
let @xmath9 with meas@xmath28 and meas(@xmath11 .
let @xmath29 and @xmath30 .
for each @xmath31 , we define the sets @xmath32 \gamma^\epsilon_1 & = \gamma_1\times(-\epsilon , \epsilon),\quad \gamma^\epsilon_e=\gamma_e\times(-\epsilon , \epsilon),\quad \gamma^\epsilon_s=\gamma_s\times(-\epsilon , \epsilon).\end{aligned}\ ] ] let @xmath33 be a generic point on @xmath34 and let @xmath35 and @xmath36 .
we assume that for each @xmath22 , we are given a function @xmath37 of class @xmath6 .
we then define the map @xmath38 by @xmath39 at each point of the surface @xmath40 , we define the normal vector @xmath41 for each @xmath42 , we define the mapping @xmath43 by @xmath44 it can be shown that there exists an @xmath45 such that the mappings @xmath46 are @xmath47 diffeomorphisms for all @xmath48 .
the set @xmath49 is the reference configuration of the shell . for @xmath50
, we define the sets @xmath51 & \hat{\gamma}^\epsilon_e = \phi(\gamma^\epsilon_e),\quad \hat{\gamma}^\epsilon_s=\phi(\gamma^\epsilon_s),\quad \hat{\gamma}^\epsilon_{ed}=\hat{\gamma}^\epsilon_e\cup \hat{\gamma}^{\pm\epsilon}\end{aligned}\ ] ] and we define vectors @xmath52 and @xmath53 by the relations @xmath54 which form the covariant and contravariant basis respectively of the tangent plane of @xmath55 at @xmath56 .
the covariant and contravariant metric tensors are given respectively by @xmath57 the christoffel symbols are defined by @xmath58 note however that when the set @xmath34 is of the special form @xmath59 and the mapping @xmath60 is of the form ( [ eq : a2 ] ) , the following relations hold : @xmath61 the volume element is given by @xmath62 where @xmath63 it can be shown that there exist constants @xmath64 and @xmath65 such that @xmath66 for @xmath67 .
let @xmath68 and @xmath69 be the elastic , piezoelectric and dielectric tensors respectively .
we assume that the material of the shell is _ homogeneous and isotropic_. then the elasticity tensor is given by @xmath70 where @xmath71 and @xmath72 are the lam constants of the material .
these tensors satisfy the following coercive relations .
there exists a constant @xmath73 such that for all symmetric tensors @xmath74 and for any vector @xmath75 , @xmath76 & \hat{\cal e}^{kl,\epsilon}t_kt_l \geq c{\sum_{j=1}^3t^2_j}.\label{eq : a5}\end{aligned}\ ] ] moreover we have the symmetries @xmath77 then the eigenvalue problem consists of finding @xmath78 such that @xmath79 \hat{\sigma}^\epsilon(\hat{u}^\epsilon , \hat{\varphi}^\epsilon)\nu = 0 \mbox { on } \hat{\gamma}^\epsilon_n\\[.1pc ] \hat{u}^\epsilon=0 \mbox { on } \hat{\gamma}^\epsilon_0 \end{array } \right\},\label{eq : aa1}\\[.2pc ] \left .
\begin{array}{lcl } { \rm div}\hat{d}^\epsilon(\hat{u}^\epsilon , \hat{\varphi}^\epsilon)=0 \mbox { in } \hat{\omega}^\epsilon\\[.1pc ] \hat{d}^\epsilon(\hat{u}^\epsilon , \hat{\varphi}^\epsilon)\nu=0 \mbox { on } \hat{\gamma}_{s}^\epsilon\\[.1pc ] \hat{\varphi}^\epsilon=0 \mbox { on } \hat{\gamma}^\epsilon_{ed}. \end{array } \right\},\label{eq : aa2}\end{aligned}\ ] ] where @xmath80 \hat{d}_k & = \hat{p}^{kij,\epsilon}\hat{e}_{ij}^\epsilon + \hat{\cal e}^{kl,\epsilon}\hat{e}_l,\label{eq : aa4}\end{aligned}\ ] ] where @xmath81 and @xmath82 .
we define the spaces @xmath83 \hat{\psi}^\epsilon & = \{\hat{\psi}\in h^1(\hat{\omega}^\epsilon ) , \hat{\psi}|_{\hat{\gamma}_{ed}^\epsilon}=0\}.\label{eq : aa6}\end{aligned}\ ] ] then the variational form of systems ( [ eq : aa1 ] ) and ( [ eq : aa2 ] ) is to find @xmath84 such that @xmath85 where @xmath86 & \quad\ + \int_{\hat{\omega}^\epsilon}\hat{\cal e}^{ij,\epsilon}\hat{\partial}_i^\epsilon { \hat{\varphi}}^\epsilon\hat{\partial}_j^\epsilon\hat{\psi}^{\epsilon } { \rm d}\hat{x}^\epsilon\nonumber\\[.2pc ] & \quad\ + \int_{\hat{\omega}^\epsilon } \hat{p}^{mij,\epsilon}(\hat{\partial}^\epsilon_m { \hat{\varphi}}^\epsilon \hat{e}_{ij}^\epsilon(\hat{v}^\epsilon)- \hat{\partial}_m^\epsilon\hat{\psi}^\epsilon \hat{e}_{ij}^\epsilon(\hat{u}^\epsilon)){\rm d}\hat{x}^\epsilon,\label{eq : aa8}\end{aligned}\ ] ] @xmath87 @xmath88 since the mappings @xmath89 are assumed to be @xmath47 diffeomorphisms , the correspondences that associate with every element @xmath90 , the vector @xmath91 and with every element @xmath92 , the function @xmath93 induce bijections between the spaces @xmath94 and @xmath95 , and the spaces @xmath96 and @xmath97 respectively , where @xmath98 \psi^\epsilon & = \{\psi^\epsilon\in h^1(\omega^\epsilon)|\psi^{\epsilon } = 0 \ { \rm on } \ \gamma^\epsilon_{ed}\}.\label{eq : a7}\end{aligned}\ ] ] then we have @xmath99 & \hat{e}_{ij}(\hat{v})(\hat{x}^\epsilon ) = e^\epsilon_{k\|l}(v^\epsilon ) ( g^{k,\epsilon})_i(g^{l,\epsilon})_j,\label{eq : aa10}\end{aligned}\ ] ] where @xmath100 then the variational form ( [ eq : aa7 ] ) posed on the domain @xmath34 is to find @xmath101 such that @xmath102 where @xmath103 & \quad\ + \int_{\omega^\epsilon } { \cal e}^{ij,\epsilon}\partial_i^\epsilon { \varphi}^\epsilon\partial_j^\epsilon\psi^\epsilon\sqrt{g^\epsilon } { \rm d}x^\epsilon\nonumber\\[.2pc ] & \quad\ + \int_{\omega^\epsilon } p^{mij,\epsilon}(\partial^\epsilon_m{\varphi}^\epsilon e_{i\|j}^\epsilon(v^\epsilon)\nonumber\\[.2pc ] & \quad\ -\partial_m^\epsilon\psi^\epsilon e_{i\|j}^\epsilon(u^\epsilon))\sqrt{g^\epsilon}{\rm d}x^\epsilon,\label{eq : a10}\end{aligned}\ ] ] @xmath87 @xmath104 & a^{pqrs , \epsilon } = \hat{a}^{ijkl , \epsilon}(g^{p,\epsilon})_i \cdot ( g^{q,\epsilon})_j \cdot ( g^{r,\epsilon})_k \cdot ( g^{s,\epsilon})_l,\label{eq : aa12}\\[.2pc ] & { \cal e}^{pq,\epsilon } = \hat{\cal e}^{ij,\epsilon}(g^{p,\epsilon})_i \cdot ( g^{q,\epsilon})_j,\label{eq : aa13}\\[.2pc ] & p^{pqr,\epsilon } = \hat{p}^{ijk,\epsilon}(g^{p,\epsilon})_i \cdot ( g^{q,\epsilon})_j \cdot ( g^{r,\epsilon})_k .
\label{eq : aa14}\end{aligned}\ ] ] using the relations ( [ eq:1a2 ] ) , ( [ eq : a4 ] ) and ( [ eq : a5 ] ) , it can be shown that there exists a constant @xmath73 such that for all symmetric tensor @xmath74 and for any vector @xmath75 , @xmath105 & { \cal e}^{ij,\epsilon}t_it_j \geq c \sum_{i=1}^3t_i^2.\label{eq : aa16}\end{aligned}\ ] ] clearly the bilinear form associated with the left - hand side of ( [ eq : a9 ] ) is elliptic . hence by lax
milgram theorem , given @xmath106 and @xmath107 , there exists a unique @xmath108 such that @xmath109 in particular , for each @xmath110 , there exists a unique solution @xmath111 such that @xmath112 this is equivalent to the following equations . @xmath113 and
@xmath114 from relation ( 2.28 ) , it follows that the bilinear form associated with the left - hand side of ( [ eq : a15 ] ) is @xmath97-elliptic .
also for each @xmath115 , the mapping @xmath116 defines a linear functional on @xmath97 .
hence for each @xmath115 , there exists a unique @xmath117 such that @xmath118 and that @xmath119 is continuous .
in particular , it follows from ( [ eq : a15 ] ) and the above equation that @xmath120 and eqs ( [ eq : a14 ] ) and ( [ eq : a15 ] ) become @xmath121 & = \int_{\omega^\epsilon}\ ! f^\epsilon v^\epsilon \sqrt{g^\epsilon}{\rm d}x^\epsilon\ \ \ \forall v^\epsilon\in v^\epsilon,\label{eq : a22}\\[.2pc ] \int_{{\omega}^\epsilon}{\cal e}^{ij,\epsilon}{\partial}_i^\epsilon ( t^\epsilon(u^\epsilon)){\partial}_j^\epsilon{\psi}^{\epsilon } \sqrt{g^\epsilon}{\rm d}x^\epsilon & = \int_{{\omega}^\epsilon}{p}^{mij,\epsilon } { \partial}_m^\epsilon{\psi}^\epsilon { e}_{i\|j}^\epsilon({u}^\epsilon)\sqrt{g^\epsilon}{\rm d}x^\epsilon\nonumber\\[.2pc ] & \quad\ \forall \psi^\epsilon\in \psi^\epsilon.\label{eq : a23}\end{aligned}\ ] ] for each @xmath122 , there exists a unique @xmath123 such that @xmath124 \hskip -3.5pc & = \int_{\omega^\epsilon } h^\epsilon v^\epsilon \sqrt{g^\epsilon}{\rm d}x^\epsilon\quad \forall v^\epsilon\in v^\epsilon\label{eq : a24}\end{aligned}\ ] ] and that @xmath125 is continuous .
let @xmath126 denotes the bilinear form associated with the left - hand side of eq .
( [ eq : a22 ] ) . using ( [ eq : a23 ] )
, we have @xmath127 & \quad\ + \int_{{\omega}^\epsilon } { p}^{mij,\epsilon}{\partial}^\epsilon_m ( t^\epsilon(u^\epsilon ) ) { e}_{i\|j}^\epsilon({v}^\epsilon)\sqrt{g^\epsilon}{\rm d}x^\epsilon\nonumber\\[.2pc ] & = \int_{{\omega}^\epsilon}{a}^{ijkl,\epsilon}{e}_{k\|l}^\epsilon({u}^\epsilon ) { e}_{i\|j}^\epsilon({v}^\epsilon)\sqrt{g^\epsilon}{\rm d}x^\epsilon\nonumber\\[.2pc ] & \quad\ + \int_{{\omega}^\epsilon}{\cal e}^{ij,\epsilon}{\partial}_i^\epsilon ( t^\epsilon(u^\epsilon)){\partial}_j^\epsilon(t^\epsilon(v^\epsilon ) ) \sqrt{g^\epsilon}{\rm d}x^\epsilon\nonumber\end{aligned}\ ] ] @xmath128 & \quad\ + \int_{{\omega}^\epsilon}{\cal e}^{ij,\epsilon}{\partial}_i^\epsilon ( t^\epsilon(v^\epsilon)){\partial}_j^\epsilon(t^\epsilon(u^\epsilon ) ) \sqrt{g^\epsilon}{\rm d}x^\epsilon\nonumber\\[.2pc ] & = b^\epsilon(v^\epsilon , u^\epsilon).\label{eq : a25}\end{aligned}\ ] ] also , using ( [ eq : a23 ] ) and the relations ( [ eq : aa15 ] ) and ( [ eq : aa16 ] ) , we have @xmath129 & \quad\ + \int_{{\omega}^\epsilon } { p}^{mij,\epsilon}{\partial}^\epsilon_m ( t^\epsilon(u^\epsilon ) ) { e}_{i\|j}^\epsilon({u}^\epsilon)\sqrt{g^\epsilon}{\rm d}x^\epsilon\nonumber\\[.2pc ] & = \int_{{\omega}^\epsilon}{a}^{ijkl,\epsilon}{e}_{k\|l}^\epsilon({u}^\epsilon ) { e}_{i\|j}^\epsilon({u}^\epsilon)\sqrt{g^\epsilon}{\rm d}x^\epsilon\nonumber\\[.2pc ] & \quad\ + \int_{{\omega}^\epsilon}{\cal e}^{ij,\epsilon}{\partial}_i^\epsilon ( t^\epsilon(u^\epsilon)){\partial}_j^\epsilon(t^\epsilon(u^\epsilon ) ) \sqrt{g^\epsilon}{\rm d}x^\epsilon\nonumber\\[.2pc ] & \geq c\|u^\epsilon\|^2_{v^\epsilon}.\label{eq : a26}\end{aligned}\ ] ] hence @xmath130 is symmetric and @xmath95-elliptic . hence by lax milgram theorem , there exists a unique @xmath131 satisfying ( [ eq : a24 ] ) . letting @xmath132 in ( [ eq : a24 ] ) , we get @xmath133 & \quad\ + \int_{{\omega}^\epsilon } { p}^{mij,\epsilon}{\partial}^\epsilon_m ( t^\epsilon(g^\epsilon(h^\epsilon))){e}_{i\|j}^\epsilon(g^\epsilon(h^\epsilon ) ) \sqrt{g^\epsilon}{\rm d}x^\epsilon\nonumber\\[.2pc ] & = \int_{\omega^\epsilon } h^\epsilon g^\epsilon(h^\epsilon ) \sqrt{g^\epsilon}{\rm d}x^\epsilon\label{eq : a27 } .\end{aligned}\ ] ] using ( [ eq : a23 ] ) , it becomes @xmath133 & \quad\ + \int_{{\omega}^\epsilon}{\cal e}^{ij,\epsilon}{\partial}_i^\epsilon ( t^\epsilon(g^\epsilon(h^\epsilon))){\partial}_j^\epsilon ( t^\epsilon(g^\epsilon(h^\epsilon ) ) ) \sqrt{g^\epsilon}{\rm d}x^\epsilon\nonumber\\[.2pc ] & = \int_{\omega^\epsilon } h^\epsilon g^\epsilon(h^\epsilon ) \sqrt{g^\epsilon}{\rm d}x^\epsilon.\label{eq : a28}\end{aligned}\ ] ] using the relations ( [ eq : aa15 ] ) and ( [ eq : aa16 ] ) , we have @xmath134 hence @xmath135 which implies that @xmath136 is continuous .
it follows from ( [ eq : a22 ] ) and the above lemma that @xmath137 .
since the inclusion @xmath138 is compact , it follows that @xmath139 is compact . also since the bilinear form @xmath130 is symmetric
, it follows that @xmath136 is self - adjoint .
hence from the spectral theory of compact , self - adjoint operators , it follows that there exists a sequence of eigenpairs @xmath140 such that @xmath141 & \quad\ + \int_{{\omega}^\epsilon } { p}^{mij,\epsilon}{\partial}^\epsilon_m ( t^\epsilon(u^{m,\epsilon } ) ) { e}_{i\|j}^\epsilon({v}^\epsilon ) \sqrt{g^\epsilon}{\rm d}x^\epsilon\nonumber\\[.2pc ] & = \xi^{m,\epsilon}\int_{\omega^\epsilon } u^{m,\epsilon } v^\epsilon \sqrt{g^\epsilon}{\rm d}x^\epsilon\quad \forall v^\epsilon\in v^\epsilon,\label{eq : a31}\\[.2pc ] & \int_{{\omega}^\epsilon}{\cal e}^{ij,\epsilon}{\partial}_i^\epsilon ( t^\epsilon(u^{m,\epsilon})){\partial}_j^\epsilon{\psi}^{\epsilon } \sqrt{g^\epsilon}{\rm d}x^\epsilon\nonumber\\[.2pc ] & = \int_{{\omega}^\epsilon}{p}^{mij,\epsilon } { \partial}_m^\epsilon{\psi}^\epsilon { e}_{i\|j}^\epsilon({u}^{m,\epsilon})\sqrt{g^\epsilon}{\rm d}x^\epsilon
\quad \forall \psi^\epsilon\in \psi^\epsilon,\label{eq : a32}\\[.2pc ] & 0<\xi^{1,\epsilon } \leq\xi^{2,\epsilon}\leq \cdots \leq\xi^{m,\epsilon}\leq
\cdots \rightarrow \infty,\label{eq : a33}\\[.2pc ] & \int_{\omega^\epsilon}u^{m,\epsilon}_i u^{n,\epsilon}_i\sqrt{g^\epsilon}{\rm d}x^\epsilon = \epsilon^3 \delta_{mn}.\label{eq : a34}\end{aligned}\ ] ] the sequence @xmath142 forms a complete orthonormal basis for @xmath143 .
define the rayleigh quotient @xmath144 for @xmath145 by @xmath146 then @xmath147 where @xmath148 denotes the collection of all @xmath149-dimensional subspaces of @xmath95 .
we now perform a change of variable so that the domain no longer depends on @xmath22 . with @xmath150 , we associate @xmath151 .
let @xmath152 \gamma_e & = \gamma_e\times(-1 , 1),\quad \gamma_s = \gamma_s\times(-1 , 1),\\[.2pc ] \gamma_n & = \gamma_1\cup\gamma^{+}\cup\gamma^{-},\quad \gamma_{ed}=\gamma^{+}\cup\gamma^{-}\cup\gamma_e.\end{aligned}\ ] ] with the functions @xmath153 , we associate the functions @xmath154 defined by @xmath155 we assume that the shell is a shallow shell , i.e. there exists a function @xmath156 such that @xmath157 i.e. , the curvature of the shell is of the order of the thickness of the shell .
with the tensors @xmath159 , we associate the tensors @xmath160 through the relation @xmath161 we define the spaces @xmath162 \psi(\omega ) & = \{\psi\in h^1(\omega ) , \psi|_{\gamma_{ed}}=0\}.\label{eq : b12}\end{aligned}\ ] ] we denote @xmath163 .
then the variational equations ( eqs ( [ eq : a31])([eq : a34 ] ) ) become @xmath164 & \quad\ + \int_\omega p^{3kl}\partial_3{\varphi}^m ( \epsilon)e_{k\|l}(\epsilon , v)\sqrt{g(\epsilon)}{\rm d}x\nonumber\\[.2pc ] & \quad\ + \epsilon\int_\omega p^{\alpha kl}(\epsilon ) \partial_\alpha{\varphi}^m(\epsilon ) e_{k\|l}(\epsilon , v)\sqrt{g(\epsilon)}{\rm d}x\nonumber\\[.2pc ] & = \xi^m(\epsilon)\int_\omega [ \epsilon^2 u^m_\alpha(\epsilon)v_\alpha + u^m_3(\epsilon)v_3]\sqrt{g(\epsilon)}{\rm d}x \quad\mbox{for all } v\in v(\omega ) .
\label{eq : b13}\\[.2pc ] & \int_\omega { \cal e}^{33}(\epsilon)\partial_3{\varphi}^m(\epsilon ) \partial_3\psi \sqrt{g(\epsilon)}{\rm d}x\nonumber\\[.2pc ] & \quad\ + \epsilon\int_\omega[{\cal e}^{3\alpha}(\epsilon ) ( \partial_\alpha{\varphi}^m(\epsilon ) \partial_3\psi+\partial_3{\varphi}^m(\epsilon)\partial_\alpha\psi ) ] \sqrt{g(\epsilon)}{\rm d}x\nonumber\\[.2pc ] & \quad\ + \epsilon^2\int_\omega { \cal e}^{\alpha\beta}(\epsilon ) \partial_\alpha{\varphi}^m(\epsilon ) \partial_\beta\psi\sqrt{g(\epsilon ) } { \rm d}x\nonumber\\[.2pc ] & = \int_\omega p^{3kl}(\epsilon)\partial_3\psi e_{k\|l}(\epsilon , u^m(\epsilon))\sqrt{g(\epsilon)}{\rm d}x \nonumber\\[.2pc ] & \quad\ + \epsilon\int_\omega [ p^{\alpha kl}(\epsilon ) \partial_\alpha\psi e_{k\|l}(\epsilon , u^m(\epsilon ) ) ] \sqrt{g(\epsilon)}{\rm d}x \quad\mbox{for all } \psi\in \psi(\omega),\label{eq : b14}\\[.2pc ] & \int_\omega [ \epsilon^2u^m_\alpha(\epsilon ) u^n_\alpha(\epsilon ) + u^m_3(\epsilon)u^n_3(\epsilon)]\sqrt{g(\epsilon)}{\rm d}x = \delta_{mn}. \label{eq : b15}\end{aligned}\ ] ]
the following two lemmas are crucial ; they play an important role in the proof of the convergence of the scaled unknowns as @xmath165 . in the sequel
, we denote by @xmath166 various constants whose values do not depend on @xmath22 but may depend on @xmath5 .
the functions @xmath167 defined in ( [ eq : b10 ] ) are of the form @xmath168 e_{\alpha\|3}(\epsilon ; v ) & = \frac{1}{\epsilon}\{{\tilde{e}}_{\alpha 3}(v ) + \epsilon^2 e^{\sharp}_{\alpha\|3}(\epsilon ; v)\},\label{eq : c2}\\[.2pc ] e_{3\|3}(\epsilon ; v ) & = \frac{1}{\epsilon^2}{\tilde{e}}_{33}(v),\label{eq : c3}\end{aligned}\ ] ] where @xmath169 { \tilde{e}}_{\alpha 3}(v ) & = { \frac{1}{2}}(\partial_\alpha v_3+\partial_3 v_\alpha),\label{eq : c5}\\[.2pc ] { \tilde{e}}_{33}(v ) & = \partial_3 v_3\label{eq : c6}\end{aligned}\ ] ] and there exists constant @xmath170 such that @xmath171 also there exist constants @xmath172 and @xmath173 such that @xmath174 & \sup_{0<\epsilon\leq \epsilon_0}\max_{x\in\omega}|a^{ijkl}(\epsilon)-a^{ijkl}| \leq c_3\epsilon^2,\label{eq : c9}\end{aligned}\ ] ] where @xmath175 and @xmath176 for @xmath50 and for all symmetric tensors @xmath177 the proof is based on lemma 4.1 of @xcite . from relation ( [ eq : a5 ] ) and definition ( [ eq : b2 ] ) , it follows that there exists a constant @xmath178 such that for any vector @xmath75 , @xmath179 we assume that there exists functions @xmath180 and @xmath181 such that @xmath182 & \sup_{0<\epsilon\leq \epsilon_0}\max_{x\in\omega}|{\cal e}^{ij } ( \epsilon)- { \cal e}^{ij}| \leq c_7\epsilon.\label{eq : b5}\end{aligned}\ ] ] let @xmath156 be a given function and let the functions @xmath183 be defined as in ( [ eq : c4])([eq : c6 ] ) .
then there exists a constant @xmath184 such that the following generalised korn s inequality holds : @xmath185 for all @xmath186 where @xmath187 is the space defined in ( [ eq : b11 ] ) . the proof is based on lemma 4.2 of @xcite .
in this section , we show that for each positive integer @xmath149 , the scaled eigenvalues @xmath188 are bounded uniformly with respect to @xmath22 . for each @xmath196 , there exists a unique solution @xmath197 such that @xmath198 \sqrt{g(\epsilon)}{\rm d}x\nonumber\\ & \quad\ + \epsilon^2\int_\omega { \cal e}^{\alpha\beta}(\epsilon)\partial_\alpha t(\epsilon)(h ) \partial_\beta\psi \sqrt{g(\epsilon)}{\rm d}x\nonumber\\ & = \int_\omega p^{3kl}(\epsilon ) \partial_3\psi e_{k\|l}(\epsilon , h)\sqrt{g(\epsilon)}{\rm d}x\nonumber\\ & \quad\ + \epsilon\int_\omega p^{\alpha kl}(\epsilon ) \partial_\alpha\psi e_{k\|l}(\epsilon , h)\sqrt{g(\epsilon)}{\rm
d}x\quad \forall \psi\in\psi.\end{aligned}\ ] ] taking @xmath199 and @xmath200 in the above equation , we have @xmath201 \sqrt{g(\epsilon)}{\rm d}x\nonumber\\ & \quad\ + \epsilon^2\int_\omega { \cal e}^{\alpha\beta}(\epsilon)\partial_\alpha t(\epsilon)(v_\varphi ) \partial_\beta t(\epsilon)(v_\varphi ) \sqrt{g(\epsilon)}{\rm d}x\nonumber\\ & = \int_\omega p^{3kl}(\epsilon ) \partial_3 t(\epsilon)(v_\varphi ) e_{k\|l}(\epsilon , v_\varphi)\sqrt{g(\epsilon)}{\rm d}x\nonumber\\ & \quad\ + \epsilon\int_\omega p^{\alpha kl}(\epsilon ) \partial_\alpha t(\epsilon)(v_\varphi)e_{k\|l}(\epsilon , v_\varphi ) \sqrt{g(\epsilon)}{\rm d}x.\end{aligned}\ ] ] it follows from lemma 6.2 in @xcite that the bilinear form @xmath346 defined by @xmath347 is @xmath334-elliptic and symmetric . hence it is clear that @xmath345 is also @xmath334-elliptic and symmetric .
let @xmath348 , be the eigensolutions of problem ( [ eq : e56 ] ) found as limits of the subsequence @xmath349 of eigensolutions of the problem ( [ eq : b13 ] ) .
then the sequence @xmath350 comprises all the eigenvalues , counting multiplicities , of problem ( [ eq : e56 ] ) and the associated sequence @xmath351 of eigenfunctions forms a complete orthonormal set in the space @xmath352 .
99 bernadou m and haenel c , modelization and numerical approximation of piezoelectric thin shells , parts i , ii , iii : rapport de recherche , der - cs ( ecole suprieure dingnierie lonard de vinci , france ) ( 2002 ) no rr-6 , 7 , 8 | in this paper we consider the eigenvalue problem for piezoelectric shallow shells and we show that , as the thickness of the shell goes to zero , the eigensolutions of the three - dimensional piezoelectric shells converge to the eigensolutions of a two - dimensional eigenvalue problem .
= msam10 at 10pt = tibi at 10.4pt [ theo]*theorem * [ theo]lemma [ theo]remark [ theo]definition [ theo]corollary |
let @xmath1 be a compact , connected , orientable @xmath0-manifold with a torus boundary component @xmath2 .
a _ slope _ on @xmath2 is the isotopy class of an unoriented essential simple loop . for a slope @xmath3 , the manifold obtained from @xmath1 by _ @xmath3-dehn filling _ is @xmath4 , where @xmath5 is a solid torus glued to @xmath1 along @xmath2 in such a way that @xmath3 bounds a meridian disk in @xmath5 .
if @xmath3 and @xmath6 are two slopes on @xmath2 , then @xmath7 denotes their minimal geometric intersection number .
we say that a @xmath0-manifold @xmath1 is _ hyperbolic _ if @xmath1 with its boundary tori removed admits a complete hyperbolic structure of finite volume with totally geodesic boundary .
if @xmath1 has non - empty boundary , then thurston s geometrization theorem for haken manifolds @xcite says that @xmath1 is hyperbolic if and only if @xmath1 contains no essential sphere , disk , torus or annulus .
we are interested in obtaining restrictions on when a dehn filling on a hyperbolic @xmath0-manifold fails to be hyperbolic .
such a filling is said to be _
exceptional_. it is well known that if @xmath1 is hyperbolic , then there are only finitely many exceptional dehn fillings on @xmath2 @xcite , and there are a large amount of investigations on exceptional dehn fillings ( see @xcite ) . in this paper
, we deal with three specific exceptional dehn fillings .
a @xmath0-manifold is _
@xmath8-reducible _ if it contains a projective plane , and _
@xmath8-irreducible _ otherwise . if a closed @xmath0-manifold is @xmath8-reducible , then it is either the real @xmath0-dimensional projective space @xmath9 or a reducible manifold with a @xmath9-summand .
a @xmath0-manifold is _ toroidal _ if it contains an essential torus .
clearly , if @xmath10 is @xmath8-reducible or toroidal , then such a filling is exceptional .
finally a hyperbolic @xmath0-manifold contains no klein bottle , and if @xmath10 contains a klein bottle , then the filling is also exceptional ( see @xcite ) .
we consider two situations .
that is , one filling yields a @xmath8-reducible manifold , and the other gives either a toroidal manifold or a manifold containing a klein bottle . for each case
, we can find the best possible upper bound on the distance between two exceptional fillings .
let @xmath1 be a hyperbolic @xmath0-manifold with a torus boundary component @xmath2 .
let @xmath11 and @xmath12 be two slopes on @xmath2 such that @xmath13 is @xmath8-reducible and @xmath14 is toroidal . then either * @xmath15 ; or * @xmath16 and @xmath14 contains an essential torus which intersects the core of the attached solid torus in two points .
let @xmath1 be as in theorem 1.1 .
let @xmath11 and @xmath12 be two slopes on @xmath2 such that @xmath13 is @xmath8-reducible and @xmath17 contains a klein bottle .
then either * @xmath18 ; or * @xmath19 and @xmath17 contains a klein bottle which intersects the core of the attached solid torus in a single point .
the examples showing that these estimates are sharp are given in the final section .
in @xcite , gordon gave an upper bound @xmath20 for the distance between a toroidal filling and a lens space filling on a hyperbolic @xmath0-manifold with torus boundary .
our theorem 1.1 gives a partial improvement of this result in case where a lens space is the real projective @xmath0-space @xmath9 .
let @xmath1 be a hyperbolic @xmath0-manifold with torus boundary @xmath21 , and let @xmath11 and @xmath12 be slopes on @xmath21 such that @xmath13 is the lens space @xmath22 and @xmath14 is toroidal
. then @xmath23 .
the authors would like to thank cameron gordon for suggesting the problem .
in the remainder of this paper , we assume that @xmath1 is a hyperbolic @xmath0-manifold with a torus boundary component @xmath2 , and that @xmath24 are slopes on @xmath2 such that @xmath13 is @xmath8-reducible , @xmath14 is toroidal , and @xmath17 contains a klein bottle .
assume that @xmath25 .
suppose that @xmath26 and @xmath17 contain a projective plane @xmath27 , an essential torus @xmath28 and a klein bottle @xmath29 respectively .
then we may assume that @xmath27 meets the attached solid torus @xmath30 in a finite collection of meridian disks , so that @xmath31 is a punctured projective plane properly embedded in @xmath1 , each of whose boundary components has slope @xmath11 .
furthermore , we can assume that @xmath27 is chosen so that the number of boundary components @xmath32 is minimal among all projective planes in @xmath13 .
similarly , @xmath28 and @xmath29 give rise to the surfaces @xmath33 and @xmath34 respectively , and the numbers of boundary components @xmath35 and @xmath36 of @xmath33 and @xmath34 are assumed to be minimal . recall that @xmath13 is either the real projective @xmath0-space @xmath9 , or a reducible manifold with a @xmath9-summand .
therefore we may assume that @xmath14 and @xmath17 are irreducible by @xcite .
[ p2 ] @xmath37 . if @xmath38 , then @xmath1 contains a projective plane , which is impossible since @xmath1 is hyperbolic . if @xmath39 , then @xmath1 contains a mbius band , which is also impossible .
@xmath40 is incompressible and boundary - incompressibe in @xmath1 .
assume @xmath40 is compressible in @xmath1 .
let @xmath41 be a compressing disk for @xmath40 .
note that @xmath42 is orientation - preserving on @xmath40 ( and hence @xmath27 ) .
hence @xmath42 bounds a disk @xmath43 on @xmath27 .
since @xmath44 meets @xmath30 , we can create a new projective plane by replacing @xmath43 with @xmath41 , which meets @xmath30 fewer than @xmath27 .
this contradicts the minimality of @xmath27 .
therefore @xmath40 is incompressible in @xmath1 .
next , assume that @xmath40 is boundary - compressible .
then @xmath40 would be compressible , or the core of @xmath30 can be isotoped into @xmath27 as an orientation - reversing loop .
but , the latter case implies that @xmath1 is boundary - reducible .
since @xmath35 is minimal , it is clear that @xmath33 is incompressible and boundary - incompressible in @xmath1 .
we have @xmath45 , and @xmath46 , since @xmath1 can not contain a klein bottle .
@xmath34 is incompressible and boundary - incompressible in @xmath1 .
suppose that @xmath34 is compressible in @xmath1 .
let @xmath41 be a disk in @xmath1 such that @xmath47 and @xmath42 does not bound a disk on @xmath34 .
note that @xmath42 is orientation - preserving on @xmath34 .
if @xmath42 is non - separating on @xmath29 , then we get a non - separating @xmath48-sphere in @xmath17 by compressing @xmath29 along @xmath41 .
this is clearly a contradiction .
if @xmath42 bound a disk on @xmath29 , then we replace the disk with @xmath41 , and get a new klein bottle in @xmath17 with fewer intersections with @xmath49 than @xmath29 .
this contradicts the choice of @xmath29 .
thus @xmath42 is essential and separating on @xmath29 .
compressing @xmath29 along @xmath41 gives two disjoint projective planes in @xmath17 .
since @xmath17 is irreducible , this is also impossible .
thus we have shown that @xmath34 is incompressible . next , let @xmath50 be a disk in @xmath1 such that @xmath51 , @xmath52 , where @xmath53 is an essential ( i.e. not boundary - parallel ) arc in @xmath34 and @xmath54 .
if @xmath55 joins distinct components of @xmath56 , then a compressing disk for @xmath34 is obtained from two parallel copies of @xmath50 and the disk obtained by removing a neighborhood of @xmath57 from the annulus in @xmath2 cobounded by those components of @xmath56 meeting @xmath55 .
hence @xmath58 is contained in the same component @xmath59 , say , of @xmath56 .
if @xmath60 , then @xmath57 bounds a disk @xmath43 in @xmath2 together with a subarc of @xmath59 . then @xmath61 gives a compressing disk for @xmath34 in @xmath1 .
therefore @xmath62 .
then we can move the core of @xmath49 onto an orientation - reversing loop in @xmath63 by using @xmath50 .
this implies that @xmath1 contains a properly embedded mbius band , which contradicts the fact that @xmath1 is hyperbolic . from the arc components of @xmath64 , @xmath65
, we can construct two pairs of graphs @xmath66 and @xmath67 in the usual way ( see @xcite ) .
we number the components of @xmath68 as @xmath69 in the order in which they appear on @xmath2 .
similarly number the components of @xmath56 .
but the components of @xmath70 are numbered from @xmath71 to @xmath72 unusually . for simplifying the notations
, we use the symbol @xmath73 for both @xmath74 and @xmath75 , and @xmath76 for @xmath77 or @xmath78 . for a graph @xmath79 , the _ reduced graph _ @xmath80 of @xmath79
is defined to be the graph obtained from @xmath79 by amalgamating each family of parallel edges into a single edge .
neither @xmath73 nor @xmath76 has trivial loops .
this follows from the fact that @xmath81 and @xmath34 are boundary - incompressible .
we may assume that no circle component of @xmath64 or @xmath65 bounds a disk in @xmath81 or @xmath34 , because of the incompressibilities of these surfaces .
although @xmath40 and @xmath34 are non - orientable , we can establish a parity rule , which plays a crucial role in this paper .
in fact , this is a natural generalization of the usual parity rule @xcite .
first , orient all components of @xmath70 so that they are mutually homologous on @xmath2 .
similarly for @xmath68 and @xmath56 .
let @xmath82 be an edge in @xmath73 . since @xmath82 is an arc properly embedded in @xmath40 , a regular neighborhood @xmath41 of @xmath82 in @xmath40 is a disk in @xmath40 .
then @xmath83 , where @xmath55 and @xmath84 are arcs in @xmath70 with induced orientations from @xmath70 .
on @xmath41 , if @xmath55 and @xmath84 have opposite directions , then @xmath82 is called _ positive _ , otherwise _
negative_. see figure [ parity ] .
similarly , define the sign of edges in @xmath76 .
then we have the following rule .
an edge @xmath82 is positive or negative in @xmath73 if and only if @xmath82 is negative positive resp . in @xmath76 .
this follows from the fact that @xmath1 is orientable and @xmath2 is a torus .
remark that @xmath77 can have positive loops , but no negative loops . @xmath73 and
@xmath78 can have positive and negative loops . but the key point is ; [ loop ] * at most one vertex can be a base of negative loops in @xmath73 .
* at most two vertices can be bases of negative loops in @xmath78 .
let @xmath82 be a negative loop based at a vertex @xmath85 .
then @xmath86 is a mbius band .
since @xmath27 and @xmath29 can contain at most one mbius band and at most two disjoint mbius bands respectively , the conclusions follow . now , we have some basic properties of the graphs .
an edge is called an _ @xmath85-edge _ if it has label @xmath85 at its endpoint , and a _ level edge _ if both endpoints have the same label .
for example , @xmath87 can have a positive level edge , which corresponds to a negative loop in @xmath73 by the parity rule .
thus at most one label of @xmath76 can be a label of positive level edges by lemma [ loop](1 ) .
therefore , we adopt the convention that @xmath88 is the label of positive level edges in @xmath76 .
let @xmath89 or @xmath76 .
a cycle in @xmath79 is a _
scharlemann cycle _ if it bounds a disk face , and the edges in the cycle are all positive and have the same label pair @xmath90 at their two endpoints , called the _
label pair _ of the scharlemann cycle .
( in this case , the label set of @xmath79 must have at least two elements . ) in particular , a scharlemann cycle of length two is called an _ @xmath91-cycle _ in short .
when @xmath92 , a _ generalized scharlemann cycle _ in @xmath76 is defined to be either a scharlemann cycle or a cycle of positive edges whose labels are in the set @xmath93 at both ends and which bounds a disk face in @xmath94 .
in addition , if @xmath95 , then the label @xmath88 must appear around each vertex in the disk face bounded by the generalized scharlemann cycle in @xmath94 .
for example , figure 2(b ) is not a generalized scharlemann cycle .
refer to @xcite for more details .
a generalized scharlemann cycle of length two is called a _ generalized @xmath91-cycle_. then a generalized @xmath91-cycle , not an @xmath91-cycle , in @xmath76 is a triple @xmath96 of mutually parallel positive edges where @xmath97 and @xmath98 have the same label pair @xmath99 , and @xmath100 is a level edge with label @xmath88 . a generalized @xmath91-cycle , not an @xmath91-cycle , can be defined in @xmath73 ( precisely , @xmath75 ) , but then the label set is @xmath101 for some @xmath102 .
[ ss ] @xmath76 contains no generalized scharlemann cycles .
this is ( * ? ? ?
* theorem 1.1 ) .
[ positive ] * if @xmath32 is even , @xmath76 has at most @xmath103 mutually parallel positive edges .
furthermore , if there is a family of @xmath103 mutually parallel positive edges , then either the first , say , edge of the family is level , or the set of labels of one end of the family is disjoint from that of the other end . *
if @xmath32 is odd , @xmath76 has at most @xmath104 mutually parallel positive edges .
furthermore , if there is a family of @xmath104 mutually parallel positive edges , then the first , say , edge of the family is level . these follow from lemmas [ loop](1 ) and [ ss ] .
[ p - parallel ] let @xmath37 .
then @xmath76 can not contain @xmath32 mutually parallel edges .
let @xmath105 be a family of @xmath32 mutually parallel edges in @xmath76 labelled successively .
by lemma [ positive ] , all @xmath106 s are negative , and then they make orientation - preserving cycles in @xmath73 .
note that an orientation - preserving loop in a projective plane is contractible .
therefore , we can choose an innermost cycle among them .
then the construction in ( * ? ? ?
* section 5 ) implies that @xmath1 is cabled , a contradiction .
in this section , we use @xmath107 to denote the number of vertices of @xmath76 . for a label @xmath85 of @xmath76 ,
let @xmath108 be the subgraph of @xmath76 consisting of all the vertices and positive @xmath85-edges of @xmath76 .
a disk face @xmath41 of @xmath108 is called an _
@xmath85-face _ of @xmath76 .
let @xmath109 denote the subgraph of @xmath73 consisting of all the vertices of @xmath73 and the positive edges of @xmath73 .
note that @xmath109 has a disk support in @xmath27 , that is , there is a disk in @xmath27 which contains @xmath109 in its interior , since any orientation - preserving loop in a projective plane is contractible .
let @xmath110 be a subgraph of @xmath109 with a disk support @xmath41 .
a vertex of @xmath110 is a _ boundary vertex _
if there is an arc connecting it to @xmath42 whose interior is disjoint from @xmath110 , and an _ interior vertex _ otherwise . a _ generalized web _ @xmath111 is a connected subgraph of @xmath109 satisfying that * at most one boundary vertex @xmath112 of @xmath111 is a cut vertex of @xmath109 ; * each vertex of @xmath111 , except @xmath112 if it exists , has degree at least @xmath113 in @xmath111 ; * there is a disk @xmath41 in @xmath27 such that @xmath114 .
the vertex @xmath112 as in ( i ) , if it exists , is called an _ exceptional vertex _ of @xmath111 .
[ gweb ] one of the followings holds * @xmath76 contains an @xmath85-face for some @xmath115 * @xmath73 contains a generalized web .
we distinguish two cases .
* case 1 : * suppose that there is a vertex @xmath116 of @xmath73 such that more than @xmath107 negative edges are incident to @xmath85 .
remark that such negative edges are not loops because @xmath115 .
this implies that there exist more than @xmath107 positive @xmath85-edges in @xmath76 by the parity rule .
thus @xmath108 has a larger number of edges than that of vertices .
an euler characteristic calculation gives that @xmath108 contains a disk face .
* case 2 : * as the negation of case 1 , suppose that each vertex @xmath116 of @xmath73 has at least @xmath113 positive edge endpoints .
let @xmath110 be an extremal component of @xmath109 .
that is , @xmath110 is a component of @xmath109 having a disk support @xmath41 such that @xmath117 .
first , assume that @xmath110 is a single vertex
. then it must be vertex @xmath88 .
thus only negative edges are incident to vertex @xmath88 in @xmath73 .
if there is no negative loop at vertex @xmath88 , then @xmath110 can be the only extremal component of @xmath109 .
but then , @xmath76 contains a @xmath88-face @xmath50 as in case 1 , and @xmath76 has no positive level edges . then @xmath50 contains a scharlemann cycle by ( * ? ? ?
* proposition 5.1 ) , which is impossible by lemma [ ss ] .
if a negative loop is incident there , we can choose another extremal component of @xmath109 , which has more than one vertex .
thus we can assume that @xmath110 is not a single vertex .
choose a block @xmath111 of @xmath110 with at most one cut vertex .
then @xmath111 is clearly a generalized web .
in this section , we treat the case ( 1 ) of proposition [ gweb ] .
[ gene ] let @xmath92 .
if @xmath76 contains a non - zero @xmath85-face @xmath41 , then it contains a generalized scharlemann cycle in @xmath41 .
there is a possibility that @xmath42 is not a circle .
that is , @xmath42 may contain a double edge , and also more than two edges of @xmath42 may be incident to a vertex on @xmath42 .
since we will find a generalized scharlemann cycle within @xmath41 , we can cut formally the graph @xmath118 along double edges of @xmath42 and at vertices to which more than two edges of @xmath42 are incident so that @xmath42 is deformed into a circle .
( see also ( * ? ? ?
* section 5 ) . )
thus we may assume that @xmath42 is a circle .
if @xmath41 is a bigon , then the conclusion is obvious .
therefore , we assume that @xmath41 has at least three sides .
suppose that @xmath41 has a diagonal edge @xmath82 , which has two distinct labels @xmath55 and @xmath57 at its endpoints as in figure [ split](a ) . since @xmath41 is an @xmath85-face , @xmath119 and @xmath120 .
without loss of generality , we may assume that the labels appear in counterclockwise order around the boundary of each vertex , and that @xmath121 . formally , construct a new @xmath85-face @xmath43 as follows .
the edge @xmath82 divides @xmath41 into two disks @xmath122 and @xmath123 .
we can assume that @xmath122 lies on the right side of @xmath82 when @xmath82 is oriented from the endpoint with label @xmath55 .
if three labels @xmath124 appear in this order around the corners of @xmath42 , then discard @xmath123 , and insert additional edges to the left of @xmath82 , and parallel to @xmath82 , until we first reach label @xmath85 at one or both ends of an additional edge .
see figure [ split](b ) .
the new @xmath85-face @xmath43 is the union of @xmath122 and some additional bigons .
if three label @xmath125 appear in this order , then discard @xmath122 , and insert additional edges to the right of @xmath82 as above .
then @xmath43 is the union of @xmath123 and some additional bigons .
remark that there is no generalized scharlemann cycle among additional edges and @xmath82 .
repeat the above process for every diagonal edge which is not level , then get a new @xmath85-face @xmath50 and a graph @xmath126 in @xmath50 .
all diagonal edges of @xmath126 are level , and all boundary edges are @xmath85-edges , especially label @xmath85 can appear on both ends of a boundary edge .
such boundary edges are called level @xmath85-edges to distinguish them from level @xmath88-edges .
see figure [ gamma ] . ] from now on , we assume that there is no scharlemann cycle in @xmath126 .
@xmath126 contains a level @xmath88-edge .
suppose not .
then @xmath126 contains a scharlemann cycle by ( * ? ? ?
* lemma 5.2 ) .
let @xmath127 be the disk face of @xmath128 , which is not a bigon .
we will show that for any edge of @xmath126 on @xmath129 , it has labels only on @xmath93 at both ends .
there are consecutive vertices @xmath130 on @xmath129 such that @xmath131 is not a base of a level @xmath88-edge for @xmath132 , but @xmath133 and @xmath134 are base of level @xmath88-edges .
possibly , @xmath135 .
see figure [ gamma](b ) .
let @xmath136 be the @xmath85-edge connecting @xmath131 with @xmath137 for @xmath138 .
let @xmath139 ( @xmath140 , resp . )
be the label of the end of @xmath136 at @xmath131 ( @xmath137 , resp . ) and @xmath141 the family of mutually parallel edges containing @xmath136 .
the number of edges in @xmath141 is denoted by @xmath142 .
also , let @xmath143 be the edge of @xmath141 lying on @xmath129 .
we define @xmath144 and @xmath145 similarly as above .
[ f ] * if @xmath32 is even , then @xmath146 .
furthermore , if @xmath147 , then either the @xmath85-edge @xmath136 is level , or its label on the other end is @xmath148 or @xmath149 . * if @xmath32 is odd , then @xmath150 .
furthermore , if @xmath151 , then @xmath136 is a level @xmath85-edge .
these follow from lemma [ positive ] and the definition of @xmath141 .
[ one ] if @xmath152 , or @xmath153 for some @xmath154 , then each @xmath143 has labels on @xmath93 .
suppose @xmath155 .
then @xmath156 , since @xmath157 can not have @xmath158 edges , except the case @xmath95 , by claim [ f ] .
thus @xmath159 and @xmath160 , since there can not be a level edge or an @xmath91-cycle .
when @xmath95 , we have the same conclusion . for , @xmath161 or @xmath48 , and if two then @xmath162 by claim [ f ] .
the same argument runs until we got that @xmath163 and @xmath164 for all @xmath102 .
then it is easy to see that @xmath143 has labels on @xmath93 at both ends ( see figure [ pm1](a ) ) . in particular , when @xmath95 , the corner on @xmath129 at @xmath131 contains label @xmath88 .
similarly for the case where @xmath165 .
suppose that @xmath153 as in figure [ pm1](b ) .
assume @xmath166 .
by claim [ f ] , @xmath167 , and so @xmath168 . if @xmath169 , then it contains either a level edge or an @xmath91-cycle .
thus @xmath170 .
hence @xmath171 and @xmath172 .
this implies that @xmath155 . if @xmath173 , then a similar argument shows that @xmath161 and @xmath174 , which implies @xmath165 . when @xmath95 , then @xmath152
. then we get the desired result by claim [ one ] .
therefore , we assume @xmath175 . furthermore , we can assume that @xmath176 and any @xmath177 by claim [ one ] .
if @xmath178 for all @xmath102 , then each @xmath143 has labels on @xmath93 .
note that @xmath179 and @xmath180 .
consider @xmath157 .
since @xmath157 contains neither level edge nor @xmath91-cycle , @xmath181 .
this implies @xmath182
. then @xmath183 .
thus we have @xmath184 .
hence @xmath185 or @xmath186 for all @xmath102 .
the result immediately follows from this observation .
thus we can assume that @xmath187 and @xmath188 for some @xmath189 .
( possibly , @xmath190 . )
then @xmath191 .
see figure [ red ] .
also , @xmath192 by claim [ f ] . then @xmath193 .
thus we have @xmath194 and @xmath195 for @xmath196 , and @xmath197 . * * case 1:**__@xmath32 is even . _
_ then @xmath198 , since @xmath199 .
see figure [ red](a ) . thus @xmath200 or @xmath85 by claim [ f ] . for , if @xmath201 then @xmath202 contains an @xmath91-cycle .
indeed , @xmath200 , since @xmath203 .
hence @xmath204 , and @xmath205 or @xmath85 again .
if @xmath206 , then @xmath205 as above . thus we can conclude that @xmath147 @xmath207 , and that @xmath208 , and @xmath185 , @xmath209 for @xmath210 .
also @xmath211 or @xmath148 , and hence @xmath212 or @xmath186 , and @xmath213 .
since @xmath214 and @xmath180 , we see @xmath215 and so @xmath216 .
thus @xmath217 for @xmath196 , and @xmath218 for @xmath219 and @xmath220 or @xmath88 . *
* case 2:**__@xmath32 is odd . _ _ then @xmath221 , and @xmath222 similarly for the case where @xmath32 is even .
see figure [ red](b ) .
also , @xmath223 by claim [ f](2 )
. then we see that @xmath151 and @xmath195 for @xmath224 .
thus @xmath209 and @xmath185 for @xmath224 . since @xmath222 and @xmath180 , we have @xmath215 and @xmath225
. then @xmath226 and therefore @xmath227 .
by claim [ f](2 ) , the only possibility is @xmath228 .
thus we have shown that @xmath129 gives a generalized scharlemann cycle .
this completes the proof of theorem [ gene ] .
[ web ] let @xmath92 .
then @xmath76 can not contain a non - zero @xmath85-face .
this follows immediately from lemma [ ss ] and theorem [ gene ] .
[ propk ] * @xmath73 can not contain a scharlemann cycle .
* @xmath73 can not contain a generalized @xmath91-cycle .
* @xmath73 has at most two level edges with different labels .
* suppose @xmath230
. then @xmath73 has at most @xmath231 @xmath232 , resp .
mutually parallel positive edges if @xmath36 is even odd , resp .. furthermore , if @xmath73 has a family of @xmath231 @xmath232 , resp .
mutually parallel positive edges , then the first and last edge the first or last edge , resp . of the family are level , when @xmath36 is even odd , resp .. if @xmath73 contains a scharlemann cycle , then we can get a new klein bottle in @xmath17 which meets @xmath49 fewer than @xmath29 ( ( * ? ? ? * theorem 6.4 ) ) .
theorem 6.6 in @xcite implies ( 2 ) . here , we need the fact that the distance between two dehn fillings creating projective planes is at most one @xcite .
( 3 ) is a restatement of lemma [ loop](2 ) . for ( 4 ) , if not , such a family contains a generalized @xmath91-cycle .
[ pge3 ] if @xmath60 , then @xmath233 .
assume for contradiction that @xmath234 because of lemma [ p2 ] .
let @xmath235 and @xmath236 be the vertices of @xmath73 , where @xmath235 can be a base of negative loops . since @xmath235 and @xmath236 have the same degree @xmath237 , if @xmath235 has a loop , then so does @xmath236
. then there would be a trivial loop .
thus we can see that @xmath73 has no loops .
then @xmath73 consists of at most two families of mutually parallel edges ; one is a family of positive edges , and the other is that of negative edges .
if @xmath230 , then @xmath73 contains more than @xmath36 mutually parallel negative edges by lemma [ propk](4 ) .
then an easy euler characteristic calculation shows that @xmath78 contains a @xmath88-face and no level edges . as in case 2 of the proof of proposition [ gweb ] , this gives a contradiction . if @xmath238 , then @xmath73 has at most two positive edges .
otherwise , there would be two edges which are parallel in both @xmath73 and @xmath78 .
but this implies that @xmath1 is cabled by ( * ? ? ?
* lemma 2.1 ) .
thus @xmath73 contains at least four negative edges .
similarly , @xmath78 contains a @xmath88-face and no level edges , a contradiction .
[ nowebk ] if @xmath230 , then @xmath73 can not contain a generalized web .
assume for contradiction that @xmath73 contains a generalized web @xmath111 , possibly with an exceptional vertex @xmath112 among boundary vertices of @xmath111 .
let @xmath41 denote a disk support of @xmath111 .
lemma [ propk](3 ) guarantees the existence of a label @xmath85 such that @xmath73 contains no positive level @xmath85-edges .
consider @xmath239 consisting of all vertices and @xmath85-edges of @xmath111 .
since every boundary vertex of @xmath111 , except @xmath112 , has degree at least @xmath240 , it has at least two edges attached with label @xmath85 .
we remark that @xmath239 may be disconnected .
choose an innermost component @xmath79 of @xmath239 ( in @xmath41 ) , and let @xmath241 be its block with at most one cut vertex of @xmath79 .
let @xmath242 and @xmath243 be the numbers of vertices , edges , and disk faces of @xmath241 , respectively .
( we view @xmath241 as the graph in a disk . )
also let @xmath244 and @xmath245 be the numbers of interior vertices , boundary vertices of @xmath241 and a cut vertex of @xmath79 in @xmath241 , respectively .
hence @xmath246 and @xmath247 or @xmath186 .
since @xmath241 has neither a level @xmath85-edge nor a generalized @xmath91-cycle , each face of @xmath241 is a disk with at least @xmath0 sides .
thus we have @xmath248 .
combined with @xmath249 because it has only disk faces , we get @xmath250 . on the other hand
we have @xmath251 because each boundary vertex of @xmath241 , except a cut vertex of @xmath79 , has at least two edges attached with label @xmath85 .
these two inequalities give us that @xmath252 , a contradiction .
consider the case where @xmath238 and @xmath253 when @xmath229 . recall that @xmath73 contains a generalized web @xmath111 by proposition [ gweb ] , theorem [ web ] and lemma [ pge3 ] .
we remark that at most two positive edges can be parallel in @xmath73 . for
, if there are three mutually parallel positive edges , then @xmath73 contains either an @xmath91-cycle or a pair of positive level edges with the same label .
the former is impossible by lemma [ propk ] . in the latter case ,
such two edges are also parallel in @xmath78 . but
this implies that @xmath1 is cabled ( * ? ? ?
* lemma 2.1 ) .
also , if there are two parallel positive edges , then both edges must be level . a positive level edge with label @xmath186 ( or @xmath48 ) in @xmath73
is called a _ @xmath186-edge _ ( _ @xmath48-edge _ , resp . ) and a positive non - level edge is called a _ mixed edge_. [ pair ] @xmath111 contains a pair of parallel edges . in particular , these edges are a @xmath186-edge and a @xmath48-edge .
any interior vertex of @xmath111 has degree at least six , since @xmath254 .
also , any boundary vertex of @xmath111 , except an exceptional one , has degree at least four .
therefore @xmath111 contains a pair of parallel edges by ( * ? ? ?
* lemmas 2.3 , 3.2 ) .
then these edges are both level as above .
[ 3level ] let @xmath85 be a vertex of @xmath111 .
* there are no three @xmath102-edges at @xmath85 for each @xmath255 .
* there are no three positive level edges connecting @xmath85 to mutually distinct vertices , not @xmath85 .
\(1 ) suppose that three positive @xmath186-edges , say , are incident to @xmath85 . since all negative loops based at vertex @xmath186 in @xmath78 are mutually parallel , there are at least @xmath256 negative loops at vertex @xmath186 .
this is impossible by lemma [ p - parallel ] .
\(2 ) let @xmath257 and @xmath258 be the vertices of @xmath78 . by lemma [ pair ]
, @xmath111 contains a pair of parallel ( positive ) edges @xmath259 .
here , @xmath136 is an @xmath102-edge for @xmath260 , and @xmath136 connects a vertex @xmath55 to another @xmath57 ( possibly , @xmath261 ) .
then @xmath136 gives a negative loop based at @xmath131 with the label pair @xmath262 .
remark that all negative loops at @xmath131 are mutually parallel . if @xmath111 has another @xmath102-edge @xmath82 connecting @xmath85 to @xmath112 , then @xmath136 and @xmath82 are parallel , and so @xmath263 .
this means that any @xmath102-edge at @xmath85 goes to either vertex @xmath264 or @xmath265 .
let @xmath266 denote the number of negative loops at vertex @xmath102 in @xmath78 .
without loss of generality , we can assume that @xmath267 .
[ mark ] let @xmath85 be a vertex of @xmath111 .
if a level edge @xmath82 and a pair of parallel edges are incident to @xmath85 successively in @xmath111 , then @xmath82 is a @xmath186-edge .
furthermore , there are no consecutive pairs of parallel edges at @xmath85 in @xmath111 .
let @xmath259 be these pair edges , where @xmath136 is an @xmath102-edge .
assume for contradiction that @xmath82 is the consecutive @xmath48-edge next to @xmath98 .
as in the proof of [ 3level](2 ) , we may assume that @xmath82 connects @xmath85 to @xmath268 and @xmath98 and @xmath269 connect @xmath85 to @xmath270 for some @xmath3 .
let @xmath271 and @xmath272 be the endpoints of these edges .
let @xmath273 be a properly oriented arc from @xmath11 to @xmath12 along @xmath70 , and @xmath274 similarly .
see figure [ key](b ) . in @xmath78 , @xmath98 is a negative loop at vertex @xmath186 , and its endpoint @xmath12 has label @xmath85 , and @xmath272 has @xmath270 .
also , @xmath82 and @xmath269 are contained in the family of mutually parallel negative loops at vertex @xmath48 in @xmath78 , and @xmath11 and @xmath275 appear at the same end of this family .
otherwise , the family contains more than @xmath32 edges , contradicting lemma [ p - parallel ] .
let @xmath276 be the boundary component of @xmath34 with label @xmath102 ( @xmath260 ) .
we may assume that @xmath277 is oriented counterclockwise as in figure [ key](c ) .
then the subarc @xmath278 , with the induced orientation from @xmath277 , contains at most @xmath279 edge endpoints . consider the annulus part of @xmath2 between @xmath59 and @xmath277 , containing @xmath273 and @xmath274 .
then we see that the two oriented subarcs @xmath278 and @xmath280 contain the same number of points in @xmath281 .
see figure [ key](d ) .
but , the subarc @xmath280 contains at least @xmath282 edge endpoints .
the second conclusion immediately follows from the first one .
[ mix ] @xmath111 has the following properties . *
every disk face is bounded by mixed edges and a non - zero even number of level edges .
* there is no disk bounded by a cycle consisting of mixed edges , except at most one edge , and containing no vertex in its interior .
\(1 ) if a disk face is bounded by only mixed edges , then the boundary cycle is a scharlemann cycle , which is impossible by lemma [ propk](1 ) .
if the boundary cycle contains an odd number of level edges , we can not put the labels correctly .
\(2 ) suppose that there is such a disk @xmath41 .
since its interior contains no vertex of @xmath111 , there is an outermost disk face of @xmath111 whose boundary edges consist of mixed edges and at most one diagonal edge ( or exceptional one ) , contradicting ( 1 ) .
suppose that @xmath111 has no interior vertex .
then , @xmath111 has a boundary vertex @xmath85 of degree two in @xmath283 by [ 24 , lemma 3.2 ] .
then there are two consecutive pairs of parallel edges incident to @xmath85 in @xmath111 .
this contradicts lemma [ mark ] .
thus @xmath111 has an interior vertex .
lemma [ 3level](1 ) implies that each interior vertex has at least two mixed edges .
consider a maximal path consisting of mixed edges starting at an interior vertex .
by lemma [ 3level](2 ) , both ends will reach some boundary vertices .
thus we have a consecutive sequence of mixed edges between two boundary vertices .
then we can choose an outermost disk @xmath41 so that its interior contains no vertex and no mixed edges , and its boundary consists of a consecutive sequence of mixed edges and a consecutive sequence of boundary edges .
furthermore , we can assume that @xmath41 does not contain an exceptional vertex of @xmath111 . in @xmath41
, every diagonal edge of @xmath283 connects an interior vertex and a boundary vertex of @xmath111 . for
, if there is a diagonal edge connecting two interior vertices , then it contradicts lemma [ mix](2 ) .
also , if there is a diagonal edge connecting two boundary vertices , then between these two boundary vertices there must be a vertex which has degree two in @xmath283 , contradicting lemma [ mark ] .
recall that boundary vertices has degree at least four in @xmath111 . see figure [ last ] .
let @xmath284 be the boundary vertices from left to right .
there are at least three boundary vertices by lemma [ mix](2 ) and each vertex @xmath285 , has diagonal edges ( so level edges ) incident to by lemma [ mark ] .
thus all boundary edges between @xmath286 and @xmath287 are mixed edges , and the other boundary edges are level edges by lemma [ mix](1 ) . furthermore vertices @xmath286 and @xmath287 have a pair of parallel edges and a consecutive level edge @xmath82 . by lemma [ mark ] , @xmath82 is a @xmath186-edge .
but there must be another vertex @xmath288 between @xmath286 and @xmath287 which has three consecutive level edges , so that the level edge not of the pair is a @xmath48-edge .
this contradicts lemma [ mark ] . if @xmath46 , then @xmath289 . by lemma [ pge3 ] , @xmath290 . if @xmath230 , then proposition [ gweb ] , theorems [ web ] and [ nowebk ] give a contradiction
we have just shown that the case @xmath238 is impossible .
if @xmath62 , then @xmath291 . assume that @xmath292 .
the reduced graph @xmath293 is a subgraph of the graphs shown in figure [ reduced ] . ] if @xmath32 is even , there are at most @xmath103 positive edges by lemma [ positive ] , and at most @xmath158 mutually parallel negative edges by lemma [ p - parallel ] .
thus @xmath294 , and then @xmath295 and @xmath296 .
similarly , if @xmath32 is odd , we have that @xmath295 and @xmath92 .
it follows that @xmath78 contains at least two positive edges .
first , assume that @xmath293 is a subgraph of figure [ reduced](a ) .
this family of mutually parallel positive edges in @xmath78 contains a generalized @xmath91-cycle , contradicting lemma [ ss ] .
next , assume that @xmath293 is a subgraph of figure [ reduced](b ) .
since each family of mutually parallel negative edges contains at most @xmath158 edges , we can assume that the labels are as in figure [ b ] .
then the family of mutually parallel positive edges contain an @xmath91-cycle with label pair @xmath297 . ]
thus we have proved theorem 1.2 .
an _ extended @xmath91-cycle _ in @xmath73 is the quadruple @xmath299 of mutually parallel positive edges in succession and @xmath300 forms an @xmath91-cycle .
[ sst ] let @xmath301 .
* @xmath73 can not contain an extended @xmath91-cycle .
* @xmath73 has at most @xmath0 labels which are labels of @xmath91-cycles .
* @xmath73 has at most @xmath302 mutually parallel positive edges .
\(1 ) is shown in ( * ? ? ?
* lemma 2.10 ) .
\(2 ) if there are four such labels , there are two @xmath91-cycles with disjoint label pairs
. then @xmath14 contains a klein bottle as in the proof of ( * ? ? ?
* lemma 3.10 ) .
( recall that @xmath14 is irreducible . ) by theorem 1.2 , @xmath14 must contain a klein bottle @xmath303 which meets the core @xmath304 of the attached solid torus @xmath305 in a single point , since @xmath253 .
then the torus @xmath306 is essential in @xmath14 . otherwise @xmath14 would be either a lens space containing a klein bottle or a prism manifold , both of which are not toroidal .
this essential torus meets @xmath304 in two points .
this contradicts the minimality of @xmath35 .
\(3 ) follows from ( 1 ) and ( 2 ) , since a family of @xmath307 mutually parallel positive edges contains either an extended @xmath91-cycle or two @xmath91-cycles with disjoint label pairs .
[ pge3 t ] if @xmath301 , then @xmath92 .
this follows from the same argument as in the proof of lemma [ pge3 ] .
we use lemma [ sst](3 ) instead of lemma [ propk](4 ) .
[ nowebt ] if @xmath301 , then @xmath73 can not contain a generalized web .
assume for contradiction that @xmath73 contains a generalized web @xmath111 , possibly with an exceptional vertex @xmath112 of @xmath109 .
let @xmath41 denote a disk support of @xmath111 .
let @xmath85 be a label of @xmath73 .
suppose that @xmath73 has no @xmath91-cycle with a label @xmath85 .
consider @xmath239 , consisting of all vertices and @xmath85-edges of @xmath111 as in the proof of theorem [ nowebk ] .
choose an innermost component @xmath79 of @xmath239 ( in @xmath41 ) , and let @xmath241 be its block with at most one cut vertex of @xmath79 . since @xmath241 has neither @xmath91-cycle with label @xmath85 nor extended @xmath91-cycle , each face of @xmath241 is a disk with at least @xmath0 sides .
then the same calculation as in the proof of theorem [ nowebk ] gives a contradiction .
therefore @xmath73 has @xmath35 labels which are labels of s - cycles .
note that the existence of an s - cycle in @xmath73 guarantees that @xmath35 is even ( * ? ? ?
* lemma 2.2 ) and so @xmath308 .
this contradicts lemma [ sst ] ( 2 ) . if @xmath301 , then @xmath289 . by lemma [ pge3 t ] , @xmath290 .
then proposition [ gweb ] , theorems [ web ] and [ nowebt ] give a contradiction .
if @xmath309 , then @xmath291 .
the reduced graph @xmath310 is a subgraph of the graph shown in figure [ 5 - 1 ] ( * ? ? ?
* lemma 5.2 ) . ] here , @xmath311 denotes the number of edges in each of the families of mutually parallel edges .
then @xmath312 .
we have that @xmath313 and @xmath314 for @xmath315 by lemmas [ positive ] and [ p - parallel ] .
thus , @xmath316 .
this inequality implies that @xmath295 and all @xmath317 s are non - zero .
all non - loop edges of @xmath77 are negative .
assume not .
let @xmath85 and @xmath112 be the vertices of @xmath77 . if @xmath32 is even , then @xmath318 , which is absurd . if @xmath32 is odd , then @xmath319 .
thus @xmath95 . but
then , each of six families at @xmath85 , say , consists of two edges , and we have an @xmath91-cycle in the family of loops at vertex @xmath85 . without loss of generality
, we can assume that @xmath320 . let @xmath321 . since @xmath322 , we see @xmath323
. then @xmath324 .
hence @xmath325 .
thus the loop family around @xmath85 , say , contains a generalized @xmath91-cycle .
if @xmath326 , then @xmath327 . assume that @xmath328 .
since @xmath326 , @xmath77 can have only positive edges , and so @xmath73 has only negative edges . note that @xmath77 has at most three families of mutually parallel edges by [ 8 , lemma 5.1 ] .
let @xmath329 , @xmath330 and @xmath331 be such three families , and @xmath332 , @xmath333 and @xmath334 be the number of edges of each family . without loss of generality , we can assume that @xmath329 has at least one edge .
first assume that @xmath335 is even .
if @xmath336 , then @xmath329 contains a generalized s - cycle by the symmetry of labels around the vertex .
thus @xmath337 .
similarly we have @xmath338 , @xmath339 , and therefore @xmath340 . since @xmath335 is
even and @xmath37 , we have @xmath341 and @xmath342 .
then @xmath77 has two or three level edges with different labels , contradicting lemma 2.6(1 ) .
assume now that @xmath335 is odd .
since @xmath343 is integral , @xmath32 must be even .
thus @xmath344 by lemma [ positive ] , and similarly for @xmath345
. then @xmath346 .
hence we have @xmath347 and @xmath348 .
this implies that @xmath77 contains two scharlemann cycles of length three , which is a contradiction .
we have thus completed the proof of theorem 1.1 .
in this section we give examples showing that our estimates are sharp .
* theorem 4.2 ) shows that there is a hyperbolic manifold @xmath1 , admitting two dehn fillings @xmath13 and @xmath14 such that @xmath349 and @xmath14 is toroidal , and @xmath16 .
in fact , @xmath14 contains an essential torus which bounds @xmath350 and @xmath351 , where @xmath352 is a seifert fibered manifold with orbifold a disk with two cone points of index @xmath3 and @xmath6 .
also , @xmath351 contains a klein bottle , and so does @xmath14 .
it is not hard to see that @xmath14 contains an incompressible torus hitting the attached solid torus @xmath305 twice , and a klein bottle hitting @xmath305 once from the construction using tangles . | we study the situation where we have two exceptional dehn fillings on a given hyperbolic @xmath0-manifold .
we consider two cases that one filling creates a projective plane , and the other creates an essential torus or a klein bottle , and give the best possible upper bound on the distance between two fillings for each case . |
figure 1 : velocity plots of lyman @xmath112 ( left ) , and all the metal lines ( right ) in the absorption system towards qso 1009 + 2956 ( @xmath113 , v=16 ) .
zero velocity corresponds to the redshift @xmath114 , of the blue component .
the vertical dotted lines also show the red component at 2.503704 .
the histogram represents the observed counts in each pixel normalized to the quasar continuum .
the smooth curve show the best fitting voigt profiles convolved with the instrumental resolution ( @xcite ) .
the lyman lines and si iii ( 1206 ) are in the lyman alpha forest region where there is additional absorption which we do not fit . on the nights of december 27 , 28 1995
, we obtained 5.9 hours of spectra with the hires echelle spectrograph on the 10-m w. m. keck telescope ( @xcite ) .
two exposures of 2.5 and 2 hours covered 3540 5530 , while a third 1.4 hour covered 3165 4370 . a 1.14 arcsec slit produced spectra with resolution of 8 kms@xmath4 .
each exposure was accompanied by dark , quartz lamp , and throium - argon arclamp exposures .
a standard star was observed to trace the echelle orders and remove the blaze response in the spectra .
we used our standard data reduction ( @xcite ) .
figure 2 : the lick spectrum of qso 1009 + 2956 shows the lyman limit due to the absorber at @xmath0 . on november 28 , 1995 we used the kast spectrograph on the lick 3-m telescope to obtain 1.9 hours of integration covering 3100 - 5950 .
the spectra were calibrated to vacuum heliocentric wavelengths , and optimally extracted .
the smooth line shows the convolved voigt proflies of the higher order lyman lines calculated with the parameters in table 1 .
the optical depth , measured by the ratio of flux blueward and redward of 3200 , constrains the total column density of neutral hydrogen .
figure 3 : a representation of the flow of information in the standard cosmological model .
boxes show measurements , and circles show theories and derived quantities .
most references to quantities shown are discussed in the text .
the range for the hubble constant , @xmath115 , is taken from mould et al .
( @xcite ) .
when two numbers are shown , one is for @xmath116 and the other is @xmath117 .
when @xmath83 is used , we add or subtract all three errors to enlarge the range .
figure 4 : the bbns predicted primordial abundances of d and @xmath67he as a function of @xmath2 and @xmath95 ( @xcite ) .
the lower two rectangles are defined by the @xmath17 random plus systematic errors on d / h towards q1009 + 2956 and q19371009 .
the shaded region is from our mean d / h value , with @xmath17 random errors from the quadratic sum of our random error plus the @xmath106 nuclear error .
we then add on the systematic error .
the upper limit shows d / h towards q0014 + 8118 .
figure 5 : as figure 4 but for @xmath72 , the mass fraction of @xmath66he .
dashed rectangles show the bounds ( 1@xmath106 statistical plus systematic errors ) from recent measurements of @xmath66he ( @xcite,@xcite ) .
the lack of intersection with the shaded region illustrates the inconsistency of d / h with helium-4 measurements .
the lower rectangle represents the estimates of @xmath66he deduced by pagel et als . , and it includes in its range a systematic underestimation of @xmath72 of 0.004 .
figure 6 : as figure 4 but @xmath1li / h the spite and thorburn @xmath1li plateau from measurements of population ii stars are shown as dashed and dot - dashed lines respectively .
( @xcite,@xcite ) . significant depletion of the surface @xmath1li abundance is the likely source of the discordances .
( @xcite ) .
the dotted line shows upper limits from lithium measurements in population i stars . | standard big bang nucleosynthesis ( bbns ) promises accurate predictions of the primordial abundances of deuterium , helium-3 , helium-4 and lithium-7 as a function of a single parameter .
previous measurements have nearly always been interpreted as confirmation of the model ( @xcite ) . here
we present a measurement of the deuterium to hydrogen ratio ( d / h ) in a newly discovered high redshift metal - poor gas cloud at redshift @xmath0 .
this confirms our earlier measurement of d / h ( @xcite ) , and together they give the first accurate measurement of the primordial d abundance , and a ten - fold improvement in the accuracy of the cosmological density of ordinary matter .
this is a high density , with most ordinary matter unaccounted or dark , which is too high to agree with measurements of the primordial abundances of helium-4 and lithium-7 .
since the d / h measurement is apparently simple , direct , accurate and highly sensitive , we propose that helium requires a systematic correction , and that population ii stars have less than the primordial abundance of @xmath1li . alternatively
, there is no concordance between the light element abundances , and the simple model of the big bang must be incomplete and lacking physics , or wrong . in the standard big bang model the primordial abundances of the light elements
depend on the unknown value of the cosmological baryon ( ordinary matter ) to photon ratio , @xmath2 ( @xcite,@xcite@xcite ) .
a measurement of primordial d / h , which is very sensitive to @xmath2 , leads to precise predictions for the abundances of the other light elements , which can be compared with observations to test the model .
unfortunately d is destroyed inside stars , which then eject gas with h but no d. as more gas is processed and ejected , the value of d / h in the interstellar medium of our galaxy drops further below primordial .
we look for primordial d / h in quasar absorption systems ( qas ) ( @xcite,@xcite ) , some of which sample gas at early epochs , and in intergalactic space where there are too few stars to destroy significant amounts of d. the absorption system at @xmath0 towards qso 1009 + 2956 was from 40 qso spectra which we obtained at lick observatory because it showed a steep lyman limit break and weak metal lines , which indicate a low @xmath3-value suitable for the detection of deuterium . of the 15 qsos which met these criteria and
were subsequently observed with the 10-m w. m. keck telescope , only 2 have yielded deuterium measurements ( @xcite ) .
we now describe a measurement of d / h , in this new qas , which agrees with our previous low value , but not with measurements of a high d / h ( @xcite,@xcite ) .
either bbns was inhomogeneous ( @xcite ) and d / h varies spatially , or more likely , the spectra showing high d / h were contaminated . in figure 1
we show nine lines in the absorption system at @xmath0 towards qso 1009 + 2956 .
all six metal absorption lines are best fit by two components separated by 11 km s@xmath4 , using the parameters given in table 1 .
the dotted lines show the expected positions of the h and d lines for the gas at the well determined redshifts of these components .
the two components adequately fit the blue side of the lyman features , but an extra h component at @xmath5 is seen on the red side which is not significant in our measurement of d / h . the d lyman-@xmath6 absorption is unsaturated and has a well determined total column density , n(d i ) . however
the amount of d in the red component is not well determined because it is blended with both the blue d and h components .
if we do not include the red d component , systematic under - absorption would occur at that velocity , so we set ( d / h)@xmath7=(d / h)@xmath8=(d / h)@xmath9 .
all lines from the same gas will have the same turbulent velocity dispersions @xmath10 , but the thermal widths will depend on the mass of each ion : @xmath11 , where @xmath12 is the proton mass .
we use the @xmath3-values of si , c , and h in table 1 to obtain the temperatures and @xmath10 of the two components .
these values are consistent with a single @xmath10 for all lines , and allow us to calculate that the red component of d should have @xmath3(d i ) @xmath13 = 15.4 which we use as a constraint on the fit because we can not measure this value from the d line .
the widths of the h and d lines are dominated by thermal motions , so the lines are accurately fit by voigt profiles ( @xcite,@xcite ) .
a simultaneous fit to the lyman @xmath6 , @xmath14 , and @xmath15 lines in the keck spectrum and the lyman continuum absorption in the lick spectrum ( figure 2 ) gives d / h = @xmath16 ( @xmath17 random photon and fitting errors ) .
a systematic increase in d / h comes from the chance superposition of weak h absorption at the expected position of d. to estimate this , we made noise - less model spectra with d and h lines at the redshifts given by the metal lines .
we used the @xmath3 values and n(h i ) from table 1 , but we vary n(d i ) .
for each n(d i ) we made @xmath18 spectra , and to each of these we added random lines to simulate the forest , using the known distribution of @xmath19 , n(h i ) and @xmath3-values for forest lines ( @xcite ) .
we then calculated the likelihood that the data came from the model spectra for each d / h .
assuming that all values of d / h are equally likely _ a priori _ , the expected or mean d / h is log ( d / h)@xmath20 , where the errors are the standard deviations of the likelihood distribution above and below the mean .
the correction from the monte carlo simulations changes our value from a formal upper limit to a measurement of d / h .
additional systematic errors in d / h from fitting the continuum level are estimated to be @xmath21log ( d / h ) = 0.06 .
we varied the continuum by 2% near ( figure 1 ) , and by 10% near the lyman continuum ( figure 2 ) , and for each combination we re - fit the spectra . including the uncertainties due to systematics , we find @xmath22 where the errors represent the @xmath17 random error followed by the systematic errors from the continuum level .
the blue side of the blue d line appears unblended , and is fit with @xmath3(di ) = 15.7 @xmath23 2.1 km s@xmath4 .
this is consistent the value of @xmath24 predicted from the @xmath25 and @xmath26 given by the metal and h lines .
less than 2% of h lines have @xmath3-values this small or smaller ( @xcite ) , so it is unlikely that this line is strongly contaminated with h. metal lines can be this narrow , although they are usually narrower with @xmath27 km s@xmath4 .
the d feature is unlikely to be a metal line because metal lines of this strength are nearly always accompanied by other lines , and the spectra has been searched for these lines .
the _ a posteriori _ probability of a random h line with @xmath28 and @xmath29 @xmath30 , to account for @xmath31 of the d line , and redshift within 20 of that of the metal lines is @xmath32 .
a high correlation of satellite " components on these velocity scales could increase the probability by as much as a factor of 3 ( @xcite,@xcite ) .
the residual flux below the lyman edge gives an accurate measure of all h i in this velocity region ( figure 2 , @xcite ) .
there are no other lines within 5000 of @xmath33 which have h i column densities @xmath34 @xmath30 , so that all of the lyman continuum absorption must be produced by gas in the @xmath0 absorption system .
the blue side of the , and lines are best fit if all of this h i is near the two velocity components which are seen in the metals
. there could be be additional h at velocities between the metal lines and @xmath35 , provided this gas has very low metal abundances ( [ c / h ] @xmath36 ) .
however , nearly all known qas with large n(h i ) have metal abundances [ c / h ] @xmath37 .
the column densities of the metals and neutral hydrogen give the metallicity and neutral fraction of the gas .
following donahue & shull ( @xcite ) , we model the system as an optically thin gas ionized by a typical qso photoionizing spectrum given by mathews & ferland .
we estimate the ionization parameter ( the ratio of the number of photons with energies above one rydberg to the number of atoms ) : u @xmath38 , which gives the metal abundances shown in table 1 . for a photoionization model of low metallicity components , this corresponds to a neutral hydrogen fraction of h i / h @xmath39 .
the measured d / h is consistent with normal galactic chemical evolution ( @xcite,@xcite ) ; the destruction of d in known populations of stars can account for our d / h , that in the per - solar nebula , and the current ism d / h . in the ism , d / h @xmath40 ( @xcite ) , for [ o / h ] = -0.25 .
if the destruction of d is proportional to [ o / h ] , then we would expect 0.005 of the d would be destroyed for [ o / h ] @xmath41 ( @xcite ) .
we find [ c / si ] @xmath42 in both components which is characteristic of low metallicity stars in the halo of our galaxy .
this suggests that the c and si were created in normal " supernovae .
if some additional astrophysical processes destroys d , it must do so without producing more c and si than we see , or other elements which we would have seen if they were made , and without changing the usual c / si ratio .
if primordial d / h were @xmath43 ( @xcite,@xcite ) then 87% of d must be destroyed in q1009@xmath442956 and 90% in q1937@xmath458118 .
the level of destruction would be large and similar for [ c / h ] @xmath46 to @xmath47 , but small for q0014 + 8118 with [ c / h ] @xmath36 .
redshift is apparently not a factor .
the qas towards q1937 - 1009 in which we previouly measured a low d / h is extremely similar to q1009 + 2956 , except that n(h i ) is larger , and the redshift is higher ( @xcite ) .
we used the lyman series lines up to 19 , and the absence of flux in the lyman continuum to constrain the n(h i ) , and we obtained smaller errors : log d / h @xmath48 .
the best estimate of the primordial d / h is obtained by taking the average of our two measurements weighted by the squares of their random errors : @xmath49 where the first error is the random error on the weighted mean , and the second is the larger of the systematic errors from the continuum level uncertainty .
we do not add these systematic errors because they have a similar origin and the two measurments agree to within the random errors : prob@xmath50 .
in linear units , the mean is @xmath51 in table 2 we list all published measurements of d / h in qas .
only the two discussed above are measurements .
the others were initially presented as limits or possible detections , and they will all be biased to higher than true d / h because they do not include corrections for weak h at the position of d. the qas towards q0014 + 8118 differs from our two in several ways .
since no metal lines were detected , the velocity structure of the cloud could be determined only by median filtering of the higher - order lines in the forest ( @xcite ) .
the lyman-@xmath6 feature was considerably more complex : five components were required for an adequate fit ( instead of three ) , with two components within 30 of the deuterium absorption line ( @xcite ) .
the neutral hydrogen column density of the component where deuterium is measured , log n(h i ) = 16.74 , is 4 and 10 times lower than the column densities in our two qas , which reduces the sensitivity to low d / h .
rugers & hogan reanalyzed the published spectra of q0014 + 8118 , and determined that the deuterium feature is better fit with two very narrow components separated by 21 ( @xcite ) .
they claim that it is very unlikely that there are two narrow lines by chance at the expected position of d , but we are not convinced because there are no metal lines to constrain the velocities and their model is a poor fit to the data in many places ( @xcite ) .
these data remain consistent with this d line being contaminated ( @xcite ) . towards qso 1202 - 0725 , wampler
et al .
( @xcite ) find @xmath52 at a redshift @xmath53 .
this qas has high metallicity , [ o / h ] = 0.3 , and does not look suitable for inferences of primordial d / h . towards
qso 0420 - 3851 , carswell et al .
( @xcite ) find a lower limit of @xmath54 at @xmath55 .
this qas also has a high metallicity , [ o / h ] = -1.0 .
the d i column is fairly well - determined , but the h i is high , n(h i ) @xmath56 @xmath30 , and is very uncertain .
all data in table 2 are consistent with our low d / h value , but our data are inconsistent with high d / h , unless d / h is distributed inhomogeneously ( @xcite ) .
we now discuss the baryon density implied by our low d / h measurements . in figure 3
we show how d / h fits into the standard cosmological framework . for the remainder of this paper
, we present the implications of the new d / h measurements on these quantities . the current cosmological density of baryons , @xmath57 , is given by @xmath58 where the third error is the @xmath17 from nucleosynthesis predictions ( @xcite ) .
the density of photons from the cosmic microwave background ( @xcite ) , @xmath59 gives : @xmath60 and as a fraction of the current critical density , @xmath61 , @xmath62 where the hubble constant , @xmath63 mpc@xmath4 .
since the observed density of visible baryons in stars and hot gas is about @xmath64 ( @xcite ) , most baryons ( about 94% for @xmath65 ) are unaccounted . in figures 46 , we present the predicted adundances of @xmath66he , @xmath67he and @xmath1li relative to hydrogen ( @xcite ) .
we show values of @xmath2 in the shaded region which are consistent with our d / h .
the usually accepted estimates of primordial @xmath66he come from measurements of low - metallicity extragalactic h ii regions .
even modest chemical production can significantly change the abundance of @xmath66he , so the primordial value is inferred by extrapolating @xmath66he measurements to zero metallicity ( @xcite@xcite ) .
pagel et al . reported an inferred primordial mass fraction of @xmath66he , @xmath68 , with a 95% upper bound of 0.242 including systematic errors ( @xcite ) . recently ,
thuan et al . found @xmath69 ( @xcite ) .
these @xmath66he measurements are not consistent with our d / h , which predicts @xmath70 to quantify the discordance of d / h with the @xmath66he measurements mentioned above , we form the weighted mean @xmath2 using our mean d / h and either of the @xmath66he values .
we weight with the random errors alone , where we now include nuclear uncertainties in quadrature in our d / h error .
we find prob(@xmath71 for the low @xmath72 , and prob(@xmath73 for the high @xmath72 .
if bbns is correct , this disagreement shows that there must be systematic errors which should be identified and corrected . reducing our d / h by its systematic error , gives prob(@xmath74 for the low @xmath72 shifted up by its systematic error , and prob(@xmath75 for the high @xmath72 , which does not have a quoted systematic error . even with the allowances of current systematic error estimates , d / h and y@xmath76 fail to predict the same value of @xmath2 with bbns .
@xmath67he is more complex and poorly understood because it is both created and destroyed in stars .
the ratio of the d mass fraction in the ism to primordial d is @xmath77 ( random , not systematic error ) .
if @xmath67he undergoes the same amount of destruction , then its primordial abundance would be 1/0.67 times that in the ism , which is an upper limit because @xmath67he is also created ( e.g. primordial d is burned in stars to make @xmath67he ) .
low mass stars may produce @xmath67he copiously ( @xcite ) , as is suggested by its high abundance in the ejecta of planetary nebulae ( @xcite ) .
if this is the case , only upper limits on @xmath67he , not d + @xmath67he , can be correctly applied to cosmology .
however the observed abundance in galactic h ii regions is 5 20 times lower than expected ( @xcite ) , for unknown reasons , perhaps because some of the @xmath67he made in low mass stars is later destroyed , or because observations are made in h ii regions which contain the ejecta of high mass stars which destroy @xmath67he ( @xcite ) .
the abundance of @xmath1li in stars in the disk of our galaxy ( population i ) spans three orders of magnitude , because li is both made ( e.g. by cosmic ray spallation in the interstellar medium , and in novea and supernovae ) and destroyed , but warm metal - poor halo ( population ii ) stars have similar plateau " abundances of @xmath1li in their atmospheres .
these measurements ( @xcite-@xcite ) , shown in figure 6 , are often taken as the primordial abundance , but recent data show that this is not justified .
stars with very similar temperatures and metallicities , subgiant stars in the globular cluster m92 , and turnoff stars in the halo , have significantly different @xmath78 abundances ( @xcite ) , perhaps because of different amounts of @xmath1li depletion .
ryan et al .
have found that the @xmath1li abundance depends on both temperature and metallicity in ways which were not predicted and are not understood .
if we are to obtain a primordial @xmath1li abundance we must either ( 1 ) understand why its abundance varies from star to star , and learn to make quantitative predictions of the level of depletion , or ( 2 ) make measurements in relatively unprocessed gas .
our d / h measurements imply that @xmath1li in population ii stars has been depleted by about 0.5 dex , which may be difficult to reconcile with the near constancy of @xmath1li in the warm halo stars .
we conclude that our d / h measurements are probably the first measurements of a primordial abundance ratio of any elements , and that they give about an order of magnitude improvement in the accuracy of estimates of @xmath57 . where are the baryons ?
the macho collaboration recently announced detection of baryonic dark matter in the halo of our galaxy from 7 microlensing events towards the large magellanic cloud ( @xcite ) .
the most likely mass of the machos is 0.3 - 0.5 solar masses , from the durations of the events and the velocities implied by the model of the halo of our galaxy .
this corresponds to a standard halo model mass of 1.6@xmath79 , and a large fraction of the total mass .
if the halos in all galaxies are entirely baryonic machos , they have @xmath80 ( 90% confidence , with prefered values exceeding 0.3 for h=0.75 @xcite ) . this is more than @xmath81 , which is predicted from our two measurements , so @xmath82 ( prefered value 0.16 ) of halos are baryonic .
the macho detections strongly imply a high @xmath83 : if halos are 50% baryonic machos then @xmath84 ( 90% confidence for h=0.75 ) , enough to account for all baryons .
x - ray observations reveal that baryons in the form of hot gas contribute a fraction @xmath85 of the total mass of clusters of galaxies ( @xcite ) .
the remaining mass is some combination of unseen dark baryons ( e.g. machos ) and any other non - baryonic dark matter , such as massive supersymmetric particles .
the baryon fraction in clusters is then @xmath86 .
since galaxy clusters are the largest bound structures in the universe , their @xmath87 should provide good estimates of the cosmological value , so the total mass density of the universe is then : @xmath88 where we used @xmath83 from d / h and the inequality allows for @xmath86 ( @xcite ) .
although our d / h measurements give a very high @xmath83 , it is not high enough to permit @xmath89 from matter : we have @xmath90 from @xmath91 , @xmath92 , and @xmath93 .
either the universe is open , as is now allowed by the theory of inflationary cosmology @xmath94 ( @xcite ) , or there is a cosmological constant .
we are forced to conclude , given the present state of primordial abundance measurements , that d / h gives the only reliable constraints on @xmath2 and @xmath83 .
the practice of finding consensus among all the light elements should be heavily weighted to d / h , because it is most sensitive to @xmath2 , and because of the apparent absence of systematic uncertainties arising from destruction and creation .
the discordance with other light elements does not demand alternative models to the standard big - bang , rather it begs for all measurements of primordial abundances to be made in similar pristine sites .
the search for more deuterium measurements in qas will continue , and at the current rate , @xmath95 is likely to be known to 5% by the end of the millenium . cccc ion & blue component & red component & total h i & n = 17.36 @xmath23 0.09 & n = 16.78 @xmath23 0.11 & n = 17.46 @xmath23 0.05 & b = 18.8 @xmath23 0.5 & b = 21.9 @xmath23 4.1 & d i & n = 12.84 @xmath23 0.09 & n = 12.26 & n = 12.94 @xmath23 0.06 & b = 15.7 @xmath23 2.1 & b = 15.4 & si iii & n = 12.81 @xmath23 0.06 & n = 12.55 @xmath23 0.04 & n = 13.00 @xmath23 0.03 & b = 4.9 @xmath23 0.4 & b = 4.8 @xmath23 0.8 & si iv & n = 12.50 @xmath23 0.03 & n = 12.05 @xmath23 0.08 & n = 12.63 @xmath23 0.02 & b = 4.9 @xmath23 0.6 & b = 3.9 @xmath23 1.5 & c ii & n = 12.46 @xmath23 0.12 & n = 12.18 @xmath23 0.18 & n = 12.64 @xmath23 0.08 & b = 7.0 @xmath23 3.2 & b = 3.7 @xmath23 4.2 & c iv & n = 12.81 @xmath23 0.04 & n = 12.56 @xmath23 0.06 & n = 13.00 @xmath23 0.03 & b = 5.4 @xmath23 0.6 & b = 5.6 @xmath23 1.2 & t(@xmath96 k ) & 2.1 @xmath23 0.1 & 2.4 @xmath23 0.7 & b@xmath97 & 3.2 @xmath23 0.4 & 2.3 @xmath23 1.4 & & @xmath98 & @xmath99 & ... & @xmath100 & @xmath101 & ... & @xmath102 & @xmath103 & ... ccccccc qso & @xmath104 & d / h@xmath105 & 1@xmath106 & logn(h i ) & [ c / h ] & reference ism & 0.0 & 1.6 & 0.1 & 18.2 & ... & @xcite 1009 + 2956 & 2.504 & 2.5 & 0.5 & 17.46 & 2.9 & this paper 1937@xmath451009 & 3.572 & 2.3 & 0.3 & 17.94 & @xmath452.2 , @xmath453.0 & @xcite 0014 + 8118 & 3.320 & @xmath107 & ... & 16.7 & @xmath1083.5 & @xcite,@xcite,@xcite 1202@xmath450725 & 4.672 & @xmath109 & ... & 16.7 & ... & @xcite 0420@xmath453851 & 3.086 & @xmath110 & ... & @xmath111 & -1.0 & @xcite \1 . c. j. copi , d. n. schramm , m. s. turner , _ science _ * 267 * , 192 ( 1995 ) .
\2 .
d. tytler , x - m . fan , s. burles , submitted to _ nature _ ( 1996 ) .
\3 .
r. v. wagoner , w. a. fowler , f. hoyle , _ astrophys .
j. _ * 148 * , 3 ( 1967 ) ; j. yang , m. s. turner , g. steigman , d. n. schramm , k. a. olive , _ ibid . _ * 281 * , 493 ( 1984 ) ; k. a. olive , d. n. schramm , d. thomas , t. p. walker , _ phys .
lett .
b _ * 236 * , 454 ( 1990 ) . by standard bbn , we assume spatial homogeneity , three light neutrino species with vanishing chemical potentials , and no additional light particles present in thermal abundance
. \4 .
t.p . walker , g. steigman , d. n. schramm , k. a. olive , h. s.kang , _ astrophys . j. _ * 376 * , 51 ( 1991 ) ; m. s. smith , l. h. kawano , r. a. malaney , _ astrophys . j. suppl .
ser . _ * 85 * , 219 ( 1993 ) .
\5 .
s. sarkar submitted to _ rep . on progress in physics _ ( 1996 ) .
\6 . t. f. adams , _ astron .
astrophys . _ * 50 * , 461 ( 1976 ) .
\7 .
j. k. webb , r. f. carswell , m. j. irwin , m. v. penston , _ mon . not .
r. astron .
soc .
_ * 250 * , 657 ( 1991 ) .
\8 .
our first d / h measurement is reported in ( 2 ) , our second measurement is the subject of this article .
\9 .
a. songaila , l. l. cowie , c. j. hogan , m. rugers , _ nature _ * 368 * , 599 ( 1994 ) .
\10 .
r. f. carswell , m. rauch , r. j. weymann , a. j. cooke , j. k. webb , _ mon . not .
r. astron .
soc .
_ * 268 * , l1 ( 1994 ) .
\11 .
k. jedamzik , g. m. fuller , _ astrophys . j. _ * 452 * , 33 ( 1995 ) ; k. jedamzik , g. m. fuller , g. j. mathews , _ ibid . _
* 423 * , 50 ( 1994 ) .
. s. a. levshakov & w. h. kegel , _ mon . not .
r. astr .
soc . _ * 278 * , 497 - 505 ( 1996 ) .
\13 .
e. m. hu , t. s. kim , l. l. cowie , a. songaila , _ astron . j. _ * 110 * , 1526 ( 1995 ) ; d. kirkman and d. tytler , _ astrophys . j. _ in press .
\14 .
c. c. steidel , _ ibid . _ * 74 * , 37 ( 1990 ) .
\15 .
m. donahue and j. m. shull , _ astrophys .
j. _ * 383 * , 511 ( 1991 ) ; w. d. mathews and g. ferland , _ ibid . _ * 323 * , 456 ( 1987 ) .
\16 .
m. g. edmunds , _ mon . not .
r. astron .
soc . _ * 270 * , l37 ( 1994 ) ; n. hata , r. j. scherrer , g. steigman , d. thomas , t. p. walker , s. bludman , p. langacker , _ phys .
rev .
lett . _ * 75 * , 3977 ( 1995 ) .
\17 .
j. l. linsky , a. diplas , b. e. wood , a. brown , t. r. ayres , b. d. savage , _ astrophys . j. _ * 451 * , 335 ( 1995 ) .
\18 .
g. steigman , & m. tosi , _ astrophys .
j. _ * 453 * , 173 ( 1995 ) ; g. steigman , & m. tosi , _ astrophys .
j. _ * 401 * , 150 ( 1992 ) ; b. d. fields _ ibid . _ * 456 * , 478 ( 1996 ) .
\19 .
m. rugers and c. j. hogan , _ astrophys .
j. _ in press ( 1996 ) .
\20 .
the model presented by rugers & hogan ( @xcite ) is a poor fit to the data in several places : d i , h i and blue wing and central spike , and high order lines at 3974 , 3972 , 3962.5 , 3961,3959.5 , 3954.8 . \21 .
g. steigman _ mon . not .
r. astron .
soc . _ * 269 * , l53 ( 1994 ) .
\22 .
e. j. wampler , g. m. williger , j. a. baldwin , r. f. carswell , c. hazard , r. g. mcmahon , _ astron .
astrophys .
_ in press ( 1996 ) .
\23 .
r. f. carswell _ et al .
_ , _ et al .
mon . not .
r. astr .
soc . _ * 278 * , 506 ( 1996 ) .
\24 .
j. c. mather et al .
, _ astrophys .
j. _ * 420 * , 439 ( 1994 ) .
\25 .
m. persic and p. salucci , _ mon . not .
r. astron .
soc . _ * 258 * , 14 ( 1992 ) .
\26 .
b. e. j. pagel , e. a. simonson , r. j. terlevich , m. g. edmunds , _ mon .
not r. astron .
soc . _ * 255 * , 325 ( 1992 ) .
\27 .
k. a. olive and g. steigman , _ astrophys
. j. suppl .
ser . _ * 97 * , 49 ( 1995 ) .
\28 .
thuan , t. x. , izotov , y. i. , & lipovetsky , v. a. in `` interplay between massive star formation , the ism and galaxy evolution '' , ed .
d. kunth , b.guiderdoni , m. heydari - malayeri & t. x. thuan , ( gif - sur - yvette : editions frontieres ) ( 1996 ) .
\29 .
r. t. rood , g. steigman , b. m. tinsley , _ astrophys .
j. _ * 207 * , l57 ( 1976 ) ; e. vassiliadis & p. r. wood , _ ibid . _ * 413 * , 641 ( 1993 ) .
\30 .
rood , r.t . ,
bania , t.m . & wilson , t.l .
1992 _ nature _ , * 355 * , 618 .
\31 .
d. galli , f. palla , f. ferrini , u. penco , _ ibid .
_ * 443 * , 536 ( 1995 ) .
\32 .
k. a. olive , r.t .
rood , d.n.schramm & j. truran & e. vangioni - flam 1995 _ astrophys .
j. _ , * 444 * , 680 ( 1995 ) .
\33 .
f. spite and m. spite , _ astron .
astrophys . _ * 115 * , 357 ( 1982 ) ; m. spite , f. spite , j. p. maillard , _ ibid . _ * 141 * , 56 ( 1984 ) .
\34 .
r. rebolo , j. e. beckman , p. molaro , _ ibid . _
* 192 * , 192 ( 1988 ) ; l. hobbs and c. pilachowski , _ astrophys . j. _ * 326 * , l23 ( 1988 ) .
\35 .
j. a. thorburn , _ ibid .
_ * 421 * , 318 ( 1994 ) .
\36 .
s. g. ryan , t. c. beers , c. p. deliyannis , j. a. thorburn , _ ibid .
_ * 458 * , 543 ( 1996 ) ; c. p. deliyanis , s. g. ryan , t. c. beers , j. a. thorburn , _ ibid . _
* 425 * , l21 ( 1994 ) .
\37 .
m. pratt _
et al . _ , ( macho collaboration ) , `` workshop on dark matter in the universe '' , santa monica , february ( 1996 ) \38 .
d. zaritsky & s.d.m .
white _ astrophys . j. _
, 435 , 599 ( 1994 ) .
\39 .
white , s. d. m. , navarro , j. f. , evrard , a. e. , & frenk , c. s. nature 366 , 429 ( 1993 ) ; m. markvevitch , r. mushotzky , h. inoue , k. yamashita , a. furuzawa , y. tawara , _ astrophys . j. _ * 456 * , 437 ( 1996 ) .
\40 .
we add or subtract all three errors to give the largest range for @xmath83 .
\41 .
a. linde , `` recent progress in inflationary cosmology '' , 1st resceu inter .
symp . ,
tokyo , nov .
( 1995 ) .
\42 .
r. f. carswell , j. k. webb , j. a. baldwin , b. atwood , _ astrophys .
j. _ * 319 * , 709 ( 1987 ) includes a description the vpfit code which we used to fit absorption lines .
\43 .
s. s. vogt _
et al . , proc .
spie _ * 2198 * , 362 ( 1994 ) .
\44 .
j. mould , _ et al .
_ , _ astrophys
. j. _ * 449 * , 413 ( 1995 ) .
\45 .
we are very grateful to w.m .
keck foundation for their generous gift to science .
we would like to thank r.f .
carswell and t. barlow for providing software which made our analysis possible .
we thank wayne wack , meg whittle , randy campbell , and keith baker for their assistance in obtaining the data presented here .
we thank george fuller , bernard pagel , max pettini , alvio renzini , sean ryan for important science discussions and suggestions . |
the latest generation of near and mid infrared surveys such as the vvv ( _ vista variables in the via lactea _ , * ? ? ? * ) , ukidss ( _ ukirt infrared deep sky survey _
* ) , glimpse ( _ galactic legacy infrared midplane survey extraordinaire _ ) , and wise ( _ wide - field infrared survey explorer _
* ) surveys have fostered the discovery of countless stellar nurseries .
@xcite discovered 92 star clusters using glimpse data ( 3.6 , 5.8 , 8.0 , @xmath6 ) , of which @xmath7% likely host a young stellar demographic .
similarly , a sizable fraction of the 96 clusters discovered by @xcite in the vvv survey ( @xmath8 ) house heavily obscured young populations .
infrared photometry is particularly efficient at revealing embedded sources which are otherwise obscured from optical surveys since @xmath9 ( e.g. , * ? ? ?
the total extinction tied to photometry employed here is @xmath10 for 2mass @xmath2 ( @xmath11 ) , and @xmath12}$]:@xmath13}$]:@xmath14}$]:@xmath15 } \sim0.05 a_{v}$ ] for wise photometry ( e.g. , * ? ? ?
* ) . in this study
, initial constraints are placed on young stellar objects ( ysos ) discovered in the field of the @xmath16 classical cepheid su cas . in
[ s - cl ] , optical ( dss ) and @xmath17 ( wise ) images of the cluster are compared , and its apparent extent is determined from the latter ; [ s - sed ] , seds are constructed to classify the ysos according to the prescription outlined by @xcite , and to assess whether the yso class varies as a function of radial distance from the cluster center ; [ s - ccd ] , @xmath18 color - color diagrams are constructed for cluster stars and an adjacent comparison field .
the field of the classical cepheid su cas hosts two clusters .
the first encompasses su cas , and @xcite speculated that the cepheid is a cluster member .
the connection is of particular importance since su cas is the shortest - period calibrator of the galactic @xmath19 wesenheit function .
such functions are employed to constrain @xmath20 and dark energy @xcite .
observations obtained from the abbey ridge observatory @xcite for the parent cluster of su cas shall be described in a forthcoming study .
the second cluster , detailed here , lies @xmath21 north east of su cas ( fig .
[ fig - cl ] ) .
the cluster is absent from optical images of the region ( @xmath22 ) . a nearby yso and an overdensity in 2mass data
were noted previously ( e.g. , * ? ? ?
* ; * ? ? ?
the 3.4 , 4.6 , 12 , and @xmath23 wise images unambiguously reveal a cluster centered at j2000 02:54:31.4 + 69:20:32.5 .
the cluster deviates from spherical symmetry , and exhibits an apparent diameter of @xmath24 in the north - south direction .
the cluster s extent in the east - west direction is approximately half that value ( @xmath25 ) .
the ysos were classified by assessing the spectral energy distribution according to the approach of @xcite .
the sed for each star was constructed in the canonical fashion : @xmath26 @xmath27 where @xmath28 is the wavelength of the passband ( cm ) , @xmath29 is the magnitude in that passband , @xmath30 is the speed of light ( cm / s ) , @xmath31 is the zero - magnitude flux ( jy ) , @xmath32 is a conversion factor ( @xmath33 ergs @xmath34 s@xmath35 hz@xmath35 ) , and @xmath36 is the slope of the function . class i , flat ( hereafter class f ) , and class ii ysos exhibit @xmath37 , @xmath38 , and @xmath39 accordingly ( see also * ? ? ? * ) .
the slope ( @xmath36 ) of equation ( [ eqn - fd ] ) is determined from 2mass @xmath2 and photometry extending redward of @xmath3 . however , the wavelength dependence of extinction invariably biases the perceived slope .
radiation emitted at 2mass @xmath2 exhibits increased sensitivity to extinction than wise passbands .
simulations for j025427.46 + 692012.8 ( class i ) and j025435.68 + 692222.7 ( class f ) indicate that @xmath36 varies according to : @xmath40 .
that bias may be mitigated by determining @xmath36 based solely on wise photometry since the passbands display nearly equivalent extinction ratios ( @xmath12}$]:@xmath13}$]:@xmath14}$]:@xmath15 } \sim0.5 a_{k_s}$ ] ) . in cases where @xmath2 was unavailable the determination of @xmath36 resulted from wise photometry .
the correlation between the slope ( @xmath36 ) derived including and excluding 2mass @xmath2 is described by : @xmath41 . for 53 objects detected within @xmath25 of the cluster centre : 19 ( 36% ) are class i , 21 ( 40% ) are class f , and 13 ( 25% ) are class ii objects .
beyond @xmath5 from the cluster center and the detection rates change to 11 ( 18% ) class i , 13 ( 61% ) class f , and 37 class ii objects . within @xmath4 10 ( 20% )
class i , 5 ( 10% ) class f , and 0 class ii sources lack 2mass photometry . beyond @xmath42
a total of 9 ( 15% ) class i , 6 ( 10% ) class f , and 3 ( 5% ) class ii sources were not catalogued by 2mass .
clcl j025426.03 + 691837.2 & i ( @xmath2+w ) & j025428.46 + 691851.6 & ii ( @xmath2+w ) + j025423.45 + 692011.6 & i ( @xmath2+w ) & j025441.33 + 691911.1 & ii ( @xmath2+w ) + j025423.60 + 692221.4 & i ( @xmath2+w ) & j025432.89 + 692008.7 & i ( @xmath2+w ) + j025425.93 + 692251.4 & f ( @xmath2+w ) & j025436.88 + 691911.3 & f ( @xmath2+w ) + j025437.43 + 692113.6 & i ( @xmath2+w ) & j025434.48 + 692316.0 & f ( @xmath2+w ) + j025435.68 + 692222.7 & f ( @xmath2+w ) & j025424.19 + 692051.9 & f ( @xmath2+w ) + j025439.36 + 691838.6 & f ( @xmath2+w ) & j025437.54 + 691902.6 & f ( @xmath2+w ) + j025433.11 + 691753.3 & f ( @xmath2+w ) & j025429.86 + 692221.9 & i ( @xmath2+w ) + j025425.62 + 691755.5 & ii ( @xmath2+w ) & j025437.03 + 692020.0 & ii ( @xmath2+w ) + j025427.46 + 692012.8 & i ( @xmath2+w ) & j025434.66 + 691933.9 & ii ( @xmath2+w ) + j025430.67 + 692100.6 & i ( @xmath2+w ) & j025438.85 + 691919.0 & f ( @xmath2+w ) + j025441.82 + 691945.9 & f ( @xmath2+w ) & j025430.94 + 692034.8 & i ( w ) + j025421.04 + 691932.9 & ii ( @xmath2+w ) & j025429.32 + 692000.0 & f ( w ) + j025439.47 + 692244.1 & f ( @xmath2+w ) & j025428.12 + 692057.3 & i ( w ) + j025420.10 + 691954.7 & f ( @xmath2+w ) & j025428.92 + 692122.1 & i ( w ) + j025436.28 + 692154.2 & i ( @xmath2+w ) & j025427.83 + 691953.5 & i
( w ) + j025422.57 + 691948.7 & f ( @xmath2+w ) & j025433.74 + 692139.4 & i ( w ) + j025420.65 + 692122.5 & ii ( @xmath2+w ) & j025429.20 + 692141.1 & i ( w ) + j025434.16 + 691824.0 & f ( @xmath2+w ) & j025426.00 + 691951.3 & i ( w ) + j025431.31 + 691929.6 & ii ( @xmath2+w ) & j025435.57 + 692139.0 & f ( w ) + j025428.03 + 691801.1 & ii ( @xmath2+w ) & j025425.68 + 692121.3 & f ( w ) + j025430.92 + 691919.1 & ii ( @xmath2+w ) & j025426.03 + 692207.7 & f ( w ) + j025436.07 + 691844.0 & ii ( @xmath2+w ) & j025424.17 + 691915.3 & f ( w ) + j025434.47 + 691835.3 & ii ( @xmath2+w ) & j025432.46 + 691805.3 & i ( w ) + j025435.30 + 692246.0 & f ( @xmath2+w ) & j025431.18 + 692311.2 & i ( w ) + j025424.17 + 692136.5 & f ( @xmath2+w ) & j025429.68 + 691742.5 & i ( w ) + j025420.46 + 691919.5 & ii ( @xmath2+w ) & & + [ table-1 ] the aforementioned statistics reaffirm that the yso class varies with radial distance from the cluster center . the fraction of less - evolved objects decreases as a function of increasing radial distance .
admittedly , @xmath36 may be biased by the effects of photometric contamination for stars near the crowded cluster core .
@xmath18 color - color diagrams were constructed for the cluster stars and a comparison field @xmath43 distant ( fig .
[ fig - ccd ] ) .
the sequence of ysos is absent from the comparison field .
the latter is dominated by unreddened late - type field stars according to the intrinsic @xmath18 relation of @xcite .
conversely , several cluster members are heavily obscured ( @xmath44 ) .
indeed , radiation emitted from the 10 class i sources lacking @xmath2 photometry may suffer additional extinction .
the ysos are highlighted on the color - color diagram by class ( fig .
[ fig - ccd ] ) .
identifications for the objects , which include their j2000 coordinates and classifications , are provided in table [ table-1 ] .
discrete co and hi emission indicate that the cluster may be coincident with a complex foreground to su cas @xcite , or beyond the cepheid ( @xmath45 pc , * ? ? ?
* ) at @xmath46 kpc .
the latter sets a soft upper limit since most young clusters exhibit a height @xmath47 less than that implied ( note that @xmath48 ) .
additional research is required .
countless ysos were discovered at j2000 coordinates 02:54:31.4 + 69:20:32.5 ( fig .
[ fig - cl ] ) , which is @xmath49 north east of the classical cepheid su cas .
the cluster appears non - symmetric and displays an apparent diameter of @xmath24 ( fig .
[ fig - cl ] ) .
seds constructed using 2mass and wise photometry indicate that the cluster hosts 19 ( 36% ) class i , 21 ( 40% ) class f , and 13 ( 25% ) class ii objects ( fig . [ fig - sed ] ) . at a distance beyond 3@xmath50 from the cluster center the sampling changes markedly to 11 ( 18% ) class i , 13 ( 20% ) class f , and 37 ( 60% ) class ii objects .
the statistics reaffirm that the yso class exhibits a radial dependence .
10 class i sources lacking 2mass @xmath2 photometry were classified via seds tied to wise observations .
the slope @xmath36 ( equation [ eqn - fd ] ) inferred from that approach appears less biased by the wavelength dependence of extinction since wise passbands exhibit nearly equivalent extinction ratios ( as compared to @xmath2 ) .
the extreme extinction obscures many of the cluster s ysos ( @xmath44 ) beyond optical detection ( fig .
[ fig - ccd ] ) . the latest generation of infrared surveys ( e.g. , wise , glimpse , etc .
) shall continue to foster the detection and characterization of new stellar clusters , as reaffirmed here .
additional research , which includes spectroscopic follow - up of the cluster and broader region , is required .
bonnarel , f. , fernique , p. , bienaym , o. , et al .
2000 , a&as , 143 , 33 borissova , j. , bonatto , c. , kurtev , r. , et al .
2011 , a&a , 532 , a131 flaherty , k. m. , pipher , j. l. , megeath , s. t. , et al .
2007 , apj , 663 , 1069 freedman , w. l. , & madore , b. f. 2010 , araa , 48 , 673 froebrich , d. , scholz , a. , & raftery , c. l. 2007 , mnras , 374 , 399 lada , c. j. 1987 , star forming regions , 115 , 1 lane , d. j. 2008 , jaavso , 36 , 143 liu , w. m. , padgett , d. l. , leisawitz , d. , fajardo - acosta , s. , & koenig , x. p. 2011 , apjl , 733 , l2 lucas , p. w. , et al .
2008 , mnras , 391 , 136 majaess , d. j. , turner , d. g. , lane , d. j. , & moncrieff , k. e. 2008 , jaavso , 36 , 90 mercer , e. p. , clemens , d. p. , meade , m. r. , et al .
2005 , apj , 635 , 560 minniti , d. , et al . 2010 , new astronomy , 15 , 433 nishiyama , s. , tamura , m. , hatano , h. , et al .
2009 , apj , 696 , 1407 storm , j. , gieren , w. , fouqu , p. , et al .
2011 , a&a , 534 , a94 straiys , v. , & laugalys , v. 2008 , baltic astronomy , 17 , 253 straiys , v. , & kazlauskas , a. 2010 , baltic astronomy , 19 , 1 turner , d. g. , & evans , n. r. 1984 , apj , 283 , 254 turner , d. g. 2011 , rmxaa , 47 , 127 wright , e. l. , eisenhardt , p. r. m. , mainzer , a. k. , et al . 2010 ,
aj , 140 , 1868 | _ preliminary _ constraints are placed on a cluster of ysos ( j2000 02:54:31.4 + 69:20:32.5 ) discovered in the field of the classical cepheid su cas .
wise 3.4 , 4.6 , 12 , and 22 @xmath0 images reveal that the cluster deviates from spherical symmetry and exhibits an apparent diameter of @xmath1 .
seds constructed using 2mass @xmath2 ( @xmath3 ) and wise photometry indicate that 19 ( 36% ) class i , 21 ( 40% ) class f , and 13 ( 25% ) class ii objects lie @xmath4 from the cluster center .
conversely , 11 ( 18% ) class i , 13 ( 21% ) class f , and 37 ( 61% ) class ii objects were detected for @xmath5 .
approximately 50% of the class i sources within @xmath4 were classified solely using wise photometry owing to the absence of detections by 2mass . |
the chemical evolution of galaxies during their earliest stages provides important insights for the initial physical and environmental conditions of the galaxy formation process .
the star formation history ( sfh ) of galaxies is one of the important factors determining the path of their chemical evolution because the number of stars formed ( i.e. the star formation rate ) with a given initial mass function ( imf ) controls the rate of chemical enrichment ( matteucci 2012 ) . one key observable signature of the sfh of a galaxy is the metallicity distribution function ( mdf ) of its stars .
thus , we can test possible scenarios for the early epochs of galaxy formation by comparing the observed mdf of the oldest stellar population in the galaxy with several analytical models of chemical evolution . in this context ,
the sculptor group dwarf galaxies provide an excellent test bed for investigating the early chemical evolution of dwarf satellite galaxies formed in low density environments ( i.e. outskirts of dense galaxy clusters , local group - like group environment , or field - like environment ; see section 4.4 of the present study ) .
the sculptor group , also known as the sculptor filament , is one of the closest groups of galaxies to our local group ( i.e. distance to the group center , @xmath23.9 mpc ; karachentsev et al .
five bright late - type galaxies ( ngc 55 , 247 , 253 , 300 , and 7793 ) and at least 16 dwarf satellites form a very loosely bound , field - like system stretched along the line of sight over @xmath35 mpc .
our target galaxies , eso294-g010 and eso410-g005 , are phoenix - like transition - type dwarf galaxies ( dtrans ) that morphologically resemble the gas - poor dwarf spheroidal galaxies ( dsphs ) but contain detectable amounts of neutral hydrogen ( h i ) gas ( i.e. @xmath4 ; bouchard et al . 2005 ; jerjen , h. , freeman , k. c. , & binggeli , b. 1998 ) , and show signs of recent star formation activity .
most recently , from their detailed analysis of the color - magnitude diagrams obtained using deep hst / acs imaging for the five sculptor group dwarf galaxies , lianou et al .
( 2013 ; hereafter l13 ) found population gradients in eso294-g010 and eso410-g005 in the sense that young blue main sequence stars ( age @xmath5 100 myr ) and intermediate age asymptotic giant branch stars ( age @xmath3 1 - 2 gyr ) are more centrally concentrated , while old horizontal branch stars ( age @xmath6 10 gyr ) tend to be more spatially extended .
the high quality of those hst / acs images also allowed the first detection of a significant population of rr lyrae variable star candidates beyond the local group .
da costa et al .
( 2010 ) were able to discover numerous rr lyrae candidates in both dtrans galaxies using a template light curve technique ( layden 1998 ) .
the presence of the rr lyrae stars in eso294-g010 and eso410-g005 directly confirms the existence of ancient stellar populations with ages @xmath6 10 gyr and indicates that the sculptor group dwarf galaxies share a common epoch of the earliest star formation with our local group galaxies even though these two groups of galaxies have quite different local density environments .
in addition to being a direct probe of ancient stellar populations , rr lyrae stars provide a variety of utilities to investigate a number of important astrophysical applications ; these stars are well - known reliable population ii distance indicators .
there is a correlation between the metallicities and the pulsation properties ( i.e. periods and amplitudes ) of the fundamental mode rr lyrae stars ( rr0 or rrab ) . using this relation ( alcock et al .
2000 ; sarajedini et al . 2006 ) , we can calculate the metallicities of individual rrab stars , and thus construct the mdf for a purely old stellar population .
lastly , the intrinsic colors of rrab stars at their faintest luminosity are largely independent of their other physical properties [ @[email protected] ; guldenschuh et al . 2005 ] , therefore using their observed colors at minimum light we can also estimate the line - of - sight reddening . the main goal of the present study is to examine the mdfs of legitimate rr lyrae stars in the two sculptor group dtrans galaxies , eso294-g010 and eso410-g005 , in order to investigate the early chemical evolution of these systems , especially in light of the low density environments in which they reside .
this paper is organized as follows .
section 2 provides a description of the data set and photometry process . in section 3
we report the results of our analysis including the general trends in the color - magnitude diagrams ( cmds ) of the galaxies ( sec 3.1 ) , the details of our rr lyrae detection method and the pulsation properties of the rr lyrae candidates ( sec 3.2 ) , the calculation of the metallicities of individual rr lyraes ( sec 3.3 ) , and the distance measurements ( sec 3.4 ) . in section 4 based on our estimates in the previous sections , we present our in depth discussion on the star formation histories , early chemical evolutions , the luminosity - metallicity relations of the our target dtrans , and the environmental effects on the evolution of the near field dwarf galaxies .
finally in section 5 we present a summary of our results .
the science images of eso294-g010 and eso410-g005 used in this study were taken with the advanced camera for surveys wide field channel ( acs / wfc ) on board @xmath9 @xmath10 @xmath11 ( @xmath12 ) as a part of go-10503 ( pi : da costa ) .
the observing log for the retrieved archival images is shown in table 1 .
the detailed observing strategy has been provided by da costa et al .
( 2010 ) . to briefly describe the observations , the central regions of each galaxy were placed on one of wfc chips so that the imaging covers almost the entire visual extent of the targets .
eso294-g010 was imaged 12 times in the f606w and 24 times in the f814w filter with an exposure time of 1160s during the observing baseline of @xmath3 1.86 days .
the same number of exposures were obtained for the observations of eso410-g005 with an exposure time of 1120s over the observing baseline of 4.55 days . in this way
the observations provide a good quality time - series photometry that allows us to detect genuine rr lyrae variable candidates in each galaxy .
ccccccc eso294-g010 & 00 26 33.40 & -41 51 19.0 & f606w & 12x1160s & j9bd0701*,j9bd0801*,j9bd1202 * & 859.27358 861.13659 + & & & f814w & 24x1160s & j9bd0702*,j9bd0802*,j9bd0901*j9bd1201 * & 859.40613 861.00342 + & & & & & & + eso410-g005 & 00 15 31.40 & -32 10 47.0 & f606w & 12x1120s & j9bd1301*,j9bd1401*,j9bd1802 * & 861.40425 863.21702 + & & & f814w & 24x1120s & j9bd1302*,j9bd1402*,j9bd1501*j9bd1801 * & 861.53711 866.08158 + we photometered the point sources in the cosmic - ray cleaned _
@xmath13 images with the acs module of dolphot package ( dolphin 2000 ) by following the standard procedures for dolphot . these _
@xmath13 images are calibrated fits images that were produced through the acs calibration ( calacs ) pipeline , which performs basic image reduction and data calibration including cosmic ray rejection .
pre - constructed point spread functions ( psfs ) for each acs passband were used to expedite the psf photometry . once bad pixels were removed , each acs / wfc image which contains both chips
was split into two separate image files with single - chip format .
then , all of the individual images were aligned and photometered simultaneously .
aperture corrections were calculated for each image with the default settings of dolphot .
the instrumental magnitudes were transformed to the native acs / wfc vegamag system as well as the ground - based johnson - cousins vi system using the calibration equations of sirianni et al .
( 2005 ) implemented in dolphot .
the resultant magnitudes were corrected for the loss of charge transfer efficiency ( cte ) as described in acs isr 2003 - 09 .
we performed photometric completeness tests using the artificial star feature ( `` acsfakelist '' ) in dolphot .
this experiment allows us to gauge the degree of photometric completeness , as well as photometric errors .
the final set of standard vi photometry was extracted from the resultant file by selecting only those stars with `` object type '' equal to 1 ( i.e. a good star ) .
these stars also meet the additional selection criteria set by the several photometric quality parameters : @xmath14 10 , 0.5 @xmath15 + 0.5 , and 1 @xmath16 + 1 .
the sharpness approaches zero for a star with a nearly perfect stellar profile , while for a source that is too sharp ( i.e. a cosmic ray ) this value becomes positive .
a negative value indicates a source that is extended or too broad ( i.e. a blend , cluster or galaxy ) . according to the test done by dolphin et al .
( 2004 ) , good stars in an uncrowded field should have sharpness values between -0.3 or + 0.3 .
the crowding parameter is defined in magnitude units . for an isolated star ,
the value is zero . for stars in a highly crowded field ,
this value becomes larger than unity ( positive or negative ) .
color - magnitude diagrams ( cmds ) of eso294-g010 and eso410-g005 in the standard vi system are presented in figures 1 and 2 .
our artificial star tests show that the point sources in both galaxies are well photometered to @xmath3 2 magnitude below the horizontal branch ( hb ) level with better than @xmath3 50% photometric completeness ( see the right panels of fig 1 & fig 2 ) . at the level of the hb magnitude
the photometric completeness reaches @xmath3 88% .
therefore we assume that photometric incompleteness does not significantly affect our analysis of the rr lyrae variable stars .
according to the morphological type @xmath17 ( de vaucouleurs et al .
1991 ; karachentsev et al .
2004 ) , and the classifications for dwarf galaxies ( mateo 1998 ) , both eso294-g010 ( @xmath17=3 ) and eso410-g005 ( @xmath17=1 ) satisfy the criteria of transition - type dwarf galaxies ( dtrans ) that exhibit reduced recent star formation but relatively high gas fractions . the tidal indices ( @xmath18 ; karachentsev et al . 2004 , 2013 ; also see section 4.4 . of the present study ) of these two dtrans systems indicate that they are gravitationally bound to their nearest neighbor , ngc 55 which is a nearly edge - on barred irregular galaxy . in general , given that these two dtrans galaxies have comparable mass and size , the cmds of eso294-g010 and eso410-g005 show a close resemblance to each other . both of the two dtrans feature diverse stellar constituents with wide ranges of age and metallicity : young blue main - sequence stars ( ymss ) , relatively weak but recognizable intermediate mass he core - burning blue loop stars ( bls ) , a handful of vertical extension red clump stars ( vrcs ) and asymptotic giant branch stars ( agbs ) both of which represent intermediate - age ( 1 10 gyr ) populations , dominant first ascent red giant branch stars ( rgbs ) , tightly gathered red clump stars ( rcs ) , and distinctively stretched horizontal branch stars ( hbs ) that strongly signal an ancient stellar population ( age@xmath610 gyr ) .
detailed descriptions of each characteristic stellar constituent in these two galaxies can be found in the recent work of l13 .
we employed the technique of yang et al .
( 2010 ; y10 hereafter ) and yang & sarajedini ( 2012 ; ys12 hereafter ) to identify and characterize rr lyrae stars in eso294-g010 and eso410-g005 . in summary , we start by searching all possible variable star candidates at the level of the hb brightness ( 26.5 @xmath5 v @xmath5 27.6 ) .
the variability of each star within the given range of v magnitude was evaluated by the reduced @xmath19 defined by the following formula : @xmath20 , ( 1)\ ] ] .
any data points located more than @xmath83@xmath21 from the mean magnitude were excluded from the calculation of @xmath19 . only those stars with @xmath19 values greater than 3.0 were considered as potential variable candidates [ i.e. typical non - variable stars at the hb level ( v(hb)@xmath327 ) have @xmath19 values less than 3.0 ] .
the vi time series photometry of the variable star candidates that satisfy our variability criterion was analyzed with our template light curve fitting routine , rrfit , which fits both v and i band data simultaneously yields the best fitting light curve parameters such as period , amplitude , epoch at maximum light , and mean magnitudes ( see ys12 for more details ) .
each output best - fit light curve was carefully examined by eye .
then in the vi cmds , we checked the mean color and magnitude of those rr lyrae candidates that passed our eye - inspection to make sure they are located in the expected region of the hb .
finally , in order to further ensure that our sample contains genuine rr lyrae variables , we applied marginal v - i color ranges using the blue and red edges of the empirical instability strip for rr lyrae variables from mackey & gilmore ( 2003 ) to which we added a 3@xmath21 @xmath22 error ( i.e. @xmath23 mag ) in the ( v - i ) colors at the level of the hbs .
figure 3 illustrates zoomed - in cmds with the rr lyrae candidates identified from our analysis .
both cmds show that the bulk of genuine rrl candidates are well selected within the adopted v - i color ranges for each target galaxy . after applying the above procedures , we have identified and characterized 232 rr lyrae stars ( 219 rrab and 13 rrc ) in eso294-g010 and 269 rr lyraes ( 225 rrab and 44 rrc ) in eso410-g005 .
now we go back to the reference images to examine the individual rr lyraes to see how many blended objects were included among the selected rr lyrae samples .
our examination reveals that there are 73 blended objects ( 68 rrab + 5 rrc ) among the 232 rrl samples in eso294-g010 , while we found 92 blended objects ( 73 rrab + 19 rrc ) among the 269 rrl candidates in eso410-g050 .
we further examine the pulsation properties ( see fig 9 .
period - amplitude diagrams in section 3.2.3 ) of these objects to investigate whether the blending causes any systematic bias in the mean pulsation properties of our rr lyrae samples .
we find no indication of significant systematic biases in the period - amplitude space due to the blending ; instead , we see that the blended objects follow the mean period - amplitude trend just like the other well - isolated rr lyrae stars .
therefore we assume that our analyses in this study are not negatively affected by the effects of stellar blending .
figures 45 illustrate a sample of the best - fitting light curves for the rr lyrae candidates .
the basic properties of each of these rr lyrae candidates are summarized in tables 2 - 3 .
we also attempted to search for other types of variable stars in both dtrans by expanding our search ranges in the color - magnitude ( @xmath24 ; @xmath25 ) and period ( @xmath26 days ) spaces .
then we apply the same method as we did for the rr lyrae survey .
we have uncovered a significant number of luminous variable ( lv ) candidates in both dtrans ( 22 in eso294-g010 ; 19 in eso410-g005 ) .
their photometric and pulsation properties are summarized in table 4 .
figure 6 illustrates some examples of the best - fitting light curves for the lv candidates in the two dtrans .
these stars are brighter than the rrl group by 0.5 @xmath3 3.0 mag ( see fig 3 ) .
their color - magnitude positions and period - amplitude properties appear to be consistent with those of typical anomalous cepheids , pop ii cepheids , and short - period classical cepheids ( gallart et al .
2004 ; pritzl et al . 2005 ; bernard et al .
we find that these lvs are more concentrated in the inner fields and their spatial distributions seem to be aligned with the major axis of the ellipsoid of each dtran , which might indicate that these stars belong to the young or intermediate - age populations of the dtrans ( see the bottom panels of fig 12 ) .
the derived pulsation parameters of rr lyrae variables are often subject to spurious periods ( or aliases ) that are mainly caused by the effects of undersampling . in order to estimate the degree of any aliases in our period - searching analysis
, we have performed extensive simulations and statistical tests ( y10 ; ys12 ; sarajedini et al .
input periods and amplitudes were randomly generated within an appropriate range of ab - type and c - type rr lyraes .
we then produced @xmath31000 synthetic rr lyrae light curves with these input periods and amplitudes by applying the observing windows - such as observing baseline , number of epochs , cadence , and photometric errors - that were extracted from the time series photometry of a well measured rr lyrae candidate .
finally these synthetic vi time series photometry sets were analyzed by the rrfit routine to check how well the output periods and amplitudes recover the input values .
the comparison between the input and output solutions is presented in figures 78 where we plot the distributions of the input and output periods ( @xmath27p ) , and the difference between these two values as a function of input period .
the next step involved the following statistical test . from the artificial rr lyrae lists , we randomly sample artificial rr lyrae stars with the same number as the observationally detected rr lyraes in each galaxy [ eso294-g010 : 219 ( rrab ) and 13 ( rrc ) ; eso410-g005 : 225 ( rrab ) and 44 ( rrc ) ] to examine their @xmath27p distribution .
the 1-@xmath21 error ( @xmath28 ) of this @xmath27p distribution from the best - fit gaussian can be considered as a good estimate of the standard error in the individual periods derived by our period - searching analysis . in order to improve the statistical robustness of the @xmath28 value
, this random sampling was iterated 10,000 times .
we then consider the average spread ( @xmath29 ) of these 10,000 samples as a realistic error for our period measurements .
the same approach was applied to estimate errors in the v - band amplitude .
we summarize here the main findings of our simulations . 1 . for the rrab stars in eso294-g010 , @xmath3 80% of the input periods
were well recovered within @xmath8 0.05 day , while this recovery fraction was significantly reduced to @xmath3 25% for the rrc stars . in the case of eso410-g005 , this value becomes @xmath3 70% and 30% for the rrab and rrc respectively .
the standard errors ( @xmath30 ) of the individual rr lyrae periods in our analysis are @xmath31=@xmath8 0.0214 and @xmath32=@xmath8 0.0006 days for the rr lyrae candidates found in eso294-g010 . for the eso410-g005 rr
lyrae candidates , we find @xmath31=@xmath8 0.0061 and @xmath32=@xmath8 0.0008 days .
the errors in the v band amplitudes are @xmath33 = 0.0214 mag and @xmath34 = 0.0016 mag for the eso294-g010 rr lyraes , while these values become @xmath33 = 0.0013 mag and @xmath34 = 0.0018 mag for the eso410-g005 rr lyraes .
3 . we found no significant biases in our determination of the mean periods and amplitudes for the rrab stars .
however , it is evident for both galaxies that rrc stars are subject to more significant aliasing as compared to rrab stars .
in addition there exists a systematic trend in the sense that the output periods of rrc stars tend to be longer than the input periods .
thus we conclude that the aliasing in the periods of rrc stars significantly hinders the accurate determination of the pulsation periods of these stars , and it is not possible in this study to utilize their pulsation properties to study other aspects of rr lyrae stars .
it is well known that the period - amplitude ( p - a ) relations ( i.e. bailey diagram ) of rr lyrae stars in the milky way galaxy ( both in the galactic globular clusters and the field galactic halo ) exhibit the so called the oosterhoff dichotomy : oosterhoff type i ( oo i ) systems tend to have intermediate metallicities and a mean pulsation period of @xmath3 0.55 days in the fundamental mode with a low rrc frequency ( @xmath35 ) , while oosterhoff type ii ( oo ii ) systems are generally metal - poor and have mean pulsation periods of @xmath3 0.65 days with a higher fraction of rrc stars ( @xmath36 ; castellani & quarta 1987 ) .
thus the bailey diagram can serve as a useful diagnostic tool for investigating the basic pulsation properties of the rr lyrae populations . in figure 9
we plot the p - a relations and period distributions of our rr lyrae candidates found in eso294-g010 and eso410-g005 .
the loci of oosterhoff type i and ii for the galactic globular cluster ( ggc ) systems [ straight lines ( clement & rowe 2000 ) ; quadratic relations ( cacciari et al .
2005 , zorotovic et al .
2010 ) ] are also plotted for comparison .
solid symbols [ red dots ( rrab ) ; blue triangles ( rrc ) ] indicate the well - isolated rr lyraes stars not affected by crowding , while the blended objects ( see section 3.2.1 ) are marked with open symbols . according to figure 9 ,
the blended stars are indistinguishable from the other well - isolated rr lyrae variables .
thus , we find no significant systematic biases in the period - amplitude space due to stellar blending .
we see that the bulk of the rrab candidates in these two dtrans galaxies appear to be more tightly distributed around the locus of the oosterhoff i type ggcs , while there are no significant population of oosterhoff ii rr lyraes in both galaxies .
the mean pulsation periods of the rrab candidates are @xmath37 = 0.5932 @xmath8 0.0003 ( error1 ) @xmath8 0.0006 ( error2 ) days and @xmath38 = 0.5898 @xmath8 0.0003 ( error1 ) @xmath8 0.0008 ( error2 ) for eso294-g010 and eso410-g005 respectively .
the quoted value of `` error1 '' represents the standard error of the mean and the `` error2 '' value is the error derived from our synthetic light curve simulations .
rr lyrae stars are very useful for tracing the early chemical enrichment histories of dwarf satellite galaxies .
these pulsating variables represent a purely old stellar population ( age @xmath6 10 gyr ) in a given galaxy .
hence , the advantage of using rr lyraes for a robust measurement of the metal abundance of an ancient stellar population particularly stands out in the case of studying dtrans galaxies which exhibit recent star formation and diverse stellar populations with wide ranges of ages and metallicities .
the metal abundance of ab - type rr lyraes can be calculated using the correlation between period , amplitude and @xmath39 $ ] given by alcock et al .
( 2000 ) in the following form @xmath40 = -8.85 [ log p_{ab } + 0.15 amp(v ) ] -2.60 , ( 2)\ ] ] where amp(v ) indicates the v band amplitude , and [ fe / h ] is in the zinn & west ( 1984 ) metallicity scale .
figure 10 shows the metallicity distribution functions ( mdfs ; binned histograms and the best - fit gaussians ) of the rrab stars found in the two dtrans galaxies of the present study .
the mdf of the eso294-g010 rrab stars exhibits a slightly lower metallicity peak as compared to that of the eso410-g005 rrab stars .
the mean metallicities estimated from the best - fit gaussians to the mdfs are @xmath0 \rangle=$]1.77 @xmath8 0.03 ( sem ) for eso294-g010 , and , @xmath0 \rangle=$]-1.64 @xmath8 0.03 ( sem ) for eso410-g005 respectively . the uncertainties shown as sem here represent the standard errors of the mean values .
the systematic errors are likely larger and closer to @xmath30.2 dex as demonstrated by jeffery et al .
( 2011 ) . to better understand the properties of the rr lyrae populations in these two dtrans galaxies , we now compare the rr lyrae mdfs with those of the rgb stars .
we construct mdfs of the rgb stars for both eso294-g010 and eso410-g005 by interpolating ( sarajedini & jablonka 2005 ) within a set of dartmouth isochrones ( dotter et al .
2008 ) that were generated to have metallicity values ranging from -2.5 to + 0.5 dex at a fixed age of 12.5 gyr .
we used three values of @xmath41=$]0.0 , 0.2 , and 0.4 dex to gauge the effects of @xmath42 element enhancement on the overall shape of the rgb mdf . for each galaxy ,
the rgb stars were selected within a magnitude range of @xmath43 ( i.e. @xmath4422.33 ; @xmath4522.36 , see the next section for the details ) . to account for interstellar extinction , reddening values of e(b - v )
= 0.006 and 0.014 for eso294-g010 and eso410-g005 were adopted respectively from the galactic extinction map of schlegel , finkbeiner & davis ( 1998 ) .
figure 11 illustrates the resulting mdfs of the rgb stars compared to those of the rr lyrae stars .
the mdfs are normalized to have the same total counts . in general , the overall shapes of the resultant rgb mdfs are not gaussian but tend to be skewed toward lower abundances .
we also see that the mdfs of the rgb stars are affected by the assumed values of @xmath42 element ratio in the sense that higher the @xmath42 element enhancement , the lower the overall metallicity . for eso294-g010 ,
the rrl stars appear to have a lower mean metal abundance as compared to the rgb stars at @xmath41=$]0.0 and 0.2 dex while the mdfs of these two stellar populations agree fairly well at @xmath41=$]0.4 .
in contrast with eso294-g010 , the metallicities of the rrl and rgb stars in eso410-g005 generally agree well with each other except for where @xmath41=$]0.4 dex .
as seen from the cmds of these two galaxies , their rgb stars exhibit a broad range in @xmath46 colors and noticeable features of intermediate age ( 1 @xmath5 age @xmath5 10 gyr ) populations .
it is worth mentioning here that rejkuba et al .
( 2011 ) demonstrate that using the locations of rgb stars relative to a grid of standard rgb sequences with different metallicities in order to estimate the metal abundance of a diverse stellar population can lead to systematic effects .
in particular , a metallicity bias of 0.1 0.2 dex in the metal - poor direction can be introduced by age shifts from 12 gyr to 8 gyr , and this can cause a metallicity distribution to be skewed to the metal - poor end . the comparisons presented above suggest that the rr lyrae mdfs may be more reliable for studying the early chemical enrichment history of dwarf galaxies because rr lyrae stars are not subject to the age - metallicity degeneracy , and the overall shape of the rrl mdf is not as significantly influenced by an assumed value of the @xmath41 $ ] ratio as compared with mdfs generated using rgb stars . as a result ,
our estimates of the mean metallicities of type - ab rr lyrae stars are free from the metal - poor bias induced by intermediate - age stellar populations ; furthermore , the handful number of rr lyrae candidates within the empirical instability strips in each target galaxy greatly reduces possible bias in the obtained metallicity values due to the evolutionary effects of these pulsation hb stars from the zero age hb ( zahb ) .
therefore , we consider our metallicity estimates to be faithful to the purely ancient stellar populations in eso294-g010 and eso410-g005 .
now we turn our attention to whether the metallicities of the rrab samples show any radial trends throughout our target dtrans .
previously a metallicity gradient has been reported for the rr lyrae stars in tucana , one of the milky way dtrans in the sense that the more luminous , longer period ( i.e. more metal - poor ) rr lyrae variables tend to be located at larger distances from the galactic center as compared to the fainter , shorter period ( i.e. more metal - rich ) rr lyraes ( bernard et al . 2008 ) .
this phenomenon indicates that such a metallicity gradient should have occurred in a very early stage of evolution of this isolated dtran .
we examine the spatial distributions of our rrab samples in the fields of eso294-g010 and eso410-g005 in order to see if any radial trends similar to the case of tucana exist in these sculptor group dtrans . to do this , we divide our rrab samples into three groups [ i.e. metal - rich ( @xmath39 > \langle [ fe / h ] \rangle + 0.2 $ ] dex ) , metal - poor ( @xmath39 < \langle [ fe / h ] \rangle -0.2 $ ] dex ) , and metal - intermediate ( @xmath0 \rangle -0.2 < [ fe / h ] < \langle [ fe / h ] \rangle + 0.2 $ ] ) ] according to their metallicity ranges . then we plot their locations on top of the elliptical isophote which corresponds to the visual extent of each dtran . from figure 12 , we are able to see that all three groups of rrab samples are well mixed with each other from the inner to the outer regions ; therefore we conclude that there are no significant metallicity gradients among the rrab stars for both sculptor group dtrans .
we also tested for a radial trend in the brightnesses of the rrab stars ( see the bottom panels of fig .
similar to the previous test for metallicity , we do not see any systematic difference in their spatial distributions between luminous [ @xmath47 and faint [ @xmath48 groups .
further we checked the spatial distributions of the peculiar peak at around @xmath49 dex ( blue open circles ) in the mdf of the eso410-g005 rrab samples . as we can see from the figure , their spatial distributions do not show any peculiarity .
most recently l13 presented a very detailed analysis of the spatial distributions of several different stellar populations in different age groups in these two dtrans .
their analysis revealed that the old stellar populations such as hb stars are more uniformly distributed throughout the galaxies , while the younger populations such as young blue main sequence stars or agb stars are more concentrated to the inner regions of the dtrans .
the results from our analysis of the spatial distributions of the rrab samples in eso294-g010 and eso410-g005 are consistent with those of l13 indicating that the physical and chemical processes in the early evolutionary phase of these sculptor group dtran seem to be different from the case of the isolated milky way dtran , tucana .
we adopt reddening values of e(b - v ) = 0.006 and 0.014 for eso294-g010 and eso410-g005 , respectively from the galactic extinction map of schlegel , finkbeiner & davis ( 1998 ) in the measurement of the distances to each galaxy .
we are not able to estimate the line - of - sight reddenings toward the target galaxies using the minimum - light colors of the rrab stars because of the relatively low quality of the i - band light curves .
along with the metallicity values of the individual rr lyrae stars estimated in the previous sections , we now have the foundation for the calculation of accurate distances of our target galaxies .
the mean v magnitude of the 219 rrab stars detected in eso294-g010 is @[email protected] mag .
the absolute v magnitudes of the individual rrab stars are calculated using the calibration given by chaboyer ( 1999 ) , @xmath51 + 0.93 $ ] .
this gives a mean absolute v magnitude of @xmath52 .
the quoted error is the amount of uncertainty propagated from our measurement of the metal abundances of the rrab stars using the alcock relation . by employing the extinction law , @xmath530.019 we obtained a distance modulus of @xmath54 mag which places eso294-g010 at a distance of @xmath3 1.9 @xmath8 0.1 mpc . for eso410-g005 ,
the mean v magnitude of 225 rrab stars is @xmath5026.91 @xmath8 0.01 mag .
applying once again the calibration of chaboyer ( 1999 ) for the luminosity - metallicity relation for rrab stars , and the line - of - sight extinction @xmath550.043 , we obtain @xmath560.53 @xmath8 0.07 mag .
this yields a distance modulus of @xmath57 26.33 @xmath8 0.07 mag which is equivalent to a distance of @xmath3 1.9 @xmath8 0.1 mpc .
we have also calculated the distances of the two dtrans galaxies using the i - band magnitude of red giant branch tip ( trgb ) in the vi cmds in order to check the validity of our rr lyrae distance estimate .
we employed a weighted sobel kernel @xmath58 $ ] ( madore & freedman 1995 ) to detect the trgb of each galaxy ( see fig 13 & 14 ) .
this yields @xmath59 22.33 @xmath8 0.05 mag for eso294-g010 and 22.36 @xmath8 0.05 for eso410-g005 .
applying the reddening ratio @xmath601.38 @xmath61 provided by tammann , sandage & reindl ( 2003 ) , and @xmath624.04 @xmath8 0.12 ( bellazzini et al .
2001 , 2004 ) , we obtain @xmath57 26.37 @xmath8 0.13 and 26.38 @xmath8 0.13 for eso294-g010 and eso410-g005 respectively .
the errors represent the quadrature sum of the errors in the apparent and absolute i - band trgb magnitudes .
our distance estimates for eso294-g010 using two independent methods ( rr lyraes and trgb ) agree very well with each other .
while eso410-g005 shows @xmath3 0.1 mag discrepancy , this is still consistent within the margin allowed by the errors .
the distances to these two galaxies derived in this study show excellent agreement with those measured in previous work such as l13 , da costa et al .
( 2010 ) , and the `` updated nearby galaxy catalog '' compiled by karachentsev et al .
the stellar populations of eso294-g010 and eso410-g005 share common features in the vi cmds . however there also exist several noticeable differences in their cmds indicating that these two dtrans systems might have experienced slightly different star formation histories .
the main difference in the cmd features arises from the densities of yms , bl , and blue hb ( bhb ) populations , in the sense that the yms and bl stars in eso410-g005 appear to be more populous than those in eso294-g010 , while the bhb feature seems relatively stronger in eso294-g010 than in eso410-g005 .
this indicates that eso294-g010 apparently experienced reduced recent star formation ( sf ) as compared to eso410-g005 .
our assertion is nicely supported by the recent study of weisz et al .
( 2011 , w11 hereafter ) on the star formation histories ( sfhs ) of 60 dwarf galaxies in the local volume .
they used the same acs / wfc imaging as in this study to extract the sfhs of eso294-g010 and eso410-g005 with the synthetic cmd routine developed by dolphin ( 2002 ) .
w11 s results reveal that the two dtrans galaxies exhibit comparable star formation rates ( sfrs ) during the early epoch ( look - back time @xmath610 gyr ) of their formation and evolution .
since then , eso294-g010 has remained mostly dormant and has been sustaining very weak sfrs lower than its lifetime - averaged sfr , while eso410-g005 shows a sudden enhancement of its sfrs within the most recent @xmath3 2 gyr higher than its lifetime - averaged sfr . as a result , eso294-g010 displays a bluer hb morphology and relatively weak features of luminous yms and core he burning bl stars compared to eso410-g005 . according to the standard cosmological models of galaxy formation ( white & rees 1978 ;
hernquist & quinn 1988 , 1989 ; white & frenk 1991 ) , dwarf galaxies have likely formed through infall of primordial gas into a dark matter halo , and the subsequent star formation episodes determine the paths of their own chemical evolution as typical for galaxies .
however the detailed patterns of the chemical evolution of these dwarf galaxies could be different from that of giant galaxies due to their shallower potential wells which lead to more efficient mixing and also make these systems more vulnerable to various kinds of environmental effects such as supernova driven galactic winds , ram pressure stripping , and tidal disruption .
if this is the case , what would the early chemical evolution of dwarf galaxies in a low density environment such as the sculptor group be like ? to answer this question , we analyze the rr lyrae mdfs of our target dtrans systems using simple chemical enrichment models ( searle & sargent 1972 ; pagel & patchett 1975 ; pagel 1997 ; binney & merrifield 1998 ) which depend on the initial heavy element abundance @xmath63 of the star forming gas , effective nucleosynthetic yield @xmath64 and loss or gain of star forming gas . the simplest version among the chemical evolution models is the closed - box model with an assumption of no inflow or outflow of gas during the evolution of the system .
this model predicts a mdf of a stellar population produced from a star formation episode at a certain age with a gaussian - like exponential function , @xmath65 } \sim \frac{\mathit z - z_{0}}{\mathit y } \mathit{e}^{-(z - z_{0})/y}. ( 3)\ ] ] in these equations , the yield , @xmath64 is equivalent to the mean metallicity of the system and @xmath66 $ ] is the total metal abundance of the system which accounts for the effect of alpha - capture element enhancement [ i.e. @xmath67 = [ fe / h ] + log_{10}\left(0.638\times10^{[\alpha / fe ] } + 0.362\right ) $ ] ; salaris et al . 1993 ] .
a number of studies ( tolstoy , hill , & tosi 2009 and references therein ) have found that the old and metal - poor stars ( age @xmath68 gyr ; @xmath66 < -1.6 $ ] dex ) in nearby dwarf galaxies exhibit the galactic halo - like compositions with an enhanced alpha element ratio @xmath69 \rangle$ ] @xmath3 0.3 dex .
therefore , assuming such an alpha element enhancement should be a reasonable approach .
the red solid line in figure 15 indicates the resulting fits of the simple closed box model with an assumption of a primordial abundance ( e.g. @xmath700 ) to our rr lyrae mdfs of eso294-g010 and eso410-g005 .
we see that the predicted mdfs of the simple closed box model traces the observed mdfs reasonably well ; however , in the case of eso294-g010 , the model mdf appears to slightly overestimate the number of metal - poor stars ( e.g. @xmath39 < $ ] -2.2 dex ) .
this shortage of the low - z stars , first recognized as the g - dwarf problem by van den bergh ( 1962 ) and schmidt(1963 ) , is in fact a common issue for all types of galaxies ( binney & merrifield 1998 ; harris & harris 2000 ; pagel & patchett 1975 ; wheeler et al . 1989 ) . to compensate for this deficit of metal - poor stars ,
we adopt a pre - enrichment ( @xmath71 ) hypothesis for the star forming gas in which the dwarf galaxies were formed .
the blue solid lines in figure 15 illustrate the best - fit pre - enriched closed - box models . for eso294-g010 , a pre - enrichment model with an initial metal abundance of @xmath72 ( i.e. @xmath66_{0 } \sim -2.7 $ ] dex ) and a chemical yield of @xmath73 enhances the fit to the metal - poor tail of the observed mdf as compared to the simple closed box model . in the case of eso410-g005
, a pre - enriched system with an initial metal abundance of @xmath74 ( i.e. @xmath66_{0 } \sim -3.0 $ ] dex ) and a yield of @xmath75 agrees well with the observed rr lyrae mdf featuring a comparable quality of fit to the simple closed box model .
the results of our analysis above naturally lead to the question as to the origin of the pre - enrichment source , and how the conditions in which the pre - enrichment was attained .
it has been suggested by truran & cameron ( 1971 ) that such pre - enrichment in the star forming gas can be achieved through the process of `` prompt initial enrichment '' or `` initial nucleosynthesis spike '' which includes a preferential pre - galactic or proto - galactic event by high - mass stars during the very early phase of galaxy formation .
indeed our estimates of the initial metal abundances ( @xmath76 ; @xmath77 ) from the best - fit pre - enrichment models are comparable to the initial cosmic metal enrichment levels ( @xmath78 ) attained by the end products of population iii stars ( schneider et al .
in addition to this , it can not be ruled out that the proto - galactic bodies of eso294-g010 and eso410-g005 could have been enriched by the products from more evolved systems of their nearest neighbor , ngc 55 .
both hypotheses seem reasonable to explain the early chemical enrichment histories of our target dtrans galaxies in the low density environment of the sculptor group .
another approach to explain the metal - poor tail of the observed mdf is to lift the pure closed - box assumption by introducing an early phase gas inflow ( or accretion ) from the environment .
the simplest form of this `` accretion model '' or `` accreting - box '' modification ( larson 1972 ; timmes et al .
1995 ; gibson & matteucci 1997 ) assumes that the gas inflow rate is equals to the star formation rate of the system .
this means that the mass of star forming gas @xmath79 was a constant during the early epoch of galaxy formation . in these configurations ,
the predicted mdf is simplified to the following hyperbolic function ( binney & merrifield 1998 ; harris & harris 2000 ; sarajedini & jablonka 2005 ) , @xmath80 } \sim { \mathit m_g } \frac{\mathit z}{\mathit y - z}. ( 4)\ ] ] in this formulation , during the early phase of evolution the system accumulates the heavy elements very efficiently with a high nucleosynthetic yield . as @xmath81 approaches @xmath64 the dilution effect of the infalling gas starts to catch up to and ultimately squares off the accumulation rate of the heavy elements , thus the system reaches a steady state with no further global chemical enrichment . in figure 15 , we see that the observed metal - poor tails of the rr lyrae mdfs ( @xmath39 < -1.7 $ ] dex ) of both dwarf galaxies are very well described by our simple accretion models . from the best - fit models , we find @xmath82 and @xmath83 for eso294-g010 and eso410-g005 respectively . as mentioned , this simple accretion model only accounts for the metal - poor stars formed during the very early epoch of the chemical evolution of the system .
one of the most probable scenarios to explain the global pattern of the observed mdfs of the two dtrans galaxies is the so called `` two phase scenario '' of chemical evolution ( wyse & gilmore 1993 ; harris & harris 2000 ; sarajedini & jablonka 2005 ) : a bulk of stars attributed to the metal - poor tails of the mdfs formed in a very early gas infall phase .
once eso294-g010 and eso410-g005 were enriched enough to enter a steady state ( @xmath84 ) , both galaxies entered into the closed - box - like phase and the rest of the stars associated with the remaining parts of the mdfs formed in the enriched gas deriving further chemical enrichment .
our analysis of the mdfs of rrl stars ( age @xmath610 gyr & age spread among rr lyraes in a given stellar system does not exceed @xmath3 2 gyr ; lee & carney 1999 ) also suggests that this phase transition should occur within a short period of time .
the luminosity - metallicity ( or mass - metallicity ) relation of galaxies ( l - m relation ) is one of the major predictions of chemical evolution models ( van den bergh 1962 ; schmidt 1963 ; searle & sargent 1972 ; erb et al .
2006 ) , and it has been confirmed observationally by many studies .
the current understanding of the l - m relation based on both theoretical and observational studies suggests that ( 1 ) the l - m relation exists among all kinds of galaxies within the local volume ( @xmath85 mpc ) regardless of their morphological types ( lequeux et al .
1979 ; skillman et al .
1989 ; zaritsky et al . 1994 ; mateo 1998 ; grebel et al .
2003 , g03 hereafter ; salzer et al .
2005 ) ; ( 2 ) the l - m relation also exists among galaxies at intermediate redshift ( i.e. @xmath86 ) , and it seems that the l - m relation evolves over cosmological time in the sense that the slope at lower redshifts tends to be flatter than that at higher redshifts ( i.e. the metallicity refers to the oxygen nebular emission , @xmath87 in this case ; savaglio et al .
2005 ; kobulnicky & koo 2000 ; kobulnicky et al . 2003 ; kobulnicky & kewley 2004 ; maier et al .
2004 ; shapley et al .
the l - m relation for the old stellar populations of dwarf satellite galaxies in the local group and its neighbors is well established by using mean stellar @xmath39$]-values which are based on both spectroscopic and photometric measurements for old rgb stars ( see figure 1 . of g03 ) .
this l - m relation for nearby galaxies becomes not only a very useful analytical tool for understanding the fundamental differences and a possible evolutionary link between satellite galaxies with different morphological types , but also a basic ladder for the proper interpretation of the l - m relations at high redshifts . with the mean metal abundances of the purely ancient stellar populations measured in section 3.3 by using the periods and amplitudes of the rrab stars , we plot eso294-g010 and eso410-g005 on the l - m relation of the nearby galaxies in figure 16 .
the sample of the known nearby galaxies within a distance of @xmath88mpc has been collected from the seminal work of g03 and the latest nearby galaxy catalog of karachentsev et al .
( 2013 ; k13 hereafter ) .
our compilation ( see table 5 ) includes the recently discovered ultra - faint dwarf systems ( ufds ) around the milky way , and faint andromeda dwarf galaxies ( willman et al .
2005a , b ; belokurov et al .
2006 ; zucker et al .
2004 , 2006a , b , 2007 ; irwin et al . 2007,2008 ; walsh , jerjen & willman 2007 ; mcconnachie et al . 2008 ; martin et al . 2009 ; richardson et al . 2011 ; bell et al . 2011 ; slater et al . 2011 ) , which yields a total of 79 sample dwarf satellite galaxies .
we use integrated k - band luminosity ( @xmath89 ) values ( adopted from table 2 of k13 ) as the abscissa instead of using absolute v or b band magnitudes because the @xmath89 is a better tracer for the total mass of a given galaxy ( drory et al .
2004 ) . from figure 16
, we see that all nearby galaxies , from luminous des ( @xmath90 @xmath91 ) to ufds ( @xmath92 @xmath93 ) , fall along the typical trend of the l - m relation - more luminous galaxies tend to be more metal - rich .
noticeable separations between dsphs ( @xmath94 @xmath95)and dirrs ( @xmath94 @xmath96 ) are also observed in the sense that the dsphs appear to have higher metal abundances as compared to the dirrs at the same luminosity level .
this implies that at the early epochs of formation ( age @xmath68 gyr ) , the dsphs should have relatively more active star formation ( sfs ) . as a result ,
the early chemical enrichment in the dsphs becomes much faster and efficient than in the dirrs .
this scenario of the early sf enhancement in the dsphs is echoed by the work of w11 . using the synthetic cmd analysis ( dolphin 2002 ) ,
they have drawn the average cumulative sfs for five different morphological types ( dsph , dis , dtrans , dspirals , and dtidals ; see figure 6 of w11 ) .
they have shown that on average the dsphs build up the bulk of their stellar populations ( i.e. almost 65 70 @xmath97 of their stellar masses ) at the very earliest epochs of their formation z @xmath6 2 ( i.e. age @xmath6 10 gyr ) , while star formation in the dirrs remains at the @xmath3 50 @xmath97 level during the same epochs .
this fundamental difference between the dsphs and dirrs in their early star formation and chemical enrichment histories has raised serious doubts about the scenario which involves the morphological transformation dirrs into dsphs by certain gas removal processes ( g03 ) .
the locations of our target dtrans galaxies eso294-g010 and eso410-g005 in the l - m relation agree with the typical trend of the other dtrans systems ( @xmath98 @xmath99 ) such as phoenix , tucana , cetus , lgs 3 , kkr 25 , ddo 210 , and antlia . within a luminosity range of 5 @xmath100 7.5 , the l - m relation of the sample dtrans appears to show a closer resemblance to that of the canonical dsphs than dirrs . in this regard
, it has been suggested that a bulk of the dtran galaxies can be considered to be `` @xmath101 @xmath102 '' of the dsph galaxies if the recent star formation of dtrans galaxies were largely suppressed by rapid gas loss ( g03 ) .
the most pivotal differences between dtrans and dsphs galaxies might be the degree of recent star formations ( sfs ) and the local density environments where these galaxies formed and have been evolving .
the morphology - density relation of galaxies in the local universe ( z @xmath3 0 ) has revealed how different types of galaxies tend to be arranged in cluster environments , and provides important clues about how cluster / group environments affects the gas content and global sfr in the galaxy members . in general ,
gas - deficient early - type galaxies and bulge - dominated lenticular galaxies are preferentially located in the central , densest areas of galaxy clusters .
meanwhile , disk - dominated late - types galaxies ( i.e. spirals and irregulars ) tend to be more sparsely distributed in the outer regions of these clusters .
the morphology - density relation appears to be a ubiquitous phenomenon in cluster environments regardless of their shapes and richness ( van der wel et al .
2010 , and references therein ) .
likewise , similar patterns of these morphological segregations have been observed among the dwarf satellite galaxies in the local group and other nearby galaxy groups .
figure 17 illustrates the h i gas content of our sample of dwarf galaxies normalized by their k - band luminosity as a function of the tidal index @xmath103 .
this is basically the same plot as in the 2nd panel of the fig 16 in k13 , but for our 79 volume limited ( @xmath12 mpc ) sample dwarf galaxies .
we consider this ratio to be a proxy of the gas fraction of the galaxy with respect to its total mass .
this is a reasonable approximation not only because the k - band luminosity is an excellent tracer for the total mass of the galaxy but also the @xmath104 values of galaxies in the k - band vary by only a factor of two across a wide range of galaxy morphologies and star formation histories ( i.e. the @xmath104 values in the b - band vary up to almost a factor of 10 in the nearby dwarf galaxy samples ; drory et al .
2004 and references therein ) .
k13 have provided the formula for the tidal index @xmath103 in the following form , @xmath105 where @xmath106 is the spatial distance of a neighboring galaxy and @xmath107 is its mass .
the @xmath103 is a measure of the tidal force strength on a given galaxy exerted by the five most influential neighbors .
thus , it can be a robust estimate for the local tidal field strength . in this definition ,
the high positive @xmath103 values indicate a galaxy residing in a dense environment , and the negative values of @xmath103 represent isolated galaxies or galaxies in a very low density environment . in figure 17 , we see a clear morphological segregation by local density environment and gas fraction in the senses like the following : ( 1 ) dirr galaxies are the most gas - abundant systems and there are no dirrs with a gas fraction @xmath108 less than 0.1 ; ( 2 ) with the exception of which is known to be one of the closest neighboring galaxy groups the lmc / smc pair ( @[email protected] ) and ic 10 ( @[email protected] ) , the majority of the dirr galaxies are mostly isolated in low density environments ( @xmath109 1 ) . together with their unusually high 3d velocities ( @xmath110 @xmath111 @xmath112 ; kallivayalil et al .
2006 ) this could be another piece of evidence supporting the idea that the lmc / smc pair may not be a canonical satellite system of the milky way but an infalling intruder which might be experiencing its first passage around the milky way ( besla et al .
2007 ) ; and ( 3 ) similar to the dirr galaxies , most dtran galaxies are also found to be isolated but in general they have lower h i gas fractions as compared to the dirrs .
all dsph , ufd and three de ( m32 , ngc 205 and ngc 185 ) galaxies in our sample have relatively high tidal index values @xmath113 1 indicating that these galaxies have been subjected to relatively higher local tidal fields ( i.e. they are located within a galactocentric distance of d@xmath3200 kpc to their giant hosts , milky way or m31 ) than the dtrans and dirrs galaxies . in terms of the gas content , some of the dsph systems such as carina , leo i , sextans , draco , sag dsph , and ngc 147 exhibit the lowest h i gas fractions [ @xmath1144 ] ; however , the rest of the early - type dwarfs are not completely gas - deficient but appear to have a wide range of h i gas fractions comparable to the levels of the dtran galaxies . along with the difference in the local density environments , another conspicuous distinction between dsphs and dtrans seems to be the degree of recent sf . according to w11 , within the most recent 1 gyr , the dsph galaxies exhibit largely suppressed sfrs relative to those of dtrans systems .
their estimates of the mean sfhs indicate that the typical dsph and dtran galaxies formed @xmath3 2% and 4% of their total stellar mass respectively within the last 1 gyr .
this sudden drop in the recent sfrs of the dsphs might be due to a rapid gas loss which is a quite plausible process in the high local density environments in which these dsphs reside .
similar phenomena have been observed for many larger galaxy group / cluster environments in the local universe ( z@xmath30 ) .
recent star formation in galaxies located in denser regions appear to be largely quenched , and these galaxies tend to be more early - types and red ( oemler 1974 ; davis & geller 1976 ; dressler 1980 ; postman & geller 1984 ; wetzel , tinker & conroy 2012 ) .
this implies that the effects of environment on galaxy evolution operate in a similar fashion even in the low density environments of the lg and its neighboring groups ( i.e. the sculptor / maffei groups ) .
we have presented an extensive analysis of the rr lyrae stars in two transition type sculptor group dwarf galaxies , eso294-g010 and eso410-g005 . based on the properties of the rr lyrae stars we present the following results . 1 .
we have detected numerous rr lyrae candidates in both eso294-g010 ( @xmath115219 ) and eso410-g005 ( @xmath115225 ) .
based on the bailey diagrams , the characteristics of the rr lyrae stars in these two sculptor group dtrans appear to follow the typical trend of the rr lyrae populations of local group dwarf satellite galaxies .
we construct the mdfs of the rrab stars using the period - amplitude-[fe / h ] relationship of alcock et al .
the mean metallicities estimated from the best - fit gaussians to the mdfs are @xmath116>=$]1.77 @xmath8 0.03 ( sem ) for eso294-g010 , and @xmath116>=$]-1.64 @xmath8 0.03 ( sem ) for eso410-g005 respectively .
our results represent the metallicity values for purely ancient stellar populations in both dtrans galaxies . 3 .
the distance of each dtran was calculated using the absolute v magnitudes of the rrab stars .
we find @xmath117 26.40 @xmath8 0.07 mag for eso294-g010 and @xmath57 26.33 @xmath8 0.07 mag for eso410-g005 .
we have also calculated the distances of the two dtrans using the i - band magnitudes of trgbs in order to check the validity of our rr lyrae distance estimates [ @xmath11826.37 ; @xmath11926.38 ] .
our distance estimates for eso294-g010 using two independent methods ( rrls and trgb ) agree very well with each other . in the case of eso410-g005 , there is a @xmath3 0.1 mag difference but this is consistent within the margin allowed by the errors .
we have compared the rr lyrae mdfs for eso294-g010 and eso410-g005 with several chemical evolution models . for both galaxies , the shapes of the rr lyrae mdfs
are nicely described by pre - enrichment models .
this suggests two possible channels for the early chemical evolution of these sculptor group systems : 1 ) the ancient stellar populations of our target dwarf galaxies might have formed from the star forming gas which was already enriched through `` prompt initial enrichment '' or an `` initial nucleosynthesis spike '' caused by the very first massive stars , or 2 ) this pre - enrichment state might have been achieved by the end products from more evolved systems of their nearest neighbor , ngc 55 .
we also fit a simple accretion model to the rr lyrae mdfs of each galaxy .
the gas infall hypothesis seems to explain the metal - poor tail of the rr lyrae mdfs indicating that we can not completely rule out the role of gas accretion on the early chemical evolution of the dtrans .
the l - m relation of our target dtrans eso294-g010 and eso410-g005 follow the typical trend of other local group dtrans galaxies , such as phoenix , tucana , cetus , lgs 3 , kkr 25 , ddo 210 , and antlia . within a luminosity range of 5 @xmath100 7.5
, the l - m relation of the sample dtrans appear to show a closer resemblance to that of the canonical dsphs than dirrs .
this suggests that the bulk of the dtran galaxies can be considered as `` @xmath101 @xmath102 '' of the dsph galaxies if the recent star formation of dtrans galaxies has been largely suppressed by rapid gas loss ( g03 ) .
our examination of the h i gas content as a function of the tidal index @xmath103 for 79 volume limited ( @xmath12 mpc ) dwarf galaxies reveals a clear morphological segregation , in the sense that gas - deficient dsphs tend to be located in dense areas while most gas - rich dirrs ( except lmc / smc pairs ) are isolated .
similar to the dirr galaxies , most dtran galaxies are also found to be isolated but in general they have lower h i gas fractions as compared to the dirrs .
our analysis supports the idea that the morphology - density relation appears to be a ubiquitous phenomena in cluster environments regardless of their shapes and richness ( van der wel et al .
2010 , and references therein ) , and the environmental effects on galaxy evolution operate in a similar fashion even in such low density environments of the local group or its neighboring groups .
we are grateful to the anonymous referee whose comments and suggestions improved the clarity and quality of this paper .
this work was supported by kasi - carnegie fellowship program jointly managed by korea astronomy and space science institute ( kasi ) and the observatories of the carnegie institution for science .
alcock , c. , et al .
2000 , aj , 119 , 2194 binney , j. , & merrifield , m. 1998 , galactic astronomy ( princeton , nj : princeton university press ) bell , e. f. , slater , c. t. , & martin , n. f. 2011 , apjl , 742 , 15 bellazzini , m. , ferraro , f. r. , & pancino , e. 2001 , apj , 556 , 635 bellazzini , m. , ferraro , f. r. , sollima , a. , pancino , e. , & origlia , l. 2004 , a&a , 424 , 199 belokurov , v. et al .
2006 , apjl , 647 , 111 bernard , e. j. et al .
2008 , apj , 678l , 21 bernard , e. j. et al .
2009 , apj , 699 , 1742 besla , g. , kallivayalil , n. , hernquist , l. , robertson , b. , cox , t. j. , van der marel , r. p. , & alcock , c. 2007 , apj , 668 , 949 bouchard , a. , da costa , g. s. , jerjen , h , & ott , j. 2005 , in iauc .
198 , near - field cosmology with dwarf elliptical galaxies ( les diablerets , switzerland ) , 255 cacciari , c. , corwin , t. m. , & carney , b. w. 2005 , aj , 129 , 267 castellani , v. & quarta , m. l. 1987 , a&as , 71 , 1 chaboyer , b. 1999 , in heck a. , caputo f. , eds , astrophysics and space science library vol .
237 , post - hipparcos cosmi c candles .
kluwer , dordrecht , p.111 clement , c. m. , & rowe , j. 2000 , aj , 120 , 2579 da costa , g. s. , rejkuba , m. , jerjen , h. , & grebel , e. k. 2010 , apjl , 708 , 121 davis , m. , & geller , m. j. 1976 , apj , 208 , 13 de vaucouleurs , g. , de vaucouleurs , a. , corwin , h. g. , jr . , buta , r. j. , paturel , g. , & fouqu , p. 1991
, third refe rence catalogue of bright galaxies .
volume i : explanations and references .
volume ii : data for galaxies between @xmath120 and @xmath121 .
volume iii : data for galaxies between @xmath121 and @xmath122 .
( new york , ny ( usa ) : springer ) , isbn 0 - 387 - 97552 - 7 dolphin , a. e. 2000 , pasp , 112 , 1383 dolphin , a. e. 2002 , mnras , 332 , 91 dolphin , a. e. et al .
2004 , aj , 127 , 875 dotter , a. , chaboyer , b. , jevremovi ? ,
d. , kostov , v. , baron , e. , & ferguson , j. w. 2008 , apjs , 178 , 89 dressler , a. 1980 , apj , 236 , 351 drory , n. , bender , r. , feulner , g. , hopp , u. , maraston , c. , snigula , j. , & hill , g. j. 2004 , apj , 608 , 742 erb , d. k. , shapley , a. e. , pettini , m. , steidel , c. c. , reddy , n. a. , & adelberger , k. l. 2006 , apj , 644 , 813 gallart , c. et al .
2004 , apj , 127 , 1486 gibson , b. k. , & matteucci , f. 1997 , mnras , 291 , 8 grebel , e. k. , gallagher , j. s. , iii , & harbeck , d. 2003 , aj , 125 , 1926 guldenschuh , k. a. , et al . 2005 , pasp , 117 , 721 harris , g. l. h. , & harris , w. e. 2000 , aj , 120 , 2423 hernquist , l. , & quinn , p. j. 1988 , apj , 331 , 682 henrquist , l. , & quinn , p. j. 1989 , apj , 342 , 1 irwin , m. j. et al . 2007 , apjl , 656 , 13 irwin , m. j. , ferguson , a. m. n. , huxor , a. p. , tanvir , n. r. , ibata , r. a. , & lewis , g. f. 2008 , apjl , 676 , 17 jeffery , e. j. et al .
2011 , aj , 141 , 171 jerjen , h. , freeman , k. c. , & binggeli , b. 1998 , aj , 116 , 2873 kallivayalil , n. , van der marel , r. p. , alcock , c. , axelrod , t. , cook , k. h. , drake , a. j. , & geha , m. 2006 , apj , 6 38 , 772 karachentsev , d. , et al .
2003 , a&a , 404 , 93 karachentsev , i. d. , karachentseva , v. e. , huchtmeier , w. k. , & makarov , d. i. 2004 , aj , 127 , 2031 karachentsev , i. d. , makarov , d. i. , & kaisina , e. i. 2013 , aj , 145 , 101 kirby , e. n. , simon , j. d. , geha , m. , guhathakurta , p. , & frebel , a. 2008 , apjl , 685 , 43 kobulnicky , h. a. , & koo , d. c. 2000 , apj , 545 , 712 kobulnicky , h. a. et al .
2003 , apj , 599 , 1006 kobulnicky , h. a. , & kewley , l. j. 2004 , apj , 617 , 240 larson , r. b. 1972 , nature , 236 , 21 layden , a. c. 1998 , aj , 115 , 193 lee , j .- w . , & carney , b. w. 1999 , aj , 118 , 1373 lequeux , j. , peimbert , m. , rayo , j. f. , serrano , a. , & torres - peimbert , s. 1979 , a&a , 80 , 155 lianou , s. , grebel , e. k , da costa , g. s. , rejkuba , m. , jerjen , h. , & koch , a. 2013 , a&a , 550 , a7 ( l13 ) madore , b. f. , & freedman , w. l. 1995 , aj , 109 , 1645 maier , c. , meisenheimer , k. , & hippelein , h. 2004 , a&a , 418 , 475 mackey , a. d. , & gilmore , g. f. 2003 , mnras , 343 , 747 martin , n. f. et al .
2009 , apj , 705 , 758 mateo , m. l. 1998 , ara&a , 36 , 435 matteucci , f. 2012 , chemical evolution of galaxies ( heidelberg : springer ) mcconnachie , a. w. et al .
2008 , apj , 688 , 1009 norris , j. e. et al .
2010 , apj , 723 , 1632 oemler , a. j. 1974 , apj , 194 , 10 pagel , b. e. j. , & patchett , b. e. 1975 , mnras , 172 , 13 pagel , b. e. j. 1997 , nucleosynthesis and chemical evolution of galaxies ( cambridge , uk : cambridge university press ) , isbn 0521550610 postman , m. & geller , m. j. 1984 , apj , 281 , 95 pritzl , b. j. , armandroff , t. e. , jacoby , g. h. , & da costa , g. s. 2005 , 129 , 2232 rejkuba , m. , harris , w. e. , greggio , l. , & harris , g. l. h. 2011 , a&a , 526 , 123 richardson , j. c. et al .
2011 , apj , 732 , 76 sarajedini , a. , & jablonka , p. 2005 , aj , 130 , 1627 sarajedini , a. , barker , m. k. , geisler , d. , harding , p. , & schommer ,
r. 2006 , aj , 132 , 1361 sarajedini , a. , yang , s - c . , monachesi , a. , lauer , t. r. , & trager , s. c. 2012 , mnras , 425 , 1459 salaris , m. , chieffi , a. , & straniero , o. 1993 , apj , 414 , 580 salzer , j. j. , lee , j. c. , melbourne , j. , hinz , j. l. , alonso - herrero , a. , & jangren , a. 2005 , apj , 624 , 661 savaglio , s. et al .
2005 , apj , 635 , 260 schmidt , m. 1963 , apj , 137 , 758 schneider , r. , ferrara , a. , natarajan , p. , & omukai , k. 2002 , apj , 571 , 30 searle , l. , & sargent , w. l. w. 1972 , apj , 173 , 25 schlegel , d. j. , finkbeiner , d. p. , & davis , m. 1998 , apj , 500 , 525 shapley , a. e. et al .
2005 , apj , 626 , 698 sirianni , m. , jee , m. j. , bentez , n. , & et al .
2005 , pasp , 117 , 1049 skillman , e. d. , kennicutt , r. c. , & hodge , p. w. 1989 , apj , 347 , 875 slater , c. t. , bell , e. f. , & martin , n. f. 2011 , apjl , 742 , 14 tamman , g. a. , sandage , a. , & reindl , b. 2003 , a&a , 404 , 423 timmes , f. x. , woosley , s. e. , & weaver , thomas a. 1995 , apjs , 98 , 617 tolstoy , e. , hill , v. , & tosi , m. 2009 , ara&a , 47 , 371 truran , j. w. , & cameron , a. g. w. 1971 , ap&ss , 14 , 179 van den bergh , s. 1962 , aj , 67 , 486 van der wel , a. , bell , e. f. , holden , b. p. , skibba , r. a. , rix , h. 2010 , apj , 714 , 1779 walsh , s. m. , jerjen , h. , & willman , b. 2007 , apjl , 662 , 83 weisz , d. r. et al .
2011 , apj , 739 , 5 ( w11 ) wetzel , a. r. , tinker , j. l. , & conroy , c. 2012 , mnras , 424 , 232 wheeler , j. c. , sneden , c. , & truran , j. w. , jr .
1989 , ara&a , 27 , 279 willman , b. et al . 2005a , aj , 129 , 2692 willman , b. et al .
2005b , apjl , 626 , 85 willman , b. et al .
2011 , aj , 142 , 128 white , s. d. m. , & rees , m. j. 1978 , mnras , 183 , 341 white , s. d. m. , frenk , c. s. 1991 , apj , 379 , 52 wyse , r. f. g. , & gilmore , g. 1993 , in asp conf ser .
48 , the globular clusters - galaxy connection , ed .
g. h. smith , & j. p. brodie ( santa cruz , san francisco , ca : asp ) , 727 yang , s .- c . , sarajedini , a. , holtzman , j. a. , & garnett , d. r. 2010 , apj , 724 , 799 ( y10 ) yang , s .- c . , &
sarajedini , a. 2012 , mnras , 419 , 1362 ( ys12 ) zaritsky , d. , kennicutt , r. c. , jr . , & huchra , j. p. 1994 , 1994 , apj , 420 , 87 zinn , r. & west , m. 1984 , apjs , 55 , 45 zorotovic , m. et al .
2010 , aj , 139 , 357 zucker , d. b. et al .
2004 , apjl , 612 , 121 zucker , d. b. et al .
2006a , apjl , 643 , 103 zucker , d. b. et al .
2006b , apjl , 650 , 41 zucker , d. b. et al .
2007 , apjl , 659 , 21 lrrcccccc rrab & wfc1 field & & & & & & & + 31519 & 0:26:27.47 & -41:51:54.93 & & 27.086 & 0.548 & 0.5200 & 0.8116 & + 31435 & 0:26:28.37 & -41:51:54.95 & & 26.934 & 0.438 & 0.6760 & 1.0592 & + 26874 & 0:26:28.56 & -41:52:12.85 & & 26.759 & 0.394 & 0.6471 & 0.9721 & + 29750 & 0:26:29.03 & -41:52:24.77 & & 27.011 & 0.657 & 0.4951 & 1.0102 & + 32272 & 0:26:29.42 & -41:50:51.19 & & 27.193 & 0.633 & 0.5932 & 1.0251 & + : & : & : & & : & : & : & : & + + rrc & wfc1 field & & & & & & & + 32536 & 0:26:27.76 & -41:52:09.32 & & 26.940 & 0.265 & 0.3004 & 0.6436 & + 28014 & 0:26:28.65 & -41:51:03.54 & & 26.792 & 0.546 & 0.3060 & 0.4139 & + 29510 & 0:26:31.46 & -41:51:35.01 & & 26.812 & 0.318 & 0.3914 & 0.5809 & + 32861 & 0:26:31.94 & -41:52:58.44 & & 27.044 & 0.451 & 0.3699 & 0.4827 & + 27866 & 0:26:33.32 & -41:51:34.42 & & 26.729 & 0.434 & 0.3183 & 0.7197 & bl + : & : & : & & : & : & : & : & + + rrab & wfc2 field & & & & & & & + 23778 & 0:26:28.85 & -41:49:57.87 & & 26.870 & 0.600 & 0.6373 & 0.4490 & + 23713 & 0:26:30.13 & -41:50:17.34 & & 26.823 & 0.484 & 0.5887 & 0.6919 & + 25347 & 0:26:31.04 & -41:50:02.74 & & 27.107 & 0.685 & 0.6058 & 1.0376 & + 24215 & 0:26:31.34 & -41:49:55.30 & & 26.926 & 0.666 & 0.6529 & 0.7934 & + 24335 & 0:26:31.93 & -41:50:41.47 & & 27.007 & 0.429 & 0.5483 & 1.0382 & bl + : & : & : & & : & : & : & : & + + rrc & wfc2 field & & & & & & & + 23174 & 0:26:33.36 & -41:50:29.40 & & 26.832 & 0.568 & 0.3446 & 0.5975 & + 23720 & 0:26:34.14 & -41:50:11.63 & & 26.898 & 0.643 & 0.3781 & 0.4427 & + 25657 & 0:26:38.24 & -41:49:48.61 & & 27.108 & 0.605 & 0.4025 & 0.5852 & + lrrcccccc rrab & wfc1 field & & & & & & & + 27360 & 0:15:24.34 & -32:10:42.23 & & 26.983 & 0.518 & 0.5624 & 0.9455 & + 25183 & 0:15:25.04 & -32:10:40.32 & & 26.830 & 0.636 & 0.6718 & 0.6186 & + 25200 & 0:15:25.27 & -32:11:25.31 & & 26.824 & 0.462 & 0.6851 & 0.7559 & + 24100 & 0:15:25.34 & -32:11:25.56 & & 26.797 & 0.412 & 0.5933 & 1.1051 & + 27060 & 0:15:25.54 & -32:11:31.52 & & 26.926 & 0.412 & 0.5707 & 0.6277 & + : & : & : & & : & : & : & : & + + rrc & wfc1 field & & & & & & & + 24150 & 0:15:25.57 & -32:11:24.13 & & 26.684 & 0.218 & 0.3242 & 0.5591 & + 31623 & 0:15:27.96 & -32:11:25.57 & & 27.143 & 0.394 & 0.2895 & 0.4600 & bl + 24390 & 0:15:28.44 & -32:10:58.97 & & 26.759 & 0.574 & 0.2668 & 0.4252 & + 26141 & 0:15:28.54 & -32:11:46.15 & & 26.779 & 0.340 & 0.3460 & 0.7136 & + 23520 & 0:15:28.66 & -32:11:17.69 & & 26.791 & 0.690 & 0.4048 & 0.6748 & + : & : & : & & : & : & : & : & + + rrab & wfc2 field & & & & & & & + 18627 & 0:15:29.79 & -32:09:45.39 & & 26.987 & 0.622 & 0.5446 & 0.8530 & + 15681 & 0:15:30.19 & -32:09:19.01 & & 26.725 & 0.433 & 0.6162 & 0.9297 & + 18442 & 0:15:30.40 & -32:10:05.50 & & 26.794 & 0.478 & 0.6037 & 1.2543 & + 16370 & 0:15:30.52 & -32:08:38.45 & & 26.845 & 0.617 & 0.6592 & 1.1345 & + 20519 & 0:15:30.90 & -32:09:56.61 & & 26.950 & 0.597 & 0.6162 & 1.0357 & + : & : & : & & : & : & : & : & + + rrc & wfc2 field & & & & & & & + 19124 & 0:15:31.63 & -32:10:15.45 & & 26.809 & 0.479 & 0.3146 & 0.6532 & + 16185 & 0:15:31.65 & -32:10:19.04 & & 26.709 & 0.700 & 0.2195 & 0.6250 & + 19968 & 0:15:31.99 & -32:10:28.29 & & 26.958 & 0.479 & 0.3597 & 0.7719 & bl + 19264 & 0:15:33.55 & -32:09:53.06 & & 26.862 & 0.576 & 0.2885 & 0.5935 & + 16866 & 0:15:33.68 & -32:10:26.99 & & 26.783 & 0.667 & 0.3271 & 0.5668 & bl + : & : & : & & : & : & : & : & + lrrcccccc eso294-g010 & wfc1 field & & & & & & & + 3817 & 0:26:32.12 & -41:51:49.45 & & 25.377 & 0.543 & 1.8274 & 1.0387 & bl + 14561 & 0:26:32.44 & -41:52:24.09 & & 26.038 & 0.393 & 0.4650 & 0.7667 & + 23638 & 0:26:32.62 & -41:52:22.60 & & 26.482 & 0.282 & 0.4418 & 0.5510 & + 21370 & 0:26:33.31 & -41:51:35.21 & & 26.521 & 0.561 & 0.5339 & 0.4820 & bl + 1660 & 0:26:33.52 & -41:51:34.13 & & 24.784 & 0.501 & 0.8735 & 0.3924 & bl + 5048 & 0:26:33.61 & -41:51:47.51 & & 25.492 & 0.483 & 1.4140 & 0.6971 & bl + 18994 & 0:26:33.68 & -41:51:20.84 & & 26.278 & 0.250 & 0.4283 & 0.5306 & bl + 19596 & 0:26:33.79 & -41:51:22.39 & & 26.378 & 0.480 & 0.5473 & 0.2845 & bl + 20461 & 0:26:33.80 & -41:51:42.26 & & 26.302 & 0.417 & 0.5473 & 0.7387 & bl + 1226 & 0:26:33.94 & -41:51:13.44 & & 24.503 & 0.511 & 1.9431 & 0.5900 & bl + 5366 & 0:26:34.24 & -41:51:19.52 & & 25.588 & 0.593 & 1.3464 & 0.7185 & bl + 15933 & 0:26:34.36 & -41:51:51.62 & & 26.204 & 0.416 & 0.5698 & 0.5743 & bl + eso294-g010 & wfc2 field & & & & & & & + 2754 & 0:26:32.62 & -41:50:19.13 & & 25.287 & 0.649 & 1.5201 & 0.6490 & + 2412 & 0:26:33.07 & -41:50:47.32 & & 24.933 & 0.390 & 1.1971 & 0.5833 & bl + 3249 & 0:26:33.21 & -41:50:55.11 & & 25.395 & 0.594 & 0.4683 & 0.2520 & bl + 10896 & 0:26:33.31 & -41:50:53.40 & & 26.136 & 0.482 & 0.7133 & 0.8989 & bl + 2432 & 0:26:33.69 & -41:50:59.90 & & 24.967 & 0.473 & 1.2844 & 1.1950 & bl + 14201 & 0:26:34.19 & -41:51:00.50 & & 26.209 & 0.485 & 0.6302 & 0.3131 & bl + 3942 & 0:26:34.74 & -41:50:10.73 & & 25.642 & 0.631 & 1.2321 & 0.7236 & + + eso410-g005 & wfc1 field & & & & & & & + 6202 & 0:15:28.56 & -32:11:31.48 & & 25.707 & 0.540 & 0.9203 & 0.6795 & + 1756 & 0:15:29.54 & -32:11:03.79 & & 24.746 & 0.547 & 1.5094 & 1.2808 & bl + 2067 & 0:15:29.97 & -32:11:11.94 & & 24.902 & 0.523 & 0.9324 & 0.4800 & bl + 1409 & 0:15:30.29 & -32:11:13.20 & & 24.417 & 0.390 & 1.1129 & 0.4249 & bl + 2387 & 0:15:30.50 & -32:10:52.73 & & 25.025 & 0.523 & 0.7383 & 0.5900 & bl + 973 & 0:15:30.91 & -32:10:55.99 & & 24.268 & 0.419 & 1.7204 & 1.1464 & bl + 1959 & 0:15:31.10 & -32:11:03.09 & & 24.887 & 0.652 & 1.7143 & 0.5899 & bl + 2278 & 0:15:31.49 & -32:10:53.14 & & 25.039 & 0.538 & 1.2008 & 0.6843 & bl + 1147 & 0:15:31.85 & -32:10:52.33 & & 24.207 & 0.441 & 1.1759 & 0.6793 & bl + 2803 & 0:15:32.05 & -32:10:51.68 & & 25.159 & 0.630 & 1.2684 & 0.9685 & bl + 9037 & 0:15:32.71 & -32:11:22.99 & & 25.987 & 0.606 & 0.5413 & 0.5250 & + 1553 & 0:15:32.76 & -32:10:57.75 & & 24.590 & 0.447 & 1.0492 & 0.3652 & bl + eso410-g005 & wfc2 field & & & & & & & + 1413 & 0:15:32.33 & -32:10:31.04 & & 25.022 & 0.418 & 0.8278 & 0.4415 & bl + 1023 & 0:15:32.51 & -32:10:18.59 & & 24.694 & 0.412 & 0.7997 & 0.5748 & bl + 1105 & 0:15:32.78 & -32:10:34.68 & & 24.840 & 0.570 & 1.0211 & 0.5039 & bl + 1546 & 0:15:33.46 & -32:10:51.05 & & 25.130 & 0.477 & 1.0981 & 0.4714 & bl + 1923 & 0:15:33.61 & -32:10:42.03 & & 25.261 & 0.484 & 0.7039 & 0.3242 & bl + 6160 & 0:15:33.63 & -32:09:17.19 & & 26.044 & 0.486 & 0.6571 & 0.4327 & bl + 879 & 0:15:33.99 & -32:10:22.82 & & 24.603 & 0.476 & 1.0377 & 0.8358 & + llllrccllrc mw group & & & & & & & & & & + smc & 00 52 38.0 & -72 48 01 & 0.06 & -16.5 & -1.20 @xmath8 0.40 & 1 & 8.65 & 8.85 & 3.6 & dirr + sculptor & 01 00 09.4 & -33 42 33 & 0.09 & -9.8 & -1.50 @xmath8 0.50 & 1 & 5.39 & 6.86 & 2.8 & dsph + phoenix & 01 51 06.3 & -44 26 41 & 0.44 & -9.6 & -1.90 @xmath8 0.40 & 1 & 5.22 & 6.08 & 0.8 & dtran + segue 2 & 02 19 16.0 & + 20 10 31 & 0.03 & -2.3 & - & - & - & 3.88 & 3.8 & ufd + fornax & 02 39 54.7 & -34 31 33 & 0.14 & -11.5 & -1.20 @xmath8 0.50 & 1 & 5.19 & 7.55 & 2.2 & dsph + lmc & 05 23 34.6 & -69 45 22 & 0.05 & -17.9 & -0.60 @xmath8 0.50 & 1 & 8.66 & 9.42 & 3.6 & dirr + carina & 06 41 36.7 & -50 57 58 & 0.10 & -9.0 & -1.80 @xmath8 0.30 & 1 & 2.32 & 6.54 & 2.8 & dsph + uma ii & 08 51 30.0 & + 63 07 48 & 0.03 & -3.1 & -2.44 @xmath8 0.53 & 2 & 2.80 & 4.18 & 3.9 & ufd + leot & 09 34 53.4 & + 17 03 05 & 0.42 & -6.7 & -2.02 @xmath8 0.54 & 4 & 5.63 & 4.94 & 0.8 & dirr + sexb & 10 00 00.1 & + 05 19 56 & 1.36 & -14.0 & -2.10 @xmath8 0.40 & 1 & 7.66 & 7.78 & -0.5 & dirr + segue 1 & 10 07 03.2 & + 16 04 25 & 0.02 & -0.7 & -2.70 @xmath8 0.70 & 2 & 1.12 & 3.24 & 4.3 & ufd + leoi & 10 08 26.9 & + 12 18 29 & 0.25 & -11.0 & -1.40 @xmath8 0.50 & 1 & 3.18 & 7.34 & 1.4 & dsph + sexa & 10 11 00.8 & -04 41 34 & 1.32 & -13.9 & -1.90 @xmath8 0.40 & 1 & 7.82 & 7.49 & -0.4 & dirr + sextans & 10 13 03.0 & -01 36 52 & 0.09 & -8.7 & -1.90 @xmath8 0.40 & 1 & 2.30 & 6.42 & 2.7 & dsph + uma i & 10 34 52.8 & + 51 55 12 & 0.10 & -4.8 & -2.29 @xmath8 0.50 & 2 & 3.80 & 4.87 & 2.6 & ufd + willman1 & 10 49 21.0 & + 51 03 00 & 0.04 & -1.9 & -2.10 & 5 & 2.96 & 3.72 & 3.7 & ufd + leoii & 11 13 29.2 & + 22 09 17 & 0.21 & -9.1 & -1.60 @xmath8 0.50 & 1 & 4.02 & 6.58 & 1.7 & dsph + leov & 11 31 09.6 & + 02 13 12 & 0.18 & -3.8 & - & - & 2.88 & 4.46 & 1.9 & ufd + leoiv & 11 32 57.0 & -00 32 00 & 0.16 & -4.2 & -2.58 @xmath8 0.72 & 2 & 2.78 & 4.63 & 2.0 & ufd + comai & 12 26 59.0 & + 23 54 15 & 0.04 & -3.2 & -2.53 @xmath8 0.40 & 2 & 1.63 & 4.24 & 3.7 & ufd + cvnii & 12 57 10.0 & + 34 19 15 & 0.16 & -4.1 & -2.19 @xmath8 0.54 & 2 & 4.10 & 4.59 & 2.0 & ufd + cvni & 13 28 03.5 & + 33 33 21 & 0.22 & -7.9 & -2.08 @xmath8 0.41 & 2 & 4.50 & 6.10 & 1.6 & dsph + bootesiii & 13 57 07.4 & + 26 46 30 & 0.05 & -5.9 & - & - & - & 5.29 & 3.6 & ufd + bootesii & 13 58 00.0 & + 12 50 00 & 0.04 & -1.9 & - & - & 1.62 & 3.70 & 3.8 & ufd + bootesi & 14 00 00.0 & + 14 30 00 & 0.07 & -5.5 & -2.55 @xmath8 0.37 & 2 & 2.01 & 5.15 & 3.2 & ufd + umin & 15 09 11.3 & + 67 12 52 & 0.06 & -7.1 & -1.90 @xmath8 0.70 & 2 & 4.52 & 5.80 & 3.2 & dsph + hercules & 16 31 02.0 & + 12 47 30 & 0.15 & -6.1 & -2.58 @xmath8 0.47 & 2 & 2.72 & 5.39 & 2.2 & ufd + draco & 17 20 01.4 & + 57 54 34 & 0.08 & -8.7 & -2.00 @xmath8 0.70 & 1 & 2.19 & 6.45 & 2.9 & dsph + sag dsph & 18 55 03.1 & -30 28 42 & 0.02 & -12.7 & -0.50 @xmath8 0.80 & 1 & 1.99 & 8.02 & 5.3 & dsph + sag dirr & 19 29 59.0 & -17 40 41 & 1.04 & -11.5 & -2.30 @xmath8 0.40 & 1 & 6.94 & 6.51 & -0.2 & dirr + ngc6822 & 19 44 57.7 & -14 48 11 & 0.50 & -15.2 & -1.20 @xmath8 0.40 & 1 & 8.15 & 8.34 & 0.6 & dirr + tucana & 22 41 49.0 & -64 25 12 & 0.88 & -9.2 & -1.70 @xmath8 0.20 & 1 & 4.18 & 6.62 & 0.0 & dtran + piscesii & 22 58 31.0 & + 05 57 09 & 0.18 & -4.4 & - & - & - & 4.70 & 1.9 & ufd + + m31 group & & & & & & & & & & + wlm & 00 01 58.1 & -15 27 40 & 0.97 & -14.1 & -1.40 @xmath8 0.40 & 1 & 7.83 & 7.69 & 0.2 & dirr + and xviii & 00 02 14.5 & + 45 05 20 & 1.36 & -9.1 & - & - & - & 6.60 & 0.5 & dsph + and xx & 00 07 30.7 & + 35 07 56 & 0.80 & -5.8 & - & - & - & 5.26 & 2.4 & dsph + and xix & 00 19 32.1 & + 35 02 37 & 0.93 & -8.3 & - & - & - & 6.28 & 1.9 & dsph + ic10 & 00 20 24.5 & + 59 17 30 & 0.66 & -16.0 & -1.30 @xmath8 0.40 & 1 & 8.00 & 8.47 & 1.6 & dirr + and xxvi & 00 23 45.6 & + 47 54 58 & 0.76 & -6.5 & - & - & - & 5.54 & 2.9 & dsph + cetus & 00 26 11.0 & -11 02 40 & 0.78 & -10.2 & -1.70 @xmath8 0.20 & 1 & 4.18 & 7.02 & 0.5 & dtran + and xxv & 00 30 08.9 & + 46 51 07 & 0.81 & -9.1 & - & - & - & 6.58 & 2.9 & dsph + ngc147 & 00 33 11.6 & + 48 30 28 & 0.76 & -14.8 & -1.10 @xmath8 0.40 & 1 & 3.74 & 8.21 & 2.8 & dsph + and iii & 00 35 33.8 & + 36 29 52 & 0.75 & -9.3 & -1.70 @xmath8 0.20 & 1 & 5.55 & 6.66 & 3.2 & dsph + and xxx & 00 36 34.9 & + 49 38 48 & 0.68 & - & - & - & - & - & 2.4 & - + and xvii & 00 37 07.0 & + 44 19 20 & 0.74 & -7.0 & - & - & 5.54 & 5.74 & 3.6 & dsph + and xxvii & 00 37 27.1 & + 45 23 13 & 0.83 & -7.3 & - & - & - & 5.88 & 3.1 & dsph + ngc185 & 00 38 58.0 & + 48 20 10 & 0.61 & -14.7 & -0.80 @xmath8 0.40 & 1 & 5.02 & 8.29 & 2.0 & de + ngc205 & 00 40 22.5 & + 41 41 11 & 0.82 & -16.1 & -0.50 @xmath8 0.50 & 1 & 5.60 & 8.92 & 3.6 & de + m32 & 00 42 42.1 & + 40 51 59 & 0.49 & -14.8 & -1.10 @xmath8 0.60 & 1 & 6.00 & 8.65 & 1.5 & de + and i & 00 45 40.0 & + 38 02 14 & 0.73 & -10.7 & -1.40 @xmath8 0.20 & 1 & 5.53 & 7.21 & 3.4 & dsph + and xi & 00 46 20.0 & + 33 48 05 & 0.73 & -6.2 & -1.75 @xmath8 0.18 & 3 & 5.53 & 5.42 & 2.7 & dsph + and xii & 00 47 27.0 & + 34 22 29 & 0.83 & -6.4 & - & - & 5.64 & 5.51 & 2.6 & dsph + and xiv & 00 51 35.0 & + 29 41 49 & 0.73 & -7.7 & - & - & 5.53 & 6.03 & 2.2 & dsph + and xiii & 00 51 51.0 & + 33 00 16 & 0.84 & -6.8 & -1.74 @xmath8 0.18 & 3 & 5.65 & 5.66 & 2.4 & dsph + and ix & 00 52 52.8 & + 43 12 00 & 0.79 & -8.1 & - & - & 5.60 & 6.20 & 4.0 & dsph + and xvi & 00 59 29.8 & + 32 22 36 & 0.52 & -8.2 & - & - & 5.24 & 6.23 & 1.6 & dsph + lgs 3 & 01 03 55.0 & + 21 53 06 & 0.65 & -9.3 & -1.70 @xmath8 0.30 & 1 & 5.02 & 5.96 & 1.5 & dtran + ic1613 & 01 04 47.8 & + 02 08 00 & 0.73 & -14.5 & -1.40 @xmath8 0.30 & 1 & 7.77 & 8.07 & 0.8 & dirr + and x & 01 06 33.7 & + 44 48 16 & 0.63 & -7.9 & - & - & 5.40 & 6.10 & 2.2 & dsph + and v & 01 10 17.1 & + 47 37 41 & 0.81 & -9.2 & -1.90 @xmath8 0.10 & 1 & 5.62 & 6.62 & 2.6 & dsph + and xv & 01 14 18.7 & + 38 07 03 & 0.76 & -8.7 & - & - & 5.56 & 6.43 & 2.9 & dsph + and ii & 01 16 29.8 & + 33 25 09 & 0.65 & -9.2 & -1.50 @xmath8 0.30 & 1 & 5.44 & 6.65 & 2.1 & dsph + and xxiv & 01 18 30.0 & + 46 21 58 & 0.60 & -7.0 & - & - & - & 5.77 & 1.9 & dsph + and xxii & 01 27 40.0 & + 28 05 25 & 0.79 & -6.0 & - & - & - & 5.36 & 2.3 & dsph + and xxiii & 01 29 21.8 & + 38 43 08 & 0.73 & -9.5 & - & - & - & 6.75 & 2.5 & dsph + leoa & 09 59 26.4 & + 30 44 47 & 0.81 & -11.7 & -2.10 @xmath8 0.40 & 1 & 7.04 & 6.93 & 0.1 & dirr + kkr25 & 16 13 47.6 & + 54 22 16 & 1.86 & -9.4 & -2.10 @xmath8 0.40 & 1 & 4.91 & 6.71 & -0.6 & dtran + ddo210 & 20 46 51.8 & -12 50 53 & 0.94 & -11.1 & -1.90 @xmath8 0.30 & 1 & 6.42 & 6.75 & 0.0 & dtran + ic5152 & 22 02 41.9 & -51 17 43 & 1.97 & -15.6 & -1.40 @xmath8 0.50 & 1 & 8.02 & 8.72 & -0.7 & bcd + and xxviii & 22 32 41.2 & + 31 12 58 & 0.65 & -7.7 & - & - & - & 6.04 & 1.2 & dtran + cas dsph & 23 26 31.8 & + 50 40 32 & 0.79 & -11.7 & -1.50 @xmath8 0.20 & 1 & 5.60 & 7.62 & 1.8 & dsph + pegasus & 23 28 34.1 & + 14 44 48 & 0.76 & -11.5 & -2.00 @xmath8 0.30 & 1 & 6.53 & 7.13 & 1.0 & dirr + peg dsph & 23 51 46.4 & + 24 35 10 & 0.82 & -10.7 & -1.70 @xmath8 0.20 & 1 & 4.21 & 7.22 & 1.5 & dsph + and xxi & 23 54 47.7 & + 42 28 15 & 0.86 & -9.3 & - & - & - & 6.66 & 2.2 & dsph + and xxix & 23 58 55.6 & + 30 45 20 & 0.73 & -7.5 & - & - & - & 5.96 & 2.0 & dsph + + antilia group & & & & & & & & & & + ngc3109 & 10 03 07.2 & -26 09 36 & 1.32 & -15.7 & -1.70 @xmath8 0.40 & 1 & 8.37 & 8.57 & 0.2 & dirr + antlia & 10 04 04.0 & -27 19 55 & 1.32 & -9.8 & -1.90 @xmath8 0.20 & 1 & 5.92 & 6.47 & 2.3 & dtran + + sculptor group & & & & & & & & & & + eso410-g005 & 00 15 31.4 & -32 10 48 & 1.82 & -11.6 & -1.64 @xmath8 0.03 & this study & 5.91 & 6.88 & 0.2 & dtran + eso294-g010 & 00 26 33.3 & -41 51 20 & 1.88 & -10.9 & -1.77 @xmath8 0.03 & this study & 5.48 & 6.25 & 0.5 & dtran + | we present the results of our analysis of the rr lyrae ( rrl ) variable stars detected in two transition - type dwarf galaxies ( dtrans ) , eso294-g010 and eso410-g005 in the sculptor group , which is known to be one of the closest neighboring galaxy groups to our local group . using deep archival images from the advanced camera for surveys ( acs ) onboard the hubble space telescope ( hst ) ,
we have identified a sample of rr lyrae candidates in both dtrans galaxies [ 219 rrab ( rr0 ) and 13 rrc ( rr1 ) variables in eso294-g010 ; 225 rrab and 44 rrc stars in eso410-g005 ] .
the metallicities of the individual rrab stars are calculated via the period - amplitude-[fe / h ] relation derived by alcock et al .
this yields mean metallicities of @xmath0 \rangle_{eso294 } = -1.77 \pm 0.03 $ ] and @xmath0 \rangle_{eso410 } = -1.64 \pm 0.03 $ ] .
the rrl metallicity distribution functions ( mdfs ) are investigated further via simple chemical evolution models ; these reveal the relics of the early chemical enrichment processes for these two dtrans galaxies . in the case of both galaxies ,
the shapes of the rrl mdfs are well - described by pre - enrichment models .
this suggests two possible channels for the early chemical evolution for these sculptor group dtrans galaxies : 1 ) the ancient stellar populations of our target dwarf galaxies might have formed from the star forming gas which was already enriched through `` prompt initial enrichment '' or an `` initial nucleosynthetic spike '' from the very first massive stars , or 2 ) this pre - enrichment state might have been achieved by the end products from more evolved systems of their nearest neighbor , ngc 55 .
we also study the environmental effects of the formation and evolution of our target dtrans galaxies by comparing their properties with those of 79 volume limited ( @xmath12 mpc ) dwarf galaxy samples in terms of the luminosity - metallicity relation and the h i gas content .
the presence of these rrl stars strongly supports the idea that although the sculptor group galaxies have a considerably different environment from the local group ( e.g. no giant host galaxies , loosely bound and very low local density ) , they share a common epoch of early star formation with the dwarf satellite galaxies in the local group . |
in the last two decades , laser technologies have made a giant leap forward such that laser sources of ultrarelativistic intensity level can be attained @xcite .
near - infrared laser beams are available up to intensities of @xmath0 @xmath1 and much stronger lasers will be available in near future @xcite . for such lasers the dimensionless relativistic invariant parameter of intensity @xmath2 ( @xmath3 -elementary charge , @xmath4 - electron mass , @xmath5 , @xmath6 -electric field amplitude and frequency of a laser radiation , @xmath7 -light speed in vacuum ) .
the latter represents the work of the field on the one wavelength in the units of the particle rest energy .
interaction of such lasers with the matter at extreme conditions in ultrashort space - time scales have attracted broad interest over the last few years conditioned by a number of important applications , such as generation and probing of highenergy - density plasma @xcite , ions acceleration and inertial confinement fusion @xcite , vacuum nonlinear optics @xcite , compact laser - plasma accelerators lpa , etc .
generally , the interaction of such fields with the electrons in the presence of a third body makes available the revelation of many nonlinear relativistic electrodynamic phenomena . as a third body
can serve ion and in the superintense laser fields one can observe relativistic above threshold ionization @xcite and high order harmonic generation rhhg , electron - positron pairs production on nuclei @xcite , and multiphoton stimulated bremsstrahlung ( sb ) of electrons on the ions / nuclei @xcite .
the latter is one of the fundamental processes at the interaction of superstrong laser pulses with plasma and under the some circumstances inverse - bremsstrahlung absorption may become dominant mechanism of absorption of strong electromagnetic ( em ) radiation in underdense plasma . with the advent of lasers
many pioneering papers have been devoted to the theoretical investigation of the electron - ion scattering processes in gas or plasma in the presence of a laser field using nonrelativistic csb1,qsb1,csb2,bornnr1,lf1,csb3,bunk , eik1,lf2,qsb2 as well as relativistic @xcite considerations .
the appearance of superpower ultrashort laser pulses of relativistic intensities has initiated new interest in sb in relativistic domain @xcite , where investigations were carried out mainly in the born approximation over the scattering potential .
meanwhile for ions with the large charge and for the clusters @xcite , when electron interaction with the entire dense cluster ion core that composed of a large number of ions is dominant , the born approximation is not applicable .
the theoretical description of sb in superstrong em fields and scattering centers of large charges requires one to go beyond the scope of born approximation over the scattering potential and the perturbation theory over laser field . in this context ,
when quantum effects are considerable , one can apply eikonal @xcite or generalized eikonal approximation @xcite . for the infrared and optical lasers , in the multiphoton interaction regime
, one can apply classical theory and the main approximation in the classical theory is low frequency ( lf ) or impact approximation @xcite .
lf approximation have been generalized for relativistic case in refs .
@xcite , where the effect of an intense em wave on the dynamics of sb and non - linear absorption of intense laser radiation by a monochromatic electron beam due to sb have been carried out
. regarding the absorption of an em of relativistic intensities in plasma due to inverse sb there are not investigations which could help to clarify the behavior of nonlinear absorption of an intense em radiation at very large @xmath8 .
hence , it is of interest to clear up how nonlinear sb effect will proceeds in the plasma , taking also into account initial relativism of plasma electrons . in the present paper inverse - bremsstrahlung absorption of an intense laser radiation in relativistic maxwellian plasma
is considered in the relativistic lf approximation @xcite . in this approximation
one can consider as superstrong laser fields as well as scattering centers with large charges .
the radiation power absorption in such plasma is investigated as for circularly polarized wave ( cpw ) , as well as for linearly polarized wave ( lpw ) .
we consider the dependence of the absorption coefficient on the intensity and polarization of the laser radiation , as well as on the temperature of the plasma electrons .
the organization of the paper is as follows : in sec .
ii the relativistic absorption coefficient of the em wave of arbitrary polarization and intensity due to the mechanism of sb process is presented . in sec .
iii we consider the problem numerically along with derivation of asymptotic formulas for absorption coefficient .
conclusions are given in sec .
the absorption coefficient @xmath9 for an em radiation field of arbitrary intensity and polarization , in general case of the homogeneous ensemble of electrons of concentration @xmath10 , with the arbitrary distribution function @xmath11 over momenta @xmath12 , at the inverse bremsstrahlung on the scattering centers with concentration @xmath13 , can be represented in the form:@xmath14where @xmath15 is the classical energy absorbed by a single electron per unit time from the em wave of intensity @xmath16 due to sb process on the scattering centers . for the homogeneous scattering centers @xmath17 . for the generality , we assume maxwellian plasma with the relativistic distribution function:@xmath18where @xmath19 is the boltzmann s constant , @xmath20 is the temperature of electrons in plasma , @xmath21 is relativistic energy - momentum dispersion law of electrons , @xmath22 is the mcdonald s function ; @xmath23 is normalized as @xmath24 to obtain @xmath15 for the sb process , the electron interaction with the scattering potential and em wave in the lf approximation can be considered as independently proceeding processes @xcite , separated into the following three stages , schematically depicted in fig .
1 . field free electron with energy @xmath25 and momentum @xmath26 interacts with the em wave . the exact solution of the relativistic equation of motion of an electron in a plane em wave is well known @xcite .
the energy and momentum in a plane em wave field can be written as : @xmath27@xmath28@xmath29 , \label{2}\]]@xmath30where @xmath31is the vector potential of the em wave of currier frequency @xmath6 and slowly varying amplitude @xmath32 . here
@xmath33 is the phase , @xmath34 , @xmath35 is an unit vector in the em wave propagation direction , @xmath36 are the unit polarization vectors , and @xmath37 is the polarization angle . at the second stage
the elastic scattering of the electron in the potential field takes place at the arbitrary , but certain phase @xmath38 of the em wave .
thus , taking into account adiabatic turn on of the wave ( @xmath39 ) from eqs .
( [ 1])-([3 ] ) before the scattering one can write @xmath40@xmath41 , \label{b2}\]]@xmath42where@xmath43is the integral of motion for a charged particle in the field of a plane em wave .
the mean energy of an electron in the wave - field before the scattering will be @xmath44 then it takes place elastic scattering on the scattering center @xmath45 . due to the instantaneous interaction of the electron with the scattering potential
the wave phase does not change its value during the scattering .
the electron with initial momentum @xmath46 is acquired the momentum @xmath47 ( @xmath48 ) after the scattering directly at the same phase @xmath38 of the wave , which can be defined from the generalized consideration of the elastic scattering .
thus , measuring the scattering angle @xmath49 from a direction @xmath50 , with corresponding azimuthal angle @xmath51 for the scattered momentum one can write@xmath52 = p(\psi _ { s})\widehat{r}\left [ \begin{array}{c } \sin \vartheta \cos \varphi
\\ \sin \vartheta \sin \varphi \\ \cos \vartheta \end{array}% \right ] , \label{p'}\]]where @xmath53 , @xmath54 and @xmath55 are basic rotation matrices about the @xmath56 and @xmath57 axes : @xmath58 .
\label{rot}\]]as a @xmath59 axis we take wave propagation direction @xmath35 , @xmath60 is the polar angle and @xmath61 is the azimuthal angle in the wave - polarization plane . at the third stage
the electron again interacts only with the wave , moving in the wave field with the momentum and energy defined from eqs .
( [ 1])-([3 ] ) : @xmath62@xmath63@xmath64 , \label{after2}\]]@xmath65where@xmath66the mean energy of an electron in the wave field after the scattering will be@xmath67@xmath68 .
\label{meanfin}\]]the energy change due to sb can be calculated as a difference of mean energy in the field before and after the scattering:@xmath69taking into account eqs .
( [ meanin ] ) and ( [ meanfin ] ) we obtain : @xmath70@xmath71for the energy absorbed by a single electron per unit time from the em wave due to sb process on the scattering centers one can write@xmath72where @xmath73 is the velocity of an electron in the wave - field , @xmath74 is the momentum , and @xmath75 is the differential cross section of the elastic scattering in the potential field @xmath76 . taking into account that the main contribution in the integral ( [ w1 ] ) comes from the small angle scatterings one can write @xmath77where @xmath78is the transport differential cross section . for the coulomb scattering centers with potential energy @xmath79of electron interaction with ion of charge @xmath80 , one can use relativistic cross section for elastic scattering at small angles @xcite and make integration over @xmath49 and @xmath51 to obtain : @xmath81@xmath82 \frac{\mathcal{e}(\psi _ { s})}{p^{3}(\psi _ { s})}l_{\mathrm{cb } } , \label{w3}\]]where @xmath83is the coulomb logarithm .
the latter has been obtained taking@xmath84 as a lower limit of the impact parameter , while for the upper limit we assume@xmath85 .
taking into account eqs .
( [ a1 ] ) , ( [ a ] ) , and ( [ w3 ] ) for the absorption coefficient we obtain:@xmath86@xmath82 \frac{m^{2}c^{2}\mathcal{e}% ( \psi _ { s})}{\lambda ^{3}p^{3}(\psi _ { s})}l_{\mathrm{cb } } , \label{alfa}\]]where@xmath87thus , eq . ( [ alfa ] ) represents nonlinear inverse - bremsstrahlung absorption coefficient @xmath9 for an em radiation field of arbitrary intensity and polarization , for homogeneous ensemble of electrons of concentration @xmath10 , with the arbitrary distribution function @xmath88 over momenta @xmath26**. * * note that the lf approximation in the intense laser field is applicable when @xmath89where @xmath90 is the laser radiation wavelength and @xmath91 is the debye screening length:@xmath92=7.43\times 10^{2}\times \sqrt{\frac{t_{e}[\mathrm{% ev}]}{zn_{e}[\mathrm{cm}^{-3}]}}. \label{5a1}\]]besides , for an underdense plasma one should take into account condition @xmath93 , where @xmath94 is the plasma frequency with `` effective mass '' @xmath95 of the relativistic electron in the em wave @xcite:@xmath96thus for cpw and for large @xmath97 one can write@xmath98there is also limitation on the pulse duration @xmath99 of an em wave .
the sb should be the main mechanism , which is responsible for the absorption of the laser radiation in plasma .
this condition is failed , if the influence of the strong em wave lead to development of an instability .
hence , pulse duration @xmath100 of an em wave has to satisfy the condition @xmath101 , where @xmath102 is the maximal increment of the instability of the plasma in the strong laser field . in general , analytical integration over momentum @xmath26 and scattering phase @xmath38
is impossible , and one should make numerical integration . the latter along with derivation of asymptotic formulas for absorption coefficient @xmath9 will be done in the next section .
as we are interested in superintense laser pulses of relativistic intensities , then it is convenient to represent the absorption coefficient ( [ alfa ] ) in the form of dimensionless quantities:@xmath103@xmath104here @xmath105and @xmath106 is the classical electron radius . in eq .
( [ acn ] ) the dimensionless momentum , energy , and temperature were introduced as follows : @xmath107and the dimensionless relativistic intensity parameters of em wave,@xmath108the scaled relativistic distribution function is @xmath109and coulomb logarithm@xmath110where @xmath111 . in the near - infrared and optical domain of frequencies , at the @xmath112 @xmath113 . taking into account that normalized absorption coefficient ( [ acn ] ) depends on @xmath114 and @xmath6 through logarithm @xmath115 , for the numerical simulations we will not concretize @xmath114 and @xmath116 , assuming @xmath117 .
it is well known that the kinematics of an electron in the field of a strong em wave essentially depends on the polarization of the wave @xcite . in particular , for the particle initially at rest , in the cpw , energy @xmath118 and momentum @xmath119 are constants , since @xmath120 .
meanwhile in the lpw @xmath121 oscillates and as a consequence small values of @xmath122 give the main contribution in eq .
the latter leads to more complicated behavior of the dynamics of sb at the linear polarization of a stimulating strong wave .
besides in the case of cpw , thanks to azimuthal symmetry one can make a step forward in analytical calculation and obtain explicit formula for the absorption coefficient at superstrong laser fields .
thus , taking into account azimuthal symmetry in the case of cpw , one can make integration over phase @xmath38 , which results in@xmath123@xmath124where@xmath125 , \label{b1'}\]]@xmath126at the large @xmath127 , taking into account eqs .
( [ b1 ] ) and ( [ b2 ] ) , from eq .
( [ acc ] ) one can obtain @xmath128where@xmath129the formula ( [ alfa2 ] ) shows the suppression of the sb rate with increase of the wave intensity . ignoring weak logarithmic dependence
, we see that absorption coefficient inversely proportional to laser intensity : @xmath130 . for the large @xmath97 the dependence of the absorption coefficient on temperature comes from @xmath131 in eq .
( [ alfa2 ] ) .
in particular , for initially nonrelativistic plasma @xmath132 in eq .
( [ alfa2 ] ) one can put @xmath133 , which gives@xmath134the relation for the absorption coefficient in the case of lpw is complicated and even for large @xmath8 one can not integrate it analytically . therefore , for the analysis we have performed numerical investigations , making also analytic interpolation . the results of numerical investigations of eq .
( [ acn ] ) are illustrated in figures 2 - 7 both for the cpw and lpw . )
is shown for various wave intensities . ] ( in arbitrary units ) , as a function of the plasma temperature ( in units of an electron rest energy @xmath135 ) and the dimensionless relativistic invariant parameter of the circularly polarized laser beam . ] ) is shown for various wave intensities .
the wave is assumed to be linearly polarized . ] ( in arbitrary units ) , as a function of the plasma temperature ( in units of an electron rest energy @xmath135 ) and the dimensionless relativistic invariant parameter of the linearly polarized laser beam . ] to show the dependence of the inverse - bremsstrahlung absorption rate on the laser radiation intensity in fig .
2 it is shown scaled rate @xmath136 versus relativistic invariant parameter of the wave intensity for various plasma temperatures .
the wave is assumed to be circularly polarized . as is seen from this figure the sb rate
is suppressed with increase of the wave intensity and for the large values of @xmath8 it exhibits a tenuous dependence on the plasma temperature . the behavior is also seen from fig .
3 , where total scaled rate of inverse - bremsstrahlung absorption of cpw in plasma , as a function of the plasma temperature @xmath137 is shown for various wave intensities . here for large values of @xmath8 we have a weak dependence on temperature , which is a result of laser modified relativistic scattering of electrons irrespective of the initial state of electrons .
the absorption coefficient @xmath9 decreases as @xmath138 in accordance with analytical result ( [ alfa2 ] ) . to clarify the range of applicability of the asymptotic formula ( [ analitic1 ] ) in fig
. 4 density plot of the total rate of inverse - bremsstrahlung absorption scaled to asymptotic rate @xmath139 , as a function of the plasma temperature and the relativistic invariant parameter @xmath8 is shown for cpw .
as is seen in the wide range of @xmath140and @xmath8 one can apply asymptotic formula ( [ analitic1 ] ) . in fig .
5 and 6 it is shown the total scaled rate of inverse - bremsstrahlung absorption of lpw in plasma versus the dimensionless relativistic invariant parameter of wave intensity and temperature , respectively . with the interpolation
we have seen that @xmath9 decreases as @xmath141 and exhibits a tenuous dependence on the plasma temperature .
making analogy with the case of cpw , for the large @xmath8 we interpolate @xmath9 by the following formula : @xmath142 as is seen from fig . 7 , in the case of lpw and for the moderate temperatures , with the well enough accuracy one can apply asymptotic rate ( [ analitic2 ] ) .
we have presented a theory of inverse - bremsstrahlung absorption of an intense laser radiation in relativistic maxwellian plasma in the relativistic low - frequency approximation .
the coefficient of nonlinear inverse - bremsstrahlung absorption has been calculated for relativistic maxwellian plasma .
the simple analytical formulae have been obtained for absorption coefficient at asymptotically large values of laser fields both for circularly and linearly polarized radiations .
the obtained results demonstrate that the sb rate is suppressed with the increase of the wave intensity and for large values of @xmath8 absorption coefficient @xmath9 decreases as @xmath138 for circularly and as @xmath141 for the linearly polarized one in contrast to nonrelativistic case where one has a dependence @xmath143 @xcite .
the sb rate is suppressed with increase of the plasma temperature but for the relativistic laser intensities it exhibits a tenuous dependence on the plasma temperature .
this work was supported by scs of ra .
99 yanovsky v , chvykov v , kalinchenko g , rousseau p , planchon t , matsuok t , maksimchuk a , nees j , cheriaux g , mourou g , krushelnick k 2008 _ opt .
express _ * 16 * 2109 ; mourou g a , tajima t 2011 _ science _ * 331 * 41 roth m , cowan t e , key m h , hatchett s p , brown c , fountain w , johnson j , pennington d m , snavely r a , wilks s c , yasuike k , ruhl h , pegoraro f , bulanov s v , campbell e m , perry m d , powell h 2001 _ phys .
lett . _ * 86 * 436 ; naumova n , schlegel t , tikhonchuk v t , labaune c , sokolov i v , mourou g 2009 _ phys .
lett . _ * 102 * 025002 ; daido h , nishiuchi m , pirozhkov a s 2012 _ rep
phys . _ * 75 * 056401 crawford d p , reiss h r 1994 _ phys .
a _ * 50 * 1844 ; avetissian h k , markossian a g , mkrtchian g f 2001 _ phys . rev . a _ * 64 * 053404 ; avetissian h k , markossian a g 2007 _ phys
. rev . a _ * 76 * 053407 ; klaiber m , yakaboylu e , hatsagortsyan k z 2013 _ phys . rev . a _ * 87 * 023418 miloevi d b , hu s , becker w 2000 _ phys
. rev . a _
* 63 * 011403(r ) ; avetissian h k , markossian a g , mkrtchian g f 2011 _ phys . rev . a _ * 84 * 013418 ; avetissian h k , markossian a g , mkrtchian g f 2011 _ phys . lett .
a _ * 375 * 3699 ; kohler m c , pfeifer t , hatsagortsyan k z , keitel c h 2012 _ adv .
phys . _ * 61 * 159 | the nonlinear absorption of laser radiation of relativistic intensities in the underdense plasma by a mechanism of stimulated bremsstrahlung of electrons on the ions / nuclei is investigated in the low frequency approximation .
coefficient of nonlinear inverse - bremsstrahlung absorption is studied for relativistic maxwellian plasma at asymptotically large values of laser fields and high temperatures of electrons . |
first we recall some known facts on the vassiliev spectral sequence and then proceed to explaining of the main idea of the work . [ [ section ] ] let us fix a non - trivial linear map @xmath3 .
we will consider the _ space of long knots _ , _
i. e. _ , of injective smooth non - singular maps @xmath4 , that coincide with the map @xmath5 outside some compact set ( this set is not fixed ) .
the long knots form an open everywhere dense subset in the affine space @xmath6 of all smooth maps @xmath7 with the same behavior at infinity .
the complement @xmath8 of this dense subset is called the _
discriminant space_. it consists of the maps having self - intersections or singularities .
any cohomology class @xmath9 of the knot space can be realized as the linking coefficient with an appropriate chain in @xmath10 of codimension @xmath11 in @xmath6 . following
@xcite we will assume that the space @xmath6 has a very large but finite dimension @xmath12 .
a partial justification of this assumption uses finite dimensional approximations of @xmath13 , see @xcite .
below we indicate by quotes non - rigorous assertions using this assumption and needing a reference to @xcite for such a justification . the main tool of vassiliev s approach to computation of the ( co)homology of the knot space is the _ simplicial resolution _ @xmath14 ( constructed in @xcite ) of the discriminant @xmath10 .
this resolution is also called the _ resolved
discriminant_. the natural projection @xmath15 is a `` homotopy equivalence '' between the `` one - point compactifications '' of the spaces @xmath14 and @xmath10 . by the `` alexander duality '' , the reduced homology groups @xmath16 of these compactifications `` coincide '' ( up to a change of dimension ) with the cohomology groups of the space of knots : @xmath17 where @xmath18 is a commutative ring of coefficients . in the space @xmath14
there is a natural filtration @xmath19 [ con03 ] the spectral sequence ( called _ vassiliev s main spectral sequence _ ) associated with the filtration ( [ eq02 ] ) and computing the `` borel - moore homology groups of the resolution @xmath14 '' stabilizes over @xmath20 in the first term . @xmath21 [ con04 ] ( due to vassiliev ) filtration ( [ eq02 ] ) `` homotopically splits '' , in other words , @xmath22 is `` homotopy equivalent '' to the wedge @xmath23 .
@xmath21 this conjecture would imply the stabilization of our _ main spectral sequence _ in the first term over any commutative ring @xmath18 of coefficients . due to the alexander duality , the filtration ( [ eq02 ] ) induces the filtrations @xmath24 @xmath25 in respectively the cohomology and homology groups of the space of knots . for @xmath26
the filtrations , are finite for any dimension @xmath27 .
the vassiliev spectral sequence in this case computes the graded quotient associated with the above filtrations . in the most intriguing case
@xmath28 almost nothing is clear . for the dimension @xmath29 the filtration ( [ eq05 ] ) does not exhaust the whole cohomology of degree zero .
the knot invariants obtained by this method are called the vassiliev invariants , or invariants of finite type .
one can define them in a more simple and geometrical way , see @xcite .
the dual space to the graded quotient of the space of finite type knot invariants is the _ bialgebra of chord diagrams_. the invariants and the bialgebra in question were intensively studied in the last decade , see @xcite .
the completeness conjecture for the vassiliev knot invariants is the question about the convergence of the filtration ( [ eq06 ] ) to zero for @xmath28 , @xmath29 .
the realization theorem of m. kontsevich @xcite proves that the vassiliev spectral sequence for @xmath28 , @xmath29 also computes the corresponding associated quotient ( for positive dimensions @xmath27 in the case @xmath28 even this is not for sure ) and does stabilize in the first term .
the groups of the associated graded quotient to filtrations ( [ eq05 ] ) , ( [ eq06 ] ) in the case @xmath28 , @xmath30 are some quotient groups of the groups calculated by vassiliev s main spectral sequence .
let us mention that there is a natural way to construct real cohomology classes of the knot spaces by means of configuration space integrals , that generalizes in a non - trivial way the vassiliev knot invariants obtained in three dimensions from the chern - simons perturbation theory , see @xcite
. may be this approach leads to a proof of conjecture [ con03 ] .
to compute the first term of the main spectral sequence , v. a. vassiliev introduced an _ auxiliary filtration _ in the spaces @xmath31 , see @xcite .
the _ auxiliary spectral sequence _ associated to this filtration degenerates in the first term , because its first term ( for any @xmath32 ) is concentrated at only one line .
therefore _ the second term of the auxiliary spectral sequence is isomorphic to the first term of the main spectral sequence_. the term @xmath33 of the auxiliary spectral sequence together with its differential ( of degree zero ) is a direct sum of tensor products of complexes of connected graphs .
the homology groups of the complex of connected graphs with @xmath34 labelled vertices are concentrated in the dimension @xmath35 only , and the only non - trivial group is isomorphic to @xmath36 , see @xcite .
this homology group has a nice description as the quotient by the 3-term relations of the space spanned by trees ( with @xmath34 labelled vertices ) , see @xcite .
[ [ s02 ] ] in fact v. a. vassiliev considered only case of the main and auxiliary spectral sequences , _
i. e. _ , the case corresponding to the homology of the discriminant and ( by the alexander duality ) to the of the knot space @xmath37 .
this was because a convenient description only of the homology of complexes of connected graphs was known .
it was noticed in @xcite that the cohomology of complexes of connected graphs ( with @xmath34 vertices ) admits also a very nice description as the @xmath34-th component of the lie algebras operad .
this isomorphism comes from the following observation .
let us consider the space @xmath38 of injective maps from a finite set @xmath39 of cardinality @xmath34 into @xmath0 , @xmath40 .
this space can be viewed as a finite - dimensional analogue of the knot spaces .
the corresponding discriminant ( consisting of non - injective maps @xmath41 ) has also a simplicial resolution , whose filtration ( analogous to ( [ eq02 ] ) ) does split homotopically , see @xcite .
the superior non - trivial term @xmath42 of the filtration provides exactly the complex of connected graphs with @xmath34 vertices labelled by the elements of @xmath39 .
its only non - trivial homology group is isomorphic by the alexander duality to the cohomology group in the maximal degree of the space @xmath43 . on the other hand
the above space is homotopy equivalent to the @xmath34-th space of the little cubes operads , see section [ s4 ] .
the homology operad of this topological operad is well known , see @xcite ( and section [ s4 ] ) . for different @xmath2 of the same parity this homology operad is the same up to a change of grading and for odd ( resp . even ) @xmath2 it is the poisson ( resp .
gerstenhaber ) algebras operad containing ( in the maximal degree ) the operad of lie algebras .
an analogous periodicity takes place for the spaces of knots .
the degree zero term of the main spectral sequence together with its differential ( of degree zero ) depends up to a change of grading on the parity of @xmath2 only .
obviously the same is true for the whole auxiliary spectral sequence .
the above description of the cohomology of complexes of connected graphs allows one to describe easily the vassiliev spectral sequence in the case .
the main results of these computations are explained in section [ s5 ] ( see theorems [ t54 ] , [ t58 ] , [ t59 ] , [ t510 ] ) .
the proofs in full detail will be given in @xcite , ( see also the russian version @xcite ) .
[ [ section-1 ] ] the paper is organized as follows . in sections
[ s1 ] , [ s2 ] , [ s3 ] we give some preliminaries on linear graded operads .
we give a short definition and examples ( that will be useful for us ) of linear graded operads ( section [ s1 ] ) .
we construct a graded lie algebra structure on the space of any linear graded operad ( section [ s2 ] ) .
we define a hochschild complex for any graded linear operad supplied with a morphism from the associative algebras operad to this operad ( section [ s3 ] ) .
the content of sections [ s2 ] and [ s3 ] was borrowed ( up to a slightly different definition of signs ) from @xcite .
section [ s4 ] is devoted to the may operad of little cubes and to its homology . as it was already mentioned
the m - th component of this operad is a space homotopy equvalent to the space @xmath44 of injective maps @xmath45 . in section [ s4 ] we explain how the stratification in the discriminant set of non - injective maps @xmath46 corresponds to a direct sum decomposition of the homology @xmath47 . in section [ s5 ]
we describe a natural stratification in the discriminant set @xmath10 of long singular `` knots '' .
this stratification provides a direct sum decomposition of the first term of the auxiliary spectral sequence . in the case of even @xmath2 the first term of vassiliev s auxiliary spectral sequence
is completely described by the following theorem : * theorem [ t54 ] .
* _ the first term of vassiliev s homological auxiliary spectral sequence together with its first differential is isomorphic to the normalized hochschild complex of the batalin - vilkovisky algebras operad .
_ @xmath21 unfortunately in the case of odd @xmath2 it is not possible to describe the corresponding complex in terms of the hochschild complex for some graded linear operad .
a description of this complex is given in section [ s8 ] .
nevertheless the homology of the complex in question ( _ i .
e. _ , the first term of the main spectral sequence ) over @xmath48 can be defined in terms of the hochschild homology of the poisson algebras operad in the case of odd @xmath2 ( and of the gerstenhaber algebras operad in the case of even @xmath2 ) . a precise statement is given by theorem [ t510 ] . in section [ s5 ]
we also introduce complexes homologically equivalent ( for any commutative ring @xmath18 of coefficients ) to the first term of the auxiliary spectral sequence .
these complexes simplify a lot the computations of the second term . in section [ s6 ]
we explain how the bialgebra of chord diagrams arises in our construction .
we formulate some problems concerning it .
section [ s7 ] does not contain any new results and serves rather to explain one remark of m. kontsevich .
we study there the homology operads of some topological operads , that we call _ operads of turning balls_. these homology operads make more clear the difference that we have in the cases of odd and even @xmath2 . in section [ s8 ]
we describe the first term of the auxiliary spectral sequence together with the degree @xmath49 differential both for @xmath2 even and odd .
the corresponding complex is called _ complex of bracket star - diagrams_. in section [ s9 ] we construct a differential bialgebra structure on this complex .
we conjecture that this differential bialgebra structure is compatible with the homology bialgebra structure of the space of long knots .
i would like to thank my scientific advisor v. vassiliev for his consultations and for his support during the work .
also i would like to express my gratitude to ecole normale suprieur and institut des hautes etudes scientifiques for hospitality .
i would like to thank m. kontsevich for his attention to this work .
also i am grateful to p. cartier , a. v. chernavsky , m. deza , d. panov , m. finkelberg , s. loktev , i. marin , g. racinet , a. stoyanovsky .
the definition of many algebraic structures on a linear space ( such as the commutative , associative , lie algebra structures ) consists of setting several polylinear operations ( in these three cases , only one binary operation ) , that should satisfy some composition identities ( in our example , associativity or jacoby identity ) . instead of doing this one can consider the spaces of all polylinear @xmath50-ary operations , for all @xmath51 , and the composition rules , that arize from the corresponding algebraic structure .
the natural formalization of this object is given by the notion of operad .
let @xmath18 be a commutative ring of coefficients .
a _ graded @xmath18-linear operad _
@xmath52 is a collection @xmath53 of graded @xmath18-vector spaces equipped with the following set of data : \(i ) an action of the symmtric group @xmath54 on @xmath55 for each @xmath56 .
\(ii ) linear maps ( called _ compositions _ ) , preserving the grading , @xmath57 for all @xmath58 .
we write @xmath59 instead of @xmath60 .
\(iii ) an element @xmath61 , called the _ unit _ , such that @xmath62 for any non - negative @xmath5 and any @xmath63 .
it is required that these data satisfy some conditions of associativity and equivariance with respect to the symmetric group actions , see ( * ? ? ?
* chapter 1 ) , @xcite , @xcite .
one can consider any element @xmath64 as something that has @xmath5 inputs and 1 output : the composition operation ( [ eq11 ] ) of @xmath64 with @xmath65 , @xmath66 , @xmath67 is the substitution of @xmath68 , @xmath66 , @xmath69 into @xmath5 inputs of @xmath70 : ( 50,10 ) ( -130,85)@xmath71 the resulting element has @xmath72 inputs and 1 output , _
i. e. _ , it belongs to @xmath73 .
we will give several well known examples of graded linear operads , see @xcite .
* a ) * let @xmath74 be a graded @xmath18-vector space .
we define the _ endomorphism operad _ @xmath75 .
the unit element @xmath76 is put to be the identical map @xmath77 .
the composition operations and the symmetric group actions are defined in the canonical way .
this operad is very important due to the following definition .
[ d14 ] let @xmath52 be a graded linear operad . by an _
@xmath52-algebra _ ( or _ algebra over @xmath52 _ )
we call any couple @xmath78 , where @xmath74 is a graded vector space and @xmath79 is a morphism @xmath80 of operads
. @xmath21 in other words , the theory of algebras over an operad @xmath52 is the representation theory of @xmath52 . *
b ) * the _ operad @xmath81 of lie algebras_. the component @xmath82 of this operad is trivial . the @xmath50-th component @xmath83 is defined as the subspace of the free lie algebra with generators @xmath84 , that is spanned by the brackets containing each generator exactly once .
[ ex15 ] for @xmath85 one can take the bracket @xmath86,[x_1,x_2]],x_4]$ ] as an element of @xmath87 .
@xmath21 since we work in the category of graded vector spaces , we need to define a grading on each considered space .
the grading of the spaces @xmath83 , @xmath88 , is put to be zero .
it is well known that these spaces are free @xmath18-modules , @xmath89 .
the @xmath54-action is defined by permutations of @xmath90 .
let @xmath91 , @xmath92 , @xmath66 , @xmath93 be brackets respectively from @xmath94 , @xmath95 , @xmath66 , @xmath96 .
we define the composition operations ( [ eq11 ] ) as follows .
@xmath97 @xmath98 the element @xmath99 is the unit element for this operad . note that a @xmath81-algebra structure in the sense of definition [ d14 ] is exactly the same as a ( graded ) lie algebra structure in the usual sense .
indeed , the element @xmath100)\in hom(v^{\otimes 2},v)$ ] always defines a lie bracket .
the converse is also true .
it is easy to see that the element @xmath101\in{{\mathcal { lie}}}(2)$ ] generates the operad @xmath81 , so if we put @xmath100)$ ] equal to our lie bracket , then we immediately obtain a map @xmath102 of the whole operad @xmath81 . *
c ) * by analogy with the operad @xmath81 one defines the _ operad _ @xmath103 ( resp .
@xmath104 ) _ of commutative _ ( resp .
_ associative _ ) _ algebras_. the space @xmath105 ( resp .
@xmath106 ) , for @xmath88 , is put to be one - dimensionl ( resp .
@xmath107-dimensional ) free @xmath18-module defined as the subspace of the free commutative ( resp .
associative ) algebra with generators @xmath90 , that is spanned by the monomials containing each generator exactly once .
the @xmath54-actions and the composition operations are defined in the same way as for the operad @xmath81 .
the element @xmath108 ( resp .
@xmath109 ) is the unit element @xmath110 .
there are two different ways to define the space @xmath111 ( resp .
@xmath112 ) . we can put this space to be trivial or one - dimensional . in the first case
we get the operad of commutative ( resp .
associative ) algebras without unit , in the second case the operad of commutative ( resp .
associative ) algebras with unit .
below we will consider the second situation .
[ r17 ] a commutative ( resp .
associative ) algebra structure is exactly the same as a structure of @xmath103-algebra ( resp .
@xmath104-algebra ) in the sense of definition [ d14 ] . @xmath21 * d ) * the _ operads _ @xmath113 , @xmath114 , @xmath115 _ of poisson , gerstenhaber _ and _ @xmath2-poisson algebras_. first of all , let us recall the definition of poisson , gerstenhaber and @xmath2-poisson algebras .
[ d18 ] a graded commutative algebra @xmath116 is called a _ @xmath2-poisson algebra _ , if it has a lie bracket @xmath117:a\otimes a\to a\ ] ] of degree @xmath118 .
the bracket is supposed to be compatible with the multiplication .
this means that for any elements @xmath119 @xmath120=[x , y]z+(-1)^{\tilde{y}(\tilde{x}-d)}y[x , z].~\square \eqno({{\refstepcounter{theorem}}\arabic{section}.\arabic{theorem}})\label{eq19}\ ] ] 0-poisson ( resp .
1-poisson ) algebras are called simply _ poisson _ ( resp .
_ gerstenhaber _ ) algebras .
1-poisson algebras are called gerstenhaber algebras in honor of murray gerstenhaber , who discovered this structure on the hochschild cohomology of associative algebras , see @xcite and also sections [ s2 ] and [ s3 ] .
[ ex110 ] let @xmath121 be a graded lie algebra with the bracket of degree @xmath118 .
then the symmetric ( in the graded sense ) algebra @xmath122 has a natural structure of a @xmath2-poisson algebra with the usual multiplication of a symmetric algebra and with the bracket defined by the following formula : @xmath123=\sum_{i , j}(-1)^ \epsilon a_1\dots\widehat{a_i}\dots a_k\cdot [ a_i , b_j]\cdot b_1\dots \widehat{b_j}\dots b_l , \eqno({{\refstepcounter{theorem}}\arabic{section}.\arabic{theorem}})\label{eq111}\ ] ] where @xmath124 @xmath21 [ r112 ] for a graded lie algebra @xmath121 with the bracket of degree 0 the @xmath2-tuple suspension @xmath125 $ ] is also a graded lie algebra with the bracket of degree @xmath118 .
thus , the space @xmath126)$ ] is a @xmath2-poisson algebra .
@xmath21 note , that for commutative , associative , or lie algebras the @xmath50-th component of the operad is defined as the space of all natural polylinear @xmath50-ary operations , that come from the corresponding algebra structure .
now , let us describe the spaces @xmath127 of all natural polylinear @xmath50-ary operations of @xmath2-poisson algebras . consider a free graded lie algebra @xmath128 with the bracket of degree @xmath118 and with the generators @xmath129 of degree zero , and consider ( following example [ ex110 ] ) the @xmath2-poisson algebra @xmath130 .
this is a free @xmath2-poisson algebra .
we will define the space @xmath127 as the subspace of @xmath131 spanned by the products ( of brackets ) containing each generator @xmath132 exactly once .
for instance for @xmath85 we will take the product @xmath133\cdot[[x_2,x_5]x_4]$ ] as an element of @xmath134 .
the unit element @xmath110 is @xmath135 .
the symmetric group actions and the composition operations are defined analogously to the case of the operad @xmath81 .
the space @xmath127 can be decomposed into a direct sum with the summands numbered by partitions of the set @xmath136 : @xmath137 for a partition @xmath138 of the set @xmath139 we define the space @xmath140 to be linearly spanned by products of @xmath141 brackets , such that the @xmath32-th bracket contains generators only from the set @xmath142 ( thus , each generator from @xmath142 is presented exactly once in the @xmath32-th bracket ) .
let @xmath143 be of cardinalities @xmath144 respectively , then @xmath145 this implies that the space @xmath127 is isomorphic to @xmath146 , and its poincar polynomial is @xmath147 * e ) * the operad @xmath148 ( resp .
@xmath149 , @xmath2 being odd ) of batalin - vilkovisky ( resp .
@xmath2-batalin - vilkovisky ) algebras .
[ d115 ] a gerstenhaber algebra ( resp .
@xmath2-poisson algebra , for odd @xmath2 ) @xmath116 is called a _ batalin - vilkovisky algebra _ ( resp .
_ @xmath2-batalin - vilkovisky algebra _ ) , if @xmath116 is supplied with a linear map @xmath150 of degree -1 ( resp .
@xmath118 ) @xmath151 such that \(i ) @xmath152 , \(ii ) @xmath153.$ ] @xmath21 note that ( i ) and ( ii ) imply \(iii ) @xmath154)=[\delta(a),b]+(-1)^{\tilde a + 1}[a,\delta(b)].$ ] [ ex116 ] let @xmath121 be a graded lie algebra with the bracket of degree zero .
then the exterior algebra @xmath155)$ ] is a batalin - vilkovisky algebra , where the structure of a gerstenhaber algebra is from remark [ r112 ] ; the operator @xmath150 is the standard differential on the chain - complex @xmath156 : @xmath157\wedge a_1\dots \widehat{a_i}\dots\widehat{a_j}\dots\wedge a_k , \eqno({{\refstepcounter{theorem}}\arabic{section}.\arabic{theorem}})\label{eq117}\ ] ] where @xmath158 , @xmath159 . in the same way
@xmath121 defines the @xmath2-batalin - vilkovisky algebra @xmath126)$ ] for any odd @xmath2 .
@xmath21 let us describe the @xmath50-th component of the corresponding operad , denoted by @xmath148 and @xmath149 ( @xmath2 being always odd ) .
since the spaces @xmath160 and @xmath161 are isomorphic ( in the super - sense ) up to a change of grading , we will consider now only the case @xmath162 .
obviously , the space @xmath160 of all natural polylinear @xmath50-ary operations for such algebras contains @xmath163 .
consider the symmetric algebra of the free graded lie algebra @xmath164 with the bracket of degree -1 and with the generators @xmath90 of degree zero and the generators @xmath165 of degree -1 .
this space has a structure of a batalin - vilkovisky algebra .
( in fact it is a free @xmath148-algebra with generators @xmath166 ) .
in @xmath167 we will take the subspace @xmath160 linearly spanned by all the products ( of brackets ) , containing each index @xmath168 exactly once .
for instance @xmath169\cdot[x_2,[\delta(x_4 ) , \delta(x_5)]]$ ] belongs to @xmath170 . due to relations ( i ) ,
( ii ) , ( iii ) for @xmath150 , this subspace is exactly the space of all natural polylinear @xmath50-ary operations on batalin - vilkovisly algebras .
analogously to ( [ eq113 ] ) the space @xmath160 can be decomposed into the direct sum : @xmath171 where @xmath116 is a partition of the set @xmath136 , @xmath172 is a subset of @xmath136 corresponding to indices @xmath32 presented by @xmath173 in products of brackets of @xmath160 .
note that the space @xmath160 is isomorphic to @xmath174 .
the poincar polynomial of this graded space is @xmath175 . in the sequel
we will use the following definition .
[ d119 ] for a finite set @xmath39 any pair @xmath176 consisting of a partition @xmath116 of @xmath39 and of a subset @xmath177 will be called a _ star - partition _ of the set @xmath39 .
( each point @xmath178 will be called a _
star_. ) @xmath21 note that we have the following natural morphisms of operads : @xmath179 the last arrow of is defined only if @xmath2 is odd .
in this section we define a graded lie algebra structure on an arbitrary graded linear operad . in the next section we use this structure and define the hochschild complex for a graded linear operad equipped with a morphism from the operad @xmath104 to our operad .
both these constructions , which generalize the hochschild cochain complex for associative algebras , were introduced in @xcite .
the only difference between the operations , , given below and those of @xcite is in signs .
first of all , in the paper @xcite m. gerstenhaber and a. voronov considered only linear ( non - graded ) operads , hence our case is more general . but even for the case of purely even gradings the signs are slightly different
. the difference can be easily obtained by conjugation of the operations , , by means of the linear operator that maps any element @xmath180 , @xmath51 , to @xmath181 .
let @xmath182 be a graded @xmath18-linear operad . by abuse of the language the space @xmath183 will be also denoted by @xmath52 .
a tilde over an element will always designate its grading .
for any element @xmath180 we put @xmath184 .
the numbers @xmath50 and 1 here correspond to @xmath50 inputs and to 1 output respectively .
define a new grading @xmath185 on the space @xmath52 . for an element @xmath180
we put @xmath186 .
it turns out that @xmath52 is a graded lie algebra with respect to the grading @xmath185 .
note that the composition operations ( [ eq11 ] ) respect this grading .
define the following collection of multilinear operations on the space @xmath52 .
@xmath187 for @xmath188 , where the summation runs over all possible substitutions of @xmath90 into @xmath70 in the prescribed order , @xmath189 , @xmath190 being the total number of inputs in @xmath70 going after @xmath191 .
for instance , for @xmath192 and arbitrary @xmath193 @xmath194 we will also adopt the following convention : @xmath195 one can check immediately the following identities : @xmath196 @xmath197 @xmath198 where @xmath199 .
( these signs are the same as in @xcite ) .
define a bilinear operation ( respecting the grading @xmath200 ) @xmath201 on the space @xmath52 : @xmath202 for @xmath203 .
[ d22 ] a graded vector space @xmath116 with a bilinear operation @xmath204 is called a _ pre - lie algebra _ , if for any @xmath119 the following holds : @xmath205 any graded pre - lie algebra @xmath116 can be considered as a graded lie algebra with the bracket @xmath206:=x\circ y-(-1)^{|x||y|}y\circ x. \eqno({{\refstepcounter{theorem}}\arabic{section}.\arabic{theorem}}\label{bracket})\ ] ] the description of the operad of pre - lie algebras is given in @xcite .
the following lemma is a corollary of the identity ( [ eq22 ] ) applied to the case @xmath207 .
[ l23 ] the operation defines a graded pre - lie algebra structure on the space @xmath52 .
@xmath21 in particular this lemma means that any graded linear operad @xmath52 can be considered as a graded lie algebra with the bracket .
let @xmath208 be a graded linear operad equipped with a morphism @xmath209 from the operad @xmath104 .
this morphism defines the element @xmath210 , where the element @xmath211 is the operation of multiplication .
note that the elements @xmath212 , @xmath34 are odd with respect to the new grading @xmath185 ( @xmath213 ) and @xmath214=[\pi(m_2),\pi(m_2)]=2\pi(m_2\circ m_2)=0 $ ] .
thus @xmath52 becomes a differential graded lie algebra with the differential @xmath215 : @xmath216=m\circ x-(-1)^{|x|}x\circ m , \eqno({{\refstepcounter{theorem}}\arabic{section}.\arabic{theorem}})\label{eq31}\ ] ] for @xmath217 .
the complex @xmath218 is called the _ hochschild complex _ for the operad @xmath52 .
[ ex31 ] if @xmath52 is the endomorphism operad @xmath219 of a vector space @xmath116 , and we have a morphism @xmath220 that defines an associative algebra structure on @xmath116 , then the corresponding complex @xmath221 is the standard hochschild cochain complex of the associative algebra @xmath116 .
@xmath21 [ ex31 ] due to the morphisms we have the hochschild complexes @xmath222 , @xmath223 , @xmath224 , @xmath225 .
it can be shown that the complexes @xmath222 and @xmath223 are acyclic .
@xmath21 define another grading @xmath226 on the space @xmath52 .
with respect to this grading the bracket @xmath227 $ ] is homogeneous of degree @xmath228 .
it is easy to see that the product @xmath27 , defined as follows @xmath229 for @xmath203 , together with the differential @xmath215 defines a differential graded associative algebra structure on @xmath116 with respect to the grading @xmath230 [ t34 ] @xcite the multiplication @xmath27 and the bracket @xmath227 $ ] induce a gerstenhaber ( or what is the same 1-poisson ) algebra structure on the homology of the hochschild complex @xmath231 .
@xmath21 * proof of theorem [ t34 ] : * the proof is deduced from the following homotopy formulas . @xmath232
the above formula proves the graded commutativity of the multiplication @xmath27 .
@xmath233-[x , y]*z-(-1)^{(deg(x)-1)deg(y)}y*[x , z]=$ ] @xmath234 this formula proves the compatibility of the bracket with the multiplication .
analogously to linear operads one can define topological operads , _ i. e. _ , collections @xmath235 of topological sets with \(i ) an @xmath54-action on each @xmath55 ; \(ii ) compositions @xmath236 \(iii ) a unit element @xmath61 .
we assume the same associativity and symmetric group equivariance requirements . evidently , the homology spaces @xmath237 over any field @xmath18 form a graded @xmath18-linear operad .
the same is true for any commutative ring of coefficients , if the groups @xmath238 , @xmath51 , have no torsion . usually , when one considers the homology of a topological operad , one inverses the grading , supposing that the @xmath32-th homology group @xmath239 has the degree @xmath240 .
historically one the first examples of topological operads are the little cubes operads @xmath241 , @xmath242 , see @xcite , ( * ? ? ?
* chapter 2 ) , ( * ? ? ?
* chapter 4 ) . here
@xmath243 denotes the configuration space of @xmath50 disjoint cubes labelled by @xmath136 in a unit cube .
it is supposed that the faces of the cubes are parallel to the corresponding faces of the unit cube .
the group @xmath54 acts by permutations of the cubes in configurations .
the element @xmath244 is the configuration of one cube coinciding with the unit cube .
the composition operations ( ii ) are insertions of @xmath5 configurations respectively of @xmath245 cubes into the corresponding @xmath5 cubes of a configuration of @xmath5 cubes .
[ t41 ] @xcite the homology groups of the operad @xmath246 , @xmath242 , have no torsion and form the following graded linear operad \1 ) if @xmath162 , @xmath247 is isomorpic to the associative algebras operad @xmath104 , \2 ) if @xmath40 , @xmath248 is isomorphic to the @xmath249-posson algebras operad @xmath250 .
@xmath21 [ r42 ] the operad of associative algebras has a natural filtration compatible with the operad structure .
the graded quotient associated with this filtration is the poisson algebras operad @xmath113 .
@xmath21 we will not prove theorem [ t41 ] , however we will make some explanations .
first of all , note that the space @xmath251 is contractible for any @xmath242 .
the homology classe of one point in this space corresponds to @xmath252 .
the space @xmath253 is homotopy equivalent to the @xmath249-dimensional sphere @xmath254 . for @xmath40
this gives us one operation of degree zero and one operation of degree @xmath255 .
the first operation corresponds to the multiplication @xmath256 , the second one to the bracket @xmath101\in{{\mathcal { poiss}}}_{(d-1)}(2)$ ] .
obviously , the space @xmath243 is homotopy equivalent to the configuration space of collections of @xmath50 distinct points in @xmath257 , _
i. e. _ , to the space of injective maps @xmath258 .
the latter space is an open everywhere dense subset of the vector space @xmath259 of all maps @xmath260 the complement ( called the _ discriminant _ ) @xmath261 is a union of @xmath262 vector subspaces of codimension @xmath2 .
each of these subspaces @xmath263 corresponds to a pair of distinct points @xmath264 and consists of maps @xmath265 , such that @xmath266 .
the non - complete partitions of the set @xmath136 ( by the _
complete _ partition we mean the partition into @xmath50 singletons ) are in one - to - one correspondence with the strata of the arrangement @xmath267 . to a partition
we assign a vector subspace consisting of maps @xmath268 that glue the points of each set in the partition . following the general theory of arrangements the reduced homology groups @xmath269 can be decomposed into a direct sum
, each summand being assigned to some stratum of the arrangement , see @xcite . in the case
@xmath40 this decomposition is canonical ( in the case @xmath162 it depends on the choice of a component of @xmath270 let us assign to the complete partition the degree zero homology group @xmath271 .
[ p43 ] the above decomposition of the homology groups @xmath272 , @xmath40 , coincides with the decomposition ( [ eq113 ] ) .
in the same way as the homology groups of the space of injective maps @xmath258 are decomposed into a direct sum by strata of the discriminant @xmath273 ( see section [ s4 ] proposition [ p43 ] ) , the first term of the vassiliev auxiliary spectral sequence is naturally decomposed into a direct sum in which the summands are numbered by equivalence classes of so called _
@xmath274-configurations _ defined below .
let @xmath116 be a non - ordered finite collection of natural numbers @xmath275 , any of which is not less than 2 , and let @xmath276 be a non - negative integer .
set @xmath277 .
@xmath274-configuration _ is a collection of @xmath278 distinct points in @xmath279 separated into @xmath141 groups of cardinalities @xmath144 , plus a collection of @xmath276 distinct points in @xmath279 ( some of which can coincide with the above @xmath278 points ) . for short @xmath280-configuration
are called simply _ @xmath116-configurations_. a map @xmath281 _ respects _ an @xmath274-configuration , if it glues together all points inside any of its groups of cardinalities @xmath282 , and the derivative @xmath283 is equal to 0 at all the @xmath276 last points of this configuration . for any @xmath274-configuration
the set of maps respecting it is an affine subspace in @xmath284 of codimension @xmath285 ; the number @xmath286 is called the _ complexity _ of the configuration .
two @xmath274-configurations are called _ equivalent _ if they can be transformed into one another by an orientation - preserving homeomorphism @xmath287 .
consider any @xmath274-configuration @xmath288 of complexity @xmath32 and with @xmath289 geometrically distinct points in @xmath279 .
the stratum consisting of all mappings @xmath290 respecting at least one @xmath274-configuration @xmath291 equivalent to @xmath288 , can be parameterized by the space @xmath292 of affine fiber bundle , whose base space is the space @xmath293 of @xmath274-configurations @xmath291 equivalent to @xmath288 , and the fiber over @xmath291 is the affine space @xmath294 of maps respecting @xmath291 .
note that @xmath293 is contractible being an open cell of dimension @xmath289 .
therefore this fiber bundle can be trivialized : @xmath295 [ r52 ] the corresponding stratum is not homeomorphic to @xmath292 : one map @xmath290 may respect two different @xmath274-configurations equivalent to @xmath288 , and therefore the stratum has self - intersections .
@xmath21 the auxiliary spectral sequence computing the homology groups of the term @xmath296 in our filtration ( [ eq02 ] ) uses those and only those @xmath274-configurations , that are of complexity @xmath32 .
the geometrical meaning of the first differential in the auxiliary spectral sequence is in how the strata ( or rather the base spaces of the corresponding affine bundles ( [ eq51 ] ) ) coresponding to equivalence classes of @xmath274-configurations of complexity @xmath32 bound to each other .
let first @xmath2 be even .
for any equivalence class of @xmath274-configurations of complexity @xmath32 and with @xmath289 geometrically distinct points we will assign a star - partition @xmath297 of the set @xmath298 ( see definition [ d119 ] ) and therefore a subspace @xmath299 of the operad @xmath300 ( see decomposition [ eq118 ] ) , where the partition @xmath301 of the set @xmath298 and the subset @xmath302 are defined below .
[ d53 ] a _ minimal component _ of an @xmath274-configuration is either one of its @xmath276 points , which does not coincide with none of the @xmath278 points , or one of the @xmath141 groups of points with all the stars contained there .
@xmath21 for instance , the @xmath274-configuration on the figure [ f53 ] has 3 minimal components consisting respectively of the following groups of points : 1 ) @xmath303 , @xmath304 ; 2 ) @xmath305 , @xmath306 , @xmath307;3 ) @xmath308 .
( 10,10 ) ( 172,20)@xmath27 ( 243,20)@xmath27 to any @xmath274-configuration with @xmath289 geometrically distinct points we can assign a star - partition of the set @xmath298 .
let @xmath309 be the points in @xmath279 of our @xmath274-configuration ( @xmath276 of them are marked by stars ) .
the set of minimal components of this @xmath274-configuration defines a partition @xmath301 of the set @xmath298 .
we also have the subset @xmath302 ( of cardinality @xmath276 ) of indices , corresponding to the points marked by stars .
note that different equivalence classes of @xmath274-configurations correspond to different star - partitions .
but the correspondence is far from being bijective .
the star - partitions not corresponding to @xmath274-configurations are those and only those , which contain singletons not marked by a star .
[ t54 ] see @xcite .
the first term of the vassiliev auxiliary spectral sequence for even @xmath2 together with its first differenrial is isomorphic to the subcomplex of the hochschild complex @xmath310 , linearly spanned by the summands of the decomposition ( [ eq118 ] ) , corresponding to star - partitions , which do nt contain singletons not marked by a star .
the grading corresponding to the homology of the knot space @xmath311 is minus the grading `` @xmath312 '' defined by ( [ eq33 ] ) .
@xmath21 in particular , the theorem claims that the subspace in @xmath300 spanned by the summands in question is invariant with respect to the differential @xmath215 .
consider the decomposition of the space @xmath313 , where @xmath314 is the sum over all star - partitions having exactly @xmath32 singletons not marked by a star .
the filtration @xmath315 , with @xmath316 , is compatible with the differential @xmath215 .
note , that the complex @xmath317 is exactly the complex from theorem [ t54 ] .
[ p55 ] @xcite the hochschild complex @xmath318 ( @xmath2 is even ) is a direct sum of the complexes @xmath319 and @xmath320 .
the first complex @xmath317 is homology ( and even homotopy ) equivalent to @xmath318 ; the second one @xmath320 is acyclic ( and even contractible ) .
@xmath21 an analogous statement holds for the complexes @xmath321 , @xmath2 being any integer number .
the subcomplexes spanned by the summands of the decomosition ( [ eq113 ] ) corresponding to partitions non - containing singletons will be called the _ normalized hochschild complexes _ and denoted by @xmath322 .
normalized hochschild complexes _ @xmath317 for the operads @xmath300 , @xmath2 being even , will be denoted also by @xmath323 .
[ cor56 ] the first term of the main spectral sequence ( for even @xmath2 and for any commutative ring @xmath18 of coefficients ) is isomorphic to the hochschild homology ( with inversed grading ) of the batalin - vilkovisky operad @xmath300 .
@xmath21 [ r57 ] ( due to m. kontsevich ) the operator @xmath324 has a natural geometrical interpretation as the euler class @xmath325 of the special orthogonal group @xmath326 , see section [ s7 ] .
@xmath21 if @xmath2 is odd , then the first term of the auxiliary spectral sequence is very similar to the normalized hochschild complex @xmath327 , but it does not correspond to any operad .
a description of the obtained complex , which i called the _ complex of bracket star - diagrams _ , see in section [ s8 ] . in @xcite the following theorem is proved . for
this is an immediate corollary of theorem [ t54 ] .
[ t58 ] for any @xmath1 the subspace of the first term of the auxiliary spectral sequence linearly spanned only by the summands corresponding to @xmath116-configurations forms a subcomplex isomorphic to the normalized hochschild complex @xmath328 with inversed grading .
@xmath21 there is a very nice way to simplify the computations of the second term of the auxiliary spectral sequence .
this construction works both for @xmath2 even and odd .
consider the normalized hochschild complex @xmath322 and consider the quotient of the space of this complex by the `` neighboring commutativity relations '' in other words , for any @xmath329 we set @xmath132 and @xmath330 to commute . for example
the element @xmath133\cdot[x_2,[x_4,x_5]]\in { { \mathcal { poiss}}}_{(d-1)}(5)$ ] is equal to zero modulo these relations , because it contains @xmath331 $ ] .
the space of relations is invariant with respect to the differential , thus the quotient space has the structure of a quotient complex .
denote this quotient complex by @xmath332 .
[ t59 ] @xcite for any commutative ring @xmath18 of coefficients the space @xmath333 is a free @xmath18-module . the homology space of the complex @xmath334 is isomorphic to the first term of the main spectral sequence ( = to the second term of the auxiliary spectral sequence ) .
@xmath21 * idea of the proof : * consider the filtration in the first term ( of the auxiliary spectral sequence ) by the number of minimal components of corresponding @xmath274-configurations .
the associated spectral sequence degenerates in the second term , because its first term is concentrated on the only line corresponding to @xmath116-configurations .
@xmath21 before describing the relation of the hochschild homology space of @xmath250 defined over @xmath48 with the first term of the main spectral sequence also defined over @xmath48 ( see theorem [ t510 ] ) we will give some explanations . in section [ s8 ] a structure of a differential graded cocommutative bialgebra on the first term of the auxiliary spectral sequence is defined . according to theorems [ t54 ] and [ t58 ] such a structure is defined also on the normalized hochschild complexes @xmath335 , @xmath336 , @xmath337 . the neighboring commutativity relations respect multiplication and comultiplication ( but not the bracket ) , so the complexes @xmath332 are differential bialgebras .
a motivation of the existence of such a structure is that the space of long knots is an @xmath338-space ( has a homotopy associative multiplication ) , see @xcite ; therefore its homology and cohomology spaces over any field @xmath18 are mutually dual graded respectively cocommutative and commutative bialgebras . for @xmath339 these ( co)homology bialgebras are bicommutative , see @xcite , @xcite .
this fact supports conjecture [ con03 ] , because applying theorem [ t34 ] to the operads of poisson , gerstenhaber or batalin - vilkovisky algebras over any field @xmath18 , we obtain that their hochschild homology bialgebras are bicommutative .
it follows from the milnor theorem , see @xcite , that for @xmath339 these bialgebras are polynomial .
the space of primitive elements is the space of their generators .
in fact , the bracket [ bracket ] for these operads preserves the spaces of primitive elements .
we conclude that as gerstenhaber algebras the hochschild homology spaces of these operads are symmetric algebras of the lie algebras of primitive elements ( see example [ ex110 ] with @xmath162 ) .
[ t510 ] @xcite
as a graded bialgebra the first term of the main spectral sequence over @xmath48 is isomorphic to the hochschild homology bialgebra ( with the inversed grading ) of the @xmath48-linear operad @xmath115 factorized \1 ) for even @xmath2 : by one odd primitive generator @xmath101 $ ] ( of degree @xmath340 ) ;
\2 ) for odd @xmath2 : by one even primitive generator @xmath101 $ ] ( of degree @xmath340 ) and one odd primitive generator @xmath341,x_2]$ ] ( of degree @xmath342 ) . @xmath21
the _ bialgebra of chord diagrams _ , the dual to the associated quotient bialgebra of vassiliev knot invariants , was intensively studied in the last decade . in this section
we give an interpretation of the bialgebra of chord diagrams as a part of the hochschild homology algebra of the poisson algebras operad .
consider the normalized hochschild complex @xmath322 . in this complex one
can define a bigrading by the complexity @xmath32 and by the number @xmath343 of geometrically distinct points of the corresponding @xmath116-configurations .
the differential @xmath215 is of bidegree ( 0,1 ) .
if the first grading @xmath32 is fixed , then the number @xmath289 varies from @xmath11 to @xmath344 .
so any element of the bigrading @xmath345 belongs to the kernel of @xmath215 .
the case @xmath346 corresponds to the minimal possible dimension of non - trivial homology classes of the space of long knots for the complexity @xmath32 fixed .
for example , if @xmath28 , then this dimension is equal to @xmath347 .
the part of the hochschild homology groups , that lies in the bigradings @xmath345 , @xmath348 , will be called the _ bialgebra of chord diagrams _ if @xmath2 is odd , and the _ bialgebra of chord superdiagrams _ if @xmath2 is even .
any product of brackets in @xmath349 of bidegree @xmath345 is the product of @xmath32 brackets , each of wich contains exactly 2 generators .
thus , any such product of brackets can be depicted as @xmath344 points on the line @xmath279 , that are decomposed into @xmath32 pairs and connected by a chord inside each pair .
for example , @xmath350\cdot[x_4,x_1]\cdot [ x_2,x_6]$ ] is assigned to the diagram ( 10,10 ) the so called _
4-term relations _ arise as the differential @xmath215 of products of brackets , in which all brackets except one are `` chords '' , and the only non - chord is a bracket on three elements .
in other words , these products of brackets correspond to @xmath116-configurations , with @xmath351 .
[ r62 ] sometimes one considers the bialgebra of chord diagrams factorized not only by 4-term relations , but also by 1-term relations .
the latter relations arise ( when we consider the whole first term of the auxiliary spectral sequence ) as the differential of diagrams corresponding to @xmath274-configurations , with @xmath352 , @xmath353 , and the only star does not coincide with none of @xmath278 points .
@xmath21 in the case of odd @xmath2 we need to take into account orientations of chords ( since @xmath354=-[x_{i_2},x_{i_1}]$ ] )
. this definition does not coincide with the standard one , where these orientations are not important .
this discordance of definitions can be easily eliminated .
consider the hochschild complex @xmath355 , and replace each space @xmath356 by its tensor product with the one - dimensional sign representation @xmath357 of the symmetric group @xmath54 .
this can be done in the following way .
we take a free @xmath249-poisson algebra @xmath358 with generators @xmath359 of degree one instead of the analogous algebra @xmath360 with generators @xmath90 of degree zero .
afterwards we consider the subspace @xmath361 spanned by products of brackets containing each generator @xmath362 exactly once .
obviously , the @xmath54-module @xmath363 is isomorphic to @xmath364 . defining properly the differential , see section [ s8 ] , we obtain another version of the hochschild complex for the operads @xmath250 .
this new version is interpreted in the geometry of the discriminant @xmath10 as introducing an orientation of the spaces @xmath293 of the strata ( [ eq51 ] ) not according to the usual order @xmath365 of the points on the line @xmath279 , but according to the order , that was in our product of brackets .
for instance , for the element @xmath366\cdot [ x_5,[x_2,x_1]]$ ] the orientation of the corresponding space @xmath367 of @xmath368-configurations equivalent to configuration is according to the order @xmath369 .
( 10,10 ) an advantage of the new hochschild complexes ( for operads @xmath250 ) is a simpler rule of signs in the definition of the differential , see section [ s8 ] .
let @xmath52 be a graded linear operad , equipped with a morphism from the operad @xmath104 , then any @xmath52-algebra @xmath116 is an associative algebra because of the following morphisms : @xmath370 on the other hand , the map @xmath371 defines a morphism of hochschild complexes : @xmath372 therefore the classes in the hochschild homology of @xmath52 can be considered as characteristic classes of the hochschild cohomology of @xmath52-algebras ( considered as associative algebras ) .
an interesting question is whether all the classes in the hochschild homology of @xmath148 , @xmath113 , @xmath114 , @xmath300 , @xmath250 have a non - trivial realization as characteristic classes .
consider example [ ex110 ] , where we take @xmath373)$ ] as a @xmath249-poisson algebra .
the hochschild homology space of a polynomial algebra is well known .
it is the space of polynomial polyvector fields on the space of generators . in our case
the space of generators is @xmath374 $ ] .
according to the grading rule , we get that the hochschild homology space of @xmath250 of bigrading @xmath375 is mapped to the space of homogenous degree @xmath32 @xmath289-polyvector fields , _
i. e. _ , of expressions of the form @xmath376 if @xmath2 is odd ( resp .
even ) , the tensor @xmath377 is symmetric ( resp .
antisymmetric ) with respect to the indices @xmath378 and antisymmetric ( resp .
symmetric ) with respect to the indices @xmath379 .
* problem .. [ p65 ] * find explicitly @xmath377 via the bracket on @xmath121 , for example , in the case @xmath346 of chord diagrams .
@xmath21 the answer to problem [ p65 ] can be related with the invariant tensors , defined for casimir lie algebras by chord diagrams , see @xcite .
also a problem which looks interesting is to find the gerstenhaber subalgebra in the hochschild homology of the operads @xmath113 , @xmath114 generated by the space of chord ( super)diagrams . for instance , in the case of the operad @xmath114 there is an element of bigrading ( 3,5 ) , that can not be obtained from chord diagrams by means of the multiplication and the bracket , see @xcite . in the case of the operad @xmath113
we do not know such an example .
in this section we clarify the difference between the cases of odd and even @xmath2 and give a geometrical interpretation of the operad @xmath300 ( for @xmath2 even ) .
in particular we explain remark [ r57 ] ( due to m. kontsevich ) .
theorems [ t71 ] , [ t72 ] given below are classical , see @xcite , @xcite . for
any @xmath40 let us introduce the topological _
operad @xmath380 of turning balls_. the space @xmath381 is put to be the configuration space of @xmath50 mappings of the unit ball @xmath382 into itself .
these @xmath50 mappings are supposed \1 ) to be injective , preserving the orientation and the ratio of distances ; \2 ) to have disjoint images .
the space @xmath381 can be evidently identified with the direct product of the @xmath50-th power @xmath383 of the special orthogonal group with the configuration space of @xmath50 disjoint balls in the unit ball @xmath384 .
the last space is homotopy equivalent to the @xmath50-th component @xmath243 of the little cubes operad .
the composition operations , the symmetric group actions and the identity element are defined analogously to the case of the little cubes operad .
[ t71 ] @xcite the homology @xmath385 of the turning discs operad ( balls of dimension 2 ) is the batalin - vilkovisky algebras operad @xmath148 .
@xmath21 let us describe the operad @xmath386 for any @xmath40 .
note , that the space of unary operations is the homology algebra @xmath387 .
[ t72 ] @xcite the homology bialgebra @xmath388 is the exterior algebra on the following primitive generators : \1 ) case @xmath389 : generators @xmath390 of degree @xmath391 respectively ; \2 ) case @xmath392 : generators @xmath393 of degree @xmath394 and @xmath395 respectively .
@xmath21 the generators @xmath396 , @xmath397 , are called the _ pontriagin classes _ , the generator @xmath398 is called the _
euler class_. the pontriagin classes ( and the subalgebra generated by them ) lie in the kernel of the map in homology @xmath399 induced by the natural projection @xmath400 the euler class is sent to the canonical class of dimension @xmath401 in the homology of the sphere @xmath254 .
now we are ready to describe the operads @xmath386 , @xmath40 .
we will say which objects are algebras over these operads .
[ t73 ] @xcite algebras over the operad @xmath386 are \1 ) for even @xmath2 : @xmath249-batalin - vilkovisky algebras , where the operator @xmath150 is @xmath402 ; \2 ) for odd @xmath2 : @xmath249-poisson algebras .
furthemore these algebras are supposed to have @xmath403 $ ] ( where `` @xmath404 $ ] '' denotes the integral part ) mutually super - comuting differentials @xmath405 - 1}$ ] ( of degree @xmath406 $ ] ) of the batalin - vilkovisky ( resp .
poisson ) algebra structure .
this means , that for any elements @xmath407 of the algebra and @xmath408 $ ] , one has \(i ) @xmath409)=[\delta_{4i-1}a , b]+(-1)^{\tilde a + d-1}[a , \delta_{4i-1}b]$ ] ; \(ii ) @xmath410 ; \(iii ) only for even @xmath2 : @xmath411 @xmath21
in this section we describe the first term of the auxiliary spectral sequence together with its first differential .
the corresponding complex will be called _ complex of bracket star - diagrams_. in the case of even @xmath2 this complex is isomorphic to the normolized hochschild complex of the batalin - vilkovisky algebras operad @xmath337 .
let us fix an @xmath274-configuration @xmath288 .
let @xmath412 @xmath413@xmath414 ( resp .
@xmath415 , @xmath416@xmath417 ) be the points of @xmath288 on the line @xmath279 that do not have stars ( resp . that do have stars ) .
consider the free lie super - algebra with the even bracket and with odd generators @xmath418 , @xmath413@xmath414 , @xmath419 , @xmath416@xmath417 .
we will take the symmetric algebra ( in the super - sens ) of the space of this lie super - algebra . in the obtained space we will consider the subspace @xmath420 spanned by the ( products of brackets ) , where each minimal component of @xmath288 is presented by one bracket , containing only generators indexed by the points of this minimal component and containing each such generator exactly once .
such products of brackets will be called _
bracket star - diagrams_. [ ex81 ] the space @xmath420 of the bracket star - diagrams corresponding to the @xmath274-configuration @xmath288 of the figure [ f53 ] is two - dimensional .
the diagrams @xmath421\cdot[[x_{t_2},x_{t_4^*}],x_{t_5}]\cdot x_{t_6^*},\,\ , [ x_{t_1},x_{t_3}]\cdot[[x_{t_2},x_{t_5}]x_{t_4^*}]\cdot x_{t_6^*}\ ] ] form a basis in this space .
@xmath21 if two bracket star - diagrams can be transformed into one another by an orientation preserving homeomorphism @xmath287 , then they are set to be equal .
for any equivalence class @xmath422 of @xmath274-configurations we define the space @xmath423 as the space @xmath420 , where @xmath288 is any element of @xmath422 .
the _ space of bracket star - diagrams _ is defined as the direct sum of the spaces @xmath423 over all equivalence classes @xmath422 of @xmath274-configurations .
the complexity @xmath32 and the number @xmath289 of geometrically distinct points of the corresponding @xmath274-configurations define the bigrading @xmath375 on the space of bracket star - diagrams .
remind that the complex of bracket star - diagrams ( both for @xmath2 odd and even ) is supposed to compute the first term @xmath424 of the main spectal sequence , whose @xmath425 coordinates are expressed as follows : @xmath426 @xmath427 the corresponding homology degree of the space of knots @xmath311 is @xmath428 note that the first term of the main spectral sequence is non - trivial only in the second quadrant @xmath429 .
the inequality @xmath430 for @xmath274-configurations provides the condition @xmath431 .
( 10,10 ) ( -230,100)@xmath432
the bialgebra of chord ( super)diagrams occupies the diagonal @xmath433 . to describe the differential @xmath215 on the space of bracket star - diagrams we will need some complementary definitions and notations . [ d83 ] let us permit to @xmath274-configurations , with @xmath275 , to have @xmath434 ; we demand also that one - element sets should never coincide with stars .
these @xmath274-configurations will be called _ generalized @xmath274-configurations_. the generalized @xmath274-configurations that are not @xmath274-configurations in the usual sens ( that have @xmath434 ) will be called _ special _ generalized @xmath274-configurations .
@xmath21 analogously we define the _ space of ( special ) generalized bracket star - diagrams_. [ ex84 ] @xmath435\cdot x_{t_2}$ ] is a special generalized star - diagram .
@xmath21 note that the space of generalized bracket star - diagrams is the direct sum of the space of bracket star - diagrams with the space of special generalized bracket star - diagrams .
[ d85 ] we say that a ( generalized ) @xmath274-configuration @xmath288 can be _ inserted _ in a point @xmath436 of another ( generalized ) @xmath437-configuration @xmath291 , if @xmath288 does not have common points with @xmath291 except possibly the point @xmath436 .
@xmath21 [ d86 ] we say that a ( generalized ) bracket star - diagram can be _ inserted _ in the point @xmath436 of another ( generalized ) bracket star - diagram , if it is the case for their @xmath274-configurations .
@xmath21 let @xmath116 and @xmath438 be two generalized bracket star - diagrams , such that @xmath116 can be inserted in the point @xmath439 ( or @xmath440 ) of @xmath438 , define the element @xmath441 ( resp .
@xmath442 ) of the space of generalized bracket star - diagrams . up to a sign @xmath443
is defined by replacing @xmath444 ( in the diagram @xmath438 ) for @xmath116 .
the sign is defined as @xmath445 , where @xmath446 is the parity of @xmath116 ( the parity of the number of geometrically distinct points ) , @xmath50 is the number of generators of the form @xmath447 before @xmath444 in @xmath438 .
in other words : we put the bracket containing @xmath448 on the first place , then by means of antisymmetry relations we put @xmath448 on the first place in the bracket ( and therefore in the diagram ) ; we replace @xmath444 for @xmath116 ; and we do all these manipulations in the inverse order .
it is easy to see that these two definitions give the same sign .
[ ex87 ] @xmath449\cdot [ x_{t_1^*}x_{t_0}]\bigl| _ { x_{t_0}=[x_{t_4}x_{t_5 } ] \cdot x_{t_6^*}}=(-1)^{(3 - 1)\cdot 3}[x_{t_2}x_{t_3^*}]\cdot [ x_{t_1^*},[x_{t_4}x_{t_5}]\cdot x_{t_6^ * } ] . ~\square\ ] ] note that if @xmath116 has more than 1 minimal components , then the element @xmath443 contains multiplications inside brackets .
therefore it is no more a ( generalized ) bracket star - diagram . to express this element as a sum of ( generalized ) bracket star - diagrams
we will use the formula .
now we are ready to define the differential @xmath215 on the space of bracket star - diagrams .
let @xmath116 be a bracket star - diagram , and let @xmath450 be one of its points without a star , then we define @xmath451 where @xmath452 is the projection of the space of generalized bracket star - diagrams on the space of bracket star - diagrams , that sends the space of special generalized bracket star - diagrams to zero ; the points @xmath453 are respectively @xmath454 and @xmath455 for a very small @xmath456 .
[ r89 ] the formula can be made more precise : @xmath457 let now @xmath458 be one of the points of @xmath116 having a star .
we define @xmath459}\bigr ) , \eqno({{\refstepcounter{theorem}}\arabic{section}.\arabic{theorem}}\label{eq811})\ ] ] where @xmath452 is the same projection , the points @xmath460 , @xmath461 are respectively @xmath462 and @xmath463 for a very small @xmath456 .
[ r812 ] the formula can be made more precise : @xmath464}.~\square \eqno({{\refstepcounter{theorem}}\arabic{section}.\arabic{theorem}}\label{eq813})\ ] ] the differential @xmath215 on the space of bracket star - diagrams is the sum of the operators @xmath465 and @xmath466 over all points @xmath450 , @xmath413@xmath414 , and @xmath458 , @xmath416@xmath417 , of the corresponding @xmath274-configurations : @xmath467 it is easy to see that @xmath468 .
[ r815 ] @xmath469}\right ) - ( x_{t_-}-x_{t_+})\cdot a , \eqno({{\refstepcounter{theorem}}\arabic{section}.\arabic{theorem}}\label{eq816})\ ] ] where @xmath470 ( resp .
@xmath471 ) is less ( resp .
greater ) than all the points of the diagram @xmath116 on the line @xmath279 .
@xmath21 let us fix an @xmath274-configuration @xmath288 and consider the free lie super - algebra with the even bracket and with the even generators @xmath418 , @xmath413@xmath414 , and the odd generators @xmath472 , @xmath416@xmath417 , where @xmath450 , @xmath413@xmath414 , and @xmath458 , @xmath416@xmath417 , are the points of our @xmath274-configuration @xmath288 .
let us take the exterior algebra ( in the super - sens ) of the space of this lie super - algebra . by convention
the parity of an element @xmath473 is @xmath474 the sum of the parities of @xmath475 , @xmath476 , plus the number @xmath477 of the exterior product signs . in the obtained space
we will consider the subspace @xmath420 spanned by the analogous products of brackets ( see the previous subsection [ s81 ] ) .
these products of brackets will be also called _
bracket star - diagrams_. the _ space of bracket star - diagrams _ is defined analogously to the case of odd @xmath2 .
we also accept definitions [ d83 ] , [ d85 ] , [ d86 ] .
note that the parity of a ( generalized ) bracket star - diagram is the number of stars plus the number of the exterior product signs .
this parity is opposite to the parity of the corresponding homology degree .
let @xmath116 and @xmath438 be two generalized bracket star - diagrams , such that @xmath116 can be inserted in the point @xmath436 of @xmath438 .
let us define @xmath478 . to do this we replace @xmath444 in @xmath438 by @xmath116 , an we multiply the obtained expression by @xmath479 , where @xmath480 is equal to zero ( resp . to one ) if the point @xmath439 has no star ( resp .
if the point @xmath440 has a star ) ; @xmath481 ( resp .
@xmath482 ) is the number of the exterior product signs ( resp .
of the generators corresponding to stars ) before @xmath448 in @xmath438 .
[ ex817 ] @xmath449\wedge [ x_{t_1^*}x_{t_0}]\bigl| _ { x_{t_0}=[x_{t_4}x_{t_5 } ] \wedge x_{t_6^*}}=(-1)^{(2 - 0)\cdot ( 1 + 2)}[x_{t_2}x_{t_3^*}]\wedge [ x_{t_1^*},[x_{t_4}x_{t_5}]\wedge x_{t_6^ * } ] . ~\square\ ] ] let @xmath116 be a bracket star - diagram , let @xmath450 be a point of @xmath116 without a star .
define @xmath483 where @xmath452 is the projection from the space of generalized bracket star - diagrams to the space of bracket star - diagrams .
[ r819 ] the formula can be made more precise : @xmath484 let @xmath458 be a point of @xmath116 having a star , we define @xmath485}\right).~\square \eqno({{\refstepcounter{theorem}}\arabic{section}.\arabic{theorem}}\label{eq821})\ ] ] [ r822 ] the formula can be made more precise : @xmath486}.~\square \eqno({{\refstepcounter{theorem}}\arabic{section}.\arabic{theorem}}\label{eq823})\ ] ] the differential @xmath215 on the space of bracket star - diagrams is defined by the formula analogously to the case of odd @xmath2 .
[ r824 ] @xmath487}\right ) - ( x_{t_-}-x_{t_+})\wedge a. \eqno({{\refstepcounter{theorem}}\arabic{section}.\arabic{theorem}}\label{eq825})\ ] ] @xmath21
in this section we define the structure of differential bialgebras on the complexes of bracket star - diagrams . we conjecture that this structure is compatible with the corresponding bialgebra structure on the homology space of the long knots space ;
the corresponding conjectures are given in @xcite .
let @xmath488 be a bracket star - diagram , @xmath489 be a real number . define a digram @xmath490 as the diagram obtained from @xmath488 by the translation of @xmath279 ( @xmath490 is equal to @xmath488 )
: @xmath491 let now @xmath116 and @xmath438 be two bracket star - diagrams , we define their product @xmath492 as the diagram @xmath493 in the case of odd @xmath2 , and as the diagram @xmath494 in the case of even @xmath2 , @xmath489 being a very large positive number .
this product resembles the product in the space of long knots : it follows from theorem [ t34 ] and corollary [ cor56 ] that the homology algebra of the differential algebra of bracket star - diagrams is commutative in the case of even @xmath2 for any commutative ring of coefficients .
this is also true over @xmath48 in the case of odd @xmath2 , see theorem [ t510 ] .
now we will define a comultiplication on the space of bracket star - diagrams . in the case of
odd @xmath2 the coproduct @xmath496 of any diagram @xmath497 , where @xmath475 , @xmath476 , are brackets , is defined as follows : @xmath498 where @xmath499 .
e.getzler , j.d.s.jones , _ operads , homotopy algebra and iterated integrals for double loop spaces _ , preprint hep - th/9403055 , departement of mathematics , massachusetts institute of technology , march 1994 .
v.a.vassiliev , _ topology of two - connected graphs and homology of spaces of knots _ , in differential and symplectic topology of knots and curves ( s.tabachnikov , ed . ) , ams translations , ser . 2 , * 190 * , ams , providence ri , 253286 , 1999 . | * keywords : * discriminant of the space of knots , bialgebra of chord diagrams , hochschild complex , operads of poisson gerstenhaber batalin - vilkovisky algebras .
this paper is a more detailed version of @xcite , where the first term of the vassiliev spectral sequence ( computing the homology of the space of long knots in @xmath0 , @xmath1 ) was described in terms of the hochschild homology of the poisson algebras operad for @xmath2 odd , and of the gerstenhaber algebras operad for @xmath2 even . in particular
, the bialgebra of chord diagrams arises as some subspace of this homology .
the homology in question is the space of characteristic classes for hochschild cohomology of poisson ( resp .
gerstenhaber ) algebras considered as associative algebras .
the paper begins with necessary preliminaries on operads .
also we give a simplification of the computations of the first term of the vassiliev spectral sequence .
we do not give proofs of the results . |
borocarbide superconductors _ _
rt__@xmath1b@xmath1c , where _ r _ is a rare - earth element and _ t _ is a transition metal element , mainly ni , ( see , e.g. , refs . and further refs .
therein ) have been studied intensively during the last decade . however , in spite of their much more simple , less correlated , 3d electronic structure compared with high-@xmath12 materials , the superconductivity in borocarbides is still under debate . based on the interpretation of many magnetic , muon spin resonance , spectroscopic , specific heat , etc .
measurements @xcite it is widely accepted that the superconducting ( sc ) state has an @xmath13-wave symmetry and the pairing is mediated by the electron - phonon ( el - ph ) interaction . on the other hand for the magnetically ordered borocarbide superconductors the role of the crystal - electric - field ( cef ) splitting of the 4@xmath14-shell states of the @xmath15 ion and their excitations are of fundamental importance to understand both ordering phenomena .
additionally , the el - ph interaction responsible for the sc state in these compounds has also been analyzed within the multiband eliashberg theory @xcite with emphasis on the complex fermi surface with different contribution to the sc state by different orbitals on different fermi surface sheets @xcite . in the family of borocarbide superconductors
is distinguished : it exhibits a remarkable reentrant superconductivity and a variety of magnetic orderings @xcite .
the reentrant behavior occurs in a temperature range just above the commensurate antiferromagnetic order at about @xmath165.2k and below the temperature @xmath17 k , which separates the low - temperature magnetic commensurate phase from the high - temperature paramagnetic phase by a not completely resolved yet intermediate incommensurate magnetic phase . in the same temperature range between @xmath9 and @xmath18 incommensurate structures do occur : an _ a_-axis - modulated one and a spiral structure along the _ c_-axis . by point - contact ( pc ) spectroscopic investigations
@xcite both the sc order parameter and the pc electron - boson(phonon ) interaction ( eb(p)i ) spectral function @xmath19 can be determined in studying the first and second derivatives of the @xmath20 characteristic of pcs .
the measurement of the second derivative of the @xmath20 of pcs provides straightforward information on the pc eb(p)i function @xmath21 @xcite .
the knowledge of @xmath21 for conducting systems is a touchstone to the phonon - mediated superconductivity which is governed mainly by the value of the el - ph - coupling parameter @xmath22 . moreover , the comparison of the experimentally determined @xmath23 with the results of model calculations could be used to distinguish between different scenarios .
thus the pc spectroscopy could be helpful to illuminate many details of the epi in honi@xmath1b@xmath1c as well as to resolve other possible quasiparticle interactions , e.g. , mediated by magnons and cef excitations .
in addition to this the sc gap determines the behavior of the @xmath20 curve for the pc in the energy range of a few mev .
hence , this region is widely used to determine the sc gap from a routine fitting by the well - known blonder - tinkham - klapwijk ( btk ) equations @xcite .
all this potential information should be used to deepen the present sparse knowledge of the sc properties and to help to elucidate the mechanism(s ) of superconductivity of the title material . up to now , honi@xmath1b@xmath1c has been investigated by pc spectroscopy in @xcite .
the first three papers are devoted to investigations of the sc gap . in general ,
standard isotropic bcs ( eliashberg)-type behavior of the gap was found below t@xmath24 , while between @xmath9 and @xmath28.5k the sc features in pc spectra are strongly suppressed and speculation about some different sc state @xcite or gapless
superconductivity @xcite were undertaken .
the pc ebi spectra for honi@xmath1b@xmath1c were studied and analyzed in ref.@xcite . there
, the main attention was devoted to the study of the low energy part of the spectra and to the investigation and understanding of the so called soft mode at about 4mev in the electron - quasiparticle spectrum .
however , the measured spectra were featureless above 20mev , although a number of pronounced phonon peaks are well resolved at higher energies by neutron spectroscopy @xcite .
in this study we present more detailed pc ebi spectra of honi@xmath1b@xmath1c in comparison with those mentioned in the refs . given above including also our recent data presented at two international conferences sces04 and m2s - htsc - viii @xcite .
additionally , to explain our findings with respect to the sc gap we provide support for the separation of magnetism and superconductivity on different fermi surface sheets in magnetic borocarbides @xcite .
we have used single crystals of honi@xmath1b@xmath1c grown by a floating zone technique with optical heating @xcite .
the residual resistivity is @xmath25 cm and the residual resistivity ratio rrr=13 .
final series of measurements were done for sample with improved quality with rrr=19 .
the sample becomes superconducting at about 8.5k .
pcs were established both along the c axis and in perpendicular direction by standard needle - anvil or shear methods
@xcite . as a counter electrode cu or ag thin wires ( @xmath26=0.15 mm ) were used .
the spectra with resolved phonon features like in fig.[hof1 ] were registered by touching of the sharpened cu wire on a shallow bright caving of about 1@xmath271mm@xmath28 on the honi@xmath1b@xmath1c crystal surface ( see inset in fig.[hof1 ] ) .
the electron probe microanalysis ( wdx mode ) has shown that the cavity surface has single phase microstructure and stoichiometric composition well corresponding to the 1:2:2:1 ratio , while excess of ni has been observed in other places on that crystal surface . ) ] of several honi@xmath1b@xmath1c cu pcs .
the dotted curve shows an example of the tentative behavior of the background for the dash - dot curve spectrum . a magnetic field of 0.5 t
is applied to suppress superconductivity .
the arrows mark a magnetic ( m ) peak at about @xmath293mv , a cef peak , and positions of other resolved peculiarities on the curves maxima .
the inset shows a sketch of the pc configuration ( see also the text for more details).,width=359 ]
according to the theory of pc spectroscopy @xcite the second derivative @xmath30 of the @xmath31 curve of the ballistic contact at low temperatures , where @xmath32 , is determined by the pc epi function @xmath19 , @xmath33 where @xmath34 is the pc diameter , e is the electron charge and @xmath35 , roughly speaking , measures the interaction of an electron with the phonon branches .
this interaction is affected by large angle scattering ( back - scattering ) processes @xcite of electrons in the constriction of the pc .
thus @xmath19 is a kind of transport epi functions which select phonons with a large momentum or umklapp scattering . in practice @xmath36
can be extracted from the measured rms signal of the second @xmath37 ( @xmath38 d@xmath39/d@xmath40 ) harmonic of a small alternating voltage @xmath41 ( @xmath42 d@xmath43/d@xmath44 is the first harmonic signal ) superimposed on the ramped dc voltage @xmath43 , so that from ( [ pcs ] ) : @xmath45 in the case of a heterocontact between two metals the pc spectrum represents a sum of the contributions from both metals 1 and 2 weighted by the inverse fermi velocities @xcite : @xmath46 where @xmath47 is the relative volume occupied by metal 1 in the pc .
thus , using of the heterocontacts enables also us to estimate the relative strength of the ebi in the investigated material as compared to some standard or well known case . to study the sc gap we have measured @xmath48 characteristics of n - c - s pcs ( here n denotes a normal metal , c is the constriction and s is the superconductor under study ) .
then utilizing the generally used blonder - tinkham - klapwijk equations @xcite fits of the measured curves have been performed . as fit parameters the sc gap @xmath49 and the dimensionless barrier parameter @xmath50
have been used .
a possible additional smearing of the d@xmath51d@xmath44 curves was also taken into account by including the parameter @xmath52 , usually interpreted as a broadening of the quasiparticle dos in the superconductor due to finite life - time effects . with subtracted background ( 3 upper solid curves ) .
the arrows mark the position of the cef peak and the
magnetic peak at about 4mv , best visible for the two upper curves .
the bottom dotted curve shows the pc spectrum of cu .
the vertical dashed line marks the position of the first maximum in the cu spectrum .
the dashed curve presents sum of the pc spectrum of cu with a function scaled by the phonon dos from panel ( b ) .
( b ) the neutron phonon dos for luni@xmath1b@xmath1c @xcite.,width=302 ] our pc spectra of honi@xmath1b@xmath1c demonstrate clear maxima at about 3 , 10 , 16 and 22mv , a smeared maximum at 34mv , and a hump around 50mv ( fig.[hof1 ] ) .
the maxima above 15mv correspond well to the phonon dos maxima of the nonmagnetic sister compound luni@xmath1b@xmath1c @xcite [ see fig.[hof2](b ) ] derived from inelastic neutron scattering data .
only the high energy part of the pc spectrum is remarkably smeared .
fig.[hof2](a ) displays the pc ebi function , obtained from the pc spectrum after subtraction of the background . according to the neutron data @xcite
there is a gap around 40mev in the phonon dos , which separates the acoustic and the optic branches . in this energy region
a minimum occurs in our pc spectra .
b@xmath1c cu contact with well resolved high energy maxima around 50 and 100mv .
pc resistance was decreased from 10.6@xmath53 ( spectrum in the insert ) to 9.3@xmath53 ( main panel ) between successive measurements . in the main panel
also averaged for two contacts spectrum ( solid line ) is shown , while in the inset averaged for the positive and negative bias spectrum of the same contact is shown by solid line .
temperature is 3k and magnetic field is applied to suppress superconductivity.,width=359 ] we have also succeeded to receive pc spectra with well resolved high energy maxima around 50 and 100mv ( see fig.[hof2a ] ) .
the position of the 50-mv maximum coincides with the maximum in the phonon dos for luni@xmath1b@xmath1c shown in fig.[hof2]b , even the shoulder at about 60mv is resolved in the pc spectra ( this feature is more clear at the negative bias ) . in the phonon dos for the nonmagnetic sister compound yni@xmath1b@xmath1c
a maximum at around 100mev has been observed in ref . .
it has been attributed to the high - frequency b - c bond stretching vibration in accord with theoretical results obtained in frozen phonon calculations at 106 mev for the zone - center raman active mode of yni@xmath1b@xmath1c @xcite .
the low - energy part ( @xmath54mev ) of the pc spectra in fig.[hof2a ] shows a less detailed structure as compared to the spectra from fig.[hof1 ] . here
maxima only at about 4 , 10 and 15mv are seen .
also a shoulder in the range of 22mev peak is visible which is more pronounced at negative polarity and as shown for the spectra in the inset .
however , in interpreting these spectra we have to take into account for possible contribution of cu derived phonons since cu has been used as a counter electrode ( needle ) .
the dashed curve in fig.[hof2](a ) presents the sum of the cu pc spectrum and the phonon dos from fig.[hof2](b ) .
it is seen that contributions from cu may be appreciable for the spectrum of the 41@xmath53 contact in the range between 1520mev , while the two other spectra have pronounced maxima at 16 and 22mev , i. e. at energies higher than the maxima in the cu spectrum at 15 and 20mev .
moreover , all honi@xmath1b@xmath1c cu spectra contain no visible contribution from the weaker cu peak at about 27mev , therefore we conclude that a contribution of cu is unsubstantial at least for the two upper spectra in fig.[hof2](a ) .
apparently , the low fermi velocity in nickel borocarbides @xcite as compared to the noble metals accentuate the intensity of honi@xmath1b@xmath1c features according to eq.([het ] ) .
contrary , the pc spectra in fig.[hof2a ] have likely a contribution from cu derived phonons in the region 1520mv , which masks the expected honi@xmath1b@xmath1c phonon maxima so that the 22mev maximum looks like a broad shoulder . in fig.[hof2a](inset ) also the asymmetry of the spectra versus bias voltage polarity is clearly seen in the energy region of cu phonons , which is characteristic for heterocontact , if the contribution from both electrodes to the spectrum is comparable @xcite . ) ] of several honi@xmath1b@xmath1c ag pcs with a dominant 10-mv peak ( the @xmath55 marks the cef peak , the @xmath56 mark the position of the main phonon peaks from fig.[hof2]b , the black triangle marks the
magnetic peak at about 4mev ) : ( a ) two spectra with a dominant 10-mev peak . in the first case
the superconductivity is suppressed by magnetic fields , in the second case by high temperature .
the latter curve is measured on a cleaved small piece of a honi@xmath1b@xmath1c single crystal .
the bottom curve shows the pc spectrum of ag .
( b ) the modification of the spectrum with a dominant 10-mv peak for various applied magnetic fields.,width=302 ] the most prevalent type of honi@xmath1b@xmath1c pc spectrum is shown in fig.[hof3 ] . here
a maximum close to 3mv and another one around 10mev are dominant . the latter maximum
might be connected with cef excitations , observed in this range by inelastic neutron scattering studies @xcite .
the maximum around 3mev is associated with a magnetically ordered state , since it disappears above the nel temperature mainly in the range 810k ( see dashed curve in fig.[hof3](a ) ) and it can be suppressed by a magnetic field of a few tesla ( see fig.[hof3](b ) ) .
a detailed discussion of the origin of this peak will be given in the next section .
at the energy of about 10mev a contribution from the first peak in the phonon dos could also be expected .
for luni@xmath1b@xmath1c the first phonon peak is observed at about 9mev ( see fig.[hof2](b ) ) .
for the lighter ( about 6% ) ho ions the peak position may be somewhat shifted to higher energies .
note also that the measured phonon dispersion curves @xmath57 for honi@xmath1b@xmath1c @xcite have a tendency that @xmath58 around 1012mev pointing to an enhanced phonon dos ( or maximum ) in this region .
however , we have observed that the 10-mv maximum is considerably affected by a magnetic field ( see fig.[hof3](b ) ) , therefore a contribution of cef excitations to the 10-mev peak seems to be appreciable . unfortunately , due to
magnetic forces it was difficult to keep the mechanical contact stable in magnetic fields above 5 t , especially when the field direction was out of the @xmath59-axis .
therefore the influence of the magnetic field on the 10-mev peak ( to separate cef and phonon contribution to this peak ) could be investigated only qualitatively so far .
the spectra discussed here demonstrate weak and smeared phonon maxima above 10mv ( see arrows in fig.[hof3](a ) ) in contrast to the spectra shown in fig.[hof2](a ) , but they have a pronounced 10-mv peak ( see fig.[hof3 ] ) , which points to the importance of cef excitations in the transport properties of honi@xmath1b@xmath1c .
besides , the electron - cef interaction may be also important for a complete understanding of the mechanism of superconductivity in this compound .
we emphasize that the spectra with the dominant 10-mv peak were qualitatively the same independent of the counter electrode ( ag or cu ) which discards remarkable contribution to the pc spectra from these metals ( see the pc spectra of cu and ag in figs.[hof2]a and [ hof3]a ) . from the measured ebi function
( see fig.[hof2](a ) ) the ebi parameter @xmath60 was calculated , which was found to be smaller than 0.1 only . however , there are large uncertainties in the determination of @xmath61 if heterocontacts are used .
first , this is because the relative volume @xmath47 occupied by the investigated sample in a pc can deviate from 1/2 [ see ( [ het ] ) ] .
secondly , the deviation from the ballistic regime in pcs has to be corrected by a pre - factor @xmath62 in ( [ het ] ) , where @xmath63 is the elastic electron mean free path and @xmath34 is the pc diameter .
however , @xmath63 is difficult to evaluate for pc .
further , the available pc theory and all its equations such as eq.([pcs1 ] ) have been derived in the framework of the free electron model and a single - band fermi surface .
then , paying attention that the contribution of cu in the pc spectra of honi@xmath1b@xmath1c - cu heterocontacts in fig.[hof2](a ) is hardly to be resolved we can conclude that the intensity of the ebi function in honi@xmath1b@xmath1c has to be larger than that in cu , where @xmath640.25 @xcite .
this provides a complementary confirmation of the moderate strength of the ebi in honi@xmath1b@xmath1c with a lower bound of 0.25 for @xmath61 @xcite .
moreover the 10-mv peak contributes about 20 - 30% to the @xmath65 value indicating that cef excitations should be taken into account to understand the sc properties of honi@xmath1b@xmath1c .
the contribution of the high frequency modes to @xmath65 can be estimated from the spectra in fig.[hof2a ] .
it amounts about 10% for each of the 50 and 100mv maxima .
note , that by estimating @xmath65 we discard the contribution of the
peak around 3mv , the nature of which is under discussion ( see below ) . ) ] of honi@xmath1b@xmath1c cu contact established along the a - axis at different temperature .
( b ) pc spectra of honi@xmath1b@xmath1c ag contact established along the a - axis with varying of the magnetic filed direction relative to the @xmath59-axis.,width=302 ] let us turn to the low energy maximum around e@xmath66 mev . as we mentioned above it
is very probably related to a magnetic ordering , since it disappears at temperatures @xmath67 810k ( see fig.[hof4](a ) ) and it can be suppressed by an external magnetic field ( fig.[hof4](b ) ) .
figure[hof4](b ) shows that by increasing the angle between the applied magnetic field and the @xmath59-axis , the height of the maximum goes down and its position changes .
thus , this maximum is more robust , if the field is applied along the @xmath59-axis .
this is in line with the result reported in ref . that for a field along the @xmath59-axis the antiferromagnetic phase with @xmath68k is robust up to quite large fields , while for the perpendicular directions
this antiferromagnetic phase is already suppressed above @xmath69=0.5 t at @xmath70=3k . for our pcs
both @xmath71 and the magnetic field which suppresses the
magnetic peak are higher .
in particular , especially the magnetic field is enhanced which reaches here up to 5 t ( see fig.[hof3](b ) ) . in this context
it should be mentioned that inside a pc a remarkable ( uniaxial ) pressure can be created .
this might cause an increase of @xmath72 with a ratio up to d@xmath72/d@xmath738k / gpa @xcite . for completeness
it should be mentioned that spectra with a dominant 3 mev peak ( its position may vary between 2.5 and 5 mev ) and the practically absence of any other maxima were observed on uncleaved as - grown surfaces of honi@xmath1b@xmath1c .
later on the electron probe microanalysis showed deviations from the stoichiometry on this surface , which can modify both t@xmath74 and the characteristic magnetic field .
as mentioned in ref .
crystallographically oriented thin plates of the nonsuperconducting honibc compound have been observed in selected parts of as - grown honi@xmath1b@xmath1c crystals .
it is well - known that the 1:1:1:1 compound honibc exhibits long - range magnetic order below @xmath7510k @xcite .
therefore , the observation of a magnetic peak persisting up to 10k may be connected with admixture of 1:1:1:1 phase .
thus , the origin of the
magnetic peak in the pc spectra has to be still under debate and we will consider further scenarios .
it could be connected with a transition to the thermal regime by a voltage rise , which produces peak - like features in the pc spectra at e@xmath76k@xmath77 ( here @xmath78 is a magnetic transition temperature ) as it was shown for ferromagnetic metals @xcite . however , this is not likely the case , because an increase of temperature in the pc would drastically smear the phonon maxima in the spectra as well . on the other hand ,
it was shown in ref . that the reabsorbtion of nonequilibrium phonons generated by electrons in a pc leads to a lattice heating up to a temperature of e@xmath43/4 ( and the appearance of a background in the pc spectra ) , however
, here the electron gas of the conduction electrons should remain cold preventing a smearing of the pc spectra whereas the local ho 4@xmath14 moments might be affected due to a strong enough magnetoelastic coupling @xcite .
we suppose that such a `` heating '' by nonequilibrium phonons suppresses the magnetic order .
the destruction of the antiferromagnetic order in honi@xmath1b@xmath1c leads to a sharp increase of the resistivity at @xmath72 ( see , e.g. , fig.3 in ref . ) , which manifests itself in the pc spectra as a peak - like feature .
indeed , the position of the peak around 2.5 mev ( see fig.[hof4 ] ) corresponds to the temperature e@xmath79 7k @xmath80 . for the n - c - s metallic junctions ( here
n denotes a normal metal , c is the constriction and s is a superconductor ) under consideration also the sc energy gap @xmath81 can be investigated ( see , e.g. , @xcite , chapter 3.7 ) .
the sc gap manifests itself in the @xmath48 characteristic of a n - c - s contact as pronounced minima around @xmath82 at a temperature well below t@xmath83 .
such @xmath48 curves are presented in fig.[hodt ] . using the btk equations @xcite to fit @xmath48 , the sc gap @xmath81 and its temperature dependence
are established ( see fig.[hodt ] ) .
it is seen that @xmath84 has a bcs - type dependence , however the temperature where the gap vanishes [email protected] is well below [email protected] of the bulk crystal , i.e. , the gap disappears close to t@xmath24 .
similar observation of a gap which vanishes at few kelvin below t@xmath83 were presented in @xcite both for poly- and single crystals of honi@xmath1b@xmath1c . in ref .
@xcite the authors supposed a gapless state between t@xmath87 and t@xmath83 based on the observation of an abrupt suppression of sc features in @xmath48 when crossing t@xmath87 and the failure of a btk fit above t@xmath87 @xcite .
note , that in the mentioned temperature range honi@xmath1b@xmath1c exhibits @xcite a specific magnetic spiral structure .
according to @xcite the helical magnetic order weakens the superconductivity and even nonmagnetic impurities do suppress superconductivity in systems with coexisting superconductivity and antiferromagnetism @xcite .
curves ( symbols ) of honi@xmath1b@xmath1c cu contact ( @xmath88 ) established along a - axis with varying temperature .
pc spectra for this contact are shown in fig.[hof4](a ) .
solid lines are btk fitting curves .
bottom panel : temperature dependence of the sc energy gap @xmath81 and smearing
parameter @xmath52 at @xmath69=0 t obtained by a btk - like fitting of the curves shown in the upper panel at constant tunneling
parameter @xmath890.5 .
solid smooth line is the bcs curve.,width=302 ] we would like also to accentuate that the same gap value with a similar temperature and magnetic field behavior has been found for a cleaved surface of the same honi@xmath1b@xmath1c single crystal with an ag counter electrode , that is for some arbitrary direction .
this observation together with analogous data for @xmath81 from @xcite both for poly- and single honi@xmath1b@xmath1c crystals demonstrates at least a small gap anisotropy and a reasonable physical meaning of t@xmath87 .
therefore we adopt the fermi surface separation
scenario for the coexistence of magnetism and superconductivity in magnetic borocarbides , i.e. their coexistence on different fermi surface sheets ( fss ) , proposed in refs.@xcite .
in other words , we suggest that the superconductivity in the commensurate antiferromagnetic phase survives only at a special ( nearly isotropic ) fss with no admixture of ho 5 @xmath34 states .
the latter mediate the magnetic exchange interaction between the localized ho 4@xmath14 moments and a part of the delocalized conduction electrons for which a standard @xmath13-wave superconductivity is easily to be destroyed or strongly suppressed .
such a magnetic link between the ho 4@xmath14 moments provided by conduction electrons is necessary for the rkky interactions to explain the observed magnetic orderings at relative high temperatures . in order to proceed with a semi - quantitative analysis of our pc data
, we may now adopt the well understood standard isotropic single band ( isb ) model to analyse the residual superconductivity . from the measured gap value @xmath90=0.95 mev and the estimated @xmath91 ( or the slightly different value of 5.8 k as discussed below ) remarkable strong coupling corrections
can be estimated .
starting for instance from a relation first derived by geilikman and kresin @xcite @xmath92 , \label{delta}\ ] ] one estimates a relatively low value for the logarithmically weighted phonon frequency @xmath93 10 mev . using the well - known allen - dynes expression for the critical temperature we may estimate the effective electron - boson coupling constant @xmath94 as 0.934 adopting a typical value for the coulomb pseudopotential @xmath95=0.13 .
then , from the estimated upper critical field @xmath96 0.8 t one estimates in the clean limit the effective fermi velocity @xmath97 according to @xcite @xmath98=0.154\frac{t_{c}\left[\mbox{k}\right ] \left(1+\lambda\right)^{1.1 } } { \sqrt{h_{c2}(0)\left[\mbox{t}\right ] } } , \label{vfermi}\ ] ] as a lower bound @xmath99 m / s . a similar value @xmath100 m / s follows , if @xmath1010.93 t is estimated from the slope of @xmath102 near @xmath103 using the well - known whh - relation between the former two quantities .
these estimates may be helpful to elucidate which of the fss does bear the residual superconductivity . compared with the nonmagnetic borocarbides ( @xmath104 to 1.4 ) the obtained coupling
constant is surprisingly less suppressed and the low value @xmath93 10 mev indicates that most of the phonon modes above 10mev shown in fig.[hof2 ] should be regarded as relevant for those nonactive fss s where the superconductivity has been extruded by the rare earth induced magnetism . since in the scenario under consideration the pairing comes mainly from the interaction with low - frequency bosonic modes , the observed peak near 10mev and its decomposition into contributions from different bosons deserves special interest . in particular , assuming that the remaining superconducting fss couples dominantly to the modes which contribute to the @xmath105 peak , a non - negligible contribution @xmath106 from the attractive interaction with crystal - field excitations ( i.e. from the inelastic asymmetric coulomb scattering as proposed by fulde @xcite ) would be rather challenging . regarding the exact frequency values of the relevant bosonic modes , the large sensitivity of eq.([delta ] ) to the actual @xmath12 value should be stressed . in this context
the rather similar values of @xmath107 , @xmath96 0.75 to 0.9 t @xcite and 2@xmath108 3.81 for dyni@xmath1b@xmath1c are noteworthy ( see ref .
fig.6 ) . here
@xmath109 and @xmath94= 0.939 ( for @xmath110 ) can be derived from eq.([delta ] ) and the allen - dynes @xmath103-formula .
such a similarity between honi@xmath1b@xmath1c and dyni@xmath1b@xmath1c seems to be natural because the same commensurate antiferromagnetic phase is probed for both closely related systems which exhibit within the lda ( local density approximation ) nearly the same electronic structure @xcite . also similar phonon spectra
are expected . in this context , in comparing the two sc phases which coexist with the same commensurate antiferromagnetic phase in honi@xmath1b@xmath1c and dyni@xmath1b@xmath1c , the violation of the frequently used de gennes - scaling for magnetic borocarbides @xcite is striking .
this scaling employs the scaling factor dg defined as @xmath111 where @xmath112 is the land - factor and @xmath113 denotes the magnetic moment at the rare earth site . according to that standard scenario an @xmath114 sc state would be expected for the low - temperature sc phase of honi@xmath1b@xmath1c . however , by comparing the sc transition temperatures and gap values of the magnetically adjacent borocarbides contradictory results are derived .
the pc - data for erni@xmath1b@xmath1c @xcite yield @xmath115 1.7 mev ( ignoring a weak anisotropy of @xmath116 ) , the gap of dyni@xmath1b@xmath1c amounts to @xmath117 1.0mev and @xmath118 with @xmath119 is expected .
this is in contrast with the measured value of 0.95mev for honi@xmath1b@xmath1c .
anyhow , the measured relation @xmath120 suggests that the residual superconductivity below 6k is not affected by the usual magnetic pair breaking .
our fss - related coexistence picture allows also a new interpretation of the specific heat data reported by michor _
@xcite : using the estimates for dyni@xmath1b@xmath1c given above , strong coupling corrections for the relative specific heat jump @xcite @xmath121,\ ] ] can be used to estimate the partial density of states @xmath122 of the electronic subgroup involved in the sc transition .
the observed reduced sc specific heat jump of @xmath123 70 mj / molk for dyni@xmath1b@xmath1c , that corresponds to a partial density of states of about @xmath124 4.16 mj / molk@xmath28 , only , to be compared with about @xmath125 10 mj / molk@xmath28 for the total density of states as derived from lda band structure calculations for nonmagnetic and magnetic borocarbides @xcite .
it is interesting to compare @xmath4 with the superconducting coupling constant @xmath126 derived above as well as with superconducting and high - temperature resistivity data which yield @xmath127 for related systems in order to understand to which extent @xmath4 probes the el - ph coupling for all bands .
recent pc measurements for dyni@xmath1b@xmath1c @xcite and yni@xmath1b@xmath1c @xcite in the best junctions yield a significant larger ( smaller ) experimental coupling constant @xmath128 ( 0.1 ) , respectively , compared to the present case with @xmath129 for honi@xmath1b@xmath1c .
however , very often the coupling constants derived from point - contact spectra should be considered as lower bounds for the coupling constants relevant for superconductivity , say @xmath130 ( 0.5 to 0.7)@xmath126 , which can be derived from different eliashberg functions @xmath131 extracted by the inversion of usual superconductor tunneling data or from an analysis of the sommerfeld constant of the linear specific heat in the normal state .
since a value of @xmath132 for the case of the non - magnetic borocarbides and @xmath133 0.8 - 0.9 for magnetic borocarbides @xcite seems to be realistic , we adopt for ideal ballistic contacts @xmath134 0.7 to 0.8 for all ni-1221 borocarbides under consideration .
in fact , the coupling constants , measured by point - contact spectroscopy in the normal state can be regarded as effective single - band properties , i.e.for the multiband system under consideration generalizing eq.(1 ) , each band ( fss ) produces a weighting factor given by the ratio of the corresponding conductances and the fermi velocities @xmath135 which in a free electron - like picture reads @xmath136 , \label{lambda}\ ] ] where the band index
@xmath1371 , .. has been introduced . in the present case following our coexistence scenario
we divide the borocarbides under consideration into a magnetic ( @xmath138=1 ) and a superconducting subsystem ( @xmath1372 ) with @xmath139 used in calibrating the point - contact spectra and adopt @xmath140 , as suggested by the lda - fplo calculations , one estimates @xmath141 = 0.9 and 0.1 , respectively . then eq.(8 )
can be rewritten as @xmath142 . for @xmath143 a moderate contribution of about 25% for the superconducting subsystem to the total @xmath4 follows .
the latter is dominated by the quenched magnetic subsystem for honi@xmath1b@xmath1c and other magnetic borocarbides .
anyhow , the 10 mev peak can be attributed completely to the residual superconducting subsystem .
a similar result was obtained for a specific heat analysis of honi@xmath1b@xmath1c @xcite , where the peak near 10mev was estimated to contribute about 25% to the total coupling constant of 1.2 .
related problems and more detailed further comparisons with dyni@xmath1b@xmath1c , erni@xmath1b@xmath1c , and other magnetic borocarbides will be considered elsewhere . for completeness and for comparison with eq .
( [ lambda ] ) we show the corresponding multiband expressions for the renormalization of the electronic specific heat ( the sommerfeld constant @xmath144 ) and the resistivity @xmath145 where @xmath146 denotes the partial electronic density of states at the fermi level . finally , ignoring interband scattering for the high - temperature resistivity where @xmath147 one obtains from a straightforward generalization of the well - known ziman - formula @xmath148 where @xmath149 denotes the plasma frequency .
thus , in all discussed cases the effective one - band coupling constants are differently decomposed into the partial coupling constants depending on which quantity is probed experimentally .
curves ( symbols ) for the same honi@xmath1b@xmath1c cu contact as employed in fig .
[ hodt ] but exposed to an external magnetic field .
the solid lines are btk fitting curves .
bottom panel : the magnetic field dependencies of the sc energy gap and @xmath52 at constant @xmath890.5 obtained by btk - like fits of the curves shown in the upper panel.,width=302 ] lets turn to the magnetic field measurements shown in fig.[hodh ] .
in contrast to the pc data reported in ref . , where @xmath81 initially increases ( @xmath150 ) in magnetic fields before it drops off to zero , in our case @xmath151 exhibits a conventional behavior . decreasing with overall negative curvature ( see fig.[hodh ] ) .
similarly , the excess current @xmath152 , which is proportional to the gap maximum area in @xmath153 ( see fig.[iex ] and the figure caption therein ) , exhibits a temperature dependence ( not shown ) like @xmath84 from fig.[hodt ] , that is @xmath154 has also a negative curvature .
( we remind the reader , that the excess current of an s - c - n contact is governed by @xmath81 @xcite or the sc order parameter @xcite . )
contrary , @xmath155 decreases nearly linearly with the magnetic field ( fig.[iex ] ) .
this is in contrast with the observed pronounced positive curvature of @xmath155 in mgb@xmath1 @xcite and yni@xmath1b@xmath1c @xcite , where at least two bands are superconducting .
thereby , the observed behavior of @xmath155 is in line with our scenario that superconductivity survives on separate ( one or similar ) fss .
in addition to the `` vortex '' model proposed in @xcite , the @xmath156 dependence might be also related to a complex vortex structure for a system with normal and superconducting fss s as we suppose in our case . to the best of our knowledge
this vortex problem has not been studied so far theoretically .
behavior versus magnetic field ( symbols ) for pc data used also in fig.[hodh ] . dashed straight line connects two edge symbols .
the inset explains the determination of the excess current , which is proportional to the integral intensity of @xmath153 maximum ( the shaded area ) after subtracting of the normal state @xmath153 ( dashed curve ) measured at 0.5 t .
, width=321 ]
we have carried out investigations of the ebi spectral function and the sc energy gap in honi@xmath1b@xmath1c by pc spectroscopy .
we succeeded to measure pc ebi spectra of honi@xmath1b@xmath1c with distinct phonon features above 15mv along with additional low energy maxima of non - phonon nature below 15mev .
for the first time the high energy maxima at about 50mev and 100mev are clearly observed .
hence , the high energy phonons can not be disregarded by analyzing thermodynamic , electron transport and sc phenomena in the nickel borocarbides .
the phonon maxima in the pc spectra correspond well to those in the phonon dos spectra of borocarbides measured by inelastic neutron scattering . on the other hand
the maximum at about 10mev contains likely also an essential contribution from cef excitations , while the low - lying peak around 3mev is suppressed by temperature and magnetic field rise .
moreover , the peak intensity depends on the angle between the magnetic field and the @xmath59-axis .
all these observations point to a
magnetic origin of this peak , i.e. , the peak reflects possibly a transition from the low temperature magnetically ordered state to dis- ( or low)- ordered states driven by nonequilibrium phonons generated by accelerated electrons .
the sc gap exhibits in general a
conventional ` single - band like standard bcs behavior below t@xmath157k with @xmath158t@xmath159 pointing to an intermediately strongly coupled sc state .
the large similarity of this low - temperature phase with the behavior of dyni@xmath1b@xmath1c where superconductivity and a simple commensurate antiferromagnetic state coexist , clearly points to a nonstandard ( in the sense of the frequently used de gennes scaling ) suppression of the former compared with nonmagnetic borocarbides .
we addressed the specific role of selected fermi surface sheets and cef - excitations in this particular case . between t@xmath87 and the upper critical temperature [email protected] the sc signal in @xmath48
is drastically suppressed , which explains the difficulties to derive the sc order parameter ( or a sc gap ) by fitting the @xmath48 data .
hence the sc state between t@xmath87 and t@xmath83 , in the region of peculiar magnetic order , is seemingly ` unconventional ` . due to its weakness
there is a large sensitivity to both external effects ( stress or pressure by pc creation , high current density in pc ) and to the intrinsic properties ( small deviation from stoichiometry , a specific sc state at the surface caused by , e.g. , peculiar magnetic ordering ) .
further investigations are desirable to settle the nature of the sc state in this special temperature range to obtain finally a similar level of understanding as for the low - temperature sc phase coexisting with the commensurate antiferromagnetic state we have achieved here . in a more general context one may only agree with the conclusion drawn in a very recent review paper @xcite as to ' the @xmath160ni@xmath161b@xmath161c series will continue to provide insight to superconductivity and its interaction and response to local moment magnetism for years , and possibly decades to come . ''
the support by the deutsche forschungsgemeinschaft ( sfb 463 ) and the state foundation of fundamental research of ukraine are acknowledged .
one of the authors ( o.e.k . ) was supported by the daad through a nato - grant .
we are indebted to l. v. tyutrina for the help in fitting the @xmath162 data and w. lser for a critical reading of the manuscript .
discussions with k .- h .
mller , m. schneider and h. rosner are acknowledged .
10 k .- h .
mller , v. n. narozhnyi , rep .
phys . * 64 * , 943 ( 2001 ) ; k .- h .
mller , g .
fuchs , s .- l .
drechsler and v. n. narozhnyi , in : " magnetic and superconducting properties of rare earth borocarbides of the type rni@xmath161b@xmath161c " , handbook of magnetic materials , ( ed
. k. h. j. buschow ) , elsevier north - holland , vol .
14 , ( 2002 ) , pp .
199 - 305 .
drechsler , s. v. shulga , k .- h .
mller , g. fuchs , j. freudenberger , g. behr , h. eschrig , l. schultz , m.s .
golden , h. von lips , j. fink , v. n. narozhnyi , h. rosner , p. zahn , a. gladun , d. lipp , a. kreyssig , m. loewenhaupt , k. koepernik , k. winzer , k. krug , physica c * 317 - 318 * , 117 ( 1999 ) .
h. rosner , s .- l .
drechsler , k. kopernik , i. opahle , and h. eschrig , in rare earth transition metal boricarbides ( nitrides ) : superconducting , magnetic and normal state properties ( eds .
mller and v. narozhnyi ) , kluwer academic publishers dordrecht , nato science series `` ii .
mathematics , physics and chemistry '' , v. 14 ( 2001 ) pp .
71 - 82 .
a part of the data reported here has been presented within an invited talk at the m@xmath163s - htsc conference in dresden ( july 9 - 14 , 2006 ) ; to be published as yu .
g. naidyuk , o. e. kvitnitskaya , i. k.yanson , g. fuchs , k. nenkov , a. wlte , g. behr , d.souptel , and s .- l .
drechsler , in physica c ( 2007 ) ; doi:10.1016/j.physc.2007.03.165 ; see also cond - mat/0609770 .
kulik , a.n .
omelyanchuk , r.i .
shekhter , fiz .
nizk . temp .
* 3 * , 1543 ( 1977 ) [ sov .
j. low temp .
phys . * 3 * , 740 ( 1977 ) ] ; i. o. kulik , fiz .
temp . * 18 * , 450 ( 1992 ) [ sov . j. low temp
. phys . * 18 * , 302 ( 1992 ) ] . in discussing the value of @xmath164 derived from a pc spectrum , we should always to bear in mind that it is a transport coupling constant which in general differs from the thermodynamic counter parts @xmath126 , although for some metals ( see e.g. ref . )
@xmath165 holds accidentally . using equations for
the conductivity of n - c - s contacts with magnetic impurities ( see s.i .
beloborodko , low temp .
phys , * 29 * , 650 ( 2003 ) ) , calculations of the @xmath48 curve display a smooth variation of the @xmath48 shape ( minimum ) at the crossover to the gapless state ( private communication by n. l. bobrov ) h. doh , m. sigrist , b.k .
cho , and s .-
lee , phys .
lett.*83 * , 5350 ( 1999 ) and h .- j .
choi , h. doh , e .-
choi , s .-
lee , m. ohashi , and n. mori , physica b * 327 * , 438 ( 2003 ) . in these papers
a simple phenomenological ginzburg - landau model for the coexistence of magnetism and superconductivity in ho@xmath166dy@xmath167ni@xmath1b@xmath1c has been proposed .
their ni derived
band corresponds to our isotropic fss and their non - ni derived band corresponds to our fss s ( which however do contain also ni derived states ! ) where the magnetism quenches the superconductivity due to an admixture of ho(dy ) 5@xmath34 states @xcite . within this model below 4k far from the fluctuation region near 5k for honi@xmath1b@xmath1c , where the two magnetic phases still compete , the sc states of both compounds do almost coincide ( see fig . 3 of ref .
this and the reported pressure dependencies are in qualitative accord with our microscopic picture leading to rather similar low - temperature sc states . however , we strongly differ in the discription of the magnetically quenched band complex and its coupling to the `` surviving '' one . in our opinion
, for the former the intrinsic el - ph phonon coupling is not weak but the interband coupling should be very weak by reasons of symmetry , at least in the pure unmixed limiting cases whithout a possible disorder induced locally broken symmetry . only this way the well - known strong coupling properties of nonmagnetic borocarbides can be reproduced in switching off the magnetic couplings .
finally , we note that both scenarios should be distinguished from another phenomenological scenario proposed in ref . which is based on the assumtion that the maximal reduction of the anisotropic spin - flip scattering achieved in the antiferromagnetically ordered state or equivalently @xmath168 should be the necessary condition for the coexistence of local moment magnetism and superconductivity .
a more detailed comparison and critical discussion of all three approaches will be given elsewhere .
i. k. yanson and yu .
g. naidyuk , fiz .
nizk . temp . *
30 * , 355 ( 2004 ) [ low temp . phys . *
30 * , 261 ( 2004 ) ] ; yu .
g. naidyuk , o. e. kvitnitskaya , i. k. yanson , s. lee , and s. tajima , solid state commun . * 133 * , 363 ( 2005 ) . | point - contact ( pc ) spectroscopy measurements on antiferromagnetic ( af ) ( @xmath05.2k ) honi@xmath1b@xmath1c single crystals in the normal and two different superconducting ( sc ) states ( @xmath28.5k and @xmath3k ) are reported .
the pc study of the electron - boson(phonon ) interaction ( eb(p)i ) spectral function reveals pronounced phonon maxima at 16 , 22 and 34mev .
for the first time the high energy maxima at about 50mev and 100mev are resolved .
additionally , an admixture of a crystalline - electric - field ( cef ) excitations with a maximum near 10mev and a ` magnetic ` peak near 3mev are observed .
the contribution of the 10-mev peak in pc epi constant @xmath4 is evaluated as 20 - 30% , while contribution of the high energy modes at 50 and 100mev amounts about 10% for each maxima , so the superconductivity might be affected by cef excitations .
the sc gap in honi@xmath1b@xmath1c exhibits a standard single - band bcs - like dependence , but vanishes at @xmath3k@xmath5 , with @xmath6k@xmath7 .
the strong coupling eliashberg analysis of the low - temperature sc phase with @xmath3 k @xmath8 @xmath9 , coexisting with the commensurate af structure , suggests a sizable value of the epi constant @xmath10 0.93 .
we also provide strong support for the recently proposed by us ' fermi surface ( fs ) separation' scenario for the coexistence of magnetism and superconductivity in magnetic borocarbides , namely , that the superconductivity in the commensurate af phase survives at a special ( nearly isotropic ) fs sheet without an admixture of ho 5d states . above @xmath11
the sc features in the pc characteristics are strongly suppressed pointing to a specific weakened sc state between @xmath11 and @xmath12 . |
the last decade has witnessed an explosive growth of high - quality data for supernovae both from the space and ground observatories with spectacular results , and new perspectives for the use of sne ia as cosmological yard sticks and for constraining the physics of supernovae .
one of the most important new developments in observational supernova research was to establish the long - suspected correlation between the peak brightness of sne ia and their rate of decline , @xmath1 , by means of modern ccd photometry ( phillips 1993 ) .
sne ia have provided new estimates for the value of the hubble constant ( @xmath10 ) based on a purely empirical procedure ( hamuy et al . 1996ab , riess , press & kirshner , 1996 ) , and on a comparison of detailed theoretical models with observations ( hflich & khokhlov 1996 , hereafter hk96 ; nugent et al .
the values obtained are in good agreement with one another .
more recently , the routine successful detection of supernovae at large redshifts , z ( e.g. perlmutter et al .
1995 , 1997 ; riess et al . 1998
; garnavich et al .
1998 ) , has provided an exciting new tool to probe cosmology .
this work has provided results that are consistent with a low matter density in the universe and , most intriguing of all , yielded hints for a positive cosmological constant @xmath11 of @xmath12 .
it is worth noting that the differences in the maximum magnitude between @xmath11=0 and 0.7 is @xmath13 for redshifts between 0.5 to 0.8 .
these results prompted the quest for the nature of the the dark energy , i.e. cosmological equation of state .
current candidates include a network of topological defects such as strings , evolving scalar fields ( i.e. quintessence ) , or the classical cosmological constant .
for a recent review , see ostriker & steinhardt ( 2001 ) , and perlmutter , turner & white ( 1999b ) . to separate between the candidates by sne ia , the required accuracy has to be better than @xmath14 to @xmath15 ( albrecht & weller 2000 ) .
the results on @xmath16 and future projects to measure the cosmological equation of state depend on the empirical @xmath1 which is calibrated locally .
this leaves systematic effects as the main source of concern .
indeed , there is already some evidence that sne ia undergo evolution .
it has been argued , that the local sn ia sample covers all the possible variations that may come from different progenitors , different explosion mechanisms and environments , etc .. for local sne ia , the observational and statistical characteristics depend on their environment .
they occur less often in ellipticals than in spirals , and the mean peak brightness is dimmer in ellipticals ( branch et al .
1996 ; wang , hflich & wheeler 1997 ; hamuy et al .
2000 ) . in the outer part of spirals
the brightness is similar to ellipticals while , in more central regions , both intrinsically brighter and dimmer sne ia occur ( wang et al .
these dependencies show us that sne ia likely depend on the underlying population and may undergo evolution .
if the evolution realizes then we have to know it and take it into account going back in time .
otherwise we can not _ safely _ use a local calibration . in principle
, more distant sample could come from younger and more metal - poor progenitors , or the dominant explosion scenario may change .
there is general agreement that sne ia result from some process of combustion of a degenerate white dwarf ( hoyle & fowler 1960 ) . within this general picture ,
three classes of models have been considered : ( 1 ) an explosion of a co - wd , with mass close to the chandrasekhar mass , which accretes mass through roche - lobe overflow from an evolved companion star ( whelan & iben 1973 ) . the explosion is then triggered by compressional heating near the wd center .
( 2 ) an explosion of a rotating configuration formed from the merging of two low - mass wds , caused by the loss of angular momentum due to gravitational radiation ( webbink 1984 , iben & tutukov 1984 , paczyski 1985 ) . ( 3 ) explosion of a low mass co - wd triggered by the detonation of a helium layer ( nomoto 1980 , woosley et al .
1980 , woosley & weaver 1986 ) .
only the first two models appear to be viable .
the third , the sub - chandrasekhar wd model , has been ruled out on the basis of predicted light curves and spectra ( hflich et al . 1996b , nugent et al .
1997 ) .
delayed detonation ( dd ) models ( khokhlov 1991 , woosley & weaver 1994 , yamaoka et al .
1992 ) have been found to reproduce the optical and infrared light curves and spectra of typical sne ia reasonably well ( hflich 1995 , hereafter h95 ; hflich , khokhlov & wheeler 1995 , hereafter hkw95 ; hk96 ; fisher et al .
1998 ; nugent et al . 1997 ;
wheeler et al . 1998 ; hflich et al .
2000 ; lentz et al . 2001 ; gerardy et al .
this model assumes that burning starts as subsonic deflagration and then turns to a supersonic , detonative mode of burning . due to the one - dimensional nature of the model ,
the speed of the subsonic deflagration and the moment of the transition to a detonation are free parameters .
the moment of deflagration - to - detonation transition ( ddt ) is conveniently parameterized by introducing the transition density , @xmath17 , at which ddt happens .
the amount of @xmath18ni , @xmath19 , depends primarily on @xmath20 ( h95 ; hkw95 ; umeda et al .
1999 ) , and to a much lesser extent on the assumed value of the deflagration speed , initial central density of the wd , and initial chemical composition ( ratio of carbon to oxygen ) .
models with smaller transition density give less nickel and hence both lower peak luminosity and lower temperatures ( hkw95 , umeda et al .
in dds , almost the entire wd is burned , i.e. the total production of nuclear energy is almost constant .
this and the dominance of @xmath20 for the @xmath21 production are the basis of why , to first approximation , sne ia appear to be a one - parameter family .
the observed @xmath1 can be well understood as a opacity effect ( hflich et al .
1996b ) , namely , as a consequence of the rapidly dropping opacity at low temperatures ( hflich , khokhlov & mller 1993 , khokhlov , mller & hflich 1993 ) . less
ni means lower temperature and , consequently , reduced mean opacities because the emissivity is shifted from the uv towards longer wavelengths with less line blocking .
a more rapidly decreasing photosphere causes a faster release of the stored energy and , as a consequence , steeper declining lcs with decreasing brightness .
the dd models thus give a natural and physically well - motivated origin of the @xmath1 relation of sne ia within the paradigm of thermonuclear combustion of chandrasekhar - mass co white dwarfs .
nonetheless , variations of the other parameters lead to some deviation from the @xmath1 . e.g. a change of the central density results in an increased binding energy of the wd and a higher fraction of electron capture close to the center which reduce the @xmath21 production ( hflich et al .
1996b ) . because dd - models allow to reproduce the observations , we use this scenario to test the influence of the underlying stellar population on the explosion .
we note that detailed analyses of observed spectra and light curves indicate that mergers and deflagration models such as w7 may contribute to the supernovae population ( hflich & khokhlov 1996 , hatano et al .
in particular , the classical `` deflagration '' model w7 with its structure similar to dd models has been successfully applied to reproduce optical light curves and spectra ( e.g harkness , 1987 ) .
the evidence against pure deflagration models for the majority of sneia includes ir - spectra which show signs of explosive carbon burning at high expansion velocities ( e.g. wheeler et al .
1998 ) and recent calculations for 3-d deflagration fronts by khokhlov ( 2001 ) which predict the presence of unburned and partial burned material down to the central regions .
currently , pure deflagration models may be disfavored for the majority of sneia but , clearly , they can not be ruled out either .
previously , hflich , wheeler & thielemann ( 1998 , hereafter hwt98 ) studied evolutionary effects induced by the progenitor .
they calculate differences in the lc and nlte - spectra as a function of parameterized values of the integrated c / o ratio @xmath22 and metallicity of the exploding wd .
this study showed that a change of @xmath22 alters the energetics of the explosion which results in an off - set of the brightness - decline relation .
most prominently , this effect can be identified by a change in the fiducial rise - time to decline relation @xmath23 .
the offset in @xmath1 is given by @xmath24 where @xmath25 is the dispersion in the rise time of the fiducial light curve .
aldering et al .
( 2000 ) showed that @xmath23 are identical within @xmath26 for the local and distant sample lending strong support for the notion that we need a positive @xmath16 .
a change in the metallicity z causes a change in the burning conditions at the outer layers of the wd and it alters the importance of the line blanketing in the blue to the uv . based on detailed calculations , effects of similar order have been found for both the delayed dd and the deflagration scenario ( hwt98 , lentz et al .
recent studies showed the additional effect that z will influence the final structure of the progenitor and the resulting lcs ( umeda et al .
1999 , domnguez et al .
2000 , hflich et al .
however , the former two studies were restricted to the progenitor evolution whereas the latter included the connection between the progenitor and the lc but it was restricted to a progenitor of @xmath27 and two metallicities , z=0.02 and 0.004
. a more comprehensive study may be useful to eliminate potential problems due to evolution of the progenitors for the determination of the cosmological equation of state , and it may provide a direct link to the progenitors of sn ia . in this work
we connect @xmath0 and and the initial metallicity of the wd to the light curves and spectral properties of sne ia for the entire range of potential progenitors . in section 2 we discuss the evolutionary properties of our models . in section 3 ,
the results are presented for the explosion , nucleosynthesis , the light curves and spectral properties . in the final ,
concluding section , our model calculations are related to observations , and we discuss constraints for the progenitors and implications for the cosmological equation of state .
co white dwarfs are the remnants of the evolution of low and intermediate mass stars ( becker & iben 1980 ) .
their progenitors are stars less massive than @xmath28 , which is the lower stellar mass for which a degenerate carbon ignition occurs after the central helium exhaustion .
the precise value of @xmath28 depends on the chemical composition ( see domnguez et al . , 1999 , for a recent evaluation of this mass limit ) .
it ranges between 6.5 and 8 @xmath2 . on the base of updated theoretical stellar models of intermediate mass stars , domnguez et al ( 1999 ) found final co core masses in the range 0.55 - 1.04 @xmath2 , in good agreement with semi - empirical evaluations of the wd masses ( see e.g. weidemann 1987 ) . in this paper
, we use co wd structures obtained by evolving models with main sequence masses @xmath0 between 1.5 and 7 @xmath2 and metallicities z between @xmath29 and 0.02 . in the following
a label identifies a particular progenitor model , namely @xmath30 for a progenitor with a ms mass of @xmath31 and z= @xmath32 . ]
these models have been obtained by means of the frascati raphson - newton evolutionary code ( franec ) , which solves the full set of equations describing both the physical and chemical evolution of a star by assuming hydrostatic and thermal equilibrium and a spherical geometry ( chieffi & straniero 1989 , chieffi , limongi & straniero , 1998 ) . for a detailed description of the adopted input physics
see straniero , chieffi & limongi ( 1997 ) and domnguez et al ( 1999 ) . because we are interested in the final chemical structure of a co wd , let us recall the main properties and the major uncertainties of the evolutionary phases during which the co core forms . in table 1
we show some properties of our models . in columns
1 to 9 we give : ( 1 ) the initial composition ( z and y ) , ( 2 ) the model name , ( 3 ) the main sequence mass ( in @xmath2 ) , ( 4 ) the mass of the co core at the beginning of the tp phase ( in @xmath2 ) , ( 5 ) the c abundance ( mass fraction ) at the center , ( 6 ) the mass ( in @xmath2 ) of the homogeneous carbon - depleted central region , ( 7 ) the averaged c / o ratio within the final , @xmath331.37 @xmath2 , co white dwarf after accretion , and ( 8) the @xmath18ni mass ( in in @xmath2 ) synthesized during the explosion .
the c and o profiles ( mass fraction ) of the co core for selected thermally pulsing models are shown in figs .
[ z22 ] to [ corate ] .
in particular , fig .
[ z22 ] shows the changes induced by a different initial mass , while figs .
[ m3 ] and [ m5 ] illustrate the effect of the metallicity .
the internal c and o profiles of a wd are generated in three different evolutionary phases of the progenitor , namely : i ) the central he burning , ii ) shell he - burning during the early asymptotic - giant - branch ( agb ) phase , and iii ) shell he - burning during the thermally pulsing agb phase . as illustrated in the figures , they produce three distinct layers .
the central he - burning produces the innermost homogeneous layer .
this phase is initially dominated by the carbon production via the 3@xmath34 reaction occurring in the center of a convective core .
once sufficient @xmath35c is synthesized , the @xmath36 reactions becomes competitive with 3@xmath34 .
carbon is partially burned into @xmath37o .
since the opacity of a c - o mixture is larger than that of a he mixture , the extension ( in mass ) of the convective core increases with time .
when the he mass fraction in the convective core is reduced down to @xmath330.1 , the he - burning is mainly controlled by @xmath36 , and most of the oxygen in the convective core is synthesized during the late he - burning .
the final abundances in the innermost region of a wd is strongly dependent on the duration of the last @xmath38 of the entire he - burning lifetime . in column 7 of table 1 , we report the size ( in solar masses ) of this innermost homogeneous region , which corresponds to the maximum extension of the convective core .
the intermediate region of the final c / o structure is characterized by a rising carbon abundance .
it is produced during the early - agb when the he - burning shell advances in mass until it approaches the h - rich envelope .
the amount of carbon ( oxygen ) left behind increases ( decreases ) due to the progressive growth of the temperature in the shell which favors the 3@xmath34 reactions with respect to the @xmath34 capture on @xmath35c .
in addition , the short lifetime does not allow a substantial conversion of carbon into oxygen .
finally , a thin external layer is built up during the thermally pulsing agb . at the beginning of a thermal pulse ,
the large energy flux is locally produced by the 3@xmath34 reactions .
it induces the formation of a convective shell that rapidly overlaps the whole inter - shell region . owing to the large he reservoir
, a huge amount of carbon is produced at the base of the convective shell .
after few years ( 10 - 100 yr depending on the core mass ) the convective shell disappears and a quiescent he - burning takes place .
it is during this longer phase that the @xmath36 reactions convert a certain part of the carbon produced during the pulse into oxygen .
the c / o ratio left below the he - rich layer by the he burning - shell depends on the rate of the @xmath34 captures on carbon .
note that the outer blip in the carbon profile is the result of the last thermal pulse where the he has not yet fully depleted .
the size , in mass , of this third layer depends on the duration of the thermally pulsing agb phase .
although it is influenced by the assumed mass loss rate , it is generally believed that for m@xmath393 m@xmath40 the co core can not increase more than 0.1 - 0.2 m@xmath40 during the agb .
an even smaller increase of the co core is expected for larger @xmath0 .
the subsequent phase has been calculated by accreting h / he rich material on the resulting co wd .
note that we have assumed that the progenitor ejects its h / he - rich envelope prior to the onset of the accretion epoch .
the accreted matter has a final c / o ratio of @xmath331 . when the star reaches a mass close to 1.37 m@xmath40 , ignition occurs close to the center .
@xmath41 has been adjusted to enforce that the thermonuclear runaway occurs at the same central density @xmath42 in all models . in fig .
[ z22 ] , the chemical structures of our models are shown as a function of @xmath0 ( for z=0.02 ) . for low stellar masses , the core he - burning
happens under lower central temperatures ( see e.g. domnguez et al .
this favors the @xmath34 captures on @xmath35c , which are in competition with the 3@xmath34 resulting in a slightly smaller central c / o ratio for low @xmath0 .
however , the size of the region of central he - burning , @xmath43 , is increasing with @xmath0 . in a star with 7 m@xmath40 ,
the maximum size of the convective core is about 0.7 m@xmath40 while , in the 1.5 m@xmath40 model , it is only 0.25 m@xmath40 .
this is the dominating factor for changes in the mean c / o ratio which , in general , produces the monotonic relation that @xmath22 decreases with increasing @xmath0 ( table 1 ) . in figs .
[ m3 ] and [ m5 ] , the chemical structures are given for various z for stars with 3 and 5 @xmath2 , respectively .
the sizes of the innermost homogeneous region is not a monotonic function of z. this is due to the peculiarity of very low metallicity intermediate mass stars ( see chieffi et al .
2001 ) . in these stars ( z@xmath44 ) , the central hydrogen burning proceeds via the pp - chain instead of the cno - cycles , as it happen for larger metallicity .
it results in a smaller he - core ( ezer & cameron 1971 , tornamb & chieffi 1986 ) . for solar metallicity ,
the higher opacities and the steeper temperature gradients produce a smaller he core ( becker & iben 1979 , 1980 ) . in all cases
the dependence on z of the final averaged c / o ratio ( column 7 of table 1 ) is small compared to the effect of main sequence mass .
the variation with z ranges between 5 and 10@xmath45 .
for the final chemical structure , the most important uncertainty is due to the ambiguity in nuclear reaction rate of @xmath36 .
the innermost region is also sensitive to the treatment of turbulent convection which may affect the duration of the late central he - burning lifetime and the size of the convective core .
the rate of the @xmath36 at astrophysical energies is not well established ( see e.g. buchmman 1996 , 1997 ) .
the cross section around the gamow peak is dominated by ground state transitions through four different processes : the e1 amplitudes due to the low - energy tail of the @xmath46 resonance at @xmath47 and to the subthreshold resonance at @xmath48 , and the e2 amplitudes due to the @xmath49 subthreshold resonance at @xmath50 and to the direct capture to the @xmath51 ground state , both with the corresponding interference terms . besides ground state transitions ,
also cascades , mainly through the e2 direct capture to @xmath52 @xmath53 and @xmath54 @xmath55 states , have to be considered .
obviously a higher rate ( @xmath56 factor of 2 ) of this reaction reduces the carbon abundance left by both the core and the shell he - burning .
some indications in favor of an high value for this reaction rate comes from the rise times to maximum light in sne ia ( hwt98 ) , and from recent studies of pulsating wds ( metcalfe , winget & charboneeau 2001 ) .
concerning turbulent convective mixing , it only affects the region of the wd structure produced during the core he - burning .
the two major uncertainties are related to the possible existence of breathing pulses ( see castellani et al .
1985 , caputo et al . 1986 ) and the possible existence of a sizeable convective core overshoot . as recently pointed out by imbriani et al .
( 2001 ) , since breathing pulses increase the late core he - burning lifetime , they significantly reduce the central c / o . in this regard
, they mimic the effects of a high @xmath36 rate . on the contrary , convective core overshoot does not affect the central c / o , but it may enlarge the size of the convective core and , in turn , may produce a larger central , c - depleted region . in the models presented in this paper , we have neglected breathing pulses and convective core overshoot .
all the models , except the 3p0z13lr , have been obtained by means of a high rate for the @xmath36 reaction ( as given by caughlan et al . , 1985 ) .
the 3p0z13lr model has been obtained by adopting the alternative low rate presented by caughlan & fowler ( 1988 ) . a comparison between the two models with different @xmath36 reaction rate is shown in fig .
[ corate ] . changing the @xmath36 rate from the high to the low value
drastically alters the chemical profiles of the progenitor .
the carbon abundance increases by about a factor two . at the time of the explosion , the average composition of the wd changes from oxygen - rich ( @xmath57 ) to carbon - rich ( @xmath58 ) !
the consequences for the lcs are discussed below .
although a variation in the assumed convective mixing scheme may change the quantitative result of our analysis , the overall conclusions and tendencies can not be significantly altered because its influence is limited to the innermost part of the pre - explosive structure .
spherical dynamical explosions and corresponding light curves are calculated .
we consider delayed detonation ( dd ) models , because these have been found to reproduce the optical and infrared light curves and spectra of sne ia reasonably well ( hflich 1995b ; hkw95 ; hk96 ; nugent et al . 1997 ; wheeler et al .
1998 ; lentz et al .
2000 ; hflich et al .
2000 ; gerardy et al .
model parameters have been chosen which allow to reproduce light curves and spectra of typical type ia supernovae .
for our set of models , the differences can be attributed to changes in the progenitor structure of the co - wd . as reference
, we use the explosion of a progenitor with 5 @xmath2 at the main sequence and solar metallicity ( model 5p0z22 ) . at the time of the explosion of the wd , its central density is 2.0@xmath59 g/@xmath60 and
its mass is close to 1.37@xmath2 .
the transition density @xmath20 from deflagration to detonation is chosen to be 2.3@xmath61 g/@xmath60 .
explosion models are calculated using a one - dimensional radiation - hydro code ( hk96 ) that solves the hydrodynamical equations explicitly by the piecewise parabolic method ( colella & woodward 1984 ) .
nuclear burning is taken into account using an extended network of 606 isotopes from n , p to @xmath62 ( thielemann , nomoto & hashimoto 1996 and references therein ) .
the propagation of the nuclear burning front is given by the velocity of sound behind the burning front in the case of a detonation wave and in a parameterized form during the deflagration phase calibrated by detailed 3-d calculations ( e.g. khokhlov 1995 , 2000 ; niemeyer & hillebrandt 1995 ) .
we use the parameterization as described in domnguez & hflich ( 2000 ) . for a deflagration front at distance @xmath63 from the center , we assume that the burning velocity is given by @xmath64 , where @xmath65 and @xmath66 are the laminar and turbulent velocities with @xmath67}\ ] ] here @xmath68 is the atwood number , @xmath69 is the characteristic length scale , and @xmath70 and @xmath71 are the densities in front of and behind the burning front , respectively . the quantities @xmath34 and @xmath69
are directly taken from the hydro at the location of the burning front and we take @xmath72 .
the transition density is treated as a free parameter .
the description of the deflagration front does not significantly influence the final structure of the explosion ( domnguez & hflich 2000 ) .
the total @xmath18ni production is governed by the pre - expansion of the wd and , consequently , is determined by the transition density @xmath20 , at which the burning front switches from the deflagration to the detonation mode ( h95 ) . from the physical point of view , @xmath20 should be regarded as a convenient way to adjust the amount of material burned during the deflagration phase .
the value @xmath20 can be adjusted to produce a given amount of @xmath21 .
this code includes the solution of the frequency - averaged radiation transport implicitly via moment equations , expansion opacities , and a detailed equation of state ( see sect .
as expected from previous studies ( see introduction ) , the overall density , velocity and chemical structures are found to be rather insensitive to the progenitor , including the production of elements .
although explosions and light curves have been calculated for the entire set of stellar cores , we will concentrate our detailed discussion on the extreme cases and the reference model .
results for intermediate models can be understood accordingly and interpolated using the quantities given in table 1 . the final density , velocity and chemical structures and detailed production of isotopes are shown in figs .
[ rho ] , [ chem ] and [ isotope ] for the extreme cases in metallicity ( z=0.02 and z=10@xmath73 , 5 m@xmath40 ) and the extremes in @xmath0 ( 1.5 m@xmath40 to 7 m@xmath40 , z=0.02 ) . between 0.511 to 0.589 @xmath74 of @xmath21 are produced ( table 1 ) .
the production of individual isotopes varies only by about 10 % ( fig .
[ isotope ] ) . for the reference model @xmath75 ,
the final element abundances are given in table 2 .
variations in the final density and velocity structure are correspondingly small ( fig . [ rho ] ) . in delayed detonations , almost the entire wd is burned .
the total release of nuclear energy depends mainly on the fuel , i.e. on the integrated c / o ratio @xmath22 ( hwt98 ) . however , as usual for delayed detonation models , the deflagration phase is key for our understanding of the final results . during the deflagration phase ,
about @xmath76 of fuel are burned in our models ( figs .
[ rho ] & [ chem ] ) . in all explosions but the progenitors with @xmath77 with z=0.02 and @xmath78 with z=0.001
, the deflagration front will propagate in the carbon - depleted layers .
the amount of total energy produced during the deflagration phase and the binding energy of the progenitor determines the pre - expansion of the outer layers and , consequently , the overall chemical structure .
the binding energy of the wd is dominated by the central density @xmath42 at the time of the explosion .
note that the c / o ratio has little influence on the structure of the wd because the pressure is dominated by degenerate electrons and the electron to nucleon ratio @xmath79 is identical for @xmath80 and @xmath81 .
the total energy production during the deflagration phase is governed by @xmath20 and by the nuclear energy release per gram , i.e. the composition .
both @xmath82 and @xmath20 have been kept the same in all models .
variations can be understood by the change of the mean c / o ratio and the mass @xmath83 of the central , carbon depleted region .
the latter influences the temperature and , consequently , the laminar speed and the atwood number ( eq .
1 ) . at the central layers ,
all the material is burned up to iron - group elements ( fig .
[ chem ] ) . some additional variation in the total @xmath21-mass
is caused by the c / o ratio of the matter burned during the detonation phase .
for all models , the transition between @xmath21- and si - rich layers is between 0.58 and 0.98 @xmath2 . for @xmath84 and , to a much lesser extend , for @xmath85 , this transition region overlaps with layers of lower c - depleted ( fig .
1 ) . from the nuclear physics ,
the transition between complete and incomplete si burning occurs in a narrow temperature range around @xmath86 . a locally lower c / o ratio results in lower burning temperatures .
consequently , the @xmath87/si boundary is shifted inwards and @xmath19 is reduced .
overall , differences between the models remain small because the nuclear energy production for c- and oxygen burning differs only by @xmath88 .
an increasing @xmath0 for a given metallicity leads to changes in @xmath89 and @xmath22 by 33% and -25% , respectively ( table 1 ) .
increasing @xmath0 from 1.5 to 7.0 @xmath2 alters the expansion velocity of a given mass element ( fig .
[ rho ] , right panel ) and it results in a shift of the chemical interfaces between complete , incomplete si and explosive c burning by @xmath90 ( fig .
[ chem ] ) .
if we decrease the metallicity from solar ( @xmath91 ) to @xmath92 for stars with @xmath93 , @xmath89 and @xmath22 change by as little as + 10 % and -7 % , respectively .
this means that the density and velocity structure of the chemical structures are virtually indistinguishable ( fig .
[ rho ] ) . for stars with @xmath0 @xmath94 , variations in @xmath89
increase up to 30 % but , still , variations in @xmath22 remain at a level of 10 % .
the overall energetics , density and velocity structure remain mostly unchanged but the pre - expansion and , consequently , the chemical interfaces between different regimes of burning change by @xmath95 . for all @xmath0 ,
the most noticeable difference is the increasing @xmath96 production with z in the layers of incomplete si burning which changes the spectra in the blue , and in the uv ( see hwt98 and below ) . as mentioned in sect .
2.3 , there is some indirect evidence for a high cross section of this key reaction but a low rate can not be ruled out either .
the consequences of a low rate are strong .
e.g. , for a progenitor with @xmath98 and @xmath99 , the low rate suggested by caughlan & fowler ( 1988 ) will increase @xmath89 and @xmath22 from 0.26 to 0.51 and 0.74 to 1.22 when compared to our favorite rate ( caughlan et al .
the explosion becomes more energetic by about 20 % and the deflagration front propagates faster .
the result is an increase of the @xmath100 production by about 10 % and a shift in the chemical interfaces by about + 2500 km / sec .
based on the explosion models , the subsequent expansion , bolometric and monochromatic light curves are calculated ( hflich et al .
1998 , and references therein ) .
the lc - code is the same used for the explosion , except that @xmath101 ray transport is included via a monte carlo scheme and nuclear burning is neglected . in order to allow a more consistent treatment of the expansion
, we solve the time dependent , frequency averaged radiation moment equations . the frequency - averaged variable eddington factors and mean opacities are calculated by solving the frequency - dependent transport equations .
about one thousand frequencies ( in one hundred frequency groups ) and about nine hundred depth points are used . at each time step
, we use t(r ) to determine the eddington factors and mean opacities by solving the frequency dependent radiation transport equation in a comoving frame and integrate to obtain the frequency averaged quantities .
the averaged opacities have been calculated assuming local thermodynamics equilibrium ( lte ) .
both , the monochromatic and mean opacities are calculated in the narrow line limit .
scattering , photon redistribution and thermalization terms used in the light curve opacity calculations are taken into account . in previous works ,
the photon redistribution and thermalization terms have been calibrated for a sample of spectra using the formalism of the equivalent two level approach ( h95 ) . here ,
for increased consistency , we use the same equations and atomic models but solve the rate equations simultaneously with the light curves calculation at about every 100@xmath102 time step , on the expense of some simplifications in the nlte - part compared to h95 . for the opacities , we use the narrow line limit and for the radiation fields , we use the solution of the monochromatic radiation transport using @xmath103 frequency groups . both the old and new approach are about equivalent in accuracy with consistent results .
most noticeable , now , b - v is bluer by about @xmath104 . in the following discussion
is based on the same set of models used in the previous section . in figs .
[ lc ] and [ dep ] , we show the b and v light curves and some quantities at maximum light .
overall the different phases of light curves can be understood in the usual way including the bump at about day 35 which can be attributed to the change in the opacities between the layers of complete and incomplete burning ( domnguez 1991 , 1994 ) . for the reference model 5p0z22 , a maximum brightness @xmath105 of @xmath106
is reached at about 18.25 days after the explosion .
the color index @xmath107 is @xmath108 . as discussed in the introduction , the amount of @xmath18ni , its distribution and the expansion rate of the envelope are the dominant factors which determine the absolute magnitude at maximum and the light curve shape . with all model parameters fixed but
the progenitor mass and the metallicity , the differences of the light curves can be understood based on the previous discussion of the explosion models . by increasing @xmath0 from 1.5 to 7.0 @xmath2 for z=0.02 , both @xmath109 and @xmath105 decrease by @xmath110 consistent with a change in the @xmath21 mass by 14 % ( fig .
[ lc ] , upper panel , and fig .
[ dep ] ) .
the similarity in the density and velocity structures produces almost identical conditions at the photosphere .
b - v is insensitive to a change in @xmath0 ( @xmath111 \leq 0.01^m$ ] ) .
relative to the reference model , the rise times vary between -0.5 d ( 1p5z22 ) to + 1.2d ( 7p0z22 ) .
the decline rate @xmath4 is hardly affected .
a change in @xmath0 will result in an offset / dispersion in @xmath1 by up to @xmath112 .
interestingly , the fluxes on the radioactive tail are much more similar than could be expected from the spread in the @xmath21 masses by 14 % ( fig .
[ lc ] ) . the change in @xmath19 is almost compensated by the differences in the energy deposition of @xmath113-rays from the radioactive decay . in fig .
[ gam ] , the escape probability for hard radiation is shown as a function of time .
a significant fraction of @xmath113 photons is thermalized up to about 150 to 200 days .
the actual value of thermalization depends on the expansion rate which is decreasing with mass ( see above ) .
e.g. the fraction of thermalized @xmath113 photons for the models 1p5z22 , 5p0z22 and 7p0z22 are 24.2 % , 25% and 27 % , respectively .
the increase in the efficiency for the thermalization amounts to 11 % over the mass range and almost compensate the decrease in the @xmath21 mass .
note that the ratio between maximum and tail brightness is decreasing with @xmath105 .
this effect is opposite to the observed @xmath1 relation ( e.g. hamuy et al .
if realized in nature , a wide range in @xmath0 would increase the dispersion in @xmath114 by about @xmath112 .
the presence of this effect would reveal itself by an additional change in the expansion velocity measured by the doppler shift of lines .
e.g. at maximum light , weak lines would indicate an expansion velocity at the photosphere which is smaller by @xmath115 if we compare model 7p0z22 vs. 1p5z22 .
the discussion above applies to all metallicities because @xmath89 and @xmath22 vary in a similar range . for progenitors with @xmath116 ,
a change in the metallicity has little effects on the energetics and , consequently , on the light curves ( fig .
[ lc ] , middle panel ) .
the most important effect is the systematic decline of b - v by @xmath117 when z is changed from 0.02 to @xmath29 .
this effect can be attributed to a change of the line blending by fe in the decoupling region of photons , i.e. the atmosphere ( hwt98 , lentz et al . 2000 ) . in b and
v at maximum light , opacities are dominated by electron scattering ( hkm93 ) but iron lines are more important in b compared to v. consequently , a lower metallicity results in increase of the flux ratio between b and v. in the u - band and the uv , the opacities are dominated by lines .
a change in the line blending will cause both a change of the radius of the flux formation and the specific flux .
therefore , a decrease in z may result in either an in- or decrease of the monochromatic flux depending on the density structure .
these findings apply to all ms - masses in our sample . to some extend ,
the exception are the models with @xmath118 for which the central carbon concentration varies with metallicity and produces a change in @xmath19 by 3 % and a corresponding change in @xmath109 and @xmath105 . a low nuclear rate @xmath119 increases the explosion energy compared to our standard rate ( 3p0z13lr vs. 3p0z13 , see fig .
[ lc ] , lower panel ) .
the rise time in v is reduced by 2.7 days ( 15.3d for 3p0z13lr vs. 18.0d for 3p0z13 ) . the enhanced escape probability for @xmath113-rays ( fig .
[ gam ] ) explains the remaining differences including the increased maximum brightness to tail ratio and the moderate increase of @xmath105 and @xmath109 .
these results are consistent with previous findings which identified the importance of the c / o ratio for the change in the rise time of typical sne ia ( hwt98 ) .
using a delayed detonation model and realistic structures for the exploding white dwarf , we have studied the influence of the progenitor star on the light curves and spectral properties of type ia supernovae .
* stellar models : * we considered stars between 1.5 to 7 @xmath2 and metallicities between @xmath91 ( solar ) to @xmath120 which covers the full range of potential progenitors .
the progenitor structures are based on detailed calculations for the stellar evolution starting at the pre - main sequence up to the thermal pulses when most of the stellar envelope is ejected and a white dwarf is formed with a mass between 0.5 and 1.0 @xmath74 .
its size increases with @xmath0 and , to a lesser extend , changes with the metallicity . the subsequent accretion and burning at the surface of the wd
let it grow to @xmath121 . as a final chemical structure
, the wd shows a central region of reduced c abundance between 0.21 to 0.32 originating from the convective he - burning , a layer of increased c abundance from the he - shell burning , and a layer originating for the accretion phase .
the mean c / o ratio decreases by about 30 % over the entire mass range .
the sensitivity on the metallicity is much weaker ( @xmath122 ) , and not monotonic .
* supernovae : * our study of sne ia is based on delayed detonation models because they have been found to reproduce the monochromatic light curves and and spectra of sne ia reasonably well including the brightness decline relation @xmath1 .
deviation from a perfect relation are due to variations in the central density , properties of the deflagration front , and the progenitor structure .
all parameters but the progenitors have been fixed to produce lcs and spectra typical for normal sne ia . in this work ,
rise times to maximum light are between 17.7 to 19.4 days , @xmath123 to @xmath124 , and @xmath125 to @xmath126 .
differences between the models and light curves remain small because the nuclear energy production by burning carbon and oxygen to iron - group elements differs by as little as @xmath88 .
the change of @xmath0 is the decisive factor to change the energetics .
the @xmath87 production varies by about 14 % and the velocities of the various chemical layers differ by up to 1500 km / sec .
a change in the metallicity hardly affects the overall structure of the progenitor . as already discussed in detail in hwt98 ,
its main effect is a change in the production of @xmath127 in the outer layers of incomplete si burning .
@xmath0 alters the @xmath1 relation which may be offset by up to @xmath5 .
in addition , @xmath0 changes the flux ratio between maximum light and the radioactive tail , and it alters the photospheric expansion velocities @xmath128 measured by the doppler shift of lines . e.g. a change in @xmath129 by @xmath5 is coupled to a decrease in @xmath130 at maximum light by @xmath131 .
note that a change in the central density @xmath42 of the wd has a similar effect on @xmath129 but with the opposite sign for @xmath132 ( hflich , 2001 ) . in principle
, this allows to decide whether differences in @xmath129 between sn with similar @xmath1 are related to a change in the progenitor or the central density at the thermonuclear runaway which is sensitive to the accretion rate .
in contrast to @xmath0 , the metallicity z hardly changes the light curve shapes ( @xmath133 ) .
it alters the line blocking by iron group elements at the photosphere mainly in the uv , u and b but hardly in v ( hwt98 ) . in the models presented here ,
b - v becomes systematically bluer with decreasing z ( up to @xmath134 ) . because b - v is the basic color index used to correct for interstellar extinction
, the metallicity effect can systematically alter the estimates for the absolute brightness by up to @xmath5 .
* @xmath135 : * at the example of a progenitor with @xmath136 and z=0.001 , we have tested the influence of the low nuclear rate @xmath119 on the outcome . using the lower rate suggested by caughlan & fowler ( 1988 ) instead caughlan et al .
( 1985 ) results in more energetic explosions because @xmath22 increases by a factor of @xmath137 .
the rise times to maximum light are 15.3d instead of 18.0d . from detailed observations of nearby supernovae ,
riess et al . ( 1999 ) find the following relation between the rise time @xmath138 and the maximum brightness @xmath139}\ ] ] the theoretical lcs peak at @xmath140 and @xmath141 for @xmath142 and @xmath143 , respectively . from the empirical fit of riess
, we would expect a rise time between 17.6 and 18.4 days which favors the high rate @xmath119 of caughlan et al .
note that the uncertainties in absolute values for the rise times are @xmath56 1 to 2 days ( hwt98 ) . * constraints on the progenitor : * empirically , the @xmath1 has been well established with a rather small statistical error @xmath144 ( @xmath145 : riess et al .
1996 ; @xmath146 : schmidt et al .
( 1998 ) ; @xmath147 phillips 1999 ; @xmath148 riess et al .
1999 ; @xmath149 perlmutter et al .
1999a ) . from theoretical models , a spread of 0.3 to @xmath150 can be expected ( hflich et al . 1996 ) .
this may imply a correlation between free model parameters , namely the properties of the progenitors , the central density or the transition density @xmath20 . in this study
, we find a spread in @xmath1 of about @xmath5 for progenitors with @xmath0 between @xmath151 to @xmath152 .
this may suggest a more narrow range in @xmath0 for realistic progenitors . from the fiducial rise time ,
progenitors with @xmath0 @xmath153 are favored .
this number should be regarded as a hint because uncertainties in the lc models .
another constraint can be obtained from the observed spread in the rise - time to decline relation .
riess et al . ( 1999 ) find a spread in @xmath138 of @xmath154 days whereas our models show a spread of @xmath56 1.7 days for @xmath155 . to be consistent with the observations , the range of main sequence masses has to be reduced by a factor of two .
this range is an upper limit because additional variations in the population of sneia such as explosion scenarios or the central density of the wd at the time of ignition will likely result in lower correlations .
* the cosmological equation of state : * in the following , we want to discuss our results in context of sne ia as probes for cosmology and for the determination of the cosmological equation of state .
we limit the discussion to the effects due to different progenitors . for a discussion of other systematic effects such as grey dust , gravitational lensing , the influence of a change in the importance of different possible scenarios ( e.g merger vs. @xmath121 models ) etc .
we want to refer to the growing literature in this field ( e.g. schmidt et al .
1998 , hwt98 , perlmutter et al .
1999a ) .
recently , there has been strong evidence for a positive cosmological constant ( e.g. perlmutter et al .
1999a , riess et al .
this evidence is based on observations that sne ia in the redshift range between 0.5 to 1.2 appear to be dimmer by about @xmath156 for redshifts between 0.5 to 0.8 which is comparable to the variations produced by different progenitors . however , both the internal spread in @xmath1 ( see above ) and the similarity in @xmath1 between the local sne ia and the high - z sample ( @xmath157 , aldering et al . 2000 ) limit the likely range of models and , consequently , evolutionary effects to @xmath158 up to redshifts of @xmath159 .
in addition , we do not expect a drastic change in the metallicity between local and supernovae at @xmath160 .
taking the linear dependence of b - v on the metallicity ( fig .
[ dep ] ) and realistic ranges for z , also reddening of non - grey dust will not change the conclusion on the need for a positive cosmological constant . as discussed in the introduction ,
the quest for the nature of the dark energy is one of the central questions to be addressed in future ( e.g. white 1998 , perlmutter et al .
1999b , albrecht & weller 2000 , ostriker & steinhardt 2001 ) . for @xmath161
, we can expect both very low metallicities and a significant change of the typical @xmath0 .
from this study , systematic effects due to different progenitors are limited to @xmath162 .
therefore , without further corrections for the progenitor evolution , some of the alternatives for the nature of the dark energy may be distinguished without correction for evolution . however , for a more detailed analysis , an accuracy of about @xmath163 ( weller & albrecht 2001 ) is required .
in this paper , we have shown how a combination of spectral and lc data or different characteristics of the lc can help to achieve this goal .
* limitations : * finally , we also want to mention the limitations of this study . qualitatively , our results on the z dependence agree with a previous study ( hflich et al .
2000 ) which was based on a progenitor with @xmath164 calculated by nomoto s group ( umeda et al .
the relation between @xmath22 and the offset in the @xmath1 relation has been confirmed .
however , the influence of z on @xmath22 was found to be about twice as large , and the central c concentrations are systematically larger .
the differences point towards a general problem .
the details of the central structure and evolution in the convective he burning core depend sensitively on the treatment of convection , semi - convection , overshooting , and breathing pulses ( lattanzio 1991 , schaller et al .
1991 , bressan et al .
1993 , vassiliadis & wood 1993 ) . for a detailed discussion ,
see domnguez et al .
in particular , the central c concentration may vary between 0.1 and 0.5 .
we want to note , that our value is consistent with direct measurements of the central c / o ratio found by the analysis of pulsational modes of wds ( metcalfe et al .
these uncertainties will affect the efficiency to separate the contribution of @xmath0 and z , respectively , i.e. the reason for @xmath22 but not the observable relations for lcs and spectra .
we provide limits on the size of evolutionary effects due to @xmath0 and metallicity of the progenitor ( @xmath9 ) .
other evolutionary effects may be due to a systematic change in the dominant progenitor scenario ( e.g. mergers vs. single degenerate ) or the typical separation in binary systems which contribute to the sne ia at a given time .
the latter may alter the accretion rates and , consequently , the central densities of the wd at the time of the thermonuclear runaway .
we did not consider the effect of the progenitor and of its pre - conditioning just prior to the explosion on the propagation of the burning front and , in particular , on the deflagration to detonation transition or , alternatively , the phase of transition from a slow to a very fast deflagration ( hillebrandt , 1999 private communication ) .
although the model parameters have been chosen to allow for a representation of typical " sne ia , more comprehensive studies and detailed fitting of actual observations are needed e.g. to detangle effects due to a change in the ignition density vs. the progentor .
in particular , observations of local sne ia have to be employed to narrow down and test for the proposed range of flux ratio between maximum and tail , and its relation to the expansion velocities .
* acknowledgments : * this work has been supported by nasa grant nag5 - 7937 , the murst italian grant cofin2000 , by the mec spanish grant pb96 - 1428 , by the andalusian grant fqm-108 and it is part of the italy - spain integrated action ( murst - mec agreement ) hi1998 . the calculations for the explosion and light curves were done on a beowulf - cluster financed by the john w. cox - fund of the department of astronomy at the university of texas .
albrecht a. , weller j. 2000 , american astronomical society , 197 , 61 aldering g. , knop , p. , nugent p. 2000 , 119 , 2110 becker s.a . , iben i. jr . 1979 , 232 , 831 becker s.a .
, iben i. jr . 1980 , 273 , 111 branch d. , romanishing w. , baron e. 1996 , 465 , 73 & e467 , 473 bressan a. , fagotto f. , bertelli g. , chiosi c. 1993 , a&as 100 , 647 buchmann l. 1996 , 468 , l127 buchmann l. 1997 , 479 , l153 caputo f. , castellani , v. , chieffi a. , pulone l. , tornamb a. 1989 , 340 , 241 caughlan g.r , fowler a.m. , harris m.j . , zimmerman b.a . 1985 , at .
data nucl .
dat a tables 32 , 197 caughlan g.r , fowler a.m. 1988 , atomic data nucl .
table , 40 , 283 chieffi a. , dominguez i. , limongi m. , straniero o. 2001 , apj , in press chieffi a. , straniero o. 1989 , 71 , 47 chieffi a. , limongi m. , straniero o. 1998 , 502 , 737 collela , p. , & woodward , p.r . 1984 , j.comp.phys .
, 54 , 174 domnguez i. , chieffi a. , limongi m. , straniero o. 1999 , 524 , 226 domnguez i. , & hflich p. 2000 , apj 528 , 854 domnguez i. , hflich p. , straniero o. , wheeler c.j .
2000 , in : _ nuclei in the cosmos _ , eds . n. pranzos & s. harissulos , edition frontiers , paris , p. 259
domnguez i. 1994 , in : abstracts of the conferences on thermonuclear supernovae , eds .
p. ruiz - lapuente , r. canal & j. isern .
domnguez i. 1991 , phd thesis , university of barcelona , isbn 84.7875.736.8 ezer d. , cameron a. 1972 , apss 14 , 399 fisher a. , branch d. , hflich p. , khokhlov a. 1998 , 494 , 47 garnavich p.m. et al .
1998 , apj 509 , 74 gerardy , c.l , hflich p. , fesen r. , & wheeler j.c .
2001 , apj in preparation hamuy m. , phillips m.m , schommer r.a .
suntzeff n.b , maza j. , aviles a. 1996a , 112 , 2391 hamuy m. , phillips m.m .
, suntzeff n.b . ,
schommer r.a . , maza j. , aviles a. 1996b , 112 , 2398 hamuy m. et al .
2000 , 120 , 1479 harkness r. 1987 , in : 13th texas symposium on relativistic astrophysics , world scientific publishing co. , p. 413
hatano k. , branch d. , lentz e.j . ,
baron e. , filippenko a.v .
, garnavich p. 2000 , apj 543l , 49 hillebrandt w. 1999 , private communication at the aspen workshop on sne ia hflich p. 2001 , , in preparation hflich p. , nomoto , k. , umeda h. , wheeler j.c .
2000 , 528 , 590 hflich p. , wheeler j.c .
, thielemann f.k 1998 , 495 , 617 [ hwt98 ] hflich p. , khokhlov a. 1996 , 457 , 500 [ hk96 ] hflich , p. , khokhlov a. , wheeler j.c . ,
phillips m.m . ,
suntzeff n.b . ,
hamuy m. 1996a , apj 472l , 81 hflich , p. ; dominik , c. ; khokhlov , a. ; mller , e. ; wheeler , j.c . 1996b , @xmath165 texas symposium on relativistic astrophysics , annals of the new york academy of science 759 , 348 hflich , p. 1995 , 443 , 89 [ h95 ]
hflich p. , khokhlov a. , wheeler j.c .
1995 , 444 , 211 [ hkw95 ] hflich p. , khokhlov a. , wheeler j.c .
1993 , a&a , 270 , 223 hoyle p. , fowler w.a . 1960 , 132 , 565 iben i.j , tutukov a.v . 1984 ,
apjs 54 , 335 iben i.j 1972 , 178 , 433 ivanov v.d . , hamuy , m. , pinto , p.a . ,
2000 , 542 , 588 iwamoto , k. , brachwitz , f. , nomoto , k. , kishimoto , n. , umeda , h. , hix w.r .
, thielemann f.k .
1999 , apjs 125 , 439 khokhlov a. 2001 , , in press & astro - ph/0008463 khokhlov a. 1995 , 449 , 695 khokhlov , a. , mller , e. , hflich p. 1993
, a&a 270 , 223 khokhlov a. 1991 , 245 , 114 lattanzio j. 1991 , apjs 76 , 215 lentz e.j .
, baron e. , branch d. , hauschildt p. , nugent p.e .
2000 , apj 530 , 966l metcalfe t.s .
, winget d.e . , charbonneau p. 2001 , , submitted niemeyer , j. c. , & hillebrandt , w. 1995 , apj 452 , 779 nomoto , k. 1980 , apj 248 , 798 nomoto k. , thielemann f .- k .
, yokoi k. 1984 , 286 , 644 nugent , p. , baron , e. , hauschildt , p. , & branch .
, d. 1997a , 485 , 812 ostriker p. , steinhardt p.j .
2001 , scientific american , jan .
2001 , p. 47
paczyski b. 1985 , in _ cataclysmic variables and low - mass x - ray binaries _ , eds .
lamb & j. patterson , ( dordrecht : reidel ) p. 1
perlmutter , s. , et al .
1999a , apj 517 , 565 perlmutter s. , turner m.s . ,
white m. 1999b , physical review letters 83 , 670 perlmutter c. et al .
1997 , 483 , 565 perlmutter c. et al .
1995 , apj 440 , l95 phillips , m.m .
1993 , 413 , 105 phillips , m.m . ,
lira p. , sunzeff n.b .
, schommer r.a . , hamuy m. , maza j. 1999 , 118 , 1766 renzini a. , fusi pecci f. 1988 , 26 199 riess a.g . , press w.h .
, kirshner r.p .
1995 , 438 , l17 riess a.g . ,
press w.h . , kirshner r.p .
1996 , 473 , 88 riess a.g . , et al .
1998 , 116 , 1009 riess a.g . ,
et al . 1999 , 118 ,
2675 saha a. et al .
1997 , apj 486 , 1 schaller g. , shaerer d. , meynet g. , maeder a. 1992 , a&as 96 , 269 schmidt , b. p. , et al .
1998 , apj 507 , 46 straniero , o. chieffi , a. and limongi m. 1997 , 490 , 425 thielemann f.k .
, nomoto k. , hashimoto m. 1996 , 460 , 408 tornamb a. , chieffi .
a. 1986 , , 220 , 529 umeda , h. , nomoto , k. 1999 , 513 , 861 vassiliadis e. , wood p.r .
1993 , 413 , 641 wang l. , hflich p. , wheeler j.c .
1997 , 483 , 29 webbink , r. f. 1984 , apj 277 , 355 weller j. , albrecht a. 2001 , _ opportunities for future supernova studies of cosmic acceleration _
, astro - ph/0008314 whelan j. , iben i.jr .
1973 , apj 186 , 1007 wheeler j. c. , hflich p. , harkness r. p. , spyromilio j. 1998 , apj 496 , 908 white m. 1998 , apj 506 , 495 weidemann v. 1987 , a&a 188 , 74 woosley s.e . , weaver t.a .
1994 , 423 , 371 woosley s.e . , weaver t.a .
1986 , 24 , 205 woosley s. e. , weaver t.a .
, taam r.e .
1980 , in : type i supernovae , ed .
j.c.wheeler , austin , u.texas , p.96 yamaoka h. , nomoto k. , shigeyama t. , & thielemann f. 1992 , apj 393 , 55 llcccccc z=0.02 & 1p5z22 & 1.5 & 0.55 & 0.21 & 0.27 & 0.75 & 0.589 y=0.28 & 3p0z22 & 3.0 & 0.57 & 0.21 & 0.28 & 0.76 & 0.584 & 5p0z22 & 5.0 & 0.87 & 0.29 & 0.46 & 0.72 & 0.561 & 7p0z22 & 7.0 & 0.99 & 0.28 & 0.70 & 0.60 & 0.516 & & & & & & & z=10@xmath166 & 1p5z13 & 1.5 & 0.59 & 0.24 & 0.31 & 0.76 & 0.587 y=0.23 & 3p0z13 & 3.0 & 0.77 & 0.26 & 0.39 & 0.74 & 0.567 y=0.23 & 5p0z13 & 5.0 & 0.90 & 0.29 & 0.58 & 0.66 & 0.541 & 6p0z13 & 6.0 & 0.98 & 0.29 & 0.71 & 0.60 & 0.522 & & & & & & & & low rate & & & & & & & 3p0z13lr & 3.0 & 0.76 & 0.51 & 0.38 & 1.22 & 0.620 & & & & & & & z=10@xmath167 & 3p0z14 & 3.0 & 0.80 & 0.27 & 0.41 & 0.73 & 0.568y=0.23 & 5p0z14 & 5.0 & 0.90 & 0.29 & 0.58 & 0.65 & 0.541 & 6p0z14 & 6.0 & 0.99 & 0.28 & 0.72 & 0.59 & 0.511 & & & & & & & z=10@xmath73 & 5p0z00 & 5.0 & 0.89 & 0.32 & 0.49 & 0.70 & 0.549 y=0.23 & 7p0z00 & 7.0 & 0.99 & 0.31 & 0.59 & 0.62 & 0.525 ccccccc he & c & o & ne & na & mg & si 6.62e-04 & 1.19e-02 & 9.18e-02 & 5.34e-03 & 4.91e-05 & 1.82e-02 & 2.61e-01 & & & & & & p & s & cl & ar & k & ca & sc 2.51e-05 & 1.59e-01 & 4.00e-06 & 3.28e-02 & 1.99e-06 & 3.44e-02 & 1.02e-08 & & & & & & ti & v & cr & mn & fe & co & ni 1.61e-05 & 9.07e-04 & 6.56e-04 & 2.68e-02 & 6.57e-01 & 6.15e-03 & 6.47e-02 | detailed stellar evolution calculations have been performed to quantify the influence of the main sequence mass @xmath0 and the metallicity z of the progenitor on the structure of the exploding wd which are thought to be the progenitors of sne ia . in particular , we study the effects of progenitors on the brightness decline relation @xmath1 which is a corner stone for the use of sne ia as cosmological yard - stick . both the typical @xmath0 and z can be expected to change as we go back in time .
we consider the entire range of potential progenitors with 1.5 to 7 @xmath2 and metallicities between z=0.02 to @xmath3 .
our study is based on the delayed detonation scenario with specific parameters which give a good account of typical light curves and spectra .
based on the structures for the wd , detailed model calculations have been performed for the hydrodynamical explosion , nucleosynthesis and light curves .
the main sequence mass has been identified as the decisive factor to change the energetics of the explosion and , consequently , dominates the variations in the rise - time to decline relation of light curves .
@xmath0 has little effect on the color index b - v . for similar decline rates @xmath4 , the flux at maximum brightness relative to the flux on the radioactive tail decreases systematically with @xmath0 by about @xmath5 .
this change goes along with a reduction of the photospheric expansion velocity @xmath6 by about 2000 km / sec .
a change in the central density of the exploding wd has similar effects but produces the opposite dependency between the brightness to tail ratio and @xmath6 and , therefore , can be separated .
the metallicity alters the isotopic composition of the outer layers of the ejecta .
selective line blanketing at short wavelengths decreases with z and changes systematically the intrinsic color index b - v by up to @xmath7 , and it alters the fluxes in the u band and the uv .
the change in b - v is critical if extinction corrections are applied .
the offset in the calibration of @xmath1 is not monotonic in z and , in general , remains @xmath8 .
we use our results and recent observations to constrain the progenitors , and to discuss evolutionary effects of sne ia with redshift .
the narrow spread in the fiducial rise - time to decline relation in local sne ia restricts the range of main sequence masses to a factor of 2 . the upper limit of 1 day for the difference between the local and distance sample support the need for a positive cosmological constant .
the size of evolutionary effects are small ( @xmath9 ) but are absolutely critical for the reconstruction of the cosmological equation of state . |
quasi - one dimensional ( 1d ) spin - chain systems are of current interest due to their unconventional magnetic properties.@xcite it is well established that an ideal 1d antiferromagnetic ( afm ) spin system does not show long - range ordering above @xmath6 k due to strong quantum spin fluctuations .
moreover , the magnetic properties of such spin - chains are strongly dependent on their spin - values.@xcite half - integer spin chains are gapless and a small amount of interchain coupling induces long - range afm order,@xcite while integer spin chains have a singlet ground state and gapped excitations making them stable against magnetic order.@xcite in addition , the presence of anisotropy leads to more complex behavior and a richer phase diagram . in this context , the compounds belonging to the material class @xmath7v@xmath1o@xmath2 ( @xmath8 = sr , ba and @xmath9 = cu , ni , co , mn ) have recently attracted attention to study the role of anisotropy , spin - value , and interchain interactions on the 1d magnetic properties.@xcite these compounds contain screw - chains of @xmath9o@xmath10 octahedra along the @xmath11 axis , separated by nonmagnetic vo@xmath12 ( v@xmath13 ; 3@xmath14 , @xmath15 ) tetrahedra , resulting in a quasi 1d spin - structure [ fig .
[ fig : crysstruc]].@xcite the @xmath16 = 1 afm chain compound srni@xmath1v@xmath1o@xmath2 has a non - magnetic singlet ground state with a gap between the singlet and triplet 1@xmath17 excited states .
the presence of substantial interchain interactions ( @xmath18 0.03 ; where @xmath19 is the number of neighbors in other chains ) significantly modifies the low energy magnetic excitations and reduces the gap values at the afm zone centers .
@xcite the presence of uniaxial anisotropy ( along the @xmath11 axis ) causes a zero field splitting of the triplet states and results in unusual field - dependent magnetic behavior.@xcite further , the substitution of nonmagnetic ions such as mg@xmath20 at the ni@xmath20 site induces 3d long - range afm order.@xcite on the other hand , the @xmath16 = 1/2 xxz ising system baco@xmath1v@xmath1o@xmath2 undergoes long - range magnetic ordering at 5.4 k in zero magnetic field.@xcite application of magnetic field suppresses the afm order and results in a field - induced order to disorder transition.@xcite below 1.5 k , a novel quantum magnetic state with incommensurate propagation vector appears for magnetic fields @xmath21 4 t. in the present investigation , we have studied the magnetic correlations of two iso - structural quasi-1d half - integer spin - chain compounds sr@xmath0v@xmath1o@xmath2 [ with _ m _ = co ( effective @xmath16 = 1/2 ) , mn ( @xmath16 = 5/2 ) ] to understand the effect of interchain - interactions , anisotropy , and spin - values on the magnetic ground state properties .
it is shown that both these compounds develop long - range magnetic order but have different antiferromagnetic ground states .
previous reports on bulk magnetic properties showed that the co- and mn - compounds order antiferromagnetically below @xmath4 5 k and @xmath4 43 k , respectively.@xcite for the co - compound the magnetic easy axis is the @xmath11 axis . an interesting field
induced order - disorder quantum phase transition was also reported at @xmath4 3.7 and @xmath4 6.5 tesla for @xmath22 and @xmath23 , respectively , at 2 k which indicates the strongly anisotropic nature of the magnetism.@xcite it reveals the presence of strong quantum fluctuations which affect the nel afm state under magnetic field and results in the quantum phase transition . on the other hand ,
no such field induced magnetic transition was reported up to 7 tesla in the mn - compound .
the magnetic ground state of the mn - compound was predicted to be a classical long - range antiferromagnet with weak anisotropy.@xcite the magnetic structures of these compounds , which are necessary to understand their magnetic ground states and microscopic magnetic properties , are currently unknown .
( color online ) crystal structure of sr@xmath0v@xmath1o@xmath2 ( _ m _ = co , mn ) .
( a ) the propagation of edge shared screw - chains of @xmath9o@xmath10 octahedra along the @xmath11 axis ; for sake of clarity , only two chains ( out of four chains per unit cell ) are shown .
the connections of the chains via vo@xmath12 tetrahedra are also shown .
the sr atoms are omitted for simplification .
( b ) the projection of the four screw - chains per unit cell on to the _ ab _ plane .
( c ) the geometrical arrangement of the @xmath24 ions in the @xmath25 planes .
the solid and dashed lines are the @xmath26 connections along the @xmath27 direction and the diagonal directions , respectively.,width=313 ] in this paper , we report a detailed investigation of the magnetic correlations in both the co- and mn - compounds by neutron diffraction and dc - magnetization .
the magnetic ions co@xmath20 and mn@xmath20 have different electronic configurations ( co@xmath20 ; 3@xmath28 , effective @xmath29 and mn@xmath20 ; 3@xmath30 , @xmath31 ) resulting in different spin values , different ionic radii ( 0.745 @xmath32 for co@xmath20 and 0.83 @xmath32 for mn@xmath20)@xcite and different anisotropy ( co@xmath20 is anisotropic ; ising type and mn@xmath20 is isotropic ; heisenberg type ) .
antiferromagnetic ordering has been found below @xmath4 5.2 k and @xmath4 42.2 k for the co- and mn - compounds , respectively .
interestingly , our low temperature neutron diffraction study confirms that they have two completely different collinear afm structures , characterized by the propagation vectors ; @xmath3 = ( 0 0 1 ) and @xmath3 = ( 0 0 0 ) for the co- and mn - compounds , respectively . for the mn - compound , afm chains ( along the @xmath11 axis ) are arranged ferromagnetically within the @xmath25 plane , whereas , for the co - compound , afm chains are arranged ferro-/antiferromagnetically along the @xmath27 direction .
a critical exponent study was performed to investigate the nature of magnetic ordering in these compounds .
neutron diffraction also reveals the presence of short - range spin - spin correlations over a wide temperature range above @xmath5 for both compounds .
polycrystalline samples of srco@xmath1v@xmath1o@xmath2 were synthesized by using the solid - state reaction method .
first the high - purity reagents of srco@xmath33 , coc@[email protected]@xmath1o , and v@xmath1o@xmath34 are taken as the starting materials in the molar ratio of 1:2:1 .
the reagents were weighed separately and mixed with ethanol , then ground carefully and homogenized thoroughly in an agate mortar .
the mixture was packed into an alumina crucible and calcined at 800 @xmath35c in air for 72 h with several intermediate grindings .
finally , the product was pressed into pellets and sintered at 900 @xmath35c in air for 36 h. single crystals of srco@xmath1v@xmath1o@xmath2 were grown using the floating zone technique at the crystal laboratory , helmholtz - zentrum berlin fr materialien und energie ( hzb ) , berlin , germany . for this , the powder was pressed into a feed rod using a cold isostatic press .
the floating zone furnace was used to melt the tip of the feed rod , which recrystallized on a seed crystal to achieve single - phase growth , the molten zone was then moved along the feed rod at a rate of 1 mm / hour to obtain a large single crystal .
the crystals were grown in a mixture of high quality flowing oxygen ( 20@xmath36 ) and argon ( 80@xmath36 ) at ambient pressure .
the resulting crystals were cylindrical with a diameter of @xmath4 4 mm and length of @xmath4 10 cm .
polycrystalline samples of srmn@xmath1v@xmath1o@xmath2 were synthesized by a two - step solid - state reaction method . in the first step , as a precursor for the main synthesis , srv@xmath1o@xmath10 was synthesized from the mixture of high purity srco@xmath33 and v@xmath1o@xmath34 with a 1:1 molar ratio .
the chemicals were mixed carefully and homogeneously using a mortar and pestle in ethanol .
the mixed chemicals were then put in a pt - crucible and calcined at 550 @xmath35c in air for total 24 h with intermediate grindings . for further reaction , the calcined powder was finally heated at 600 @xmath35c for 48 h. in the second step , the as - prepared srv@xmath1o@xmath10 and high purity mno were mixed in a 1:2 molar ratio and the mixture was ground homogeneously in an agate mortar .
the mixture was then pressed into pellets and sintered at 850 @xmath35c in ar flow for total of 144 h with intermediate grindings and repelletizations .
the as - prepared powder samples were characterized by powder x - ray diffraction ( xrd ) using a lab x - ray machine ( bruker d8 advance ) .
the xrd measurements were performed at 300 k using cu @xmath37 radiation over the scattering angular range 2@xmath38@xmath39 .
the phase purity of the single crystals were also confirmed by powder xrd .
a part of the crystal was ground into powder for this study .
the backscattering x - ray laue patterns revealed the good single - crystalline nature of the samples .
temperature and field - dependent magnetization measurements were carried out using a physical properties measurement system ( 14 t ppms ; quantum design ) at the laboratory for magnetic measurements , hzb . the temperature - dependent static susceptibility [ @xmath40(t ) ]
measurements were performed over a temperature range of 2400 k for the co - compound and 2900 k for the mn - compound , respectively , under an applied magnetic field of 1 tesla .
all measurements were performed in the warming cycle after cooling the samples in zero field .
the isothermal magnetization measurements were performed at the base temperature 1.8 k over @xmath41 tesla in the zero field cooled condition .
single crystal measurements were carried out on a small piece of the crystal ( 3 @xmath42 2 @xmath42 1 mm@xmath43 ) with edges parallel to the principal crystallographic axes . for the crystal structure investigation ,
neutron diffraction patterns for both compounds were recorded at 300 k by using the high resolution powder diffractometer e9 ( @xmath44 = 1.7982 @xmath32 ) , at hzb , germany .
for the magnetic structure investigation , low temperature neutron diffraction patterns were measured by using the time - of - flight diffractometer v15 , at hzb , germany .
v15 is equipped with four movable detector banks which were positioned at 30 , 60 , 90 , and 150 degrees .
the incident wavelength band 0.77.8 @xmath32 was used for the measurements .
low temperature measurements were performed in a standard orange cryostat .
the diffraction data were analyzed by the rietveld method using the fullprof program.@xcite
cccc & site&srco@xmath1v@xmath1o@xmath2&srmn@xmath1v@xmath1o@xmath2 + @xmath45@xmath46 & & 12.2710(1 ) & 12.4527(1 ) + @xmath47@xmath46 & & 8.4192(1 ) & 8.6853(1 ) + + sr & 8@xmath48 ( 0,0,@xmath19 ) & & + @xmath49 & & 0 & 0 + @xmath50 & & 1.41(12 ) & 0.95(11 ) + + @xmath9 ( co / mn ) & 16@xmath51 ( @xmath52,@xmath53,@xmath19 ) & & + @xmath54 & & 0.3281(12 ) & 0.3282(8 ) + @xmath55 & & 0.3311(9 ) & 0.3308(6 ) + @xmath49 & & 0.2147(18 ) & 0.2107(11 ) + @xmath50 & & 0.62(12 ) & 0.66(9 ) + + v & 16@xmath51 ( @xmath52,@xmath53,@xmath19 ) & & + @xmath54 & & 0.2615(3 ) & 0.2638(9 ) + @xmath55 & & 0.0783(4 ) & 0.0786(4 ) + @xmath49 & & 0.0942(9 ) & 0.0776(3 ) + @xmath50 & & 0.31(9 ) & 0.27(2 ) + + o1 & 16@xmath51 ( @xmath52,@xmath53,@xmath19 ) & & + @xmath54 & & 0.1457(4 ) & 0.1399(4 ) + @xmath55 & & 0.5014(7 ) & 0.4990(7 ) + @xmath49 & & -0.0026(12 ) & -0.0087(10 ) + @xmath50 & & 1.26(12 ) & 1.22(11 ) + + o2 & 16@xmath51 ( @xmath52,@xmath53,@xmath19 ) & & + @xmath54 & & 0.3410(6 ) & 0.3419(5 ) + @xmath55 & & 0.6630(7 ) & 0.6668(6 ) + @xmath49 & & 0.4763(9 ) & 0.4622(8 ) + @xmath50 & & 0.49(16 ) & 0.61(12 ) + + o3 & 16@xmath51 ( @xmath52,@xmath53,@xmath19 ) & & + @xmath54 & & 0.1578(8 ) & 0.1499(6 ) + @xmath55 & & 0.6824(5 ) & 0.6885(4 ) + @xmath49 & & 0.7086(9 ) & 0.7009(8 ) + @xmath50 & & 0.58(17 ) & 0.90(11 ) + + o4 & 16@xmath51 ( @xmath52,@xmath53,@xmath19 ) & & + @xmath54 & & 0.3314(5 ) & 0.3266(4 ) + @xmath55 & & 0.4974(6 ) & 0.4996(6 ) + @xmath49 & & 0.1876(9 ) & 0.1727(7 ) + @xmath50 & & 1.02(12 ) & 0.80(10 ) + the crystal structures of srco@xmath1v@xmath1o@xmath2 and srmn@xmath1v@xmath1o@xmath2 are investigated from simultaneous rietveld analysis of both x - ray and neutron powder diffraction patterns , recorded at 300 k ( fig .
[ fig : xrd ] ) .
rietveld analysis confirms that both compounds crystallize in the tetragonal space group @xmath56 .
it also confirms that the mn - compound is single phase in nature .
the co - compound used for the x - ray and low temperature neutron diffraction measurements contains a small amount ( @xmath4 4 @xmath36 ) of coo impurity phase .
however , a phase pure sample of srco@xmath1v@xmath1o@xmath2 ( without coo impurity ) was used for the neutron diffraction study at 300 k. the refined values of various structural parameters such as lattice constants , fractional atomic coordinates , and isotropic thermal parameters are given in table [ tab : t1 ] .
the values of the lattice constants are found to be @xmath57 @xmath32 and @xmath58 @xmath32 for the co - compound and @xmath59 @xmath32 and @xmath60 @xmath32 for the mn - compound , respectively .
these values are in good agreement with the values reported earlier.@xcite the larger lattice constants of the mn - compound compared to the co - compound are due to their ionic radii .
the greater ionic radius of the mn@xmath20 ion ( 0.83 @xmath32 ) compared to that of the co@xmath20 ion ( 0.745 @xmath32 ) in octahedral coordination causes this increase of the lattice constants .
the isostructural compound srni@xmath1v@xmath1o@xmath2 also follows this sequence .
its lattice constants were reported to be @xmath61 @xmath32 and @xmath62 @xmath32 which are the smallest in this series due to the smaller ionic radius of the ni@xmath20 ions ( 0.69 @xmath32).@xcite ccc bonds & srco@xmath1v@xmath1o@xmath2 & srmn@xmath1v@xmath1o@xmath2 + sr - o1 ( @xmath32 ) & [email protected](10 ) & [email protected](9 ) + & [email protected](9 ) & [email protected](9 ) + sr - o2 ( @xmath32 ) & [email protected](8 ) & [email protected](6 ) + sr - o3 ( @xmath32 ) & [email protected](8 ) & [email protected](7 ) + sr - o4 ( @xmath32 ) & [email protected](7 ) & [email protected](5 ) + & [email protected](7 ) & [email protected](6 ) + + @xmath9-o1 ( @xmath32 ) & 2.162(16 ) & 2.176(11 ) + @xmath9-o2 ( @xmath32 ) & 2.015(17 ) & 2.165(12 ) + & 2.119(14 ) & 2.152(10 ) + @xmath9-o3 ( @xmath32 ) & 2.097(18 ) & 2.234(12 ) + & 2.062(17 ) & 2.110(12 ) + @xmath9-o4 ( @xmath32 ) & 2.051(13 ) & 2.128(11 ) + + v - o1 ( @xmath32 ) & 1.689(8 ) & 1.727(7 ) + v - o2 ( @xmath32 ) & 1.753(8 ) & 1.758(7 ) + v - o3 ( @xmath32 ) & 1.788(8 ) & 1.697(7 ) + v - o4 ( @xmath32 ) & 1.704(7 ) &
1.708(6 ) + + @xmath9-@xmath9 ( @xmath32 ) ( along @xmath11 axis ) & 2.874(19 ) & 2.937(14 ) + @xmath9-@xmath9 ( @xmath32 ) ( along @xmath27 axis ) & 6.428(15 ) & 6.523(11 ) + & 6.452(13 ) & 6.543(12 ) + the crystal structure of these compounds consists of screw chains along the @xmath11 axis [ fig .
[ fig : crysstruc ] ] .
the screw chains are formed by corner sharing @xmath9o@xmath10 octahedra and are connected via vo@xmath12 tetrahedra by sharing corners [ fig .
[ fig : crysstruc](a ) ] .
the @xmath9o@xmath10 octahedra and vo@xmath12 tetrahedra are found to be slightly distorted as evident from the differences in the co / mn - o and v - o bond lengths [ table [ tab : t2 ] ] .
there are four screw chains per unit cell which are centered around ( 1/4,1/4 ) , ( 1/4,3/4 ) , ( 3/4,1/4 ) , and ( 3/4,3/4 ) in the @xmath25 plane [ fig .
[ fig : crysstruc](b ) ] .
two diagonal chains rotate clockwise while the other two chains rotate anti - clockwise when propagating along the @xmath11 axis which results in a complex interchain interaction geometry within a given @xmath25 plane [ fig .
[ fig : crysstruc](c ) ]
. furthermore the interaction geometry rotates 90@xmath35 in the next layer ( shifted by @xmath11/4 ) along the @xmath11 axis .
interchain interactions are also possible , via superexchange pathways ( through the vo@xmath12 tetrahedra ) , from mn / co ions in a chain to the mn / co ions which are shifted by @xmath11/4 in the neighboring chains [ fig .
[ fig : crysstruc ] ] .
the peculiar crystal structure , therefore , provides many possible interchain super - exchange interaction pathways with comparable strengths .
the distances between magnetic ions along the chain direction are 2.874 @xmath63 0.019 @xmath32 for the co - compound and 2.937 @xmath63 0.014 @xmath32 for the mn - compound , respectively . within a given @xmath25 ( along @xmath27 axis )
plane , the distances between magnetic ions are 6.428 @xmath63 0.015/6.452 @xmath63 0.013 @xmath32 for the co - compound and 6.523 @xmath63 0.011/6.543 @xmath63 0.012 @xmath32 for the mn - compound , respectively .
the different interchain distances for the two compounds are also due to the ionic size effect .
figures [ fig : sus](a ) and [ fig : sus](b ) show the susceptibility ( @xmath40 ) vs. temperature curves , measured under a magnetic field of 1 t , for srco@xmath1v@xmath1o@xmath2 and srmn@xmath1v@xmath1o@xmath2 , respectively .
for srco@xmath1v@xmath1o@xmath2 , with decreasing temperature a broad hump with maximum around 30 k followed by a sharp peak at @xmath4 5.2 k [ fig .
[ fig : sus](a ) ] has been found for the powder sample .
the single crystal susceptibilities show that the broad peak @xmath4 30 k is present for both parallel ( @xmath64 @xmath65 @xmath11 ) and perpendicular ( @xmath64 @xmath65 @xmath48 ) field directions . with decreasing temperature , below @xmath4 5.2 k ,
the parallel susceptibility ( @xmath64 @xmath65 @xmath11 ) shows a sudden decrease , whereas , a sharp peak appears in the perpendicular susceptibility ( @xmath64 @xmath65 @xmath48 ) at @xmath4 5.2 k [ fig .
[ fig : sus](a ) ] .
the broad hump is characteristic of the onset of 1d short - range spin - spin correlations and the sharp peak corresponds to the onset of 3d long - range afm ordering .
the presence of short - range spin - spin correlations is confirmed from the neutron diffraction study ( see section e ) .
a large difference between parallel and perpendicular susceptibilities has been observed that persists up to 400 k. this indicates the presence of a large paramagnetic anisotropy .
it is also found that the @xmath66 curves do not follow curie - weiss behavior below 400 k which may be due to low - lying crystal field excitations as well as the presence of short - range correlations .
it should be noted that the @xmath66 curves do not show a rapid jump @xmath4 20 k [ fig .
[ fig : sus](a ) ] which was reported by he _ et .
al_.,@xcite and was explained on the basis of weak ferromagnetic moments due to canted antiferromagnetic ordering . in agreement with our results , the susceptibility curves reported by lejay _ et .
al_.@xcite from a single crystal study do not show this jump .
in addition , no ferromagnetism is evident in our low temperature neutron diffraction results ( discussed in section c ) .
it is known that for the free co@xmath20 ( 3@xmath28 ) ion having 7 electrons the total orbital and spin angular momenta are @xmath67 = 3 and @xmath16 = 3/2 , respectively , according to hund s rules .
a cubic crystal field splits the 7 orbital levels into 1 orbital singlet and 2 orbital triplets .
the lowest state is a triplet @xmath68 .
the effective orbital angular momentum within the triplet @xmath68 is @xmath67 = 1 . in octahedral ligand field and with spin - orbit coupling the degeneracy of the 12 states in @xmath68
is partially lifted and results in 6 kramers doublets .
the lowest kramers doublet can be regarded as an effective @xmath16 = 1/2 with large anisotropy .
the exchange interaction between true co@xmath20 spins ( @xmath16 = 3/2 ) may be regarded as a molecular field which lifts the degeneracy of the kramers doublets . to estimate the intrachain interaction in srco@xmath1v@xmath1o@xmath2
, the parallel susceptibility curve is fitted with the bonner - fisher model for uncoupled ising chains ( ref . )
@xmath69 where @xmath70 is the land factor parallel to the ising axis , @xmath71 stands for the boltzmann constant and @xmath72 is the nearest neighbor intrachain exchange constant .
the fitted values of the parameters are found to be @xmath72 = 45.35 @xmath63 0.04 k and @xmath70 = 9.90 @xmath63 0.01 .
the values of @xmath72 = 40.6 k and @xmath70 = 9.42 were reported by he _
al_.@xcite the fitted curve is shown by the solid line .
the discrepancy between the observed and fitted curves indicates that the present compound deviates from the pure ising limit and may be better defined as an xxz - system . for isolated chains ,
the perturbation from the pure ising limit gives rise to a continuum of soliton - pair excited states .
this is in contrast to the doubly degenerate ground states and highly degenerate set of first excited states of energy 2@xmath73 in the pure ising limit.@xcite the continuum states may be split into discrete levels by interactions with adjacent chains as reported for cscocl@xmath33.@xcite the mixing of crystal - field levels of the co@xmath20 ion with the spin excitation states may also be expected which leads to a decrease of the energy values of the excited states .
for srmn@xmath1v@xmath1o@xmath2 , with decreasing temperature the susceptibility curve shows a broad maximum @xmath4 170 k , characteristic of short - range 1d magnetism , followed by a kink @xmath4 42 k corresponding to the 3d long - range afm transition [ fig .
[ fig : sus](b ) ] .
the nature of the susceptibility curve is similar to that reported earlier by niesen _ et .
_ for this compound.@xcite below t@xmath74 , the susceptibility curve showed a slight increase , which was ascribed to the transverse susceptibility.@xcite to estimate the strength of the intrachain interactions , the susceptibility curve above 90 k is fitted with a general expression of the susceptibility of heisenberg spin - chains predicted from pad approximations @xcite
@xmath75 where the first temperature - independent term @xmath76 is due to van - vleck paramagnetism as well as diamagnetic core susceptibility and @xmath72 is the nearest - neighbor intrachain super - exchange interaction
. a good agreement between observed and calculated susceptibility curves is found .
the fitted values of the parameters are @xmath72 = 37.57 @xmath63 0.02 k , @xmath77 = 1.9999 @xmath63 0.0009 , and @xmath76 = @xmath78 emu mole@xmath79 oe @xmath79 .
for the fitting , the values of the coefficients @xmath80 and @xmath81 are adopted from ref . .
the field dependence of the magnetizations for both compounds , measured at 1.8 k are shown in fig .
[ fig : mag ] . for srco@xmath1v@xmath1o@xmath2 , field induced phase transitions under the critical fields of @xmath4 3.7 and @xmath4 6.5 t for @xmath22 and @xmath23 ( @xmath82 ) , respectively , have been found .
same critical field values ( @xmath83 = 3.7 t and @xmath84 = 6.5 t ) were reported earlier by he _
al._@xcite for the powder sample , as expected , a broad transition extending from 3.7 t to 6.5 t is found for both positive and negative magnetic fields [ inset of fig .
[ fig : mag](a ) ] .
a similar field induced magnetic ( order - disorder ) transition was reported in the related compound baco@xmath1v@xmath1o@xmath2 for @xmath22 and was attributed to the quantum phase transition characteristic of the @xmath16 = 1/2 1d xxz antiferromagnet with ising - like anisotropy .
@xcite here , the magnetic field induces a reentrant phenomenon from the 3d long - range nel state into a 1d quantum spin - liquid state .
this transition was theoretically predicted first for gapped spin systems as a function of doping of the magnetic site by non - magnetic impurities .
@xcite the magnetization of srco@xmath1v@xmath1o@xmath2 shows an almost linear field dependence below the transition fields which indicates the absence of a ferromagnetic contribution in agreement with the neutron diffraction results ( discussed later ) .
no hysteresis and remanent magnetization were found at 1.8 k. there is also no tendency towards saturation up to 14 t. on the other hand , for srmn@xmath1v@xmath1o@xmath2 no such field induced transition has been found . in this case , the magnetization shows almost linear behavior with magnetic field .
the value of magnetization is small @xmath85/mn@xmath20 at 14 t as compared to the fully polarized magnetization of 5 @xmath86/mn@xmath20 which confirms the presence of strong afm interactions .
this is in agreement with neutron diffraction results where an afm ordering has been found for this compound ( discussed later ) .
the magnetization studies clearly reveal that two compounds have very different magnetic properties despite their similar crystal structures . in order to investigate the nature of the ground state for these two iso - structural compounds ,
we have carried out low temperature neutron diffraction measurements at several temperatures below and above the corresponding nel temperatures .
for this purpose , we employed the time - of - flight diffractometer v15 which provides good resolution over the low @xmath87 ( high @xmath88 ) region .
the neutron diffraction patterns of srco@xmath1v@xmath1o@xmath2 at 8 k ( paramagnetic state ) and 1.5 k ( magnetically ordered state ) are shown in fig .
[ fig : neu - scvo ] [ ( a)(b ) ] and fig . [
fig : neu - scvo][(c ) ] , respectively . owing to the presence of a small coo impurity in this sample ( discussed earlier ) , the neutron diffraction patterns at 8 k are refined with a model which included three phases .
the phases are ( i ) nuclear phase for srco@xmath1v@xmath1o@xmath2 and ( ii ) nuclear as well as ( iii ) magnetic phases for coo .
the coo orders antiferromagnetically below 295 k.@xcite first we refined the diffraction pattern from the detector bank at 150 deg [ fig .
[ fig : neu - scvo](a ) ] which covers the @xmath88-range from 13.7 @xmath32 .
good agreement between observed and calculated patterns has been obtained [ fig . [
fig : neu - scvo](a ) ] . the tetragonal crystal structure with space group @xmath56 ( as found at room temperature ) reproduces the low temperature nuclear phase of srco@xmath1v@xmath1o@xmath2 .
the refined values of the lattice parameter are found to be @xmath89 @xmath32 and @xmath90 , respectively .
the crystal structure for coo is monoclinic with space group @xmath91.@xcite the lattice parameters are found to be @xmath48 = 5.168 @xmath63 0.018 , @xmath51 = 3.004 @xmath63 0.043 , @xmath11 = 3.008 @xmath63 0.023 , and @xmath92 = 125.5 @xmath63 0.8 deg which are in good agreement with the reported values .
@xcite the afm structure with propagation vector _
k _ = ( 0 1 1/2 ) was used for the magnetic phase of coo as reported by jauch _
@xcite the refined value of the ordered magnetic moment is 3.9 @xmath63 0.1 @xmath86/co@xmath20 which also agrees well with the reported value of 3.98 @xmath86/co@xmath20 at 10 k.@xcite the obtained values of structural parameters and magnetic moment were used to refine the diffraction patterns from the detector bank at 60 deg which covers high _ d_-value region ( 2.57.2 ) and important for the magnetic scattering study of srco@xmath1v@xmath1o@xmath2 . the refined diffraction pattern from this detector bank at 8 k
is depicted in fig .
[ fig : neu - scvo](b ) and shows good agreement between observed and calculated patterns . the diffraction pattern at the base temperature of 1.5 k ( magnetic ordered state ) is shown in fig . [ fig : neu - scvo](c ) .
appearance of a set of additional magnetic bragg peaks has been observed which are at positions ( @xmath93 ) with @xmath94 is an odd integer , that is , on the nuclear peaks forbidden by the body - centered symmetry @xmath95 of the lattice .
all magnetic peaks can be indexed with a propagation vector @xmath3 = ( 0 0 1 ) with respect to the tetragonal symmetry . in order to find the magnetic structure that is compatible with the space group symmetry , we performed representation analysis by using version 2k of the program sarah - representational analysis.@xcite the analysis was done by using the crystal structure above the transition and the propagation vector of the magnetic ordering .
firstly it involves the determination of the space group symmetry elements , that leave the propagation vector @xmath3 invariant : these form the little group @xmath96 . the magnetic representation of a crystallographic site
can then be decomposed in terms of the irreducible representations ( irs ) of @xmath96 : @xmath97 where @xmath98 is the number of times that the ir @xmath99 of order @xmath100 appears in the magnetic representation @xmath101 for the chosen crystallographic site .
ccccccccc irs & & & & atoms & & & & + & co1 & co2 & co3 & co4 & co5 & co6 & co7 & co8 + @xmath102 & 1 & 1 & -i & -i & -1 & -1 & -i & -i + + @xmath103 & 1 & 1 & -i & -i & 1 & 1 & i & i + + @xmath104 & 1 & 1 & i & i & -1 & -1 & i & i + + @xmath105 & 1 & 1 & i & i & 1 & 1 & -i & -i + + @xmath106 & 1 0 & -1 0 & 1 0 & -1 0 & 0 -1 & 0 1 & 0 -1 & 0 1 + & 0 1 & 0 -1 & 0 -1 & 0 1 & -1 0 & 1 0 & 1 0 & -1 0 + for srco@xmath1v@xmath1o@xmath2 , the crystal structure is described by the space group @xmath107 ( sp . gr .
this space group involves two centering operations and eight symmetry operations ( appendix [ g0_elements ] ) .
all the eight symmetry operations leave the propagation vector @xmath3 invariant or transform it into an equivalent vector . for the propagation vector @xmath3 = ( 0 0 1 ) , the irreducible representations of the propagation vector group @xmath96 are given in table [ tab : t3 ] .
there are five possible irs .
the magnetic reducible representation @xmath101 for the co site ( 16@xmath51 site ) can be decomposed as a direct sum of irs as @xmath108 cccccccccc irs & & + & & co1 & co2 & co3 & co4 & co5 & co6 & co7 & co8 + @xmath102 & @xmath109 & ( 200)&(-200)&(000)&(000)&(-200 ) & ( 200)&(000 ) & ( 000 ) + & @xmath110 & ( 020)&(0 - 20)&(000)&(000)&(020)&(0 - 20)&(000)&(000 ) + & @xmath111 & ( 002)&(002)&(000)&(000)&(002)&(002)&(000)&(000 ) + + @xmath103 & @xmath109 & ( 200 ) & ( -200 ) & ( 000 ) & ( 000 ) & ( 200 ) & ( -200 ) & ( 000 ) & ( 000 ) + & @xmath110&(020 ) & ( 0 - 20 ) & ( 000 ) & ( 000 ) & ( 0 - 20 ) & ( 020 ) & ( 000 ) & ( 000 ) + & @xmath111&(002 ) & ( 002 ) & ( 000 ) & ( 000 ) & ( 00 - 2 ) & ( 00 - 2 ) & ( 000 ) & ( 000 ) + + @xmath104 & @xmath109 & ( 000 ) & ( 000 ) & ( 020 ) & ( 0 - 20 ) & ( 000 ) & ( 000 ) & ( 0 - 20 ) & ( 020 ) + & @xmath110&(000 ) & ( 000 ) & ( -200 ) & ( 200 ) & ( 000 ) & ( 000 ) & ( -200 ) & ( 200 ) + & @xmath111&(000 ) & ( 000 ) & ( 002 ) & ( 002 ) & ( 000 ) & ( 000 ) & ( 00 - 2 ) & ( 00 - 2 ) + + @xmath105 & @xmath109 & ( 000 ) & ( 000 ) & ( 020 ) & ( 0 - 20 ) & ( 000 ) & ( 000 ) & ( 020 ) & ( 0 - 20 ) + & @xmath110&(000 ) & ( 000 ) & ( -200 ) & ( 200 ) & ( 000 ) & ( 000 ) & ( 200 ) & ( -200 ) + & @xmath111&(000 ) & ( 000 ) & ( 002 ) & ( 002 ) & ( 000 ) & ( 000 ) & ( 002 ) & ( 002 ) + + @xmath106 & @xmath109 & ( 100 ) & ( 100 ) & ( 010 ) & ( 010 ) & ( 000 ) & ( 000 ) & ( 000 ) & ( 000 ) + & @xmath110&(010 ) & ( 010 ) & ( -100 ) & ( -100 ) & ( 000 ) & ( 000 ) & ( 000 ) & ( 000 ) + & @xmath111&(001 ) & ( 00 - 1 ) & ( 001 ) & ( 00 - 1 ) & ( 000 ) & ( 000 ) & ( 000 ) & ( 000 ) + & @xmath112&(000 ) & ( 000 ) & ( 000 ) & ( 000 ) & ( -100 ) & ( -100 ) & ( 010 ) & ( 010 ) + & @xmath113&(000 ) & ( 000 ) & ( 000 ) & ( 000 ) & ( 010 ) & ( 010 ) & ( 100 ) & ( 100 ) + & @xmath114&(000 ) & ( 000 ) & ( 000 ) & ( 000 ) & ( 001 ) & ( 00 - 1 ) & ( 001 ) & ( 00 - 1 ) + & @xmath115&(000 ) & ( 000 ) & ( 000 ) & ( 000 ) & ( -100 ) & ( -100 ) & ( 0 - 10 ) & ( 0 - 10 ) + & @xmath116&(000 ) & ( 000 ) & ( 000 ) & ( 000 ) & ( 010 ) & ( 010 ) & ( -100 ) & ( -100 ) + & @xmath117&(000 ) & ( 000 ) & ( 000 ) & ( 000 ) & ( 001 ) & ( 00 - 1 ) & ( 00 - 1 ) & ( 001 ) + & @xmath118&(100 ) & ( 100 ) & ( 0 - 10 ) & ( 0 - 10 ) & ( 000 ) & ( 000 ) & ( 000 ) & ( 000 ) + & @xmath119&(010 ) & ( 010 ) & ( 100 ) & ( 100 ) & ( 000 ) & ( 000 ) & ( 000 ) & ( 000 ) + & @xmath120&(001 ) & ( 00 - 1 ) & ( 00 - 1 ) & ( 001 ) & ( 000 ) & ( 000 ) & ( 000 ) & ( 000 ) + according to the landau theory for second order phase transitions , only one representation can be involved in a critical transition . with this constraint , therefore , five different magnetic structures are possible with @xmath3 = ( 0 0 1 ) .
the basis vectors of the co position [ 16@xmath51 ( @xmath52 , @xmath53 , @xmath19 ) ; ( 0.3315 , 0.3301 , 0.2233 ) ] for five irs , calculated using the projection operator technique , implemented in sarah,@xcite are given in table [ tab : t4 ] .
the labeling of the propagation vector and the irs follows the scheme used by kovalev.@xcite the atomic sites are labeled following the convention given in table [ tab : t3 ] .
the representations @xmath121 , @xmath122 , @xmath123 , and @xmath124 are one dimensional and are repeated three times each in eq .
[ magnetic_representation_scvo ] .
there are three basis vectors corresponding to each of these representations .
moreover , the basis vectors for some of the sites are purely imaginary .
the representation @xmath125 is two dimensional and repeated six times .
it , therefore , corresponds to a magnetic structure of twelve basis vectors .
the refinement of the magnetic structure was tested for each of these five irs , and only @xmath125 gave satisfactory agreement .
it may be noted that the two dimensionality of @xmath125 permits two physically equivalent descriptions of the magnetic structure .
the basis vectors for @xmath125 indicate that the moment components along all @xmath48 , @xmath51 , and @xmath11 directions are refinable .
the simultaneous refinement of all three moment components gives the values @xmath126 = 0.12 @xmath63 0.08 @xmath86 , @xmath127 = @xmath1280.09 @xmath63 0.08 @xmath86 , and @xmath129 = 2.25 @xmath63 0.05 @xmath86 .
it is apparent that the moment components along the @xmath48 and @xmath51 axes are weak and in the order of the error bar .
therefore , the moments align predominantly along the @xmath11 axis which is the magnetic easy axis [ fig .
[ fig : mag](a ) ] and the maximum possible deviation from the @xmath11 axis is less than 5 @xmath36 . the fitted pattern is shown in fig .
[ fig : neu - scvo](c ) which shows good agreement with the observed pattern . here a four phases model ( nuclear and magnetic phases for both srco@xmath1v@xmath1o@xmath2 and coo ) was used for the refinement .
the @xmath130-factor was found to be 5.08 @xmath36 for the magnetic phase of srco@xmath1v@xmath1o@xmath2 . the corresponding magnetic structure ( weak moment components along the @xmath48 and @xmath51 axes are ignored )
is shown in fig .
[ fig : mag - struc - scvo ] where the co - spins align antiferromagnetically along the chain ( @xmath11 axis ) . within the basal plane ( @xmath25 plane ) , the spins form ferromagnetic / antiferromagnetic zigzag lines along the @xmath27 axis .
it may be noted that the observed magnetic structure is the same as that reported for the related compound baco@xmath1v@xmath1o@xmath2.@xcite further , to verify the magnetic structure , simulated annealing ( sa ) analysis @xcite was performed using the fullprof suite program.@xcite the details of the sa method and the results are given in appendix [ sa_scvo - smvo ] .
the solution of sa is the same as the @xmath125 magnetic structure that was obtained from representation analysis with magnetic moments aligned along the @xmath11 axis [ table [ tab : t9 ] ] .
the refined value of the magnetic moment @xmath4 2.25 @xmath86/co@xmath20 at 1.5 k is substantially smaller than the value 3 @xmath86/co@xmath20 expected for @xmath16 = 3/2 considering a spin only contribution .
therefore , it may be concluded that the co moments are not fully ordered even at 1.5 k. this reduction is probably due to strong quantum fluctuations expected in this quasi-1d system .
in fact , the presence of quantum fluctuations is evident from the field - induced transition from the 3d long - range afm ordered state to the 1d quantum disordered state.@xcite nevertheless , all magnetic bragg peaks are found to be instrumental resolution limited , thus confirming that the afm ordering is long - range at 1.5 k. the neutron diffraction patterns of srmn@xmath1v@xmath1o@xmath2 at 60 k ( paramagnetic state ) and 1.5 k ( magnetically ordered state ) are shown in fig . [
fig : neu - smvo ] ( a)(c ) .
the diffraction pattern from the detector bank at 150 deg [ fig . [ fig : neu - smvo ] ( a ) ] at 60 k can be fitted by the nuclear phase alone with the tetragonal space group @xmath56 , as found in the room temperature diffraction study , confirming the paramagnetic state .
the refinement shows good agreement between observed and calculated patterns .
the refined values of the lattice parameter are @xmath48 = 12.3768 @xmath63 0.0004 @xmath32 and @xmath11 = 8.6397 @xmath63 0.0003 @xmath32 .
the obtained values of the structural parameters were used to refine the diffraction patterns from the detector bank at 60 deg which covers high _ d_-value ( 2.59.5 @xmath32 ) region and important for the magnetic scattering study . the refined diffraction pattern from this detector bank is depicted in fig . [ fig : neu - smvo](b ) which also shows good agreement between observed and calculated patterns . the diffraction pattern measured at 1.5 k ( in the magnetic ordered state )
is shown in fig .
[ fig : neu - smvo](c ) .
appearance of a set of additional magnetic bragg peaks as well as the increase of the intensity of few nuclear bragg peaks have been observed . unlike srco@xmath1v@xmath1o@xmath2
, the magnetic peaks for this compound are at positions ( _ hkl _ ) where @xmath94 is an even integer , that is , on the nuclear peaks allowed by the body - centered symmetry _
i _ of the lattice , however , forbidden by the space group symmetry @xmath56 .
therefore , the propagation vector _
k _ = ( 0 0 0 ) can be used for the magnetic phase . for this case , we have also carried out representation analysis to find out the symmetry allowed magnetic structures . as discussed earlier in the case of srco@xmath1v@xmath1o@xmath2 ,
the space group @xmath107 involves two centering operations and eight symmetry operations ( appendix [ g0_elements ] ) .
all the eight symmetry operations leave the propagation vector @xmath3 invariant . for the propagation vector @xmath3 = ( 0 0 0 ) ,
the irs of the propagation vector group @xmath96 are given in table [ tab : t5 ] .
the decomposition of the magnetic representation @xmath101 in terms of the non - zero irs of @xmath96 for the mn site ( 16@xmath51 site ) is @xmath131 the associated basis vectors for the mn site [ 16@xmath51 ( @xmath52 , @xmath53 , @xmath19 ) ; ( 0.3328 , 0.3184 , 0.2226 ) ] for all five irs are given in table [ tab : t6 ] .
the representations @xmath121 , @xmath122 , @xmath123 , and @xmath124 are one dimensional and repeated three times . the representation @xmath125 is two dimensional and repeated six times .
the refinement of the magnetic structure was tested for each of the irreducible representations , and only @xmath125 gave satisfactory agreement .
the basis vectors for @xmath125 indicate that the moment components along all @xmath48 , @xmath51 , and @xmath11 directions are refinable .
the refinement confirms that the moments are aligned in the @xmath25 plane with @xmath129 component being zero or very weak ( should be less than 5 @xmath36 if any ) .
calculated pattern considering any arbitrary moment direction within the @xmath25 plane fits the observed data ( @xmath130 = @xmath132 ) .
however , the refinement with the moment along the @xmath48 or @xmath51 axis and the other components fixed to zero gives a marginally better fit ( @xmath130 = @xmath133 ) .
therefore , it is not possible to determine the moment direction uniquely within the @xmath25 plane from the present powder data .
single crystal neutron diffraction is required to determine the moment direction .
nevertheless , this powder diffraction investigation confirms that the moments are arranged parallel to each other within the @xmath25 plane and antiparallel to each other along the @xmath11 axis .
the calculated pattern considering moments entirely along the @xmath48 or @xmath51 axis is shown along with observed pattern in fig .
[ fig : neu - smvo](c ) . a schematic magnetic structure , considering the moments along the @xmath51 axis ,
is shown in fig .
[ fig : mag - struc - smvo ] .
ccccccccc irs & & & & atoms & & & & + & mn1 & mn2 & mn3 & mn4 & mn5 & mn6 & mn7 & mn8 + @xmath102 & 1 & 1 & 1 & 1 & 1 & 1 & 1 & 1 + + @xmath103 & 1 & 1 & 1 & 1 & -1 & -1 & -1 & -1 + + @xmath104 & 1 & 1 & -1 & -1 & 1 & 1 & -1 & -1 + + @xmath105 & 1 & 1 & -1 & -1 & -1 & -1 & 1 & 1 + + @xmath106 & 1 0 & -1 0 & 0 1 & 0 -1 & 1 0 & -1 0 & 0 1 & 0 -1 + & 0 1 & 0 -1 & -1 0 & 1 0 & 0 -1 & 0 1 & 1 0 & -1 0 + cccccccccc irs & & + & & mn1 & mn2 & mn3 & mn4 & mn5 & mn6 & mn7 & mn8 + @xmath102 & @xmath109 & ( 100)&(-100)&(010)&(0 - 10)&(100 ) & ( -100)&(0 - 10 ) & ( 010 ) + & @xmath110 & ( 010 ) & ( 0 - 10 ) & ( -100 ) & ( 100 ) & ( 0 - 10 ) & ( 010 ) & ( -100 ) & ( 100 ) + & @xmath111 & ( 001 ) & ( 001 ) & ( 001 ) & ( 001 ) & ( 00 - 1 ) & ( 00 - 1 ) & ( 00 - 1 ) & ( 00 - 1 ) + + @xmath103 & @xmath109 & ( 100 ) & ( -100 ) & ( 010 ) & ( 0 - 10 ) & ( -100 ) & ( 100 ) & ( 010 ) & ( 0 - 10 ) + & @xmath110&(010 ) & ( 0 - 10 ) & ( -100 ) & ( 100 ) & ( 010 ) & ( 0 - 10 ) & ( 100 ) & ( -100 ) + & @xmath111&(001 ) & ( 001 ) & ( 001 ) & ( 001 ) & ( 001 ) & ( 001 ) & ( 001 ) & ( 001 ) + + @xmath104 & @xmath109 & ( 100 ) & ( -100 ) & ( 0 - 10 ) & ( 010 ) & ( 100 ) & ( -100 ) & ( 010 ) & ( 0
- 10 ) + & @xmath110&(010 ) & ( 0 - 10 ) & ( 100 ) & ( -100 ) & ( 0 - 10 ) & ( 010 ) & ( 100 ) & ( -100 ) + & @xmath111&(001 ) & ( 001 ) & ( 00 - 1 ) & ( 00 - 1 ) & ( 00 - 1 ) & ( 00 - 1 ) & ( 001 ) & ( 001 ) + + @xmath105 & @xmath109 & ( 100 ) & ( -100 ) & ( 0 - 10)&(010 ) & ( -100 ) & ( 100 ) & ( 0 - 10 ) & ( 010 ) + & @xmath110&(010 ) & ( 0 - 10 ) & ( 100 ) & ( -100 ) & ( 010 ) & ( 0 - 10 ) & ( -100 ) & ( 100 ) + & @xmath111&(001 ) & ( 001 ) & ( 00 - 1 ) & ( 00 - 1 ) & ( 001 ) & ( 001 ) & ( 00 - 1 ) & ( 00 - 1 ) + + @xmath106 & @xmath109 & ( 100 ) & ( 100 ) & ( 000 ) & ( 000 ) & ( 100 ) & ( 100 ) & ( 000 ) & ( 000 ) + & @xmath110&(010 ) & ( 010 ) & ( 000 ) & ( 000 ) & ( 0 - 10 ) & ( 0 - 10 ) & ( 000 ) & ( 000 ) + & @xmath111&(001 ) & ( 00 - 1 ) & ( 000 ) & ( 000 ) & ( 00 - 1 ) & ( 001 ) & ( 000 ) & ( 000 ) + & @xmath112&(000 ) & ( 000 ) & ( 010 ) & ( 010 ) & ( 000 ) & ( 000 ) & ( 0 - 10 ) & ( 0 - 10 ) + & @xmath113&(000 ) & ( 000 ) & ( -100 ) & ( -100 ) & ( 000 ) & ( 000 ) & ( -100 ) & ( -100 ) + & @xmath114&(000 ) & ( 000 ) & ( 001 ) & ( 00 - 1 ) & ( 000 ) & ( 000 ) & ( 00 - 1 ) & ( 001 ) + & @xmath115&(000 ) & ( 000 ) & ( 0 - 10 ) & ( 0 - 10 ) & ( 000 ) & ( 000 ) & ( 0 - 10 ) & ( 0 - 10 ) + & @xmath116&(000 ) & ( 000 ) & ( 100 ) & ( 100 ) & ( 000 ) & ( 000 ) & ( -100 ) & ( -100 ) + & @xmath117&(000 ) & ( 000 ) & ( 00 - 1 ) & ( 001 ) & ( 000 ) & ( 000 ) & ( 00 - 1 ) & ( 001 ) + & @xmath118&(100 ) & ( 100 ) & ( 000 ) & ( 000 ) & ( -100 ) & ( -100 ) & ( 000 ) & ( 000 ) + & @xmath119&(010 ) & ( 010 ) & ( 000 ) & ( 000 ) & ( 010 ) & ( 010 ) & ( 000 ) & ( 000 ) + & @xmath120&(001 ) & ( 00 - 1 ) & ( 000 ) & ( 000 ) & ( 001 ) & ( 00 - 1 ) & ( 000 ) & ( 000 ) + [ bp ] for further verification of the magnetic structure , we have performed an sa analysis ( see appendix [ sa_scvo - smvo ] ) .
the solution of the sa gives a magnetic structure where spins are preferably arranged antiferromagnetically along the @xmath11 axis and ferromagnetically within a given @xmath25 plane .
the solution also indicates that the moments are preferably aligned perpendicular to the @xmath11 axis and mostly along the @xmath48 or @xmath51 axis .
the solution of sa , therefore , agrees well with the magnetic structure obtained from the representation analysis ( corresponds to the @xmath125 ) .
the refined value of the magnetic moment is found to be 3.99 @xmath63 0.01 @xmath86/mn@xmath20 at 1.5 k. this value is quite small compared to the fully ordered moment of 5 @xmath86 ( mn@xmath20 : 3@xmath30 ; @xmath31 ) .
the reduced ordered moment even at 1.5 k ( t / t@xmath74 @xmath134 0.036 ) suggests the presence of quantum fluctuations due to the quasi-1d magnetic interactions .
however , all magnetic peaks are found to be instrumental resolution limited suggesting 3d long - range afm order at 1.5 k. both the magnetization and neutron diffraction studies reveal that the two iso - structural compounds srco@xmath1v@xmath1o@xmath2 and srmn@xmath1v@xmath1o@xmath2 have different magnetic ground states . unlike srmn@xmath1v@xmath1o@xmath2
, the magnetic structure of srco@xmath1v@xmath1o@xmath2 , which consists of a ferro-/antiferromagnetic arrangement of afm chains within the tetragonal plane ( @xmath25 plane ) is unusual for a tetragonal structure .
two explanations for this situation may be considered as ( i ) the presence of competing interchain interactions along the diagonal ( @xmath135 ) direction within the @xmath25 plane and/or ( ii ) a structural transition from the tetragonal to an orthorhombic symmetry which accompanies the magnetic ordering . for the former case , the competition between afm @xmath136 ( along the principle axes @xmath48 and @xmath51 ) and afm @xmath137 ( along the diagonal direction in the @xmath25 plane ) results in magnetic frustration for a square - lattice ( tetragonal symmetry ) . for a frustrated square - lattice system ,
the @xmath138-@xmath139 model ( where @xmath138 and @xmath139 are the exchange interactions along the side and diagonal of the square ) , predicts a collinear stripe - like afm structure for @xmath139/@xmath138 @xmath21 0.6 .
the collinear stripe structure is occurred by an order by disorder phenomenon.@xcite however , it should be noted that the spin - geometry within a given @xmath25 plane for the studied compounds is more complex [ fig .
[ fig : crysstruc](c ) ] than a simple square - lattice .
furthermore , the dominant interaction is the intrachain afm interaction which is expected to be an order of magnitude greater than the interchain interactions .
moreover , the crystal structure suggests the possibility of a complex interchain interaction topology with many competing interactions ; not only within the @xmath25 plane but also with an out - of - plane component .
the screw chains also may allow second - nearest - neighbor intrachain interactions along the @xmath11 axis . to investigate the interchain interactions in detail , inelastic neutron scattering studies on single crystal sample
are required .
the reduction of the magnetic symmetry can also be explained by a finite magnetoelastic coupling within the @xmath25 plane . in this case , the effective @xmath136 interactions along two principle axes become unequal due to lattice distortions .
however , the lattice distortions , if any , are rather weak and , therefore , not evident in the present neutron diffraction data which has moderate resolution .
note that , neither the sign nor the magnitude of the change of @xmath136 as a function of distance are known . to search for a possible distortion ,
high resolution diffraction studies as a function of temperature by using either neutron or synchrotron radiation are desired . in order to understand further the nature of magnetic orderings ,
we have carried out a temperature - dependent neutron diffraction study .
figures [ fig : criti - expt ] ( a ) and ( b ) show the temperature dependence of the normalized square root of the integrated intensities of a few selected magnetic bragg peaks of srco@xmath1v@xmath1o@xmath2 and srmn@xmath1v@xmath1o@xmath2 , respectively .
the integrated intensities were obtained from gaussian function fit of the magnetic satellite peaks . the fits
confirm that the positions of the magnetic peaks do not change with temperature proving that there is no change of the @xmath3-vectors for both compounds .
the intensity of each magnetic peak was normalized with respect to its intensity at the base temperature 1.5 k. these measurements allowed us to determine the critical exponent @xmath92 for the temperature induced phase transition from the long - range afm ordered state to the paramagnetic state .
the order parameter ( i.e. , the ordered afm moment _ m _ ) can be expressed as @xmath140 .
the ordered magnetic moment is proportional to the square root of the integrated intensity of the magnetic peaks .
thus the critical exponent @xmath92 can be obtained by fitting the square root of the integrated intensity to the following equation @xmath141 where _ a _ is a proportionality constant and @xmath5 is the 3d afm ordering temperature . for srco@xmath1v@xmath1o@xmath2 ,
the fit of the above equation to the observed data over a limited temperature range near @xmath5 ( @xmath142 ) is shown by solid red curve in fig .
[ fig : criti - expt](a ) .
the fitted values of @xmath92 and @xmath5 are found to be 0.33 @xmath63 0.03 and 5.21 @xmath63 0.03 k , respectively .
the fitted value of @xmath92 = 0.33 is close to the value @xmath143 @xmath144 predicted for the 3d ising model .
@xcite the slightly higher value of @xmath92 may be due to the deviation from the pure ising limit as evident in our temperature - dependent susceptibility study .
for srmn@xmath1v@xmath1o@xmath2 , the critical exponent @xmath92 and @xmath5 are 0.36 @xmath63 0.03 and 42.2 @xmath63 0.02 k , respectively , obtained from the fit of the data in the temperature range @xmath145 [ fig .
[ fig : criti - expt](b ) ] .
the fitted value of @xmath92 agrees well with the theoretically expected value of 0.367 for the 3d heisenberg model.@xcite now , we focus on the nature of the magnetic correlations above and around @xmath5 . the magnetic patterns at 20 , 15 , 10 , 6 , 5 , and 4 k for srco@xmath1v@xmath1o@xmath2 and at 150 , 100 , 75 , 60 , 50 , 45 , 40 , and 30 k for srmn@xmath1v@xmath1o@xmath2 are shown in figs .
[ fig : diffscat ] ( a ) and ( b ) , respectively .
the magnetic patterns were obtained by subtraction of the paramagnetic background at 30 k for the co- and at 200 k for the mn - compound , respectively . for both compounds ,
the presence of diffuse magnetic scattering due to short - range spin - spin correlations is evident over a wide temperature range above and below @xmath5 .
this is in agreement with the @xmath40(t ) study where broad peaks , indicating the short - range spin - spin correlations , appeared centered at @xmath4 30 k for the co - compound and at @xmath4 170 k for the mn - compound , respectively [ fig .
[ fig : sus ] ] . for the co - compound , below 20 k
, a broad diffuse peak appears at @xmath146 @xmath32 . on lowering the temperature ,
the peak intensity increases continuously . at @xmath147 6 k
, an additional broad peak appears at @xmath148 @xmath32 .
when the temperature is lowered further , the broad peaks gradually transform into a set of bragg peaks below @xmath149 k. the complete transformation occurs at @xmath4 4 k resulting in a coexistence of both broad diffuse peaks and sharp bragg peaks over the temperature range 5.24 k. for the mn - compound , with decreasing temperature below 150 k a broad diffuse peak appears at @xmath150 @xmath32 . with
further lowering of temperature , the peak intensity increases continuously and gradually transforms into bragg peaks below @xmath151 k. the broad diffuse peak is found to coexist below @xmath5 down to @xmath4 30 k. from this diffuse scattering study , it is evident that the transition from short - range to long - range magnetic ordering is not sharp but rather occurs gradually over a wide temperature range in both compounds .
the short - range spin - spin correlations are expected to be one dimensional ( due to the strong intrachain interactions ) and appear when the interaction strength becomes comparable to the thermal energy . with further lowering of temperature ,
the thermal energy becomes comparable to the interchain interaction strength and 3d long - range magnetic ordering starts to occur at @xmath5 . to investigate the exact nature of the short - range correlations , neutron diffraction studies on single crystal are required
nevertheless , this study confirms the presence of short - range correlations over a wide temperature range above and below @xmath5 for both compounds .
in summary , we have investigated magnetic correlations in two iso - structural compounds sr@xmath0v@xmath1o@xmath2 with different magnetic ions co@xmath20 and mn@xmath20 by using dc - magnetization and neutron diffraction .
the crystal structures of both compounds have tetragonal symmetry with space group @xmath56 .
however , an expansion of unit cell volume from the co - compound to the mn - compound has been found due to the ionic radii effect .
both magnetization and neutron diffraction studies reveal that two compounds have different magnetic properties and different afm ground states . for the mn - compound ,
afm chains ( along the _ c _ axis ) are ordered ferromagnetically within the _ ab _ plane , whereas , for the co - compound , afm chains are ordered ferro-/antiferromagnetically along the _ a / b _ direction .
reduced ordered moments have been found for both compounds at base temperature 1.5 k revealing the presence of strong quantum fluctuations due to the quasi-1d magnetic interactions .
critical exponent study indicates that the co- and mn - compounds belong to the ising and heisenberg universality classes , respectively . for both compounds ,
the presence of short - range spin - spin correlations have been found over a wide temperature range due to the quasi - one - dimensional magnetic interactions .
the symmetry elements of the space group @xmath153 ( group g@xmath152 ) is given in table [ symmetry elements_of_g0_table ] .
the simulated annealing ( sa ) analysis was performed using the fullprof program suite@xcite to verify the magnetic structures obtained from the representation analyses .
this is a very general method to determine the magnetic structure .
it uses an adaption of the metropolis algorithm ( a monte carlo method ) to generate sample states of a thermodynamic system .
it does not depend on the initial parameters and is therefore ideal for solving complex and/or completely unknown structures . in detail
, the sa method is based on the concept of minimizing a @xmath154 function @xmath155(@xmath156 ) with respect to a given configuration vector @xmath156.@xcite for the input of the sa analysis , the integrated intensities of the magnetic peaks were determined from the 1.5 k patterns by using the mixed rietveld and profile matching methods in the fullprof program . here , all structural parameters for the nuclear phases were fixed ( please note that for srco@xmath1v@xmath1o@xmath2 , additional parameters for nuclear and magnetic phases of coo were also fixed ) and the magnetic contributions were introduced as a new phase using the le bail fit mode .
the sa analysis was , then , done without considering any symmetry constrain .
the only constrain was the same magnitude of the moments for all eight independent magnetic sites within a unit cell .
this gives total 17 refinable parameters ; one magnetic moment parameter and 8@xmath422 = 16 angular parameters .
one of the solutions each for srco@xmath1v@xmath1o@xmath2 and srmn@xmath1v@xmath1o@xmath2 are listed in table [ tab : t9 ] and table [ tab : t10 ] , respectively .
the solution for srco@xmath1v@xmath1o@xmath2 corresponds to a magnetic structure in which the moments are arranged antiferromagntically along the @xmath11 axis and ferro / antiferro - magnetically along the @xmath27 axis with moments aligned along the @xmath11 axis .
the solution for srmn@xmath1v@xmath1o@xmath2 corresponds to a magnetic structure where moments make an almost parallel arrangement within a given @xmath25 plane and an antiparallel arrangement of such planes along the @xmath11 axis .
the moments are aligned mostly along the @xmath27 axis .
36ifxundefined [ 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 10.1088/0953 - 8984/1/19/001 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.50.1153 [ * * , ( ) ] link:\doibase 10.1016/0375 - 9601(83)90631-x [ * * , ( ) ] link:\doibase 10.1103/physrevb.21.2001 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.78.1787 [ * * , ( ) ] link:\doibase 10.1103/physrevb.53.2583 [ * * , ( ) ] link:\doibase 10.1103/physrevb.62.8921 [ * * , ( ) ] link:\doibase 10.1103/physrevb.87.224423 [ * * , ( ) ] link:\doibase 10.1002/zaac.19865340320 [ * * , ( ) ] @noop * * , ( ) link:\doibase 10.1103/physrevb.69.220407 [ * * , ( ) ] link:\doibase 10.1103/physrevb.72.172403 [ * * , ( ) ] link:\doibase 10.1103/physrevb.73.212406 [ * * , ( ) ] link:\doibase 10.1016/j.jmmm.2011.05.056 [ * * , ( ) ] link:\doibase 10.1103/physrevb.86.024408 [ * * , ( ) ] @noop ( ) , link:\doibase 10.1103/physrevb.66.014428 [ * * , ( ) ] link:\doibase 10.1088/1742 - 6596/51/1/021 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.99.087602 [ * * , ( ) ] link:\doibase 10.1107/s0567739476001551 [ * * , ( ) ] @noop `` , '' \doibase http://dx.doi.org/10.1016/j.jcrysgro.2006.05.039 [ * * , ( ) ] \doibase http://dx.doi.org/10.1016/j.jcrysgro.2011.01.016 [ * * , ( ) ] link:\doibase 10.1103/physrev.135.a640 [ * * , ( ) ] link:\doibase 10.1103/physrevb.52.15992 [ * * , ( ) ] http://stacks.iop.org/0953-8984/25/i=6/a=065601 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.93.217204 [ * * , ( ) ] link:\doibase 10.1103/physrevb.64.052102 [ * * , ( ) ] \doibase http://dx.doi.org/10.1016/s0921-4526(99)01722-6 [ * * , ( ) ] @noop _ _ ( , ) link:\doibase 10.1103/physrevb.83.064421 [ * * , ( ) ]
\doibase http://dx.doi.org/10.1016/0921-4526(93)90108-i [ * * , ( ) ] link:\doibase 10.1103/physrevlett.97.157201 [ * * , ( ) ] @noop _ _ ( , ) @noop _ _ ( , ) link:\doibase 10.1126/science.220.4598.671 [ * * , ( ) ] | magnetic correlations of two iso - structural quasi - one - dimensional ( 1d ) antiferromagnetic spin - chain compounds sr@xmath0v@xmath1o@xmath2 ( _ m _ = co , mn ) have been investigated by magnetization and powder neutron diffraction .
two different collinear antiferromagnetic ( afm ) structures , characterized by the propagation vectors , @xmath3 = ( 0 0 1 ) and @xmath3 = ( 0 0 0 ) , have been found below @xmath4 5.2 k and @xmath4 42.2 k for the co- and mn - compounds , respectively .
for the mn - compound , afm chains ( along the _ c _ axis ) order ferromagnetically within the _ ab _ plane , whereas , for the co - compound , afm chains order ferro-/antiferromagnetically along the _ a / b _ direction .
the critical exponent study confirms that the co- and mn - compounds belong to the ising and heisenberg universality classes , respectively . for both compounds ,
short - range spin - spin correlations are present over a wide temperature range above @xmath5 .
the reduced ordered moments at base temperature ( 1.5 k ) indicate the presence of quantum fluctuations in both compounds due to the quasi-1d magnetic interactions . |
probably the most significant result of the gromov analysis of classical networks @xcite is existence of a congestion core . under a network protocol that sends the packets along least cost paths
, the _ core _ can be qualitatively defined as a point where most of the geodesics ( least cost paths ) converge , creating packet drops , high retransmission rates , and other nuisances under the tcp - ip protocol @xcite .
existence of the core has been experimentally observed @xcite and mathematically proved @xcite if the network is gromov hyperbolic , subject to some highly technical conditions related to the gromov boundary @xcite .
a gromov hyperbolic network can intuitively be defined as a network that `` looks like '' a negatively curved riemannian manifold ( e.g. , a saddle ) when viewed from a distance .
see , e.g. , @xcite for a precise definition .
next to classical networks , one can envision quantum networks : the nodes are spins that can be up @xmath4 ( not excited ) or down @xmath5 ( excited ) and the links are quantum mechanical couplings of the xx or heisenberg type . given some random source - destination pair @xmath6 , a valid question is whether some spin @xmath7 could act as a `` core , '' that is , a spin that could be excited no matter what the source and the destination are . for a linear chain , one would expect such a congestion core in the center as classically any excitation in one half of the chain would have to transit the center of the chain to reach the other half . in this work ,
we demonstrate that quantum - mechanically the transmission of excitations does not need to occur this way , and in fact the center @xmath7 of an odd - length spin chain can act as an `` anti - core , '' excitation of which is avoided . this anti - core , " or `` anti - gravity '' center as it was originally called , was first observed in @xcite . the anti
- core @xmath7 was defined as a point of high inertia , @xmath8 , @xmath9 , as opposed to the classical congestion core that has minimum inertia owing to the negative curvature of the underlying space ( * ? ? ?
* theorem 3.2.1 ) .
the inertia quantifies how difficult communication to and from the anti - core is . as it has been done along this line of work , a pre - metric @xmath10 based on the _ information transfer capacity ( itc ) _ ( see sec . [
s : homogeneous ] ) is employed .
unlike standard quantum mechanical distances @xcite , @xcite , this distance " measure aims to quantify not how distant two fixed quantum states are , but how close to a desired target state a quantum state can get under the evolution of a particular hamiltonian .
the initial and target states are typically orthogonal . in this paper , we provide an _ analytical _ justification of the numerically observed anti - core phenomenon in spin chains with xx coupling , starting with finite - length chains , extending the itc concept to semi- and bi - infinite cases ( sec .
[ s : infinite_chains ] ) , and finally _ proving _ that @xmath11 , @xmath12 in sec .
[ s : anti_core ] .
we further show that by adding a bias on the central spin its _ `` anti - core '' _ property can be made stronger in the sense that the probability of transmission of the excitation to and from it is infinitesimally small ( sec .
[ s : engineered ] ) .
the remaining nagging question is why was it observed in @xcite that spin chains appear gromov - hyperbolic and have an anti - core , while classical networks are gromov hyperbolic with the opposite core ?
this will be clarified in sec .
[ s : engineered ] by means of a spin chain example , showing that its gromov boundary has _ only one _ point , while classical networks need to have _ at least two points _ in their gromov boundary for the core to emerge .
we consider a linear array of two - level systems ( spin @xmath13 particles ) with uniform coupling between adjacent spins ( homogeneous spin chain ) made up of an odd number @xmath14 of physically equally spaced spins with coupling hamiltonian @xmath15 here we shall be primarily interested in the case of xx coupling , for which @xmath16 .
the factor @xmath17 is the pauli matrix along the @xmath18 or @xmath19 direction of spin @xmath1 in the array , i.e. , @xmath20 where the factor @xmath21 occupies the @xmath1th position among the @xmath14 factors and @xmath21 is one of the single spin pauli operators @xmath22 it is easily seen that @xmath23 is _ real _ and symmetric .
the @xmath24 hamiltonian commutes with the operator @xmath25 which counts the total number of excitations .
the hilbert space can therefore be decomposed into subspaces corresponding to the number of excitations .
define @xmath26 to be the quantum state in which the excitation is on spin @xmath1 .
the single excitation subspace @xmath27 is spanned by @xmath28 .
restricted to this subspace , the hamiltonian in this natural basis takes the form @xmath29 for xx coupling ( @xmath16 ) , @xmath30 becomes the @xmath31 toeplitz matrix @xmath32 made up of zeros on the diagonal , ones on the super- and subdiagonal and zeros everywhere else .
table [ t : eigenstructure ] shows the eigenvalues and eigenvectors of @xmath30 . in anticipation of letting @xmath33 ,
define @xmath34 and the preceding sums can be rewritten as @xmath35}_{\mathrm{max}}(i',j ' ) } & = 2\sum_{k'=-n,\mathrm{even}}^{+n } \left|f\left ( 2 \pi x_{k'}i'\right ) g\left ( 2\pi x_{k ' } j'\right)\right|\left(x_{k'+2}-x_{k ' } \right)\\ & \quad + 2 \sum_{k'=-n,\mathrm{odd}}^{+n } \left|\bar{f}\left ( 2 \pi x_{k'}i ' \right ) \bar{g}\left ( 2\pi x_{k ' } j ' \right)\right| \left(x_{k'+2}-x_{k ' } \right ) .
\end{aligned}\ ] ] letting @xmath33 yields @xmath36}_{\mathrm{max}}(i',j')}\\ & = 2\int_{-1/4}^{+1/4 } |f(2\pi xi')g(2\pi x j')| dx + 2\int_{-1/4}^{+1/4}|\bar{f}(2\pi xi')\bar{g}(2\pi x j')| dx.\end{aligned}\ ] ] in order to make the integrations more straightforward and to follow a procedure that parallels the one of appendix [ s : semi_infinite_chains ] , it is convenient to change the integration limits by making use of the periodicity of the integrands as functions of @xmath37 .
observe that both @xmath38 and @xmath39 have decompositions in terms of sines or cosines of arguments @xmath40 .
write the generic term as @xmath41 . if @xmath42 is
even , observe that @xmath43 . if @xmath42 is odd , @xmath44 . in either case
, @xmath45 and @xmath46 have period @xmath47 . with this property ,
the previous integrals can be rewritten as @xmath48 observe that @xmath49 , as easily seen from a cauchy - schwarz argument .
also observe that @xmath50 ; indeed , if @xmath51 , table [ t : in_terms_of ] reveals that the integrands are of the form @xmath52 or @xmath53 , from which the assertion is trivial .
the next step is to make @xmath54 , @xmath55 , @xmath56 , @xmath57 more manageable by expressing them as @xmath58 , @xmath59 , @xmath60 , @xmath61 if @xmath62 , @xmath63 , @xmath64 , @xmath65 are sines and by @xmath66 , @xmath67 , @xmath68 , @xmath69 if they are cosines . in the preceding ,
@xmath70 and @xmath71 are odd and even , resp .
, square waves of unit amplitude and of period @xmath72 , with fourier decompositions @xmath73 if @xmath74 and @xmath75 are even , and if we let @xmath33 along the even number subsequence of @xmath76 , we need to take @xmath77 and @xmath78 , as seen from table [ t : in_terms_of ] , together with @xmath79 and @xmath80 .
if on the other hand , we let @xmath81 along the odd number subsequence of @xmath76 , we need to take @xmath82 and @xmath83 , together with @xmath84 and @xmath85 .
however , because of the symmetry of formula , both subsequences yield the same result : @xmath86 from table [ t : in_terms_of ] , we could let @xmath33 along the even number subsequence of @xmath76 , in which case we need to take @xmath82 and @xmath78 .
if we let @xmath81 along the odd number subsequence of @xmath76 , we need to take @xmath77 and @xmath87 . in either case , because of the symmetry of , the result is the same and is given by @xmath88 from the above , it follows that all cases share a few quadrature integrals : @xmath89 the right - hand side of the last equality involves expressions like @xmath90 . to simplify the notation
, we wrote such expressions as @xmath91 . here we proceed from the general formula for @xmath95 , utilize the quadrature integrals of the preceding section , and derive , first , an infinite series representation of the asymptotic maximum transfer probability and , finally , a representation in terms of special functions .
since the general formula is in terms of function @xmath62 , @xmath63 , @xmath64 , @xmath65 that depend on whether @xmath74 and @xmath75 are even or odd ( see section [ s : consistency ] ) , it is necessary to examine each case in particular . from section [ s : consistency ]
, it follows that the case where both @xmath74 and @xmath75 are even and the case where both @xmath74 and @xmath75 are odd are the same . from the same section [ s : consistency ] , the case where @xmath74 is
even and @xmath75 odd is the same as the case where @xmath74 odd and @xmath75 even , but is different from the preceding one .
so , there are essentially two cases to be distinguished .
[ [ both - i - and - j - even - or - both - i - and - j - odd ] ] both @xmath74 and @xmath75 even or both @xmath74 and @xmath75 odd ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ where @xmath97 @xmath98 and @xmath10 are functions taking value @xmath99 or @xmath72 , and complementary in the sense that @xmath100 .
next , we find that @xmath101 where @xmath102 observe that @xmath103 , @xmath104 are relatively prime ; hence they could not be both even .
the developments follow closely the semi - infinite chain case , except that , here , @xmath103 and @xmath104 are not restricted to be positive .
hence we have to consider several cases : here we have to make a distinction between the two cases : both @xmath74 and @xmath75 odd and both @xmath74 and @xmath75 even .
we start with the easy case where in this case indeed , both @xmath109 is even .
this along with @xmath110 is even yields @xmath111 putting everything together , we find , using partial fraction decompositions , @xmath112 the case where is more complicated . if @xmath74 and @xmath75 have the same power of @xmath114 in their prime number factorization , then @xmath115 is even and the preceding formula holds .
if the powers of @xmath114 are different , then @xmath115 is odd and it is easily verified that @xmath116 with the above , we get @xmath117 despite the extra difficulties created by the @xmath98 and @xmath10 functions and the various signs , the pattern remains the same as before : the indicators are nonvanishing only if @xmath126 for some even @xmath110 .
since @xmath74 is odd and @xmath75 is even , @xmath128 does not contain any positive power of @xmath114 in its prime divisors ; therefore , @xmath103 remains odd and @xmath104 remains even ; in other words , @xmath129 is odd . from this observation ,
tedious but elementary manipulations lead to the following : @xmath116 e. jonckheere , f. c. langbein , and s. g. schirmer .
curvature of quantum rings . in _ proceedings of the 5th international symposium on communications , control and
signal processing ( isccsp 2012 ) _ , rome , italy , may 2 - 4 2012 .
e. jonckheere , s. schirmer , and f. langbein . geometry and curvature of spin networks . in _
ieee multi - conference on systems and control _ , pages 786791 , denver , co , september 2011 .
available at arxiv:1102.3208v1 [ quant - ph ] . | the purpose of this paper is to exhibit a quantum network phenomenon the anti - core that goes against the classical network concept of congestion core .
classical networks idealized as infinite , gromov hyperbolic spaces with least - cost path routing ( and subject to a technical condition on the gromov boundary ) have a congestion core , defined as a subnetwork that routing paths have a high probability of visiting . here
, we consider quantum networks , more specifically spin chains , define the so - called maximum excitation transfer probability @xmath0 between spin @xmath1 and spin @xmath2 , and show that the central spin has among all other spins the lowest probability of being excited or transmitting its excitation .
the anti - core is singled out by analytical formulas for @xmath3 , revealing the number theoretic properties of quantum chains . by engineering the chain ,
we further show that this probability can be made vanishingly small . |
[ [ adiabatic - elimination ] ] adiabatic elimination
+ + + + + + + + + + + + + + + + + + + + + + + + in order to derive the effective equations for diagonal elements of the density matrix ( eq .
( 2 ) of the main text ) , we represent @xmath15 in the fock basis as @xmath124 .
the @xmath125 and @xmath126 are vectors that describe the distribution of the atoms on the lattice sites , i.e. @xmath127 where @xmath128 is the number of atoms on site @xmath129 .
this representation is advantageous , since the decoherence free subspace of the dissipator corresponds to the density matrices with only nonvanishing diagonal entries .
this decoherence free subspace dominates the long - time dynamics .
quantum fluctuations around the decoherence free subspace can be included perturbatively in the hopping term using the adiabatic elimination method @xcite .
this leads , to second order in the kinetic term of the hamiltonian , to a virtual tunneling process of a particle to the neighbouring site , then it dissipates ( at a rate @xmath22 ) or dephases ( due to the interaction @xmath130 ) , and eventually tunnels another time .
for example , for the case of the tunneling of an atom from site @xmath131 with @xmath132 atoms to site @xmath129 with @xmath128 atoms , we can compute the time dependence of the off - diagonal terms by integrating @xmath133 \rho^{{\boldsymbol{n}}+{\boldsymbol{e}}^1_{l , l'}}_{{\boldsymbol{n}}}(t ) \nonumber \\ & - j\sqrt{(n_l+1)n_{l ' } } \left [ \rho^{{\boldsymbol{n}}}_{{\boldsymbol{n}}}(t ) - \rho^{{\boldsymbol{n}}+{\boldsymbol{e}}^1_{l , l'}}_{{\boldsymbol{n}}+{\boldsymbol{e}}^1_{l , l'}}(t ) \right ] , \tag{1s } \end{aligned}\ ] ] which is derived from eq . ( 1 ) of the main paper . here
we have used the vector @xmath134 which is @xmath135 everywhere except at the neighboring positions @xmath129 and @xmath131 where it takes respectively the values @xmath136 and @xmath137 .
the vector @xmath134 connects two configurations which are related just by the tunneling of a single atom , the only relevant dynamical process that can change the distribution of the atoms in the lattice .
the integration of eq .
( 1s ) is well approximated by [ 2,3 ] @xmath138}{\left[(n_l - n_{l'}+1)u+i\hbar\gamma\right ] } , \tag{2s } \label{eq : off - diago } \end{aligned}\ ] ] since @xmath139 or @xmath130 are much larger than @xmath140 .
the off - diagonal elements can thus be described only by the diagonal ones . when the number of atoms at site @xmath129 changes by @xmath136 and those at the neighboring site @xmath131 by @xmath137 , the evolution of the diagonal terms of the density matrix , belonging to the dissipation - free space , is derived using eq .
( [ eq : off - diago ] ) and eq .
this results in the following set of equations : @xmath141 where @xmath38 , @xmath41 is the filling , @xmath13 is the number of nearest neighbors , @xmath36 with @xmath37 . in the thermodynamic limit @xmath142 , keeping the filling @xmath12 constant , one can separate the diagonal density matrix @xmath143 $ ] . using translation invariance , i.e. @xmath144 for all @xmath6 ,
one obtains a non - linear equation for the reduced single site density matrix which is eq.(2 ) of the main text .
+ in ref .
@xcite , the steady state density matrix @xmath145 had been identified . in order to obtain the one - site reduced density matrix elements @xmath146
, the number of possible configurations on a lattice needs to be taken into account . on a generic lattice ,
the number of configurations in which @xmath10 atoms can be distributed between @xmath11 sites is given by the binomial coefficient @xmath147 .
@xmath27 corresponds to the dimension of the hilbert space .
thus , the one - site reduced density matrix elements are @xmath148 where , given translation invariance , we have simplified the notation using @xmath149 . in the limit @xmath150 and @xmath151 with @xmath41 ( the filling of the lattice ) kept constant , using stirling s approximation @xmath152 , leads to @xmath153 this result is independent of the dimensionality of the system . introducing a rescaled parametrization for the occupation , @xmath154 , we obtain , in the limit of large fillings and using @xmath155 , that @xmath156 this shows that the probability for configurations with a large occupation on a single site are exponentially suppressed . + from eq .
( 2 ) of the main paper , using @xmath45 and @xmath157 , where @xmath158 , it is possible to derive the evolution equation in the continuum limit . taking @xmath159 we can rewrite eq.(2 ) as @xmath160 \nonumber \label{eq : evoeqsupp } \end{aligned}\ ] ] which expanded to second order gives @xmath161 \nonumber \\ & + \left [ \frac{\partial^2p(y)}{\partial y^2}p(x)+ \frac{\partial^2p(x)}{\partial x^2}p(y ) - 2\frac{\partial p(y)}{\partial y}\frac{\partial p(x)}{\partial x } \right ] \nonumber \\ & \times \mathcal{t}_c(x , y ) \bigg\ } dy \nonumber \end{aligned}\ ] ] this can be rewritten in a more compact and elegant form @xmath162 dy \nonumber \end{aligned}\ ] ] and after some manipulation takes the form of the diffusion equation ( 3 ) .
+ from the definition of @xmath56 it follows that the probability to find a certain configuration is normalized to unity , i.e. @xmath163 and it is constant in time .
further , the average rescaled filling @xmath164 is also constant in time , which corresponds to the conservation of particles .
the conservation of both can be shown explicitely : @xmath165_0^{\infty}=0 \nonumber\end{aligned}\ ] ] and @xmath166\right\ } dx \nonumber \\ & = x\left[d(x,\tau)\frac{\partial p(x,\tau)}{\partial x } - f(x,\tau ) p(x,\tau)\right]_0^{\infty } \nonumber \\ & -\!\!\int_0^{\infty}\!\ ! \left[d(x,\tau)\frac{\partial p(x,\tau)}{\partial x } - f(x,\tau ) p(x,\tau)\right]dx \nonumber \\ & = -\!\!\int\!\!\!\!\int_0^{\infty}\!\ !
\left\{\frac{xy\left[p(y,\tau)\partial_x p(x,\tau ) -
\partial_y p(y,\tau ) p(x,\tau ) \right]}{(x - y)^2+\epsilon^2}\right\}dxdy \nonumber \\ & = 0 \nonumber \end{aligned}\ ] ] where we have used the definitions of @xmath50 and @xmath51 given in the main paper ( eq .
+ by using the ansatz @xmath77 and the scaling @xmath167 , eq .
( 3 ) reduces to @xmath168\frac { d g } { d \eta } + \frac { d^2 g } { d \eta^2 } = 0 \tag{9s } \label{eq : diffequg}\ ] ] where the dot stands for the derivative with respect to @xmath7 and @xmath169 . since @xmath78 is only a function of @xmath170 then @xmath171 where @xmath172 and @xmath173 are two constants .
the only physically relevant solution is such that @xmath174 and @xmath175 . using this insight , eq .
( [ eq : diffequg ] ) becomes @xmath176 which has the solution @xmath177 given the boundary conditions , @xmath178 and @xmath179 , we obtain eq . | we study the dynamics of a strongly interacting bosonic quantum gas in an optical lattice potential under the effect of a dissipative environment .
we show that the interplay between the dissipative process and the hamiltonian evolution leads to an unconventional dynamical behavior of local number fluctuations .
in particular we show , both analytically and numerically , the emergence of an anomalous diffusive evolution in configuration space at short times and , at long times , an unconventional dynamics dominated by rare events .
such rare events , common in disordered and frustrated systems , are due here to strong interactions .
this complex two - stage dynamics reveals information on the level structure of the strongly interacting gas .
unconventional , non - exponential , relaxation dynamics of a perturbed system towards equilibrium has attracted a lot of interest over decades . already in 1847 , kohlrausch @xcite observed a stretched exponential decay in time @xmath0 , i.e. @xmath1 with @xmath2 and @xmath3 a positive constant , of the discharge of capacitors fabricated from glasses .
since then , such a decay has been observed in many systems such as molecules and polymers @xcite , spin glasses @xcite , nano - sized magnetic particles @xcite , and certainly amorphous silicon @xcite . a broad variety of theoretical approaches has been developed to explain the mechanism of this unconventional relaxation dynamics @xcite . in many of these approaches ,
e.g. the treatment of the griffiths phase in disordered spin systems @xcite , rare configurations have been identified to play a key role .
these configurations have an exponentially small probability to occur , and therefore contribute minimally to the short - time dynamics .
however , because their relaxation time scale is very long , these rare configurations can dominate the long time evolution .
rare configurations play an important role in the relaxation dynamics of glasses , where they give rise to stretched exponential decays .
we will thus refer to this dynamics induced by rare events as ` glass - like ' in the following . in this work
, we uncover that also in a quantum many body systems , as the bose - hubbard model , the dissipative coupling to a markovian , i.e. memory - less , environment can cause glass - like dynamics .
we show that the long time behaviour in these systems can be dominated by rare configurations .
these rare configurations are characterized by a large number of atoms occupying a single lattice site . increasing the number of atoms on the largely occupied site is associated to a long time scale , since the energetic cost of modifying this kind of configurations is very large .
due to this long time scale , these rare configurations dominate the long time dynamics inducing an unconventional dynamics of stretched exponential form as shown , for the case of local number fluctuations @xmath4 ( where @xmath5 is the number operator of atoms on site @xmath6 ) in fig .
[ fig : figa ] . vs square - root of rescaled time @xmath7 for the interaction over dissipative coupling ratio @xmath8 . numerical results of eq . are shown for different fillings , @xmath9 in the direction of the arrow , as solid lines and corresponding stretched exponential fits as dashed lines . the analytical result ( [ eq : compequa ] ) of the diffusion equation is shown as a ( black ) dot - dashed line . ] additionally , the glass - like dynamics is preceded by an algebraic relaxation process due to the interplay of many energetically close configurations with low particle fluctuations . therefore ,
both unconventional dynamics of this open quantum many body system are signatures of the complex structure of its configuration space and energy spectrum that the dissipative term forces to explore .
the non - exponential decay is in contrast to the typical evolution found for quantum many body systems coupled to a markovian environment . in these systems ,
the decay is often dominated by an exponential dynamics , as e.g. , the counterintuitive zeno effect @xcite , or the relaxation to a desirable state driven by an artificially engineered environment @xcite . only recently ,
first signs of an intriguing slowing down of the heating dynamics for interacting bosonic @xcite and fermionic @xcite gases and an algebraic decay for a specially designed environment which imprints coherence @xcite have been predicted .
however , up to now , the understanding of the large variety of dynamical behaviors that occur in an interacting many body system and of its origin is still a great challenge .
we study the heating of @xmath10 ultracold bosonic atoms in an optical lattice of @xmath11 sites with filling @xmath12 and connectivity ( number of nearest neighbors per site ) @xmath13 , described by the following master equation @xcite : @xmath14+\mathcal{d}\left(\hat{\rho}\right ) .
\label{eq : master}\ ] ] the first term describes the unitary evolution of the density matrix @xmath15 .
this evolution is governed , in the single bloch band limit , by the bose - hubbard hamiltonian @xmath16 where @xmath17 denotes pairs of neighboring sites @xcite .
the operators @xmath18 and @xmath19 are bosonic creation and annihilation operators on site @xmath6 and @xmath20 counts the number of atoms .
the dissipator @xmath21 models the dissipative coupling to a markovian environment via the local density with strength @xmath22 .
this can be due to a noisy potential both in space and time added to the optical lattice @xcite .
we have restricted the description to the lowest bloch band of the optical lattice potential .
the validity of this approximation is discussed in the conclusions . in the following
, we study in detail the heating dynamics of a system , initially in its ground state with respect to @xmath23 , under the joint action of dissipation and the hamiltonian evolution .
we concentrate on the strongly interacting regime @xmath24 .
the dissipator causes the off - diagonal elements of the density matrix , in the following always represented in the fock basis , to decay towards the decoherence free subspace .
this consists of all possible diagonal density matrices @xmath25 . in the presence of the hopping term
, the heating process drives the system to a unique steady state @xmath26 , the highest entropy state @xcite . here , @xmath27 is the dimension of the hilbert space at fixed atom number @xmath10 , and @xmath28 the identity operator .
the approach of this steady state , can be described for @xmath29 , by adiabatically eliminating @xcite the small off - diagonal elements .
a closed set of classical rate equations for the diagonal elements of @xmath25 is obtained @xcite .
the diagonal configurations that are connected are those for which a particle is moved from a site with occupation @xmath30 to one of its neighbors with occupation @xmath31 .
the process occurs via virtual hopping to and from an off - diagonal element of the density matrix @xcite . to study this dynamics
, we use a separable and translationally invariant ansatz @xmath32 $ ] where @xmath6 runs over all the lattice sites and @xmath33 over all the possible occupations of each site .
the probability distribution @xmath34 of the single site occupation evolves as @xmath35 \label{eq : evoeq } \end{aligned}\ ] ] where @xmath36 with @xmath37 and @xmath38 @xcite
. a typical evolution of the occupation number distribution can be acquired by studying fig .
[ fig : fig1 ] .
( solid lines , eq . ( 2 ) ) in a semi - logarithmic plot , versus @xmath33 for large rescaled times @xmath7 between 0.1 and 50 ( not equidistant ) in the direction of the ( red ) arrow .
the inset shows the same evolution at shorter times @xmath7 between 0.0002 and 0.1 ( not equidistant ) in the direction of the ( red ) arrow in a linear plot .
parameters : @xmath39 , @xmath8 .
the ( red ) dashed lines show the ( analytical ) asymptotic limit . ] at short times , but still @xmath40 , the very narrow initial distribution around the average filling @xmath41 broadens almost symmetrically ( see inset of fig .
[ fig : fig1 ] ) .
after the rapid broadening a new regime with an asymmetric evolution sets in , in which the tail of the distribution slowly converges towards the expected asymptotic distribution @xmath42 @xcite .
this means that the probability for states with larger filling is exponentially suppressed , i.e. these states are rare .
note that @xmath43 is exactly the single site reduced density matrix of the full asymptotic density matrix @xmath44 @xcite . to obtain analytical insight into the very different regimes of the evolution ,
we take the continuum limit of eq .
( [ eq : evoeq ] ) for large @xmath41 . the continuous on - site occupation number distribution @xmath45 and thus @xmath46 .
hence one derives the non - linear integro - differential equation @xcite @xmath47 .
\label{eq : diffequ}\ ] ] here @xmath48 and @xmath49 .
the peculiar form of @xmath50 and @xmath51 stems from the configuration dependent rates and triggers a wide range of rich phenomena .
note that the structure of ( [ eq : df ] ) ensures that both the total probability ( @xmath52 ) and the average population ( @xmath53 ) are conserved quantities @xcite .
further , it can be checked that the asymptotic solution of ( [ eq : diffequ ] ) is @xmath54 , which is the continuum limit of the steady state in the large-@xmath41 limit @xcite .
the continuum description is justified for @xmath41 large and finite @xmath55 , assuming that @xmath56 varies smoothly enough on scales of the order of @xmath57 . in the present case ,
the strongest variations of distributions are due to the initial state , especially for a low filling @xmath41 . after this initial stage
@xmath56 smoothens out rather quickly and the continuum description is highly accurate over a wide time range . in the following
we solve analytically the diffusion equation ( [ eq : diffequ ] ) in the short time and in the long time limits focusing on the evolution of the particle distribution and the local density fluctuations @xmath58 .
_ short time relaxation : _ within the diffusion equation ( [ eq : diffequ ] ) , initially the distribution @xmath59 is strongly peaked and symmetric around the value @xmath60 .
for such a distribution the force is negligible compared to the diffusion function .
the diffusion equation at @xmath61 can be approximated by @xmath62 $ ] .
this leads to a dynamics which , in the analytically solvable limit @xmath63 , is given by an anomalous diffusion of the form @xmath64 .
@xmath65 is the gamma function @xcite . versus rescaled time @xmath7 for @xmath8 .
numerical results ( solid lines ) for various fillings , @xmath66 in the direction of the arrow , are obtained solving eq . .
the approximate analytical solution @xmath67 of the diffusion equation is represented by the dashed line . ] using this analytical solution for the particle distribution , the local number fluctuations @xmath68 exhibit a power - law relaxation with @xmath67 .
this analytical result is in excellent agreement with the numerical results shown in fig .
[ fig : compressibility ] obtained by solving eq .
( [ eq : evoeq ] ) .
deviations are found at small fillings and short times , where the approximation @xmath69 is not justified .
however , for larger fillings , for example @xmath70 , the time - regime in which the power - law decay appears is already large .
physically , this very rapid initial broadening of the particle distribution translates into the fast creation of small particle fluctuations around the average value caused by the heating .
these fluctuations arise via virtual excitations of low energetic cost of order o(u ) which thus can be reached rapidly .
this behavior is similar to the dynamics observed in a double well potential @xcite . _
long time relaxation : _ the obtained short time solution breaks down as the distribution approaches the reflective boundary at @xmath71 @xcite ; the distribution is no longer symmetric around @xmath60 , and the combined action of the force term with the diffusion drives the system towards its large time asymptotics @xmath72 .
physically , the exponential suppression of large values of @xmath73 corresponds to the rareness of the states with a high number of particles accumulated on a single site .
therefore , naively one expects that their effect is overwhelmed by the much more numerous states at low filling .
however , the rare states are associated with a decaying small diffusion function and force given by @xmath74 leading to the slow occupation of the states with large @xmath73 . due to these large time scales ,
these rare states are found to dominate the long time dynamics despite their exponentially suppressed probability to occur .
the underlying quantum mechanical process behind this slow diffusion is the large energy cost of the virtual states via which the diffusion processes at high @xmath73 take place .
the form of the forcing term @xmath75 in the large @xmath73 limit brings connections to other intriguing physical problems .
one example is the emergence of non - ergodic and super - aging behavior for diffusion in a logarithmic potential , but with a constant diffusion function @xmath76 @xcite .
this can be realized in another category of experiments with dissipative cold atoms in optical lattices @xcite .
approaching the asymptotic solution ( see fig . [
fig : fig1 ] ) , we use the convenient ansatz @xmath77 .
the evolution of the function @xmath78 is shown in fig .
[ fig : fig2 ] and suggests a scaling form @xmath79 with @xmath80 . here
@xmath81 and @xmath82 are some functions of @xmath7 to be determined . versus @xmath73 for large rescaled times
@xmath83 $ ] in the direction of the red arrow .
inset : plot of @xmath84 versus @xmath85 ( blue line ) and fit with an error function ( red dashed line ) .
( b ) @xmath86 is plotted versus @xmath87 ( blue line ) and compared to a linear fit ( red dashed line ) .
( c ) @xmath88 is plotted versus @xmath89 ( blue line ) and compared to a linear fit ( red dashed line ) .
parameters : @xmath90 , @xmath91 . ] adopting this ansatz leads to @xcite @xmath92 \right\ } \label{eq : erf}\ ] ] where @xmath93 is the error function @xcite .
figs .
[ fig : fig2](b - c ) show that , at long enough times , the numerical data and the proposed analytical @xmath7-dependence @xmath94 and @xmath95 match accurately .
note that the numerical results still show deviations from the exact analytical prefactors .
we verified that these deviations become smaller with increasing filling .
thus we conclude that the analytical form is applicable in the large @xmath41 limit . for large enough times , the obtained solution gives a very good approximation in the entire range of @xmath73 .
the reason for this is the fast initial relaxation at low values of @xmath73 .
thereafter , only small relative changes occur at small @xmath73 .
these changes in the probability distribution are mainly connected to the variations at large @xmath73 via particle number conservation .
thus , we can use the obtained solution to calculate the local particle fluctuations @xmath58 .
the corresponding integral can be solved analytically giving @xmath96 where @xmath97 depends algebraically on @xmath7 .
we thus have shown analytically the emergence of the stretched exponential behavior .
this finding , as depicted in fig .
[ fig : figa ] , compares well to the numerical solution of eq .
( [ eq : evoeq ] ) . however , the time at which the stretched exponential occurs , and the detailed decay , depend on the filling .
in particular , the stretched exponential occurs later for larger fillings . to summarize ,
we have uncovered , in the bose - hubbard model coupled to a dissipative environment , two unconventional relaxation regimes : at short times , and large enough fillings , a power - law regime , while at large times , and any filling , a stretched exponential regime .
this last regime is dominated by rare events which correspond to the occupation of a single site with a large number of atoms .
the rare states in the tail of the distribution function @xmath98 are occupied with decreasing time scales @xmath99 .
these ingredients alone allow to estimate the main time dependence of the fluctuations employing a simple saddle point argument .
since @xmath100 , the saddle point integration @xmath101 leads to @xmath102 recovering the stretched exponential . indeed , very slow transition rates , due to the high energetic cost of the processes connecting these rare configurations , dominate the long time dynamics .
this emergent glass - like dynamics is thus a signature of the complex level structure of the bose - hubbard hamiltonian .
dissipation , forcing the system to explore its whole configuration space , including rare and energetically unfavorable configurations , manifests the complex energy levels structure of the system . in future works , we plan to investigate existence of stronger connections beyond the dominating rare events to glassy physics , e.g. the emergence of dynamical heterogeneity , aging phenomena @xcite or the physics of kinetic constraints model @xcite
. experimental observation of these relaxation regimes is within reach .
we discuss in the following possible realizations for the stretched exponential regime which is experimentally more demanding to study .
we consider a gas of @xmath103rb atoms ( mass @xmath31 ) confined to an optical lattice potential with wavelength @xmath104 . for a lattice depth of @xmath105 ( where @xmath106 ) , the ratio of the interaction over tunneling is @xmath107 with @xmath108 .
the realization of the dissipator in eq.(1 ) could be achieved by a noisy optical potential pattern , e.g. due to an additional , time dependent , speckle beam @xcite or an incommensurate superlattice @xcite with a randomly changing phase or amplitude . since the strength of @xmath22 in these setups can be tuned by the intensity of the light fields , this allows to have @xmath109 or more . to observe the stretched exponential regime , a low filling , for example @xmath110 would be advantageous ( see fig.[fig : figa ] ) . for this filling and @xmath109 , the experimental time - scales needed to identify the stretched exponential regime are of the order of @xmath111s ( @xmath112 in fig .
[ fig : figa ] ) and lattice occupations up to @xmath113 atoms per site will be occupied with a non - negligible probability .
this required time - scale is small compared to the time - scale of the three - body losses ( for occupation of @xmath114 , the three - body loss scale is approx .
@xmath115s @xcite ) and to the time - scale of secondary collisions with the background gas @xcite .
thus these processes can be neglected .
further , transitions to higher bloch bands can be due to interaction of the highest occupied sites or by the dissipation . both can be neglected for experimentally relevant time - scales as i ) the next bloch band is approximately at an energy @xmath116 which is @xmath117 times the interaction energy @xmath118 for a large filling as @xmath114 and ii ) transitions due to dissipation can be controlled via engineering the noise spectrum and tuning the lamb - dicke parameter @xcite .
for example the frequency of the noise pattern could be cut below the frequency corresponding to transitions to higher bands .
we are grateful to j.s .
bernier , g. biroli , r. bouffanais , j .-
p .
eckmann , j.b .
gong , p. hnggi , h. ott , t. prosen , and p. wittwer for fruitful discussions .
we acknowledge anr ( famous ) , snsf ( division ii , manep ) , sutd start - up grant ( srg - epd-2012 - 045 ) and the darpa - ole program for financial support . _ note added .
_ during the referee process we became aware of @xcite , which shows non - exponential relaxation dynamics .
10 r. kohlrausch , ann .
phys .
( leipzig ) * 12 * , 393 ( 1847 ) .
j. c. phillips , reports on progress in physics * 59 * , 1133 ( 1996 ) .
c. a. angell _
et al . _ ,
journal of applied physics * 88 * , 3113 ( 2000 ) .
r.v .
chamberlin , g. mozurkewich , and r. orbach , phys .
rev .
lett . * 52 * , 867 ( 1984 ) .
k. binder and a. p. young , rev .
mod .
phys .
* 58 * , 801 ( 1986 ) .
s. bedanta and w. kleemann , journal of physics d : applied physics * 42 * , 013001 ( 2009 ) .
j. kakalios , r.a .
street , and w.b .
jackson , phys .
rev .
lett . * 59 * , 1037 ( 1987 ) .
l. berthier and g. biroli , review of modern physics * 83 * , 587 ( 2011 ) .
m. goldstein , j. chem .
phys . * 51 * , 3728 ( 1969 ) .
r.g .
palmer , d.l .
stein , e. abrahams , and p.w .
anderson , physical review letters * 53 * , 958 ( 1984 ) .
f. ritort and p. sollich , advances in physics * 52 * , 219 ( 2003 ) . a.j .
bray , phys .
rev .
lett .
* 59 * , 586 ( 1987 ) .
g.k .
brennen , g. pupillo , a.m. rey , c.w .
clark , c.j .
williams , journal of physics b * 38 * , 1687 ( 2005 ) v. s. shchesnovich and v. v. konotop , phys .
rev .
a * 81 * , 053611 ( 2010 ) .
p. barmettler and c. kollath , phys .
rev .
a * 84 * , 041606 ( 2011 ) .
n. syassen _
et al . _ ,
science * 320 * , 1329 ( 2008 ) .
d. a. zezyulin , v.v .
konotop , g. barontini , and h. ott , phys .
rev .
lett . * 109 * , 020405 ( 2012 ) .
j. t. barreiro _
et al .
_ , nature * 470 * , 486 ( 2011 ) .
m. mueller _
et al . _ , new journal of physics * 13 * , 085007 ( 2011 ) . h. pichler , a. j. daley , and p. zoller , phys . rev .
a * 82 * , 063605 ( 2010 ) .
d. poletti , j .- s .
bernier , a. georges , and c. kollath , physical review letters * 109 * , 045302 ( 2012 ) .
h. pichler , j. schachenmayer , j. simon , p. zoller , a.j .
daley , phys .
rev .
a * 86 * , 051605(r ) ( 2012 ) h. pichler , j. schachenmayer , a.j .
daley , p. zoller , phys .
rev .
a * 87 * , 033606 ( 2013 ) j .- s .
bernier , p. barmettler , d. poletti , and c. kollath , phys .
rev .
a * 87 * , 063608 ( 2013 ) a. tomadin , s. diehl , and p. zoller , phys . rev .
a * 83 * , 013611 ( 2011 ) .
f. gerbier and y. castin , phys .
rev .
a * 82 * , 013615 ( 2010 ) .
d. jaksch , c. bruder , j.i .
cirac , c.w .
gardiner , p .
zoller , phys .
rev .
lett . * 81 * , 3108 ( 1998 ) .
i. bloch , j. dalibard , and w. zwerger , rev .
mod .
phys . * 80 * , 885 ( 2008 ) .
j.e .
lye , l. fallani , m. modugno , d.s .
wiersma , c. fort , m. inguscio , phys .
rev .
lett .
* 95 * , 070401 ( 2005 ) e. lye , l. fallani , c. fort , v. guarrera , m. modugno , d. s. wiersma , m. inguscio , phys .
rev .
a * 75 * , 061603(r ) ( 2007 ) j. j. garca - ripoll _
et al . _ , new journal of physics * 11
* , 013053 ( 2009 ) .
d. poletti , j .- s .
bernier , a. georges , and c. kollath , in _ proceedings of the annual international conference on optoelectronics , photonics and applied physics _ , ( gstf , singapore , 2013 ) . for more detailed information
refer to the supplementary material .
this stands in contrast to steady states which one finds using a gutzwiller - like approximation from the start @xcite , since there the suppression of the single particle correlations `` freezes '' the system in many distinct states .
m. abramowitz and i. a. stegun , _ handbook of mathematical functions with formulas , graphs , and mathematical tables _
( new york : dover publications , usa , 1960 ) . at @xmath71
the reflective boundary condition is @xmath119\!=\!0.$ ] d.a .
kessler and e. barkai , phys .
rev .
lett . * 105 * , 120602 ( 2010 ) .
y. castin , j. dalibard , and c. cohen - tannoudji , _ light induced kinetic effects on atoms , ions and molecules _ ( l. moi et al .
ets editrice , pisa , 1991 ) .
p. douglas , s. bergamini , and f. renzoni , phys .
rev .
lett . * 96 * , 110601 ( 2006 ) .
z. nussinov , p. johnson , m.j .
graf and a.v .
balatsky , phys .
rev .
b * 87 * , 184202(2013 ) .
b. olmos , i. lesanovsky , and j.p .
garrahan , phys .
rev .
lett . * 109 * , 020403 ( 2012 ) .
g.k .
campbell , j. mun , m. boyd , p. medley , a.e .
leanhardt , l. marcassa , d.e .
pritchard , w. ketterle , science * 313 * , 649 - 652 ( 2006 ) .
transitions to higher bloch bands would occur with a rate @xmath120 @xcite where @xmath121 is the lamb - dicke parameter and @xmath122 is the noise spectrum at for the transition between the lowest and the second excited bloch band . both parameters can be tuned such that @xmath123 is , at most , of the same order of three body losses .
z. cai and t. barthel , phys .
rev .
lett . * 111 * , 150403 ( 2013 ) . |
our interest is in the directional hilbert transform applied in a choice of directions of the plane .
thus , for @xmath11 , set @xmath12 this definition is independent of the length of @xmath2 , and below we shall only concern ourselves with @xmath13 .
let @xmath4 be a schwartz function with frequency support in the annulus @xmath5 , and @xmath6 .
our theorem is t.main the maximal operator @xmath14 maps @xmath8 into weak @xmath8 , and @xmath9 into @xmath9 for @xmath10 .
this theorem is a complement to a corresponding result , due to j. bourgain @xcite , for the directional maximal function , namely @xmath15 see .maxfn/ for a discussion of norm bounds for this operator . for both this operator , and @xmath16 ,
the estimate of weak square integrability is sharp , as was pointed out to us by m. christ @xcite .
this argument may summarized as follows .
begin with a schwartz function @xmath17 with frequency support in a small ball about the origin in the plane .
then consider @xmath18 .
for any point in the plane @xmath19 with @xmath20 and @xmath21 , consider the line that passes through @xmath22 and the origin .
the real part of @xmath23 will not change sign on this line , so we compute the hilbert transform with no cancellation .
that is @xmath24 .
and @xmath25 is just in weak @xmath8 .
see .example/. [ f.example ] 0cm0cm6cm4.5 cm ( 0,2 ) 45!white whereas , bourgain s argument for the maximal estimate is not difficult , the theorem above is of necessity somewhat harder , as it implies the pointwise convergence of fourier series in one dimension .
this is an observation in the style of de leeuw .
one considers the trace of the operator in frequency variables along any line in the annulus @xmath5 .
this is carleson s theorem , @xcite , but also see @xcite . as such
, we use a method which is adopted from the proof of carleson s theorem given by m. lacey and c. thiele @xcite .
perhaps the principal novelty of this paper is the suitable adaptation of the time frequency analysis of lacey and thiele to the current setting of the plane and measurable choice of directions . in comparison to the proof of lacey and thiele
, we find a rather precise analogy between the proofs on the real line , and the proofs on the plane .
some differences arise from the notion of a _ tile _ , which requires some care to define , and the proof of the `` size lemma , '' an orthogonality statement , requires a small amount of innovation .
finally , there is an outstanding question , attributable to e. m. stein @xcite , concerning the boundedness of the hilbert transform on families of lines that are determined by say a lipschitz map .
thus , for a map @xmath26 , one wishes to know if @xmath27 is a bounded operator on say @xmath8 .
positive results for analytic and real analytic vector fields are due to respectively nagel , stein and wainger @xcite , and to bourgain @xcite . in a subsequent paper , the authors @xcite will prove that the operator above is bounded on @xmath8 if @xmath28 , for any positive @xmath29 .
the results of this paper are a crucial aspect of the proof of this result .
the presentation of this paper has been substantially improved by skilled and generous referees .
[ s.model ] we begin with some conventions .
we do not keep track of the value of generic absolute constants , instead using the notation @xmath30 iff @xmath31 for some constant @xmath32 . and @xmath33 iff @xmath30 and @xmath34 .
we use the notation @xmath35 to denote the indicator function of the set @xmath36 . and
the fourier transform on @xmath37 is denoted by @xmath38 , with a similar definition on the real line .
we use the notation 0 _ a fdx:=^-1_a fdx .
d.grid a _ grid _ is a collection of intervals @xmath39 so that for all @xmath40 , we have @xmath41 .
the dyadic intervals are a grid .
a grid @xmath39 is _ central _ iff for all @xmath40 , with @xmath42 we have @xmath43 .
the reader can find the details on how to construct such a central grid structure in @xcite .
let @xmath44 be rotation on @xmath45 by an angle of @xmath46 .
coordinate axes for @xmath37 are a pair of unit orthogonal vectors @xmath47 with @xmath48 .
d.rectangle we say that @xmath49 is a _ rectangle _ if it is a product of intervals with respect to a choice of axes @xmath47 of @xmath37 .
we will say that @xmath50 is an _ annular rectangle _ if @xmath51 for an integer @xmath52 with @xmath53 , with respect to the axes @xmath47 .
the dimensions of @xmath50 are said to be @xmath54 .
notice that the face @xmath55 is tangent to the circle @xmath56 at the midpoint to the face , @xmath57 .
we say that the _ scale of _
@xmath50 is @xmath58 and that the _ annular parameter of _ @xmath50 is @xmath59 . in referring to the coordinate axes of an annular rectangle ,
we shall always mean @xmath47 as above .
it is imperative to keep in mind that the choice of basis in the definition of a tile depends upon the tile in question .
these definitions are illustrated in .tile/. also see .tree/. annular rectangles will decompose our functions in the frequency variables . but our methods must be sensitive to spatial considerations ; it is this and the uncertainty principle that motivate the next definition .
d.dual two rectangles @xmath60 and @xmath61 are said to be _ dual _ if they are rectangles with respect to the same basis @xmath47 , thus @xmath62 and @xmath63 for intervals @xmath64 , @xmath65 .
moreover , @xmath66 for @xmath65 . the product of two dual rectangles we shall refer to as a _ phase rectangle .
_ the first coordinate of a phase rectangle we think of as a frequency component and the second as a spatial component .
cm we consider collections of phase rectangles @xmath67 which satisfy these conditions .
for @xmath68 we write @xmath69 , and require that 5 e.r1 + e.r2 + [ e.r - central ] + e.r3.5 \{r r@xmath70_s0at } + e.r3 = + e.r4 \{_ss=2^l , s=}c 2 ^ -l , l .
we assume that there are auxiliary sets @xmath71 associated to @xmath72or more specifically @xmath73which satisfy these properties .
5 e.bw1 + e.bw2 2=,:=(,2 ) , +
e.bw2/3 + e.bw3 k , + e.bw4 \{@xmath74@xmath74@xmath70_s}@xmath751 , + e.bw5 recall that @xmath44 is the rotation that takes @xmath76 into @xmath77 .
thus , @xmath78 .
see .omega/. [ f.omega ] 0cm0cm8cm4.5 cm ( -.15,2 ) [ f.tile ] note that @xmath79 .
thus , @xmath80 is in @xmath81 , and @xmath82 serves as `` the angle of uncertainty associated to @xmath83 . ''
let us be more precise about the geometric information encoded into the angle of uncertainty .
let @xmath84 be as above .
choose another set of coordinate axes @xmath85 with @xmath86 and let @xmath87 be the product of the intervals @xmath88 and @xmath89 in the new coordinate axes .
then @xmath90 for an absolute constant @xmath91 .
we say that _ annular tiles _ are collections @xmath92 satisfying the conditions .r1/.bw4/ above .
the constant @xmath93 appears in .r3/ and .r4/. we extend the definition of @xmath94 , @xmath95 , @xmath96 and @xmath97 to annular tiles in the obvious way , using the notation @xmath98 , @xmath99 , @xmath100 and @xmath101 .
a phase rectangle will have two distinct functions associated to it . in order to define these functions ,
set 2 _ y f(x)&:=f(x - y),y^2 + _
@xmath74f(x)&:=e^i@xmath74xf(x),@xmath74 ^ 2 + _
r_1r_2^p f(x_1,x_2)&:= 1()^1/p f ( , ) , . in the last display , @xmath102
, @xmath103 is a rectangle , and the coordinates @xmath104 are those of the rectangle .
note that the definition depends only on the side lengths of the rectangle , and not the location . and
that it preserves @xmath9 norm . for a function @xmath105 and tile
@xmath106 set 0 @xmath107_s:=_c(@xmath70_s)_c(r_s)^2_r_s @xmath107 , where @xmath108 is the center of @xmath109 .
below , we shall consider @xmath105 to be a schwartz function for which @xmath110 is supported in a small ball about the origin in @xmath37 , and is identically @xmath111 on another smaller ball around the origin .
we introduce the tool to decompose the singular integral kernels , and a measurable function @xmath112 that achieves the maximum in our operator , up to a factor of say @xmath113 .
we consider a class of functions @xmath114 , @xmath115 , so that 5 [ e.zc-fourier ] + [ e.zc-space ] ( 1+x ) ^-1/ in these conditions , @xmath116 is a small fixed constant that need not concern us . in the top line , @xmath117 is a fixed positive constant , that is only needed to ensure that .dfs2/ below is true . for the current paper
, we will need only single choice of function @xmath118 , but for the authors subsequent paper @xcite , countably many choices of @xmath118 will be needed .
define 3e.dfs2 @xmath119_s(x):=&_s(x - yv(x))@xmath120_s ( y)dy + = & ( v(x))_s(x - yv(x))@xmath120_s ( sy)dy .
+ [ e.zc_s ] _
s(y ) : = & s _ s(s y ) .
an essential feature of this definition is that the support of the integral is contained in the set @xmath121 , a fact which can be routinely verified , for an appropriate choice of @xmath117 in .zc - fourier/. see .localize - v/. that is , we can insert the indicator @xmath122 without loss of generality .
the set @xmath123 serves to localize the vector field , while @xmath124 serves to identify the location of @xmath125 in the frequency coordinate .
[ f.localize-v ] 0cm0cm8cm4.5 cm ( 1.5,2 ) 45!white * remark : * it seems likely that we could prove our theorem using the simpler definition @xmath126
. this would be the natural analog of the decomposition used by lacey and thiele @xcite .
but , in our subsequent paper on this subject @xcite , we will need to consider a truncation of the hilbert transform , which suggests the definition of @xmath127 we have adopted above .
we will also want to rely upon facts proved in this paper , to deduce our theorems concerning smooth vector fields .
the model operators we consider are defined by 0 0c_,vf:= _ _ s . in this display , @xmath128 .
l.model assume that the vector field is only measurable .
the operator @xmath129 extends to a bounded map from @xmath8 into weak @xmath8 , and @xmath9 into itself for @xmath130 .
the norm of the operator is independent of @xmath131 .
we shall prove that @xmath129 maps @xmath8 into weak @xmath8 , with constant independent of @xmath132 and @xmath2 .
by duality , it suffices to show that for all @xmath133 of @xmath8 norm one , and sets @xmath134 of finite measure @xmath135 by dilation invariance of the @xmath129 with respect to powers of @xmath136 , we can further take @xmath137 . for the case of @xmath9 , @xmath138
, we shall demonstrate the restricted type estimate 0 @xmath139 ^ -p,2<p<@xmath140 .
we need only consider this for @xmath141 , by the weak @xmath8 bound .
moreover , this inequality follows from 3 & _ s0at ( ) + & \{(e / f)},9f < e.[e.p>2 ] observe that by dilation invariance , we can assume that @xmath142 .
we also assume that the vector field @xmath2 is defined only on @xmath143 .
the proofs of .cjabs/ and .p>2/ are given in .outline/ , and follow lines of argument that are similar to the one dimensional case , as given by lacey and thiele @xcite for the case @xmath144 , and as in the work of grafakos , tao , and terwilleger @xcite for the case of @xmath145 .
also see the survey by lacey @xcite .
* the results for @xmath145 are available essentially because functions that are in @xmath146 are also locally square integrable . in one dimension ,
the proof of convergence of fourier series for @xmath147 is closely linked to the boundedness of the hardy littlewood maximal function . in the current setting , the corresponding maximal function would be the unbounded kakeya maximal function .
the relationship between maximal functions and @xmath146 bounds for @xmath147 has a subtle role to play in our companion paper @xcite .
we outline a proof of the weak type estimate on @xmath148 for the directional maximal function @xmath149 given in .dir - max/. p.maxfn for functions @xmath150 , and @xmath151 , we have the estimate 0 ^-2 f.2.^2 it suffices to give the same bound for the model operator 0 0m_,vf:= _ .
it is a remark that this inequality is an immediate consequence of .model/. the sum defining @xmath152 is in fact unconditional convergent as a sum over tiles . as a consequence the square function below , which trivially dominates @xmath153 , satisfies the weak type inequality . 0
^1/2 there is however a more direct proof .
the fundamental fact , a simplier instance of the size lemma , .size/ above , is this .
l.size-simple fix @xmath150 and @xmath151 .
let @xmath154 be a collection of tiles which are pairwise incomparable with respect to the partial order ` @xmath155 , ' and moreover for each @xmath156 , @xmath157 .
then it is the case that 0 _
s0s ^-2 f.2.^2 this is proved in e.g. barrionuevo and lacey @xcite , or one can inspect the proof of .size/ for a number of simplifications which apply in this case . it is then clear that the maximal function below maps @xmath148 into weak @xmath148 . 0 a_1 f : = _ 0at ( ) 1 _ r_s for @xmath158 , set 0 a_t f : = _ 0at ( ) 1 _ tr_s this also maps @xmath148 into weak @xmath148 with norm at most @xmath159 .
indeed , given @xmath151 , let @xmath160 denote the maximal tiles with @xmath157 . then , it is the case that 2 \ { a_t f > } _
s0s ktr_s for an absolute constant @xmath161 .
hence , the claimed bound follows from .size - simple/. then observe that 0 0m_,vf _
k=0 ^ 2 ^ -4k a _
we show that .model/ implies our main result , .main/. indeed , the hilbert transform in the direction of @xmath162 is , in the limit , an appropriate average of the discrete sums formed in the previous section .
let @xmath163 be a smooth radial function satisfying 1 [ e.zk ] ( @xmath74)@xmath139(@xmath74 ) let @xmath164 , where we choose the dilation so that @xmath165 is the fourier parameter , as opposed to the more common definition @xmath166 where @xmath165 is the spatial parameter .
let @xmath32 be the distribution on 0 _
j=-@xmath140^@xmath1402^j @xmath120(2^j y ) where @xmath118 is a schwartz function with @xmath167 supported in a small neighborhood of @xmath168 . in particular , the distribution 0 _ 0 ^ 1 2^s k(2^s y ) ds is a non zero multiple of the distribution associated with projection onto the negative frequencies on , which is itself a linear combination of the identity operator and the hilbert transform . to prove .main/
it suffices to demonstrate the same norm estimates for the linear operators in which the distribution @xmath169 is replaced by @xmath170 , namely for 1 [ e.tv ] t_v f(x)=@xmath139_*f
( x - y v(x))k(y ) dy in which @xmath93 , and the measurable @xmath171 are arbitrary . for values of @xmath172 , let 2 3s_j , f= _
@xmath107_s + t_j , vf(x)=f(x - yv(x))2^j@xmath120(2^j y ) dy also , let @xmath173 be the operation of rotation by angle @xmath174 .
the main point concerns the operator 1 [ e.avg ] s_j,:=_y@xmath140_(y ) & _ [ 0,2 ^ -s)[0,2 ^ -s)^2 _ -@xmath175_-y + & 3_j , _ y_@xmath175 _ [ 0,2^s)[0,2^s)^2 dyd@xmath175 ds where @xmath176 ^ 2\times[0,2{\ensuremath{\pi}}]\times[-1,1]$ ] .
observe these points .
( 1 ) the integral is an average over dilations , rotations , and translations .
note that the averaging over dilations is done uniformly in all directions , and that we are implicitly averaging with respect to the haar measure on @xmath177 , with the rationale being that the dilations we are using are a representation of that group .
( 2 ) due to the assumption .r3.5/ , the limit in @xmath178 will exist .
if @xmath179 were fixed we would only need to average over all translations indexed by a fixed rectangle .
( 3 ) the rectangles @xmath179 fill a fixed percentage of the annulus , due the assumption .r4/. ( 4 ) the limit , applied to a schwartz function @xmath0 is seen to exist .
l.2convolve for each @xmath172 , we have the identity 0 s_j,@xmath139_*f = c(j,)@xmath139_*f where the constant @xmath180satisfies @xmath181 , for some absolute constant @xmath182 .
to deduce bounds for .tv/ , observe that .model/ concerns norm bounds for the sums 0 _ j c(j,)^-1 t_j , v 3_j , @xmath139 _ * f of course the coefficients @xmath183 do not appear in .model/. yet the placement of the absolute values in .cjabs/ and .p>2/ demonstrates that the sum is unconditional over tiles , so that we can impose an arbitrary bounded sequence of coefficients , as we have done here .
the same norm bounds hold for averages of these sums , such as the averages that occur in .avg/. hence , by the lemma just stated , we deduce the norm bounds for the operators in .tv/. it remains to prove our lemma . before proving .2convolve/ , we record a simple lemma on convolutions .
l.convolve let @xmath105 and @xmath184 be real valued schwartz functions on @xmath37 .
then , 2 _ [ 0,1]^2_m^2 f , _ y+m@xmath107 . _ y+m@xmath119 dy = f*@xmath185 + @xmath185(x)=@xmath119(x+u ) du . in particular , @xmath186 .
the proof is immediate .
the sum and integral in question is in fact the integral 0 _ ^2 f(z)@xmath119(x - y ) dydz and one changes variables , @xmath187 .
fix @xmath188 , and let 0 3_@xmath70f:= _ @xmath107_s recall .r3.5/. consider the limit 0 s_@xmath70:=_y@xmath140y^-2 _ [ 0,y]^2 _ -y 3 _ @xmath70_y dy this limit can be explicitly written as an integral which is as in .convolve/ with an additional dilation .
thus , this limit is convolution with respect to @xmath189 , where @xmath190 . by choice of @xmath105 , it has non zero fourier transform , and is identically one on a small ball around the origin
. hence @xmath189 has fourier transform with the same properties .
the definition of @xmath191 also incorporates an average over dilations and rotations . concerning the average over rotations ,
recall .r4/ , which states that we have an essentially maximal number of @xmath73 of a given scale . concerning the dilations ,
observe that 0 ^2_[0,t)^2 [ f * [ ^2_[0,1/t)^2 g]= [ ^1_[0,t)^2 f]*g .
these observations lead to the conclusion that 2 s_j , f&=c(j , ) @xmath185*f + @xmath185&=_-1 ^ 1_[0,2@xmath192 ^1_[0,2^t)^2_@xmath175 @xmath185 d@xmath175dt the leading constant in the top line satisfy the claimed bound @xmath193 for a positive constant @xmath194 .
the main point here is assumption .r4/. this function @xmath185 has fourier transform that is identically constant for all @xmath195 .
the value of this constant depends only on the choice of @xmath105 .
we need these notions associated to tiles and sets of tiles , all directly inspired by the proof of carleson s theorem as given by lacey and thiele @xcite .
there are however some changes in the current context , due to the setting of the plane , and our frequency analysis in an annulus .
an important heuristic is that given a tile @xmath196 , the interval @xmath197 serves as an `` angle of uncertainty '' for the rectangle @xmath198 .
the geometric consequence of this can be described as follows .
write @xmath199 in the coordinate axes @xmath200 .
there is an absolute constant @xmath161 so that the following holds .
suppose that @xmath201 is some other choice of coordinate axes with 0 , .
write @xmath202 but the choice of coordinate axes is @xmath201 . then
, we have @xmath203 and @xmath204 .
the implication is that @xmath198 and @xmath205 are essentially the same rectangles .
see .uncertain/. [ f.uncertain ] 0cm0cm5.5cm4.5 cm ( -.15,2 ) ( 4.5,0 ) there is a natural partial order on tiles , but it does take some care to define . for a fixed tile @xmath72 ,
let @xmath206 be the coordinate system associated with @xmath83 .
while we have not done so to date , at this point we insist that the origin in each such coordinate system specify the same point in the plane .
let @xmath207 be the set of the projections of all @xmath83 s onto @xmath98 and @xmath208 be the set of the projections of all @xmath83 s onto @xmath99 .
we have assumed that @xmath209 is a central grid in .r - central/. namely , recalling .grid/ , any two intervals @xmath210 in @xmath211 which intersect , satisfy either @xmath43 , or @xmath212 .
the reader can find the details on how to construct such a central grid structure in @xcite .
as all tiles lie in a single annulus , the collection of intervals @xmath208 consist of disjoint intervals of equal length that partition @xmath213 .
let @xmath214 be the minimal interval in @xmath211 containing the projection of @xmath60 onto @xmath98 , and let @xmath215 be the minimal interval in @xmath216 containing the projection of @xmath60 onto @xmath99 . for two tiles @xmath217 ,
write @xmath218 if and only if @xmath219 , @xmath220 , and @xmath221 .
this is illustrated in .partial/. we have been careful to define this partial order in such a way that it is transitive and that we have the conclusion @xmath222 we can assume that @xmath223 , as these intervals of the circle come from a grid . and
so it suffices to check that @xmath220 . by the assumption .r - central/ ,
in which @xmath224 is a small positive constant , we would have @xmath225 , and the last set is contained in @xmath226 , by the central grid property .
[ f.partial ] -1cm0cm6.5cm4 cm ( 0.3,2 ) ( 0,-.7 ) a _ tree _ is a collection of tiles @xmath227 , for which there is a ( non unique ) tile @xmath228 with @xmath229 for all @xmath230 . for @xmath231 ,
call @xmath232 a @xmath233tree if the tiles @xmath234 are pairwise disjoint .
@xmath235trees are especially important for us ; .tree/ presents an useful illustration of such a tree .
observe that for any tree @xmath236 we have , due to our comments about intervals of uncertainty , 0 r_sk r_,s .
thus , the spatial locations of the tiles are localized .
in addition , as @xmath237 , we can regard all rectangles @xmath198 as having a _ fixed _ set of coordinate axes , namely those of the top .
this is a key point of the `` tree lemma '' below .
fix a positive rapidly decreasing function @xmath238 , and for rectangle @xmath60 set 5 [ e.zqrs ]
_ r=_c(r)^1_r + s:=_s < s _ v^-1( ) @xmath238_r_s dx , + : = _ ss , + : =
_ sr_s + @xmath239()^2:=_s , + : = _ ^1/2 recall that we are to prove .cjabs/ and .p>2/. in .cjabs/ , @xmath240 , @xmath241 , and @xmath2 is defined only on @xmath143 . the first two lemmas concern density and size , and are the same form .
the third lemma , the ` tree lemma ' relates size , density and trees .
l.dense any @xmath242 is the union of @xmath243 and @xmath244 satisfying these conditions . 0 12. the collection @xmath243 is a union of trees @xmath245 , with 1 [ e.dense ] _ _ ^-1f .
l.size any @xmath242 is the union of @xmath246 and @xmath247 satisfying these conditions . 0 12. the collection @xmath246 is a union of trees @xmath248 , with 1 [ e.size ] _
l.tree for any tree @xmath232 we have the estimate 0 _ s. the first lemma has a proof which is essentially identical to the proof of the `` mass lemma '' in m. lacey and c. thiele @xcite
. we do not give the proof .
the second lemma follows the well established lines of proof , yet must introduce a new element or two to address the two dimensional setting . the complete proof is given .
the proof of the last lemma is quite close to that of m. lacey and c. thiele @xcite
. we shall give a proof .
the lemmas are combined in this way to prove .cjabs/ , and hence .model/ in the case of the weak @xmath8 estimate .
.dense/ and .size/ should be applied so that their principal estimates .dense/ and .size/ are approximately equal . the density of @xmath92 is at most a constant .
the size of @xmath92 is at most a constant times the @xmath249 norm of @xmath0 .
thus , we can take the set of all tiles @xmath92 and decompose it into sub collections @xmath250 , @xmath251 , so that @xmath252 is the union of trees @xmath253 such that 4 _ _ , + ( 1 , ^-1 ) , + ^-1/2 .
hence , it follows that 2 _ s _ & _ ( 1 , ^-1)^-1/2 + & ( ^1/2,^-1/2 ) . this estimate is summable over @xmath254 to an absolute constant
this completes the proof of .cjabs/. a small variation on this argument proves .p>2/. in this instance , the function @xmath255 , with @xmath142 , so that the size of @xmath92 is @xmath256 .
and @xmath92 is a union of sub collections @xmath250 , @xmath257 , so that @xmath252 is the union of trees @xmath253 satisfying 4 _ _ f , + ^-1 , + ( 1,()^-1/2 ) .
hence , it follows that 0 _ s _ ^1/2(^1/2,^-1/2 ) .
recall that in this instance , @xmath258 .
this estimate is summable over @xmath259 to @xmath260 .
this completes the proof of .p>2/. our proof of .model/ is complete , aside from the proofs of the three lemmas of this section .
we give the proof of .size/. we find that the proof involves for the most part a standard argument in the literature .
see for instance the proof of the energy lemma in m. lacey and c. thiele @xcite .
yet there is a point at which we will rely upon the strong maximal function , with respect to a choice of axes that is specified in a particular way by the set of tiles .
we can assume that all @xmath261 lie in a fixed half circle .
note that in 1s , that as the scale of tiles increase , the intervals @xmath262 move off in a clockwise direction , as depicted in .tree/. we take advantage of this in a specific way below .
the initial step of the proof is to construct a collection of @xmath111trees @xmath263 and use them to construct the collection of trees @xmath264 .
the process is inductive .
let @xmath265 .
initialize @xmath266 while @xmath267 , select a @xmath111tree @xmath268 such that @xmath269 @xmath270 is maximal , and @xmath271 is most clockwise .
then set @xmath272 to be the maximal ( with respect to inclusion ) tree in @xmath273 with top @xmath274 .
update @xmath275 when @xmath276 , set @xmath277 , and the process stops . to conclude the lemma
, we need to show that @xmath278 we have constructed these trees so that they satisfy a property very useful to the verification of this inequality .
suppose @xmath279 and @xmath230 and @xmath280 with @xmath281 .
then @xmath282 .
we refer to this property as `` strong disjointness . '' to see that it holds , we have @xmath283 , so that @xmath271 lies in the clockwise direction from @xmath284 .
hence @xmath232 was constructed before @xmath285 .
thus , if @xmath286 , we would conclude from .<property/ that @xmath287 .
hence , @xmath288 , and so @xmath289 would have been removed from @xmath290 and so could not be in @xmath285
. for @xmath230 , set 4 b_=(s)=\{s_+-= } , + b(s)=\{s_+-_=}. note that if @xmath296 is such that @xmath297 , then @xmath298 . and if @xmath299 , then @xmath300 .
we estimate 3 f(_+)_2 ^ 2&__+ f()_2 ^ 2 [ e.top ] + [ e.middle ] & + _ s_+f,_s._s , f(b_=(s ) ) . + [ e.bottom ] & + _ s_+f,_s._s , f(b(s ) ) . for the term
, we use this estimate for tiles @xmath303 with @xmath304 , 0 _ c(r_s)_r_s^0@xmath238(c(r_s ) ) . this with cauchy
schwarz gives 2 .middle/&^1/2 + & _
s_+^2 + & @xmath305 ^ 2 _ _ + , as required by .zs2/. as for.bottom/ , use cauchy schwarz in @xmath72 to bound it by @xmath306^{1/2 } \\{}\seq{}&\bigl [ \sigma^2\sum_{\t\in\vzt_{+}}\abs { r_\t } \sum_{s\in\s_{+}}\abs{\ip \varphi_s ,
f({\mathbf b}(s)).}^2 \bigr]^{1/2}\end{aligned}\ ] ] we show that for each @xmath263 , @xmath307 this will complete the proof of .zs2/. observe that if we set @xmath308 , we have @xmath309 this is a consequence of the grid structure on the intervals @xmath310 .
moreover , all tiles @xmath311 have @xmath312 .
this has two implications .
the first is that all rectangles @xmath83 can be regarded as rectangles with respect to a fixed set of coordinate axes , those for the top of the tree .
let @xmath313 denote the strong maximal function computed in these coordinate directions .
the second , is that the strong disjointness condition applies to each pair of tiles in the collection @xmath314 .
this yields the essential observation that the rectangles @xmath315 are pairwise disjoint , and do not intersect @xmath316 .
make a further diagonalization of the set @xmath317 .
set @xmath318 , and for @xmath319 , set @xmath320 .
let us point out that @xmath321 indeed , recalling the notation .zqrs/ @xmath322 therefore , for any function @xmath323 , @xmath324 clearly , this implies .zs6/ for @xmath325 . for @xmath319 , we can strengthen our inequality to the following .
@xmath326 we use this , together with the fact that the functions @xmath327 are equivalent to an orthonormal basic sequence .
hence , @xmath328 but the functions @xmath329 are highly concentrated in a neighborhood of @xmath316 . in particular , for any function @xmath323 , @xmath330 clearly , this completes the proof of .zs6/. we may fix the vector field , and assume that the standard basis @xmath47 are the basis for the rectangle @xmath316 . as a consequence , we can without loss of generality
assume that this is the basis for all the tiles @xmath230 , and we write @xmath331 .
set @xmath332 and @xmath333 . for this proof , we set 0 @xmath238^(p)=_c(r_s)^p_r_s,0<p<. let @xmath334 be those maximal dyadic intervals @xmath109 in @xmath335 for which for @xmath336 does not contain an interval @xmath337 for some @xmath230 .
this collection partitions .
let @xmath338 be that partition of into maximal dyadic intervals @xmath109 such that @xmath339 .
let @xmath340 .
for each rectangle @xmath341 , set @xmath342 to be those @xmath230 for which @xmath343 .
estimate , without appeal to cancellation 0 _ k dx @xmath305_k_s(k,- ) ( v(x ) ) @xmath238^(@xmath140)_r_s(x ) dx the rectangles @xmath83 are smaller than those of @xmath32 , and do not intersect it . for each @xmath346 ,
the numbers @xmath347 , and they decrease as @xmath348 increases .
the integral above is at most 2 _ k dx@xmath349@xmath305 ( , _ s_xk@xmath238^(@xmath140)_r_s(x ) ) .
this is summable over @xmath341 to at most @xmath350 .
if @xmath351 is non - empty , then @xmath352 . set 0 g_k:=k_s(k,+)v^-1(2 ) which contains the support of @xmath353 .
our assertion is that @xmath354 . to see this ,
let @xmath355 , and @xmath356 be the dyadic interval that contains @xmath357 and @xmath358 .
this interval is small than that of @xmath359 .
then , by maximality , @xmath360 contains some @xmath337 , for @xmath230 .
let @xmath289 be the tile @xmath361 for which @xmath362 .
we have @xmath363 . and since @xmath364 , we deduce the important inequality that 0 @xmath349 . if @xmath351 is a @xmath111tree , then all of the interval @xmath369 contain @xmath370 . thus , for each @xmath371 , there are @xmath372 so that @xmath373 iff @xmath374
. in particular , if @xmath373 , then we have @xmath375 .
this permits us to argue that the vector field @xmath171 can be assumed to be constant .
specifically one sees the inequality 0 ( v(x ) ) _ dy @xmath238^(2)_r_s(x ) .
this inequality is in fact a straight forward calculation , but one that depends very much on the uniform assumptions .zc - space/. we make more definitions , and in the first definition , replace the vector field by @xmath376 .
2 @xmath119_s(x):=&@xmath107_s(x - ye)@xmath120_s(y ) dy , + a_k:=&_s(k,+)f,@xmath107_s .
( v(x ) ) @xmath119_s , k0k , + b:=&_sf,@xmath107_s .
@xmath119_s . in the last line ,
it is a matter of convenience to assume that all trees @xmath377 are 1s , and that @xmath378 .
we have , by a straightforward estimate , 0 @xmath305 so that it suffices to estimate the @xmath379 norm of @xmath36 on the interval @xmath32 .
there are two points : the first is that for each @xmath161 , @xmath380 is dominated by a maximal function applied to @xmath381 .
the second is that @xmath382 obeys the the inequality 1 [ e.b < ] b.2 .
@xmath305 ^ 1/2 let us turn to the second claim .
the claim will follow immediately if we can see that the we have the universal estimate 1 [ e.bb ] _
sa_s @xmath119_s .2 .
this holds for all numerical sequences @xmath383 . indeed ,
the assumptions that we have placed on the functions @xmath384 , namely that it have fourier support in a small ball around the origin , and the uniform assumption on the fourier support of the functions @xmath385 , as expressed in .zc - fourier/ , imply that the functions @xmath386 are supported in @xmath387 . by a fixed vector @xmath376 .
] as these rectangles are either equal or disjoint , it suffices to prove the estimate .bb/ when the tree @xmath236 has @xmath388 for all @xmath389 . with this assumption ,
the assumption .zc - space/ ( with the value of @xmath390 suitably small ) implies the estimate 0 @xmath238^(2)_r_s and this is more than enough to prove .bb/ under this restrictive assumption on the tree .
let us dominate @xmath391 .
for any choice of @xmath392 , we have 0 @xmath119_s= ( @xmath4_@xmath345_--@xmath4_@xmath345_+)*@xmath119_s in which we take @xmath393 to be non negative schwartz function on the plane satisfying @xmath394 ^ 2}\le\widehat { \ensuremath{\zeta}}\le { } \ind{[-5/8,5/8]}$ ] and set @xmath395 .
the identity follows from the frequency properties of @xmath105 and of the class of functions @xmath396 as described in .zc - fourier/. from this , we conclude that 0 _ jr_j dz with the last quantity being a supremum over all rectangles @xmath397 that contain @xmath32 .
this supremum is constant on @xmath32 .
thus , 0 _ g_k dx _ xkm b(x)@xmath349_xkm b(x ) where @xmath398 denotes the strong maximal function with respect to the @xmath47 coordinates . | for a schwartz function @xmath0 on the plane and a non - zero @xmath1 define the hilbert transform of @xmath0 in the direction @xmath2 to be @xmath3 let @xmath4 be a schwartz function with frequency support in the annulus @xmath5 , and @xmath6 .
we prove that the maximal operator @xmath7 maps @xmath8 into weak @xmath8 , and @xmath9 into @xmath9 for @xmath10 .
the @xmath8 estimate is sharp .
the method of proof is based upon techniques related to the pointwise convergence of fourier series .
indeed , our main theorem implies this result on fourier series . |
we would like to show that if the le acquires a finite value for some @xmath38 , isometries @xmath183 can be chosen such that is non - vanishing in the thermodynamic limit .
we define the coefficient matrix @xmath184 of @xmath185 via @xmath43 ( here @xmath186 can be arbitrarily chosen ) .
let @xmath187 and @xmath188 denote the isometries contained in @xmath189 projecting @xmath190 into the subspaces of dimension 2 corresponding to the maximum schmidt coefficients .
for the resulting state we obtain the coefficient matrix @xmath191 , where @xmath192 is the probability which arises from keeping track only of the projection into the space of the two highest schmidt coefficients , i.e. , @xmath193 .
now , we want to show that @xmath194 where @xmath13 is the entropy of entanglement @xcite , @xmath195 the concurrence @xcite and @xmath196 some finite - valued function to be calculated below .
@xmath13 is given by the von neumann entropy of the normalized reduced density operator of the bipartite state @xmath197 , @xmath198 with @xmath199 .
the concurrence , @xmath195 , is defined for a pure state @xmath200 of two qubits as @xmath201 , where @xmath202 is the @xmath89 coefficient matrix of the state .
we drop the index @xmath203 for the variables introduced in the following and denote by @xmath204 the eigenvalues of the unnormalized 1-particle reduced density operator , @xmath205 , of @xmath206 after the measurement .
thus , we get @xmath207 and it is sufficient to show that @xmath208 in order to do so we first derive an upper bound for the lhs of .
the eigenvalues of the normalized 1-particle reduced density operator , @xmath209 , of @xmath210 are @xmath211 .
hence , we obtain @xmath212 since the sum from the second term on is maximized for @xmath213 . a lower bound for the rhs of can be found by using the entropy of entanglement of the qubit system into which has been projected , @xmath214 @xcite , where @xmath215 and @xmath216 , and thus @xmath217 for @xmath218 . in our case by definition of @xmath187 and
@xmath188 the entropy of entanglement is @xmath219 since for fixed @xmath220 the uppermost sum is minimized for @xmath221 .
therefore , we have @xmath222 which is finite for @xmath223 . in the limit
@xmath224 the ratio takes the value @xmath225 , thus verifying that there exists a finite - valued function @xmath226 such that @xmath227 .
now we apply to bound the le for arbitrary @xmath228 @xmath229 with @xmath191 . in analogy
to we obtain @xmath230 we introduce an @xmath231-net ( @xmath232 ) @xmath233 of isometries contained in @xmath189 , that is , for any such isometry @xmath234 there exists a @xmath235 with operator norm @xmath236 . by choosing @xmath231 sufficiently small we get up to orders in @xmath231 @xmath237 if the le is non - zero in the thermodynamic limit , is lower bounded by @xmath238 .
consequently , for all @xmath239 there must exist @xmath240-regions @xmath241 containing isometries @xmath242 such that @xmath243 and @xmath244 we have @xmath245 and @xmath246 , respectively , with @xmath247 where @xmath248 denotes the volume , i.e. , the number of points in the @xmath231-discretization of region @xmath249 . by choosing @xmath240 small enough
, one obtains @xmath250 , such that sum in the lhs of contains @xmath251 constant terms , and therefore @xmath252 hence , for sufficiently small @xmath240 one gets @xmath253 , i.e. , is non - vanishing for some isometries @xmath254 and @xmath255 , if there is lrle for some @xmath38
we want to demonstrate that in the limit @xmath19 the sum can be non - zero only if one can choose @xmath31 such that for all well - defined @xmath256 ( i.e. those for which @xmath257 ) the inequality @xmath258 is an equality .
formally this claim reads @xmath259 or equivalently , @xmath260 which is what we want to show in the following .
thus , we assume that for the sum @xmath261 there is a minimum integer @xmath262 such that @xmath263 after building blocks of @xmath121 terms in , each of them can be upper bounded by @xmath264 which shows that the le must fulfill @xmath265 i.e. , there is no lrle .
here we show that @xmath32 can always be chosen such that is finite whenever is fulfilled .
we denote the proportionality factor in by @xmath266 for which @xmath267 implies @xmath268 .
thus reads @xmath269 following the approach in part [ app_show_qubits ] we introduce an @xmath231-net @xmath270 in the isometries of @xmath189 and obtain up to order @xmath231 @xmath271 where we defined @xmath272 consider now the sum @xmath273 since it takes a finite value , for any @xmath239 there must exist an @xmath240-region @xmath274 containing @xmath275 such that @xmath276 the relation @xmath277 holds with @xmath278 where @xmath279 and @xmath280 are the number of points of the respective regions in the @xmath231-grid .
for sufficiently small @xmath240 we consider only the following part of the sum in @xmath281 if we set @xmath282 , this is lower bounded by @xmath283 , i.e. , @xmath284 which is small but positive for sufficiently small @xmath240 .
for @xmath5 in the main text it has been shown that the case of @xmath286 corresponds to a trivial extension of the matrices for @xmath4 , @xmath287 , to @xmath5 .
moreover , it has been noted that larger possible dimensions of @xmath105 are only 6 and 8 , the latter implying @xmath137 . for the case of @xmath285
it has been shown that @xmath171 , where @xmath130 , @xmath131 and analogous definitions with @xmath158 and @xmath159 .
the reason for the introduction of @xmath288 and @xmath289 of the same type as @xmath141 and @xmath140 is that @xmath105 is closed under any @xmath106 , which maps , e.g. , @xmath141 to @xmath290 .
for some @xmath106 @xmath291 must happen not to be in @xmath292 if @xmath293 .
we see that in this case @xmath294 and @xmath295 are new orthogonal vectors with the same norm ( due to @xmath296 ) . from them
we obtain the orthonormal vectors @xmath297^{-1/2 } a_i^{\dagger}|x)$ ] and @xmath298^{-1/2 } a_i^{\dagger}|y)$ ] .
thus , @xmath21 applied from the right maps all vectors contained in @xmath139 to vectors in @xmath299 , while preserving their relative lengths , or equivalently , the angles between them .
as noted in the main text , the existence of another @xmath176 ( supporting @xmath300 and @xmath301 ) is excluded for @xmath285 .
consequently , any @xmath21 applied from the right has to map @xmath161 to @xmath161 or @xmath162 and @xmath162 to @xmath161 or @xmath162 in such a way that all angles between vectors lying in one of those subspaces are preserved . @xmath161 and @xmath302 at this point can be arbitrary two - dimensional linear subspaces of @xmath303 , whereas we assume their intersection to be along @xmath167 .
the justification for this choice is that for any orthonormal pair @xmath304 of vectors in @xmath162 @xmath169 is the full subspace of traceless matrices contained in @xmath165 . after choosing w.l.o.g . @xmath305 and @xmath306
, we can make the ansatz @xmath307 @xmath308 and @xmath309 is a unitary .
we see that indeed @xmath310 , i.e. , @xmath311 and @xmath312 , i.e. , @xmath313 are fulfilled .
the normal vector of @xmath161 is @xmath314 , and the one of @xmath162 is of the form @xmath315 , since @xmath161 and @xmath162 intersect in the @xmath316-axis . because of @xmath317
, it follows that @xmath318 .
therefore , @xmath319 + [ r^*|1 ) + s^ * |3)](3|\big),\end{aligned}\ ] ] which results in @xmath320 .
now , the matrix @xmath321 applied from the right maps @xmath161 to @xmath161 and @xmath162 to , say , @xmath322 preserving the angles between vectors lying in one of them . as a result , there are four possible cases for the action of @xmath309 @xmath323 a ) and d ) imply @xmath324 and hence @xmath325 and @xmath326 , whereas b ) and c ) mean @xmath327 . therefore ,
in the latter case the intersection of @xmath161 and @xmath322 is the @xmath316-axis .
since it has to be mapped by @xmath309 to the intersection of @xmath161 and @xmath162 , which is likewise the @xmath316-axis , @xmath316 must be an eigenvector of @xmath309 .
then , it follows that @xmath309 has a unitary action in the @xmath328-@xmath329-plane .
hence , b ) corresponds to @xmath330 , and therefore @xmath331 , @xmath332 .
last , in c ) double application of @xmath309 would map @xmath333 , wherefore @xmath334 has to carry out a reflection of @xmath162 on the @xmath161-plane , i.e. , @xmath331 and @xmath335 .
we thus obtain either @xmath331 and @xmath332 or @xmath325 and @xmath336 and conclude that @xmath337 . the requirement that @xmath21 maps @xmath161 to @xmath161 or @xmath162 and @xmath162 to @xmath161 or @xmath162 also holds for @xmath338 , from which we deduce that @xmath309 must contain only one non - vanishing entry per row and column ( with magnitude 1 ) .
all results combined together yield @xmath180(l| + e^{i \phi_i '' } |2)(m| , \end{aligned}\ ] ] with @xmath181 , @xmath182 .
f. verstraete , m. popp , and j. i. cirac , phys .
* 92 * , 027901 ( 2004 ) .
briegel , w. dr , j. i. cirac , and p. zoller , phys .
lett . * 81 * , 5932 ( 1998 ) .
f. verstraete , m. a. martin - delgado , and j. i. cirac , phys .
lett . * 92 * , 087201 ( 2004 ) .
r. orus , and h .- h .
tu , phys .
b * 83 * , 201101(r ) ( 2011 ) ; s. o. skrvseth , and s. d. bartlett , phys . rev .
a * 80 * , 022316 ( 2009 ) ; j. k. pachos , and m. b. plenio , phys .
lett . * 93 * , 056402 ( 2004 ) ; b .- q .
jin , and v. e. korepin , phys .
a * 69 * , 062314 ( 2004 ) .
j. fiurek , and l. mita , jr .
a * 75 * , 060302(r ) ( 2007 ) ; a. serafini , g. adesso , and f. illuminati , phys . rev .
a * 71 * , 032349 ( 2005 ) ; g. adesso , and f. illuminati , phys .
95 * , 150503 ( 2005 ) . v. subrahmanyama , and a. lakshminarayanb , phys .
a * 349 * , 164 ( 2006 ) .
l. campos venuti , and m. roncaglia , phys .
* 94 * , 207207 ( 2005 ) ; s. sahoo , arxiv:1201.5620v4 .
the case of lrle needs to be differentiated from the one of diverging entanglement length introduced in @xcite : a divergence of the entanglement length only implies that the le can not decay exponentially with the spin distance , however it still may decay algebraically , c.f . @xcite .
d. prez - garca , f. verstraete , m. m. wolf , and j. i. cirac , quant . inf . comp .
* 7 * , 401 ( 2007 ) . c. schn , e. solano , f. verstraete , j. i. cirac , and m. m. wolf , phys .
lett . * 95 * , 110503 ( 2005 ) .
m. fannes , b. nachtergaele , r. f. werner , commun .
* 144 * , 443 ( 1992 ) .
m. popp , f. verstraete , m. a. martin - delgado , and j. i. cirac , phys .
a * 71 * , 042306 ( 2005 ) .
o. f. syljusen , phys .
a * 322 * , 25 ( 2004 ) ; m. popp , f. verstraete , m. a. martin - delgado , and j. i. cirac , appl .
b * 82 * , 225 ( 2006 ) ; p.
androvitsaneas , e. paspalakis , and a. f. terzis , ann .
phys . * 327 * , 212 ( 2012 ) .
w. k. wootters , phys .
lett . * 80 * , 2245 ( 1998 ) .
f. verstraete , m. a. martin - delgado , and j. i. cirac , phys .
lett . * 92 * , 087201 ( 2004 ) . note that for @xmath4 all isometries are unitaries , i.e. , according to all matrices @xmath21 have to be proportional to unitaries themselves .
h. hironaka , ann .
math . * 79 * , 109 ( 1964 ) ; b. buchberger , austria , universitt innsbruck , diss . ,
1965 . c. h. bennett , h. j. bernstein , s. popescu , and b. schumacher , phys .
a * 53 * , 2046 ( 1996 ) . | we derive a criterion to determine when a translationally invariant matrix product state ( mps ) has long - range localizable entanglement , where that quantity remains finite in the thermodynamic limit .
we give examples fulfilling this criterion and eventually use it to obtain all such mps with bond dimension 2 and 3 .
= 1 localizable entanglement ( le ) @xcite is a multipartite measure characterizing the ( maximal averaged ) bipartite entanglement present in a system .
motivated by quantum repeaters @xcite , where it provides the natural figure of merit , the le has also been applied to many - body physics problems @xcite , whereby it can reveal hidden correlations , not detectable by standard observables @xcite . in both scenarios , there is an underlying spatial structure .
hence , peculiarities are expected whenever a finite amount of entanglement between two particles with arbitrary distance can be created by properly measuring the rest of the particles .
such states are said to have long - range localizable entanglement ( lrle ) @xcite and play an important role both in the context of quantum repeaters and spin chains . in the first one , they are those for which entanglement can be established at arbitrary distances . in the latter , phase transitions are signaled by a finite value of the le between any two particles irrespective of their distance .
states with lrle thus play a crucial role in those two contexts .
but , which are those states ? as most multipartite entanglement measures , the le is very hard to determine for general states .
a notable exception is the set of matrix product states ( mps ) @xcite .
this family of states describes the ground state of gapped 1d spin chains as well as those states created by sequential generation , as it is the case of atoms in a cavity @xcite .
they are characterized by a set of rank - three tensors , each one associated to a spin .
one of the indices corresponds to the spin in the @xmath0-basis , and the other two run from @xmath1 to @xmath2 , where @xmath2 is called the bond dimension .
mps of arbitrary bond dimension are dense in the set of all multipartite states @xcite , and thus , they are very relevant to describe many - body systems . for translationally invariant systems and low bond dimensions ( @xmath3 )
the le can be determined exactly @xcite . for larger bond dimensions
, one may find relatively tight lower bounds using monte - carlo methods @xcite .
in this paper we fully characterize translationally invariant mps with lrle for arbitrary bond dimension .
in particular , we give a set of necessary and sufficient conditions for such states . as we show
, those conditions can be turned into a set of polynomial equations , and thus provide us with a precise criterion to determine if a state has lrle or not .
furthermore , we give examples of non - trivial states with that property and provide the full sets of mps with lrle for bond dimension @xmath4 and @xmath5 .
the le is defined as the maximum average entanglement that can be generated between two spins of a spin chain by measuring the remaining ones @xcite .
let @xmath6 denote the density matrix of the original state .
with probability @xmath7 the outcome of a measurement @xmath8 will be @xmath9 and the system will be in the corresponding two - particle state @xmath10 .
hence , the le is given by @xmath11 where @xmath12 is the class of allowed measurements and @xmath13 an entanglement measure .
our system of consideration is an open chain of @xmath14 spin-@xmath15 particles along with two auxiliary particles of spin @xmath16 at each of the boarders .
the @xmath14 particles of the actual chain are the ones to be measured ( measurement outcomes @xmath17 ) , and the class of allowed measurements @xmath12 is the set of local projective von neumann measurements , where the same measurement is carried out on each party ( in particular , we exclude adaptive strategies ) .
therefore , the maximization of the average entanglement is performed by choosing the optimal physical basis @xmath18 .
the question to be answered in this letter is for which translationally invariant mps a finite amount of entanglement can be localized between the two ancillas in the limit @xmath19 .
we assume the state of the system to be translationally invariant apart from boundary effects ; for this reason the rank - three tensors corresponding to the spin-@xmath15 particles are taken equal .
those consist of complex @xmath20 matrices @xmath21 ( @xmath22 ) and can be taken to be in canonical form , in which the maps @xmath23 and @xmath24 satisfy ( cf .
@xcite ) @xmath25 for some diagonal positive definite matrix @xmath26 .
our goal is to find necessary and sufficient conditions on the matrices @xmath27 to give rise to lrle for some matrices of the auxiliary particles .
those can be chosen at will and are denoted by @xmath28 , where @xmath29 is the hilbert space dimension of the individual auxiliary spins .
the initial mps is therefore @xmath30 .
spin-@xmath15 particles and one auxiliary particle of spin @xmath16 at each of the borders .
the matrices of the real particles are @xmath21 and those of the auxiliary particles @xmath31 and @xmath32 , respectively . ]
subsequent to a measurement
@xmath8 the initial state of the system @xmath33 reduces with probability @xmath34 to @xmath35 , where ( excluding cases with @xmath36 ) @xmath37 it is shown that any mps with lrle for @xmath38 also has lrle for @xmath39 ( the converse is obvious ) , and also that w.l.o.g .
we can take @xmath40 and @xmath32 as isometries .
we can thus choose @xmath39 and the concurrence @xcite as the measure of entanglement , @xmath41 , where @xmath42 is the matrix with coefficients @xmath43 .
dropping the superscripts @xmath12 and @xmath44 , the le reads @xmath45 this optimization problem is in general hard , since the sum needs to be evaluated for large @xmath14 before being able to optimize over the physical basis @xmath46 .
its maximization succeeded only in special cases , like for a chain of @xmath14 spins without auxiliary ones and measurement on all spins but those at the borders @xcite . in this case
@xmath47 , which implies that there is lrle for @xmath48 . in the following
we slightly generalize this family of states and adapt them to our system that includes the auxiliary spins at the boundary : _ example 1.block structure of unitaries .
_ @xmath49 where @xmath50 is a @xmath51 permutation matrix .
the @xmath52 are @xmath53 unitaries ( @xmath54 ) and @xmath55 .
we can realize that this mps has lrle for @xmath56 by choosing @xmath57 + @xmath58 ( where @xmath59 are basis vectors of the auxiliary qubits and @xmath60 the basis vectors of the basis in which the matrices are given ) and @xmath61 . then , due to @xmath62 and therefore @xmath63 is finite in the thermodynamic limit .
the fact that there is also lrle for @xmath64 will become clear below . in the above examples
the matrices @xmath21 exhibit a block structure of unitaries .
one may think that all states with lrle need to have this property in some basis @xmath46 .
interestingly , this is not the case as shown by the following counterexample for @xmath5 : _ example 2.non-unitary matrices .
_ @xmath65 with @xmath66 and @xmath67 ( \uparrow|$ ] @xmath68(\downarrow|$ ] .
it is a simple exercise to show that remains finite in the limit @xmath19 .
the matrices @xmath21 fulfill @xmath69 .
however , @xmath70 , and thus they are not of the form @xmath48 in any basis @xmath46 . note that the mps can numerically be verified to be injective @xcite , and thus it is the unique ground state of a local translationally invariant gapped hamiltonian .
moreover , it can be easily shown ( analytically ) that it is invariant under the @xmath71-symmetry generated by the transformation @xmath72 , @xmath73 , @xmath74 .
the question arises of how one can check whether a given mps has lrle without having to resort to evaluating for large @xmath14 numerically . in the following a necessary and sufficient criterion
will be derived , which allows to decide this based on the matrices @xmath27 directly .
we first rewrite by inserting to the right of @xmath75 a projector on the subspace spanned by the row vectors of @xmath76 , which can be written as @xmath77 , where @xmath78 is an isometry .
thus , reads now @xmath79 using the svd of @xmath80 , which is @xmath81 ( neglecting the indices @xmath82 ) , along with the inequality of arithmetic and geometric means , one obtains for the factors of @xmath83 since @xmath84 . the inequality becomes an equality if and only if @xmath85 , i.e. , @xmath80 is proportional to a unitary .
the intuitive fact that all factors but the last of have to be exactly 1 to get lrle is proven in appendix [ lemma2 ] .
therefore , a necessary condition for any mps to give rise to lrle is that there exists an isometry @xmath31 , such that for a certain basis @xmath86 @xmath87 @xmath88 denoting some @xmath89 unitary .
redefining @xmath90 shows that one can require the rhs of to be the identity . after multiplying this from the right by @xmath91 , one obtains @xmath92 which can also be written as @xmath93 as illustrated by example 1 and 2 , @xmath32
can always been chosen such that is finite , which is shown rigorously in appendix [ app_suffq ] .
criterion is thus necessary and sufficient for the emergence of lrle @xcite .
if we define @xmath94 , i.e. , @xmath95 and @xmath96 , is equivalent to @xmath97 after further defining @xmath98 , @xmath99 , the map @xmath100 and @xmath101 we see that and are equivalent to @xmath102 respectively .
thus , if we define the subspace @xmath103 , and indicate that the occurrence of lrle is equivalent to @xmath104 . by definition @xmath105
is closed under the application of any @xmath106 , which leads to the following criterion characterizing translationally invariant mps with lrle : _ theorem 1.for the mps considered here , there is lrle if and only if for a certain basis @xmath46 there exists a subspace @xmath105 of the vector space of @xmath20 matrices satisfying the following conditions _ 1 .
@xmath105 _ is closed under all linear maps @xmath106 defined as @xmath107 , i.e. , @xmath108 _ , 2 .
@xmath104 , 3 . @xmath109 _ s.t .
_ @xmath110 , @xmath111 .
[ vw ] note that if theorem 1 is fulfilled , @xmath112 is also traceless , i.e. , @xmath105 could additionally be required to contain @xmath113 and correspondingly to be equal to its adjoint , @xmath114 .
theorem 1 can be used numerically to determine , whether a given mps has lrle , since it imposes conditions on the matrices @xmath27 , which can be represented by a set of polynomial equations : the entries of @xmath27 will give rise to the coefficients of those equations , whereas all other quantities introduced below will constitute their variables to be determined numerically .
the first variables to be introduced are orthonormal basis vectors @xmath115 of @xmath105 , where @xmath116 is the dimension of @xmath105 .
condition i is equivalent to requiring that any basis vector @xmath117 of @xmath105 is mapped by any @xmath106 into @xmath105 , i.e. , for all @xmath118 @xmath119 constituting the last of the set of equations to to be solved . generally , a set @xmath120 of @xmath121 polynomial equations ( @xmath122 ) with @xmath123 variables @xmath124 can be solved by means of a grbner basis @xcite .
a grbner basis @xmath125 is a special basis in the vector space of functions involving the variables @xmath124 : it has the property that the set of equations to be solved is equivalent to the set @xmath126 ( @xmath122 ) , which , in contrast , can be solved by back - substitution while having to deal with the solution of polynomials involving only one variable at a time .
for instance , one of the new equations might involve only @xmath127 . after solving it numerically the result
can be inserted into another equation involving , e.g. , only @xmath127 and @xmath128 etc .
( cf .
gaussian elimination ) . a grbner basis can be found systematically by use of buchberger s
algorithm @xcite , which is doubly exponential in the complexity of the set of equations to be solved . in our case
, this implies a computational cost that is doubly exponential in the square of the bond dimension .
however , @xmath2 is a constant and in particular independent of the length of the spin chain . if a simultaneous solution to to is found , the mps has lrle .
if even for @xmath129 no solution is found , it does not .
we now employ theorem 1 analytically to determine the complete sets of mps with lrle for @xmath4 and @xmath5 . in the former case
, we will reproduce the finding that the matrices need to be proportional to unitaries . in the latter ,
we obtain the result that matrices of the type of example 2 are the only non - trivial matrices which give rise to lrle . in both cases
we take @xmath105 to be equal to its adjoint , i.e. , it has to contain the matrices @xmath130 , @xmath131 and @xmath132 . for @xmath4
this implies that @xmath105 is the full subspace of traceless @xmath89 matrices .
if we take any element @xmath133 , @xmath134 shows that @xmath135 is orthogonal to @xmath136 with respect to the hilbert - schmidt scalar product .
however , the orthogonal subspace of @xmath136 is spanned by the identity , i.e. , @xmath137 for any mps with lrle . for @xmath5
we consider first the case of @xmath138 .
we define @xmath139 , whereas tracelessness of @xmath140 and @xmath141 implies @xmath142 and @xmath143 , respectively ( the latter will be taken equal to 1 in the following ) .
it follows that @xmath144 . if w.l.o.g .
we set @xmath145 and @xmath146 , we observe that we retrieve the case of @xmath4 , as @xmath147 where @xmath148 is proportional to a unitary , and @xmath149 and @xmath150 are arbitrary . if @xmath105 is of larger dimension , there are other matrices @xmath151 , @xmath152 and @xmath153 ( e.g. , @xmath154 with @xmath155 etc . ) , which do not span the same space as @xmath156 and @xmath132 .
we define @xmath157 and take @xmath158 and @xmath159 as orthonormal ( which is possible due to @xmath160 ) . since @xmath161 and @xmath162 intersect , we can assume the intersection to be w.l.o.g .
along @xmath163 . as @xmath164 and @xmath153
span the full space of traceless matrices contained in @xmath165 , we are allowed to rotate the basis @xmath166 for @xmath162 such that @xmath167 .
then @xmath168 and @xmath169 might only intersect in @xmath170 , which is , however , not traceless .
consequently , @xmath105 is at least of dimension 6 . in the case of @xmath171
it is simple to construct the orthogonal complement of @xmath105 , which is @xmath172 , where @xmath173 and @xmath174 are the normal vectors of @xmath161 and @xmath162 , respectively .
this structure of @xmath175 uniquely defines @xmath161 and @xmath162 and vice versa . therefore ,
another @xmath176 would not be consistent with this @xmath175 , and the only possibility would be @xmath177 corresponding to @xmath137 ( i.e. , @xmath178 ) .
thus , @xmath179 is the only remaining case which might lead to a non - trivial mps . in this case there
exist the two subspaces @xmath161 and @xmath162 of the afore - mentioned type . based on their properties an ansatz for @xmath21
is made in appendix [ a_i_cases ] , which after an elaborate case differentiation leads to @xmath180(l| + e^{i \phi_i '' } |2)(m| , \end{aligned}\ ] ] with @xmath181 , @xmath182 , as in example 2 . within the framework of mps we have specified the localizable entanglement of a spin chain with one auxiliary spin at each of the borders .
we have shown that lrle can be detected by placing qubits at the ends of the chain .
based on that we were able to derive a theorem according to which it can be checked directly from the matrices of the mps , whether it possesses lrle .
how this can been done in practice has been indicated by eqs . - , which is a polynomial set of equations that can be solved numerically in a systematic manner .
furthermore , we provided non - trivial examples of mps , for which those equations have a simultaneous solution , determining the full sets of mps with lrle for @xmath4 and @xmath5 .
this work was supported by the spanish grants s2009/esp-1594 , mtm2011 - 26912 , the european projects quevadis and aqute and the chist - era project cqc .
we are grateful to the benasque center for science and the perimeter institute for their hospitality during the course of this work .
t. b. w. would like to thank t. s. cubitt for helpful suggestions . |
the discovery of the higgs - like particle ( @xmath9 ) by atlas @xcite and cms @xcite collaborations is a great triumph of particle physics .
the following immediate question is whether this newly discovered particle is really the higgs particle in the standard model ( sm ) . to achieve this goal
, one has to determine its couplings to all the sm particles .
however , recent results from higgs searches at the lhc have already shown a hint of new physics in the higgs diphoton decay channel .
the experimental data have shown that the diphoton decay rate is about @xmath10 times larger than the sm prediction , while the measurements in other channels including @xmath11 and @xmath12 agree with the sm higgs properties . due to the fact that the higgs particle is electric neutral , its coupling to diphoton must be induced by some charged particles running in the loops , while in the sm , the w - boson and top - quark loops give the dominant contributions @xcite . if the enhancement in the diphoton channel persists , it clearly indicates @xcite that there must be some additional new charged particles which couple to the higgs and mediate this decay process .
based on the electroweak ( ew ) @xmath13 gauge symmetry , these new particles should also contribute to the @xmath14 channel and generically lead to a shift in the @xmath6 decay width from the sm expectation . as pointed out in refs .
@xcite , the simultaneous measurements of the @xmath15 and @xmath6 channels at the lhc will provide us with valuable information about the structure of new physics . in this paper
, we focus on the simple scenario in which extra contributions to the @xmath15 and @xmath6 decay widths arise from some new scalars beyond the sm particle content .
some similar scenarios have already been explored in refs .
however , the formulae for the scalar contributions to the @xmath6 decay width used in the literature @xcite are not consistent with each other .
moreover , they can not be reduced to the corresponding rates for @xmath16 by taking the limit of @xmath4 and making the charge replacement . as a result , these numerical predictions for the @xmath6 decay width may not be reliable . clearly , it is timely important to re - calculate the scalar contributions to the decay .
this paper is organized as follows . in section [ sec:2 ]
, we show the detail calculations of the scalar contributions to the decay rate of @xmath0 . in section [ sec:3 ] ,
we give phenomenological analyses of the correlations between the @xmath15 and @xmath6 channels in several specific sets as well as a generic one of the scalar multiplets .
we summarize our results in section [ sec:4 ] .
for a general scalar particle @xmath17 with the third weak isospin charge @xmath18 and non - trivial electric charge @xmath19 , in this paper .
] the relevant lagrangian involving s is given by @xmath20 where @xmath21 is the sm higgs doublet ; @xmath22 is the @xmath23 generators , and the covariant derivative involving @xmath24 and @xmath25 gauge fields is defined as @xmath26 where @xmath27 with @xmath28 and @xmath29 ( @xmath30 being the weinberg mixing angle ) .
we note that the mass splittings at tree level among the components of the scalar @xmath17 come from the last term in eq .
( [ lagrangian ] ) , while they also receive loop - induced contributions of @xmath31(100 ) mev due to the gauge boson exchange diagrams @xcite , which can be ignored in our discussion . on the other hand , the electroweak precision measurements ,
that is , the oblique parameters would constrain the mass differences @xmath32 among the scalar multiplet @xcite to be smaller than a few @xmath31(10 ) gev @xcite . as a result
, we assume the masses of the scalar multiplet are degenerate in this paper for simplicity . the trilinear coupling between higgs and the scalar particle is @xmath33 after higgs develops the vacuum expectation value , @xmath34 . the amplitudes for the three feynman diagrams shown in fig .
[ feyndiag ] can be easily written down : ) contributing to @xmath35 .
[ feyndiag],scaledwidth=100.0% ] @xmath36[(l+q+p)^2-m_s^2]}\ , , \\
-i{\cal m}^{(2 ) } & = e^2 q g_{zss}\lambda_{hs}v \mu^{4-d}\int\frac{d^d l}{(2\pi)^d}\frac{(2l - q)^{\mu}(2l-2q - p)^\nu \epsilon^{z*}_\mu(q ) \epsilon^{a*}_\nu(p)}{(l^2- m_s^2)[(l - q)^2-m_s^2][(l - q - p)^2-m_s^2]}\ , , \\
-i{\cal m}^{(3 ) } & = -2e^2 q g_{zss } \lambda_{hs } v \mu^{4-d}\int\frac{d^d l}{(2\pi)^d}\frac{g^{\mu\nu}}{(l^2-m_s^2)[(l+p+q)^2-m_s^2 ] } \epsilon^{z*}_\mu(q ) \epsilon^{a*}_\nu(p)\ , , \end{split}\ ] ] where we have taken the dimensional regularization .
note that by changing the integration variable @xmath37 , @xmath38 can be transformed into the form that is exactly identical to @xmath39 , so that we only need to the calculate @xmath40 . with the standard procedures of the feynman parametrization and the translation of the integral momentum @xmath41 , we obtain : @xmath42g^{\mu\nu}\}\epsilon^{z*}_\mu(q ) \epsilon^{a*}_\nu(p)}{(l^2- m_s^2)[(l+q)^2-m_s^2][(l+q+p)^2-m_s^2 ] } \\ & = e^2 ( 2q ) g_{zss}\lambda_{hs}v \mu^{4-d}\gamma(3)\int dx dy\int\frac{d^d l}{(2\pi)^d } \left[\frac{4l^\mu l^\nu}{(l^2-\delta^2)^3}-g^{\mu\nu}\frac{1}{(l^2-\delta^2)^2}\right .
\\ & \left .
+ \frac{-4y(1-x - y)(p^\mu q^\nu - g^{\mu\nu}p\cdot q ) -(1 - 2x-2y)(1-x - y ) m_z^2g^{\mu\nu}}{(l^2-\delta^2)^3}\right]\epsilon^{z*}_\mu(q ) \epsilon^{a*}_\nu(p ) , \end{split}\ ] ] where @xmath43 . in the second line of eq .
( [ eq2.4 ] ) , we have used the following on - shell identities : @xmath44 with the help of the dimensional regularization , it can be proved that in the second equation in eq .
( [ eq2.4 ] ) , the first two terms exactly cancel with each other and only the last finite term is left , given by @xmath45 from this equation after the integration , the first term leads to @xmath46 while the second one vanishes , where @xmath47 , @xmath48 and the loop function @xmath49 is defined in appendix [ loop func ] . by combining the sm contributions and integrating the phase space of outgoing particles ,
we obtain the @xmath14 decay width , @xmath50 with @xmath51\,,\ ] ] where @xmath52 , @xmath53 ( @xmath54 ) , and the summation is over the different isospin components in given @xmath55 multiplets .
we point out that for the non - sm contribution to the decay rate of @xmath0 , there are extra factors of -1/2 in refs .
@xcite and -1 in ref .
@xcite in comparing with ours in eq .
( [ ampzgamma ] ) .
the different signs for the scalar contributions to @xmath56 clearly lead to different results by taking limits to @xmath57 .
the modification in the partial decay width of @xmath0 is then expressed in terms of the enhancement factor , given by @xmath58 where @xmath59 represents the decay width in the sm and the factor @xmath60 represents the degeneracy of the multiplet . for completeness and convenience for the later discussions , we also present the standard formula for the enhancement factor of the diphoton rate , @xmath61 where the loop functions @xmath62 are defined in appendix [ loop func ] .
it is straightforward to show that our formula in eq .
( [ zg ] ) for the @xmath6 decay can be retrieved to the one in eq .
( [ gg ] ) for the @xmath15 mode when taking @xmath63 and making the replacement @xmath64 . if the mass splittings among components of the multiplet are taken into account , the deviations of the contributions in eqs .
( [ zg ] ) and ( [ gg ] ) are approximately proportional to @xmath65 .
the modifications from the mass splittings are at percentage level and are negligible due to the electroweak precision measurements .
consequently , we will simply take eqs .
( [ zg ] ) and ( [ gg ] ) to analyze the qualitative behaviors in @xmath3 and @xmath0 decays .
with formulae in eqs .
( [ zg ] ) and ( [ gg ] ) at hand , we are able to study the correlations between @xmath66 and @xmath14 decay rates . according to the ew gauge symmetry , one expects that the particle responsible for the observed enhancement in the diphoton channel could also lead to the modification of the @xmath6 rate .
furthermore , it is interesting to notice that the @xmath6 amplitude depends not only on the electric charges of particles running in the loops but also on their isospins , in contrast with the @xmath15 case in which only electric charges can be probed .
consequently , a combined analysis of the decay widths in these two modes can provide us with important information on the ew charges of particles contributing to both processes . in this subsection ,
we revisit two simple cases of singlet and doublet scalars under @xmath55 with the couplings : @xmath67 where @xmath68 , corresponding to @xmath69 and @xmath70 , respectively .
the aim for this part of the investigation is twofold : ( a ) illustrating the features of the isospin and hypercharge dominations and ( b ) reexamining the results in ref .
@xcite . in the two cases in eq .
( [ 2cases ] ) , since only charged particles with @xmath71 run in the loops , the enhancements for the diphoton rate are expected to be the same . however , due to the different isospin representations , those for the @xmath6 mode are distinct . from fig .
[ zgl ] , we see that the @xmath6 rate in case i ( ii ) has a small suppression ( enhancement ) in comparison with the sm value .
( solid lines ) and @xmath72 ( dashed lines with the values in the yellow boxes ) in @xmath73-@xmath74 plane for scalars of ( a ) singlet ( t=0 and y=2 ) and ( b ) doublet ( t=1/2 and y=1 ) , where the solid ( dashed ) lines from top to bottom correspond to the values of ( a ) 1.25 ( 0.94 ) , 1.50 ( 0.9 ) , 1.75 ( 0.86 ) , 2.00 ( 0.82 ) and 2.25 ( 0.74 ) ; and ( b ) 1.25 ( 1.10 ) , 1.50 ( 1.15 ) , 1.75 ( 1.20 ) , 2.00 ( 1.25 ) and 2.25 ( 1.35 ) , respectively . , title="fig:",width=283,height=226 ] ( solid lines ) and @xmath72 ( dashed lines with the values in the yellow boxes ) in @xmath73-@xmath74 plane for scalars of ( a ) singlet ( t=0 and y=2 ) and ( b ) doublet ( t=1/2 and y=1 ) , where the solid ( dashed ) lines from top to bottom correspond to the values of ( a ) 1.25 ( 0.94 ) , 1.50 ( 0.9 ) , 1.75 ( 0.86 ) , 2.00 ( 0.82 ) and 2.25 ( 0.74 ) ; and ( b ) 1.25 ( 1.10 ) , 1.50 ( 1.15 ) , 1.75 ( 1.20 ) , 2.00 ( 1.25 ) and 2.25 ( 1.35 ) , respectively . , title="fig:",width=283,height=226 ] these features can be understood from the general formula in eq .
( [ zg ] ) .
in order to enhance the @xmath16 rate by a constructive interference , it is required that @xmath74 must be negative since @xmath75 and @xmath76 possess an opposite sign . in this case
, the @xmath0 rate depends only on the sign of @xmath77 as @xmath78 corresponds to both positive values of @xmath79 and @xmath80 in the physically interesting mass regime . as a result , @xmath72 is enhanced ( suppressed ) if @xmath77 is positive ( negative ) .
accordingly , it is easy to see that @xmath81 in case i ( ii ) . note that our result is opposite to that in ref .
@xcite , in which the scalar contributions to @xmath82 and @xmath56 are positive correlated in case i and negative correlated in case ii due to the extra minus sign in the formula .
the factor @xmath83 can be examined in a general scalar multiplet . by using the identity @xmath84
, we have @xmath85 when the isoweak charge @xmath1 of the scalar multiplet is larger than its hypercharge @xmath2 , we find @xmath86 , which indicates the enhanced behavior in the @xmath6 channel compared with the sm prediction .
in contrast , for the multiplet with @xmath87 , we get @xmath88 , leading to a suppressed rate of @xmath0 .
precisely , cases i and ii are the two simple but nontrivial archetypes of the two limits , respectively .
it is known that triplet scalars are introduced in many sm extensions , such as the type ii seesaw model @xcite .
current lhc experiments also plan to search for these triplet scalars , especially the doubly charged particle in the @xmath91 case due to its interesting signatures @xcite . instead of a special model , we examine the generic feature of these triplets . in particular
, we investigate their loop effects in the @xmath15 and @xmath6 decay channels of the higgs particle .
for simplicity , we assume the multiplets of scalars possess the same masses and no mixings between the triplet scalars and the sm higgs .
for more detail analysis one can refer to refs .
our result is shown in fig .
[ zg3 ] .
( solid lines ) and @xmath72 ( dashed lines ) in @xmath73-@xmath74 plane for the triplet scalars of ( a ) ( t , y)= ( 1,0 ) and ( b ) ( t , y)=(1,2 ) , where the solid ( dashed ) lines from top to bottom correspond to the values of ( a ) 1.25 ( 1.25 ) , 1.50 ( 1.53 ) , 1.75 ( 1.75 ) , and 2.00 ( 2.00 ) ; and ( b ) 1.25 ( 1.05 ) , 1.50 ( 1.10 ) , 1.75 ( 1.15 ) , and 2.00 ( 1.20 ) , respectively . , title="fig:",width=283,height=226 ] ( solid lines ) and @xmath72 ( dashed lines ) in @xmath73-@xmath74 plane for the triplet scalars of ( a ) ( t , y)= ( 1,0 ) and ( b ) ( t , y)=(1,2 ) , where the solid ( dashed ) lines from top to bottom correspond to the values of ( a ) 1.25 ( 1.25 ) , 1.50 ( 1.53 ) , 1.75 ( 1.75 ) , and 2.00 ( 2.00 ) ; and ( b ) 1.25 ( 1.05 ) , 1.50 ( 1.10 ) , 1.75 ( 1.15 ) , and 2.00 ( 1.20 ) , respectively .
, title="fig:",width=283,height=226 ] clearly , the @xmath6 rate is enhanced in both cases , which belong to the isospin domination class discussed in the previous subsection .
furthermore , the enhancement in the @xmath92 case is much larger than that in the @xmath91 one , which implies that increasing the value of @xmath2 will decrease the @xmath0 decay rate for a given isospin multiplet .
moreover , in comparison with case i of @xmath69 in sec .
iii - a , @xmath93 for the case of @xmath94 for a given rate of @xmath3 , say , 1.5 times the sm value , is slightly larger , showing that a large @xmath95 multiplet helps to increase the @xmath0 rate .
the weak @xmath95 5-plet scalar particles are also of great phenomenological interest .
for example , a recent study in ref .
@xcite shows that with the 5-plet and @xmath91 scalar , the small majorana neutrino masses can be generated at two - loop level .
the @xmath15 and @xmath6 results are presented in fig .
[ zg5 ] .
( solid lines ) and @xmath72 ( dashed lines ) in @xmath73-@xmath74 plane for the 5-plet scalars of ( a ) ( t , y)= ( 2,0 ) and ( b ) ( t , y)=(2,2 ) , where the solid ( dashed ) lines from top to bottom correspond to the values of ( a ) 1.25 ( 1.25 ) , 1.50 ( 1.53 ) , 1.75 ( 1.75 ) , and 2.00 ( 2.00 ) ; and ( b ) 1.25 ( 1.10 ) , 1.50 ( 1.15 ) , 1.75 ( 1.25 ) , and 2.00 ( 1.50 ) , respectively .
, title="fig:",width=283,height=226 ] ( solid lines ) and @xmath72 ( dashed lines ) in @xmath73-@xmath74 plane for the 5-plet scalars of ( a ) ( t , y)= ( 2,0 ) and ( b ) ( t , y)=(2,2 ) , where the solid ( dashed ) lines from top to bottom correspond to the values of ( a ) 1.25 ( 1.25 ) , 1.50 ( 1.53 ) , 1.75 ( 1.75 ) , and 2.00 ( 2.00 ) ; and ( b ) 1.25 ( 1.10 ) , 1.50 ( 1.15 ) , 1.75 ( 1.25 ) , and 2.00 ( 1.50 ) , respectively . , title="fig:",width=283,height=226 ] similar to the @xmath95 triplet cases , the 5-plet scalars give sizable enhancements in the @xmath0 channel when the diphoton rate is lying in 1.5 to 2.0 times larger than the sm value .
explicitly , the enhancement factor @xmath72 is also about 1.5 to 2.0 ( 1.25 to 1.50 ) for the @xmath98 case .
again , the behaviors of these two cases can be expected by the rules of the isospin domination .
we note that a bit less enhancement in the @xmath91 case is due to the hypercharge cancellation . in this subsection
, we give a more comprehensive overview of how to use the combined analysis of the @xmath15 and @xmath6 modes to investigate the nature of the new scalar multiplets . by assuming that the excess of @xmath3 is originated from a constructive interference between the sm and charged scalar loops , we obtain @xmath99}{v{\cal a}^{z\gamma}_{sm } } \frac{a_0^{z\gamma}(\tau_s , \lambda_s)}{a_0^{\gamma\gamma}(\tau_s ) } \frac{\sum_{t_3 } q_s\cdot g_{zss}}{\sum_{t_3}q^2_s } \\ & = 2.71\cdot \frac{a_0^{z\gamma}(\tau_s , \lambda_s)}{a_0^{\gamma\gamma}(\tau_s ) } \frac{4t(t+1)c_w^2 - 3y^2 s_w^2}{4t(t+1)+3y^2 } \end{split}\ ] ] for a scalar with general charges of @xmath1 and @xmath2 . from eq .
( [ general ] ) , @xmath72 increases with a higher value gauge coupling to violate perturbativity for @xmath100 @xcite .
the main point in this paper is to show the quantitative behaviors of the scalars with arbitrary ew quantum numbers .
] of @xmath1 since @xmath101 and @xmath102 are both positive , whereas it decreases with a larger @xmath103 . furthermore ,
when @xmath104 , @xmath72 is saturated to a maximal value with a fixed value of @xmath105 , given by @xmath106 which is definitely positive , implying that the @xmath6 decay width must be enhanced .
this limit corresponds to the isospin domination case as discussed in sec .
iii - a . in fig .
[ changet ] , we show @xmath72 as a function of the isospin @xmath1 with @xmath107 and the scalar mass @xmath108 gev . from the figure , we find that the saturated value of @xmath109 is @xmath110 . as a function of the isospin @xmath1 with @xmath107 and @xmath108 gev . ] on the other hand , when we take the large hypercharge limit @xmath111 , that is , the hypercharge domination case , we obtain the minimum value for the factor @xmath72 .
@xmath112 in fig .
[ changey ] , we show @xmath72 as a function of the hapercharge @xmath2 with @xmath107 and @xmath108 gev . in the large-@xmath103 limit , we get the asymptotic value of @xmath113 . as a function of the hypercharge @xmath2 with @xmath107 and @xmath108 gev .
] the above features become more transparent in the large scalar mass limit , @xmath114 . in this limit
, we have that @xmath115 , and thus eq .
( [ general ] ) can be reduced to @xmath116 in most of the physically interesting regime , eq .
( [ largemass ] ) is a very good approximation with the error of ten percent for the scalar mass above 200 gev in which the ratio @xmath117 is numerically stable around @xmath118 . a remarkable feature shown in eq .
( [ largemass ] ) is that the @xmath0 decay rate for the higgs boson only depends on the ew quantum numbers assigned to the scalar multiplet for a given enhancement in the @xmath15 channel . as an illustration , fig .
[ rlm ] shows the contours of @xmath72 in terms of the general @xmath13 quantum numbers with @xmath107 for both @xmath119 and @xmath120 gev , respectively . in the plane of general @xmath13 quantum numbers by fixing @xmath107 , where the solid and dashed lines represent the results of @xmath121 and @xmath122 gev , respectively.,width=302,height=245 ] in the large mass regime ,
the maximum and minimum values of @xmath72 can be easily obtained , @xmath123 ^ 2 $ ] and @xmath124 ^ 2 $ ] , corresponding to the isospin and hypercharge dominations , respectively .
finally , we can further demonstrate that the above bounds hold when we extend our discussion to arbitrary numbers of scalar multiplets with different quantum numbers , masses @xmath125 , higgs couplings @xmath126 , and the degeneracies @xmath127 .
the formula in eq .
( [ general ] ) can be generalized to : @xmath128}{\sum_i n_i\frac{|\lambda_i|}{m_i}(2t_i+1)[4t_i(t_i+1)+3y_i^2]}.\ ] ] where we have taken @xmath129 gev .
note that eq .
( [ moregeneral ] ) is valid only for the new scalar contributions interfering with the sm part constructively .
this assumption follows by two indications : ( i ) @xmath126 should be negative in sign , so that we can replace @xmath126 with its absolute value in eq .
( [ moregeneral ] ) ; and ( ii ) contributions to the amplitudes of the @xmath16 and @xmath6 modes must be smaller than the corresponding sm parts in size to eliminate the sign ambiguity when we take the square roots of @xmath105 and @xmath72 .
we now try to derive a generic bound for @xmath72 . from eq .
( [ moregeneral ] ) and inequalities @xmath130 < 4t_i(t_i+1)c_w^2 - 3y_i^2 s_w^2 < 1.04 \left[4t_i(t_i+1)+3y_i^2\right]~~\;\ ] ] for a single multiplet , we find @xmath131 from the general bound in eq .
( [ gb ] ) , it is clear that one can deduce the allowed range of the @xmath6 decay width for a given value of @xmath105 .
since the current atlas and cms data have indicated the range @xmath132 for the diphoton mode , one predicts that @xmath133 for the @xmath6 mode , where both the upper and lower limits arise from @xmath134 .
we emphasize that the result in eq .
( [ zgb ] ) does not rely on the details of the extra scalars , such as the quantum numbers and their couplings to the higgs particle .
thus , if the future - measured diphoton rate is still enhanced about @xmath135 times over the sm value , the @xmath6 decay rate should be in the range of 0.76 to 2.05 .
otherwise , our simple scenario discussed in this paper will be excluded .
we have studied the decay width of the @xmath0 mode with extra scalar multiplets of arbitrary weak isospins and hypercharges . due to the sm gauge symmetry
, these new scalars which result in the excess of the @xmath3 decay rate would generically shift the @xmath0 decay rate as well .
the combined analysis of these two modes can provide us with valuable information about the new physics structure .
regarding the ambiguity of the formulae for @xmath93 in the literature , we have revisited some simple extensions to the sm in which new scalar multiplets are introduced .
we have found that @xmath93 only depends on the relative size of the isospin @xmath1 and the absolute value of the hypercharge @xmath2 . in particular , we have shown that the enhancement factor @xmath72 is a monotonically increasing function of @xmath1 and a monotonically decreasing one of @xmath103 with the fixed value of @xmath105 .
this observation enables us to predict that @xmath136 by imposing the observed range of @xmath137 if the scalars are heavier than 200 gev .
note that this range is irrelevant to the number of the scalar multiplets , their representations , and the couplings to the higgs particle .
our results on the @xmath0 decay clearly can be tested at the lhc .
the functions related to @xmath66 and @xmath14 are defined as follows : @xmath138,\\ a_{1/2}^{\gamma\gamma } ( x ) & = 2x^2[x^{-1}+(x^{-1}-1)f(x^{-1})],\\ a_0^{\gamma\gamma } ( x ) & = -x^2[x^{-1}-f(x^{-1})],\\ a_1^{z\gamma}(x , y ) & = 4(3-\tan^2\theta_w)i_2(x , y)+[(1 + 2x^{-1})\tan^2 \theta_w-(5 + 2x^{-1})]i_1(x , y),\\ a_{1/2}^{z\gamma } ( x , y ) & = i_1(x , y)-i_2(x , y),\\ a_0^{z\gamma}(x , y ) & = i_1(x , y),\end{aligned}\ ] ] the work was supported in part by national center for theoretical science , national science council ( nsc-98 - 2112-m-007 - 008-my3 and nsc-101 - 2112-m-007 - 006-my3 ) and national tsing - hua university ( 102n1087e1 ) , taiwan , r.o.c .
99 g. aad _ et al . _ [ atlas collaboration ] , phys .
b * 716 * , 1 ( 2012 ) [ arxiv:1207.7214 [ hep - ex ] ] .
s. chatrchyan _ et al . _
[ cms collaboration ] , phys .
b * 716 * , 30 ( 2012 ) [ arxiv:1207.7235 [ hep - ex ] ] .
m. a. shifman , a. i. vainshtein , m. b. voloshin , and v.i .
zakharov , sov .
30 * , 711 ( 1979 ) [ yad .
fiz . * 30 * , 1368 ( 1979 ) ]
. j. r. ellis , m. k. gaillard , and d. v. nanopoulos , nucl .
b * 106 * ( 1976 ) 292 .
b. l. ioffe and v. a. khoze , sov . j. part
. nucl . * 9 * , 50 ( 1978 ) [ fiz .
chast . atom .
yadra * 9 * , 118 ( 1978 ) ] .
t. g. rizzo , phys .
d * 22 * , 178 ( 1980 ) [ addendum - ibid .
d * 22 * , 1824 ( 1980 ) ] .
l. resnick , m. k. sundaresan and p. j. s. watson , phys .
d * 8 * , 172 ( 1973 ) .
f. j. petriello , jhep * 0205 * , 003 ( 2002 ) .
t. han , h. e. logan , b. mcelrath and l. -t .
wang , phys .
b * 563 * , 191 ( 2003 ) [ erratum - ibid .
b * 603 * , 257 ( 2004 ) ] [ hep - ph/0302188 ] . c. -r .
chen , k. tobe and c. -p .
yuan , phys .
b * 640 * , 263 ( 2006 ) [ hep - ph/0602211 ] .
g. cacciapaglia , a. deandrea and j. llodra - perez , jhep * 0906 * , 054 ( 2009 ) [ arxiv:0901.0927 [ hep - ph ] ]
. j. cao , z. heng , t. liu and j. m. yang , phys .
b * 703 * , 462 ( 2011 ) [ arxiv:1103.0631 [ hep - ph ] ] .
s. dawson and e. furlan , phys .
d * 86 * , 015021 ( 2012 ) [ arxiv:1205.4733 [ hep - ph ] ] . t. kitahara , jhep * 1211 * , 021 ( 2012 ) [ arxiv:1208.4792 [ hep - ph ] ] .
a. joglekar , p. schwaller and c. e. m. wagner , jhep * 1212 * , 064 ( 2012 ) [ arxiv:1207.4235 [ hep - ph ] ] ; m. hashimoto and v. a. miransky , phys .
d * 86 * , 095018 ( 2012 ) [ arxiv:1208.1305 [ hep - ph ] ] .
a. delgado , g. nardini and m. quiros , phys .
d * 86 * , 115010 ( 2012 ) [ arxiv:1207.6596 [ hep - ph ] ] . h. an , t. liu and l. -t .
wang , phys .
d * 86 * , 075030 ( 2012 ) [ arxiv:1207.2473 [ hep - ph ] ] .
n. arkani - hamed , k. blum , r. t. dagnolo and j. fan , jhep * 1301 * , 149 ( 2013 ) [ arxiv:1207.4482 [ hep - ph ] ] .
l. g. almeida , e. bertuzzo , p. a. n. machado and r. z. funchal , jhep * 1211 * , 085 ( 2012 ) [ arxiv:1207.5254 [ hep - ph ] ] .
e. bertuzzo , p. a. n. machado and r. zukanovich funchal , jhep * 1302 * , 086 ( 2013 ) [ arxiv:1209.6359 [ hep - ph ] ] .
g. moreau , phys .
d * 87 * , 015027 ( 2013 ) [ arxiv:1210.3977 [ hep - ph ] ] .
m. chala , jhep * 1301 * , 122 ( 2013 ) [ arxiv:1210.6208 [ hep - ph ] ] .
s. dawson , e. furlan and i. lewis , phys .
d * 87 * , 014007 ( 2013 ) [ arxiv:1210.6663 [ hep - ph ] ] .
k. choi , s. h. i m , k. s. jeong and m. yamaguchi , arxiv:1211.0875 [ hep - ph ] .
r. huo , g. lee , a. m. thalapillil and c. e. m. wagner , arxiv:1212.0560 [ hep - ph ] .
a. djouadi , v. driesen , w. hollik and a. kraft , eur .
j. c * 1 * , 163 ( 1998 ) [ hep - ph/9701342 ] .
a. djouadi , phys .
rept . * 457 * , 1 ( 2008 ) ; b. a. kniehl , phys .
rept . * 240 * , 211 ( 1994 ) .
m. carena , i. low and c. e. m. wagner , jhep * 1208 * , 060 ( 2012 ) ; c. w. chiang and k. yagyu , arxiv:1207.1065 [ hep - ph ] .
i. picek and b. radovcic , arxiv:1210.6449 [ hep - ph ] . w. c. huang and a. urbano , arxiv:1212.1399 [ hep - ph ] .
g. n. li , g. guo , b. ren , y. j. zheng and x. g. he , arxiv:1212.5528 [ hep - ph ] . c. han , n. liu , l. wu , j. m. yang and y. zhang , arxiv:1212.6728 [ hep - ph ] .
b. batell , s. gori and l. t. wang , jhep * 1206 * , 172 ( 2012 ) [ arxiv:1112.5180 [ hep - ph ] ] ; s. baek , p. ko and e. senaha , arxiv:1209.1685 [ hep - ph ] . w. f. chang , j. n. ng and j. m. s. wu , phys . rev .
d * 86 * , 033003 ( 2012 ) [ arxiv:1206.5047 [ hep - ph ] ] .
a. melfo , m. nemevsek , f. nesti , g. senjanovic and y. zhang , phys .
d * 85 * , 055018 ( 2012 ) ; [ arxiv:1108.4416 [ hep - ph ] ] .
a. arhrib , r. benbrik , m. chabab , g. moultaka and l. rahili , jhep * 1204 * , 136 ( 2012 ) ; [ arxiv:1112.5453 [ hep - ph ] ] .
a. g. akeroyd and s. moretti , phys .
d * 86 * , 035015 ( 2012 ) ; [ arxiv:1206.0535 [ hep - ph ] ] .
e. j. chun , h. m. lee and p. sharma , jhep * 1211 * , 106 ( 2012 ) ; [ arxiv:1209.1303 [ hep - ph ] ] .
p. s. b. dev , d. k. ghosh , n. okada and i. saha , arxiv:1301.3453 [ hep - ph ] .
l. lavoura and l. -f .
li , phys .
d * 49 * , 1409 ( 1994 ) [ hep - ph/9309262 ] .
m. baak , m. goebel , j. haller , a. hoecker , d. kennedy , r. kogler , k. moenig and m. schott _ et al .
_ , eur .
j. c * 72 * , 2205 ( 2012 ) [ arxiv:1209.2716 [ hep - ph ] ] .
m. magg and c. wetterich , phys .
b * 94 * , 61 ( 1980 ) .
j. schechter and j. w. f. valle , phys .
d * 22 * , 2227 ( 1980 ) . t. p.
cheng and l. -f .
li , phys .
d * 22 * , 2860 ( 1980 ) .
g. b. gelmini and m. roncadelli , phys .
b * 99 * , 411 ( 1981 ) .
g. lazarides , q. shafi and c. wetterich , nucl .
b * 181 * , 287 ( 1981 ) .
r. n. mohapatra and g. senjanovic , phys .
d * 23 * , 165 ( 1981 ) . j. schechter and j. w. f. valle , phys . rev .
d * 25 * , 774 ( 1982 ) .
s. chatrchyan _ et al . _
[ cms collaboration ] , eur .
j. c * 72 * , 2189 ( 2012 ) [ arxiv:1207.2666 [ hep - ex ] ] .
g. aad _ et al . _ [ atlas collaboration ] , arxiv:1210.5070 [ hep - ex ] .
c. s. chen , c. q. geng , d. huang and l. h. tsai , arxiv:1212.6208 [ hep - ph ] . | we calculate the higgs decay rate of @xmath0 by including the contributions from new scalars with arbitrary quantum numbers of the weak isospin ( @xmath1 ) and hypercharge ( @xmath2 ) in the standard model .
we find that our general formula for the decay rate of @xmath0 matches with that for @xmath3 in the limit of @xmath4 , but it is different from those in the literature . to illustrate our result , by taking the current @xmath5 excess of the @xmath3 rate measured by the lhc , we examine the corresponding shift for the @xmath6 decay channel due to the new scalar .
we show that the enhancement or reduction of the @xmath0 rate only depends on the relative size of @xmath1 and the absolute value of @xmath2 .
explicitly , we predict @xmath7 by imposing the observed range of @xmath8 , which is independent of the number of multiplets and the couplings to the higgs particle as long as the scalars are heavier than 200 gev .
this result provides a clear signature for the future lhc measurements to test physics beyond the standard model . |
the study of two - dimensional doped antiferromagnets is a central subject in quantum many - body theory .
a plethora of interesting quantum ground states and quantum phase transitions appear possible , and the results have important experimental applications to the cuprate high temperature superconductors and other layered transition metal compounds .
however the zoo of possibilities contributes to the complexity of the problem , and a widely accepted quantitative theory has not yet appeared .
this paper will present a detailed quantum theory of some simpler realizations of doped antiferromagnets .
we will consider situations in which it is possible to neglect the coupling between the spin and charge degrees of freedom and consider a theory of the spin excitations alone .
more specifically , we have in mind the following physical systems , which are the focus of intense current experimental interest : ( a ) quasi - two dimensional ` spin gap ' insulators @xcite like @xmath6 , @xmath7 or @xmath8 in which a small fraction of the magnetic transition metal ions ( cu or v ) are replaced by non - magnetic ions by doping with zn or li .
these insulators have a gap to both charge and spin excitations , but the spin gap is significantly smaller than the charge gap , validating a theory of the spin excitations alone .
moreover , additional low energy spin excitations are created in the vicinity of the dopant ions while charge fluctuations are strongly suppressed everywhere .
( b ) high temperature superconductors like @xmath5 in which a small fraction of cu has been replaced by non - magnetic zn or li @xcite .
a major fraction of the spectral weight of the spin excitiations near momentum @xmath9 resides in a spin-1 ` resonance peak ' at energy @xmath10 @xcite .
we present here a theory for the changes in this peak due to the doping by non - magnetic ions .
it may appear surprising that we are able to model this phenomenon by a theory of the spin excitations alone , but we shall describe how this is possible : under suitable conditions , the coupling between the bosonic spin-1 mode responsible for the resonance peak , and the spin-1/2 fermionic , bogoliubov quasiparticles of the d - wave superconductor is weak and , in a sense we will make precise , irrelevant .
( the alternative case in which the coupling to the quasiparticles is relevant is discussed in appendix [ quasi] here we find that qualitative features of the theory and the structure of all the scaling forms remain unchanged , and there are only quantitative changes to scaling functions . )
we will obtain a universal expression for the energy scale over which the resonance peak is broadened by non - magnetic ions , with the result depending only upon known bulk parameters .
we also predict an unusual asymmetric lineshape for the broadened peak at low temperatures , and it would be interesting to test this in future experiments .
related questions have been addressed in some earlier work .
the impurity - induced damping of bulk excitations of an antiferromagnet was addressed in ref .
: however , their analysis was restricted to the _ doublet _ spin - wave excitations of an ordered nel state well away from any quantum critical point ; our focus here is on the _ triplet _ excitations in the paramagnetic phase .
a weak - coupling analysis of the effect of zn impurities on the spin fluctuation spectrum of a traditional bcs superconductor was considered in ref . .
an elementary introduction and a summary of our results , mainly for the uniform spin susceptibility , have appeared previously @xcite .
although our results apply to disparate physical systems , they are unified by their reliance on a simple , new quantum field theory .
we find it convenient to introduce this field theory at the outset , and will attempt to keep the discussion here self - contained and accessible to experimentalists .
the discussion in this introduction will be divided into four subsections : section [ host ] will review the quantum field theory for the host antiferromagnet , section [ qimp ] will introduce the new quantum field theory of the impurity problem , our main results will be described in section [ results ] , while the outline of the remainder of the paper appears in section [ outline ] .
we shall assume that the spin fluctuations in the host antiferromagnet , prior to doping by the zn or li ions , are described by the following o(3)-symmetric quantum theory @xcite of a three - component vector field @xmath11 ( @xmath12 ) @xmath13 .
\label{sbp}\end{aligned}\ ] ] here repeated indices @xmath14 are implicitly summed over , @xmath15 is the two - dimensional spatial co - ordinate , @xmath16 is imaginary time , @xmath17 and @xmath18 is the temperature .
the field @xmath19 represents the instantaneous , local orientation and magnitude of the antiferromagnetic order parameter .
so in applications to @xmath5 , @xmath20 represents the amplitude of spin fluctuations near wavevector @xmath9 .
more generally , @xmath20 can represent collinear spin fluctuations at any commensurate wavevector , apart from wavevector @xmath21 .
so states with significant ferromagnetic fluctuations are excluded , as are spiral states ( like those on the triangular lattice antiferromagnet ) which have non - collinear spin order .
incommensurate , but collinear , spin correlations can be treated by a simple extension of the theory we consider here - this is discussed briefly in appendix [ incomm ] .
the quantum dynamics of @xmath20 is realized by the second - order time derivative term in @xmath22 ; we have rescaled the @xmath20 field to make the co - efficient of this term unity .
notice that there is no first - order berry phase term , as is found in the path integral of a single spin : these are believed to efficiently average to zero on the scale of a few lattice spacings , because the orientation of the spins oscillates rapidly in the antiferromagnet . in one - dimensional antiferromagnets , this lattice scale cancellation is not quite perfect , and a topological ` @xmath23-term ' does survive @xcite ; however no such terms appear in two dimensions the fate of the berry phases in the host antiferromagnet has been discussed at length elsewhere @xcite , and will not be important for our purposes here .
the spatial propagation of the spin excitations is associated with the two spatial gradient terms , and we have allowed for two distinct spin - wave velocities , @xmath24 , @xmath25 , for propagation in the @xmath26 and @xmath27 directions .
such anisotropies are present in the coupled ` spin - ladder ' systems like @xmath6 .
however , a simple redefinition of spatial co - ordinates @xmath28 allows us to scale away the anisotropy , and we shall assume this has been done in our subsequent discussion , unless explicitly stated otherwise .
the resulting partition function has the form @xmath29 , \label{sb}\end{aligned}\ ] ] where @xmath30 we have also generalized the action to @xmath31 spatial dimensions for future convenience .
the most important property of @xmath32 is that it exhibits a phase transition as a function of the coupling @xmath33 @xcite . for @xmath33 smaller than a critical value @xmath34 , the ground state has magnetic nel order and spin rotation invariance is broken because @xmath35 . for @xmath36 , spin rotation invariance is restored , and the ground state is a quantum paramagnet with a spin gap .
the properties of this phase transition in @xmath32 are well understood , and we review a few salient facts .
the nel state of @xmath22 is characterized by two spin stiffnesses , @xmath37 , @xmath38 , which measure the energy cost to slow twists in the direction of the nel order in the 1,2 directions respectively ( more precisely , a uniform twist by an angle @xmath23 over a length @xmath39 in the @xmath40 direction costs energy @xmath41 per unit volume ; note that @xmath42 has the dimension of energy in @xmath43 ) .
these stiffnesses are related by @xmath44 for the isotropic model @xmath32 we have a single spin stiffness @xmath45 as @xmath33 approaches @xmath34 from below , the velocities @xmath24 and @xmath25 remain constant , while @xmath46 where @xmath47 is a known exponent .
the quantum paramagnet for @xmath48 is characterized by its spin gap @xmath0 , and this vanishes as @xmath33 approaches @xmath34 from above as @xmath49 as has been discussed at length elsewhere @xcite , provided @xmath50 and @xmath18 are not too large , the energy scales @xmath0 and @xmath51 , and the velocities @xmath24 , @xmath25 , are sufficient to completely characterize the quantum dynamics of a @xmath43 antiferromagnet ; there is no need to have an additional parameter determining the strength of the quartic non - linearity @xmath52 because it reaches a universal value , determined by the zero of a renormalization group beta function , near the critical point .
the present paper will establish the remarkable fact that _ no additional parameters _ are needed to characterize the spin dynamics in the presence of a dilute concentration of impurities . as we shall specify more explicitly below , we only need to know the concentration of impurities , @xmath2 , and for each impurity a spin @xmath3 , which is integer or half - odd - integer ; the value of @xmath3 can usually be determined , a priori , by simple arguments based upon gross features of the impurity configuration .
many simple , physically relevant , lattice antiferromagnets can be explicitly shown to have a quantum phase transition described by @xmath32 .
one of the most transparent is the coupled - ladder antiferromagnet , which has been much studied in recent work @xcite .
we will consider such a model in section [ ladders ] : the relationship to @xmath32 will emerge naturally , and we will also be able to relate parameters in @xmath32 to microscopic couplings .
other models include the double - layer antiferromagnet @xcite , the antiferromagnet on the 1/5-depleted square lattice @xcite , and the square lattice antiferromagnet with frustration @xcite : our results will also apply to such antiferromagnets .
the reader might wonder why we are placing an emphasis on the quantum phase transition in @xmath32 , while most experiments on zn or li doping have taken place in an antiferromagnet or superconductor with a well - defined spin gap .
the reason is that the spin gap , @xmath0 , is often significantly smaller than the microscopic exchange interactions , @xmath53 . under such circumstances ,
it is a poor approximation to model the spin gap phase in terms on tightly bound singlet pairs of electrons .
rather , the smallness of the gap indicates that there is an appreciable spin correlation length and significant resonance between different pairings of singlet bonds .
the approach we advocate to describe such a fluctuating singlet state is to find a quantum critical point to an ordered nel state somewhere in parameter space , and to expand away from it back into the spin gap phase .
it is not necessary that the quantum critical point be experimentally accessible for such an approach to be valid : all that is required is that it be near a physically accessible point .
the value of expanding away from the quantum critical point is that detailed dynamical information can be obtained in a controlled manner , dependent only upon a few known parameters .
we now introduce _ arbitrary _
localized impurities at a set of random locations @xmath54 with a density @xmath2 per unit volume .
we place essentially no restrictions on the manner in which each impurity deforms the host antiferromagnet , provided the deformations are localized , immobile , well separated from each other , and preserve global su(2 ) spin symmetry . as we argue below , each impurity will be characterized only by a single integer / half - odd - integer , @xmath55 , which measures the net imbalance of berry phase in the vicinity of the impurity ; all other characteristics of the impurity will be shown to be strongly irrelevant .
we mentioned the essentially complete cancellations of berry phases in the host antiferromagnet earlier in section [ host ] ; this cancellation can be disrupted in the presence of an impurity @xcite , and it is only this disruption which will be important in the low energy limit .
for instance , a single zn or li impurity introduces one additional spin on one sublattice than on the other , and so in a host spin-1/2 antiferromagnet such an impurity will have @xmath56 .
as another example , consider a ferromagnetic rung bond in a spin-1/2 coupled ladder antiferromagnet : the two spins on the ends of the rung will be oriented parallel to each other and so their berry phases will add rather than cancel , leading to @xmath57 .
the above criteria determining @xmath55 are qualitative , but we can make them more precise . place the host antiferromagnet in the quantum paramagnetic phase ( @xmath58 ) and measure the response to a uniform magnetic field . in the absence of impurities , the total linear susceptibility of the antiferromagnet , @xmath59 , can be written as @xmath60 , where @xmath61 is the total area of the antiferromagnet , @xmath62 is the gyromagnetic ratio , @xmath63 is the bohr magneton , and @xmath64 is the response per unit area . as @xmath65 ,
the response is exponentially small due to the presence of a spin gap , @xmath0 ; a simple argument summing over a dilute gas of thermally excited spin-1 particles shows that @xcite @xmath66 now let us add a very small concentration of impurities .
the response @xmath59 is modified to @xmath67 \ , .
\label{defchi}\ ] ] provided the impurities are dilute enough that they do not interact with each other , then @xmath68 must take the following form at low @xmath18 : @xmath69 we will take the values of @xmath55 which appear in such an observation of @xmath68 as the definitions of impurity spins @xmath55 at the site @xmath70 : they will necessarily be integers / half - odd - integers when the host antiferromagnet has a spin gap .
it is important to note that for any fixed @xmath2 , there is an exponentially small temperature below which the impurities do interact with each other , and ( [ chiimp ] ) is no longer valid : we have assumed that @xmath18 is above such a temperature in determining the @xmath55 .
however , once the @xmath55 have been obtained , our quantum theory below will also apply in the interacting impurity regime .
we now describe the modifications to the bulk action @xmath32 by the quantum impurities with spin @xmath55 at sites @xmath71 .
the divergent curie susceptibility in ( [ chiimp ] ) is a consequence of the free rotation of the impurity spin in spin space .
let us denote the instantaneous orientation of the impurity spin at @xmath70 by the unit vector @xmath72 , where @xmath73 and @xmath74 .
then the field theory of the quantum impurity problem is @xmath75 .
\label{simp}\end{aligned}\ ] ] the first term in @xmath76 is uncompensated berry phase of the impurity at site @xmath70 : the spin @xmath55 is precisely that appearing in ( [ chiimp ] ) and @xmath77 is a ` dirac monopole ' function @xcite which satisfies @xmath78 the coupling between the impurity spin and the host spin orientation is @xmath79 and this can be either positive or negative this depends upon the sublattice location of the impurity . the main purpose of the remainder of this paper is to describe the properties of @xmath80 in its different physical regimes . , as measured @xcite by nmr on @xmath81 .
the size of each arrow represents the value of the moment .
note that there is no spin directly on the zn site , but it is expected that a net moment of @xmath82 centered at it at @xmath83 . ,
width=259 ] it is worth emphasizing that @xmath76 is to be viewed as a long - wavelength theory for the impurity at site @xmath70 , and requires some interpretation when applied to lattice models .
in particular when a non - magnetic zn or li impurity replaces a spin-1/2 cu ion , there is clearly no spin degree of freedom directly at the impurity site .
instead the spin polarization is distributed in a staggered arrangement on the neighboring cu ions , so that the total moment is expected to be @xmath56 at @xmath83 ; this has been studied in elegant experiments @xcite , and we reproduce the results of one of these observations in fig . [ stagg ] .
the unit vector @xmath84 then measures the instantaneous orientation of the collective spin polarization which is centered on a non - magnetic zn site , and the berry phase in ( [ simp ] ) is the net uncompensated contribution obtained by summing over the berry phases of all the spins on the neighboring cu ions .
the spatial correlations of the spins on the cu ions are encoded , in our theory , in those of the @xmath20 field .
some crucial properties of the partition function @xmath80 in ( [ simp ] ) follow from simple power - counting arguments .
we will perform these here at tree level , and higher loop corrections do not lead to corrections which modify the main conclusions .
we focus here on the properties of a _ single _ impurity : the consequences of interactions between the impurities will be discussed later .
we examine the behavior of @xmath85 under a rescaling transformation @xmath86 and @xmath87 centered at the impurity ( the dynamic critical exponent , @xmath88 , is unity ) .
the scaling dimensions of @xmath20 and @xmath89 are fixed by demanding invariance of the time derivative terms : this leads to @xmath90 = ( d-1)/2 \quad ; \quad \text{dim}[n_{\alpha } ] = 0 .
\label{dimphin}\ ] ] these can be used to determine the dimension of @xmath79 : @xmath91 = ( 3-d)/2 .
\label{dimgamma}\ ] ] so @xmath79 is a relevant perturbation about the decoupled fixed point in @xmath43 . following results of earlier work on related models
@xcite , we find that in an expansion in @xmath92 @xmath79 reaches one of two fixed point values @xmath93 under the renormalization group ( rg ) flow ( the value of @xmath94 depends upon @xmath55 , but we will not denote this explicitly ) .
so the low energy properties of @xmath80 are actually independent of the particular bare couplings @xmath79 , and are controlled instead by @xmath94 .
the memory of the initial values of @xmath79 is retained in the sign of the fixed point value : at the fixed point we have @xmath95 where @xmath96 , and the sign of @xmath97 is the same as that of the initial sign of @xmath79 , _
i.e. _ , @xmath79 does not change sign under the rg flow .
the initial sign of @xmath79 is determined by sublattice location of the impurity in the host lattice antiferromagnet , and so both signs are equally probable .
similar arguments can be used to show that there are no other relevant couplings between the impurity and the host antiferromagnet , consistent with the required symmetries .
the simplest possible new coupling is @xmath98 which represents variation in the strength of the exchange constants in the vicinity of the impurity .
power counting shows that @xmath99 = 2-d.\ ] ] so the @xmath100 are strongly irrelevant for small @xmath101 ( this is all that is needed , as the tree level arguments are valid only for small @xmath101 ) .
we can also consider the coupling of the impurity spin to the local uniform ( ferromagnetic ) magnetization density @xmath102 .
for the action @xmath32 , the latter is given by @xcite @xmath103 and the coupling to the impurity spin takes the form @xmath104 because @xmath102 is the density of a conserved charge @xmath105 = d , \label{diml}\ ] ] and so @xmath106 = 1-d,\ ] ] and @xmath107 is also irrelevant .
similar arguments apply to all other possible couplings between the host and the impurities consistent with global symmetries .
( it is worth mentioning , parenthetically , that if we relax the constraint of global su(2 ) symmetry , and allow for spin anisotropy , then relevant couplings are possible at the impurity site .
the most important of these are a local field @xmath108 , or a single - ion anisotropy @xmath109 : the @xmath110 , @xmath111 are easily seen to be relevant with scaling dimension 1 .
now there is one relevant coupling in the bulk ( @xmath33 ) , and one or more on the impurity ( @xmath110 , @xmath111 ) , and even the single - impurity problem has the topology of a multicritical phase diagram : separate phase transitions are possible at the impurity , the bulk , or both , and the full technology of boundary critical phenomena @xcite has to be used . for @xmath112
the single - ion anisotropy is inoperative , and if there is also no local field , we have to consider exchange anisotropies in the @xmath79 , and these can also be relevant@xcite . )
we have now assembled all the ingredients necessary to establish our earlier assertion that no additional parameters are needed to describe the dynamics of the antiferromagnet in the presence of impurities .
for the case of a single impurity ( or a non - extensive density of impurities ) our arguments show that there is only one relevant parameters at the quantum critical point the ` mass ' @xmath33 ; its scale is set by the spin gap , @xmath0 , for @xmath48 and by the spin stiffness , @xmath51 , for @xmath113 .
all other couplings , whether they are bulk ( like @xmath52 ) or associated with the impurities ( like @xmath79 ) either flow to a fixed point value or are strongly irrelevant .
we do need a parameter to set the relative scales of space and time , and this is the velocity @xmath1 .
our arguments also allow us to obtain important information for the case where the density of impurities , @xmath2 , is finite .
now @xmath2 itself is a relevant perturbation of the critical point ; however , we can use the fact that there is no new energy scale which characterizes the coupling of a single impurity to the bulk antiferromagnet to conclude that the value of @xmath2 alone is a complete measure of the strength of this perturbation .
let us restate the above important conclusion a little more explicitly .
assume we are doping the spin gap paramagnet with @xmath48 and a spin gap , @xmath0 ( very similar arguments apply also for @xmath114 with @xmath51 replacing @xmath0 ) .
for the case of a single impurity , the dynamics of the host antiferromagnet in the vicinity of the impurity is determined entirely by @xmath0 and @xmath1 with no free parameters .
a finite density of impurities , @xmath2 , is a relevant perturbation , but its impact is determined entirely by the single energy scale , @xmath115 , that can be constructed out of the above parameters , and that is also linear in @xmath2 : @xmath116 in particular , all dynamic and thermodynamic properties of the host antiferromagnet are modified only by _ a universal dependence on the ratio @xmath117 _ ; all corrections to these ( including those dependent on the magnitude of the bare coupling between the impurity and host antiferromagnet ) are suppressed by factors of @xmath118 , and only the latter are non - universal .
the leading universal modification depends also on the statistics of the values of @xmath55 and @xmath97 : in most physical applications we simply have @xmath119 independent of @xmath70 , while the @xmath97 are independently distributed and take the values @xmath120 with equal probability
. the above universal dependence on @xmath117 does not preclude the possibility that there could be a phase transition in the ground state as a function of @xmath117 : a magnetically ordered state could appear for infinitesimal doping , _ i.e. _ , at @xmath121 , or at some finite critical value of @xmath117 these two possible phase diagrams are shown in fig [ fig0 ] . for a dilute concentration of impurities @xmath2 . for @xmath48 ,
all low energy properties depend only upon the dimensionless ratio @xmath117 , where @xmath0 is the spin gap of the undoped host antiferromagnet , and @xmath115 is defined in ( [ defgamma ] ) . in ( a )
there is a phase transition to an ordered state at @xmath121 , and a quantum paramagnet is present only for @xmath122 and @xmath58 ; in ( b ) the transition to the ordered state occurs at some finite universal value of @xmath117 , and so by ( [ deltacrit],[defgamma ] ) the phase boundary obeys @xmath123 . for @xmath114
the properties are universal functions of the analogous dimensionless ratio @xmath124 ( where @xmath51 is the stiffness of the host antiferromagnet ) ; we have not shown a third possible phase diagram in which the phase boundary is present for @xmath114 with @xmath125 .
right at @xmath126 , the spacetime correlators are universal function of the dimensionless combinations @xmath127 and @xmath128 with no arbitrary scale factors . for any lattice antiferromagnet , the long - range order will disappear at sufficiently large @xmath2 when the concentration of impurities is near the percolation threshold this phase boundary has not been shown in the figure , as it is not a property of the continuum theory @xmath129 .
this latter transition was considered by chen and castro neto @xcite , along with the behavior as a function of @xmath2 for @xmath114 ; however , they did not include the berry phases , and these are expected to be important in the critical region @xcite .
, width=307 ] determination of the phase diagram requires solution of the problem of interacting quantum impurities ; in the paramagnetic phase ( @xmath58 ) , an effective action for interactions can be obtained by integrating out the @xmath20 fields from @xmath85 .
apart from the same berry phase terms in @xmath76 , the resulting action contains the term @xmath130 where the universal interaction @xmath131 is proportional to the @xmath20 propagator and depends only on the fixed point value @xmath94 ( which is an implicit function of @xmath55)@xmath131 decays exponentially for large @xmath132 or @xmath133 ; there are additional multi - spin interactions between the impurities which have not been displayed .
a notable property@xcite of ( [ mattis ] ) is that the signs of the interactions have the disorder of the `` mattis model''@xcite , _ i.e. _ , there is no frustration , and the product of interactions around any closed loop has a ferromagnetic sign .
however , unlike the classical spin glass , the fluctuating signs of the exchange interaction can not be gauged away in this quantum mattis model , as the transformation @xmath134 is not a symmetry of the berry phase term ( equivalently , it does not preserve the spin commutation relations ) .
nevertheless , the absence of frustration in such a model has been used by nagaosa _
_ @xcite and imada and iino @xcite to argue in favor of the phase diagram in fig [ fig0]a .
we reiterate that , irrespective of the correct phase diagram , all low energy properties depend only the value of the dimensionless ratio @xmath117 , which is a measure of the strength of the relevant perturbation of a finite concentration of impurities .
it is remarkable that such a measure is independent of the magnitude of the exchange coupling between the impurities and the antiferromagnet .
we also note that some of the earlier scaling arguments of imada and iino @xcite are contained in such an assertion it implies their identification of the crossover exponent of the impurity density as @xmath135 .
while it is reasonably transparent that the action @xmath136 applies to the zn doping of insulating spin gap compounds like @xmath6 , the applicability to zn doping of a good d - wave superconductor like @xmath5 has not yet been established .
actually the similarity of zn doping in a cuprate superconductor to that of spin gap compounds was already noted by the experimentalists in ref here we will sharpen their proposal .
there is a great deal of convincing experimental evidence that each zn impurity has an antiferromagnetic polarization of cu ions in its vicinity , and the net moment of this polarization is expected to be @xmath112 at @xmath83 ; at moderate temperature this moment precesses freely giving rise to a curie susceptibility , while there is evidence for spin freezing at very low temperatures , presumably from inter - impurity interactions ( see fig .
[ stagg ] , refs . and references therein ) .
the berry phase term in @xmath76 then encodes the dynamics of the collective precession of the spins on the cu ions near the zn impurity at site @xmath70 ( recall that there is no spin on the zn ion itself ) .
the impurity spin will couple to spin excitations in the host superconductor .
the most important of these is the spin-1 collective mode which gives rise to the ` resonance peak ' @xcite of the cuprate superconductors . in our approach
@xcite this collective mode is represented by the oscillation of the field @xmath20 about @xmath137 controlled by the action @xmath32 .
so our picture of the resonance peak is similar in spirit to computations @xcite which identify it with a @xmath138 bound state in the particle - hole channel in rpa - like theories of the dynamic spin response of a d - wave superconductor .
however , such theories use a weak - coupling bcs picture of the electron spin correlations and also neglect the non - linear self - interactions of the collective mode .
in contrast , in our approach the underlying spin correlations are characterized by @xmath32 , which assumes that the superconductor is close to an insulating state with magnetic long - range order ( and possibly also charge order ) : this has the advantage of allowing a systematic treatment of the strongly relevant quartic non - linearity in the @xmath20 field , and these non - linear effects are crucial to the universal nature of our results .
we also mention the approach of zhang @xcite which assumes that 3-component @xmath20 is part of 5-component `` superspin''unlike zhang , our theory does not appeal to any higher ( approximate ) symmetry group in the superspin action , beyond that expected from spin rotation invariance .
spin is also carried by the fermionic , spin-1/2 bogoliubov quasiparticles in a d - wave superconductor , and these will couple to the impurity spin .
these quasiparticles have vanishing energy at four points in the brillouin zone - @xmath139 .
as is well known , a gradient expansion the of low energy fermionic excitations in the vicinity of these points yields an action that can be expressed in terms of 4 species of anisotropic dirac fermions .
we do not wish to enter into the specific details of this action here , as only some gross features will be adequate for our basic argument .
let us represent these fermions schematically by the nambu spinors @xmath140 ; then the action has the form @xmath141 . \label{spsi}\ ] ] we have omitted all coupling constants and matrices in the nambu and spin spaces , as we do not need to know their structure here .
the important point is that this action implies the dimension @xmath142 = 1 \label{dimpsi}\ ] ] under scaling transformations .
now let us couple @xmath140 to the degrees of freedom in @xmath143
. there will be a bulk cubic coupling of the form @xmath144 or @xmath145 only if it is permitted by the momentum conservation in the host antiferromagnet .
as the @xmath20 represent spin fluctuations at the wavevector @xmath146 , such a cubic coupling is permitted only if @xmath146 equals the sum or difference of two of @xmath139 ( _ i.e. _ if @xmath147 or not ) . for simplicity , in the body of the paper
, we will assume in this is not the case , _
i.e. _ we assume @xmath148 .
indeed , in @xmath5 , @xmath149 and it appears that @xmath150 , and so a cubic term is forbidden .
however , @xmath151 is not too far from @xmath152 , and so the effect of a cubic term may be manifest at higher energies . in appendix [ quasi ]
we consider the limiting case @xmath153 , which permits a cubic term in the host d - wave superconductor at the lowest energies .
a renormalization group analysis of such a host theory was presented recently by balents _ _
et al.__@xcite , and appendix [ quasi ] combines their results with the quantum impurity theory of the present paper : we find that the scaling structure is essentially identical to that in the body of the paper , and there are only simple quantitative modifications to the fixed - point values of the couplings . for now , we assume that momentum conservation prohibits a bulk coupling between the bosonic ( @xmath20 ) and fermionic ( @xmath140 ) carriers of spin @xcite .
however these degrees of freedom can still interact via their separate couplings to the impurity spins .
the coupling of the superconducting quasiparticles to the impurities at sites @xmath71 is of the form @xmath154 . \label{spsiimp}\end{aligned}\ ] ] again , we have omitted matrices in spin and nambu space ; the @xmath155 , @xmath156 are exchange couplings between the quasiparticles and the cu spins in the vicinity of the zn impurity , and the @xmath157 , @xmath158 represent potential scattering terms from the non - magnetic zn ion itself .
models closely related to @xmath159 have been the subject of a large number of studies in the past few years @xcite and a number of interesting results have been obtained .
however , all of these earlier works have not included a coupling between the impurity spin and a collective , spin-1 , bosonic mode like @xmath20 .
one of the central assertions of this paper is that ( provided the spin gap , @xmath0 , is not too large ) such a coupling ( as in @xmath76 ) is of paramount importance for the low energy spin dynamics , while the coupling to the superconducting quasiparticles ( as in @xmath160 ) has weaker effects .
( however , when one is explicitly interested in quasiparticle properties , as in tunneling experiments , it is certainly necessary to include @xmath161 ; in stm experiments with atomic resolution @xcite , the quasiparticle tunneling can be observed directly at the zn site , and there the potential scattering terms @xmath157,@xmath158 will be especially important as there is no magnetic moment ( see fig [ stagg ] ) .
as we noted below ( [ diracmono ] ) , we are considering a long - wavelength theory of the spin dynamics , and careful interpretation is required for lattice scale effects . )
the argument behind our assertion is quite simple . using ( [ dimpsi ] )
we see that @xmath162 = -1 . \label{ingersent}\ ] ] therefore , while the couplings , @xmath79 , between the impurities and the @xmath20 had a positive scaling dimension in ( [ dimgamma ] ) , those between the fermionic quasiparticles and the impurities have a negative dimension and are strongly irrelevant .
we are therefore justified in neglecting the fermionic quasiparticles in our discussion of the impurity spin dynamics .
the above conclusion is also supported by studies @xcite of models related to @xmath159 in the context of `` kondo problems with a pseudogap in the fermionic density of the states '' : for the case where the fermionic density of states vanishes linearly at the fermi level ( as is the case for @xmath163 ) , there is no kondo screening for small and moderate @xmath155 , @xmath156 values , and the impurity spin is essentially static .
of course , the present renormalization group argument can not rule out the possibility of new physics , associated with the fermionic kondo effect , appearing at very large @xmath155 , @xmath156 : we shall assume that the bare couplings are in a regime such that this has not happened , and the scaling dimensions in ( [ ingersent ] ) continue to apply .
we will state our main results for the case of a single impurity in section [ single ] , and those for a finite density in section [ many ] .
we first consider a single impurity at the origin of co - ordinates @xmath164 with @xmath165 an important measurable response function is the impurity susceptibility defined in ( [ defchi ] ) .
its properties follow simply from a naive application of the scaling ideas we have presented above , and are summarized in fig [ fig1 ] . in @xmath43 with a single impurity at the origin of co - ordinates with @xmath119 .
the dashed lines represent crossover boundaries .
the crossover for @xmath48 occurs when @xmath166 , and is between the high - temperature ` quantum critical ' regime and the low - temperature quantum paramagnet described by a dilute gas of spin-1 @xmath20 quanta .
similarly , the crossover for @xmath113 occurs when @xmath167 , and the low - temperature regime is a locally ordered nel state whose long range order is disrupted by long - wavelength , classical , spin - wave fluctuations .
the constants @xmath168 are universal numbers , insensitive to microscopic details .
the constant @xmath169 was introduced in ref , and is determined solely by the bulk theory @xmath32 ; quantum monte carlo results were obtained in ref .
the constants @xmath170 and @xmath171 are determined by @xmath172 , and depend only on the integer / half - odd - integer valued @xmath3 .
the constant @xmath170 defines the effective spin at the quantum - critical point by @xmath173 , and @xmath174 is neither an integer nor a half - odd - integer . on the ordered side , @xmath114 , in the absence of even an infinitesimal spin anisotropy , the measured spin susceptibility will take the rotationally averaged value @xcite @xmath175 ; we have no reason to expect non - monotonic behavior of the susceptibility as a function of @xmath33 , and so this result suggests the bounds @xmath176 in @xmath43 .
, width=307 ] knowledgeable readers may be surprised by such an assertion . in the intensively studied multichannel kondo quantum impurity problem
@xcite naive scaling actually fails : even though the low energy physics is controlled by a finite coupling quantum critical point , a ` compensation ' effect @xcite causes most thermodynamic response functions to vanish in the naive scaling limit , and the leading low temperature behavior is exposed only upon considering corrections to scaling @xcite .
fortunately , our problem is simpler the bosonic @xmath20 excitations which ` screen ' the impurity are themselves controlled by a non - trivial interacting quantum field theory @xmath177 , and it is not possible to ` gauge away ' the effect of an external magnetic field on them .
armed with this reassuring knowledge , we need only know that @xmath68 has the dimensions of inverse energy to deduce its critical properties . for @xmath48
we of course have the curie response in ( [ chiimp ] ) , which is , not coincidentally , also consistent with the scaling requirements . at the critical coupling @xmath126 , we continue to have a curie - like response ( because now @xmath178 is the only available energy scale ) but can no - longer require that the effective moment be quantized : @xmath179 the universal number @xmath170 is almost certainly irrational , and we will estimate its value in sections [ rg1 ] and [ ncasusc ] ; our most reliable estimate for @xmath82 and @xmath43 is in ( [ valc1 ] ) , with @xmath180 .
turning finally to @xmath114 , the @xmath181 response is now anisotropic because of the nel order in the ground state .
we consider here only the response transverse to the nel order , @xmath182 , at @xmath83 ; further details are in section [ sec : chiperp ] .
the same scaling arguments now imply @xmath183 again @xmath171 is an irrational universal number whose value will be estimated later .
having described the response to a uniform , global magnetic field , let us consider the responses to local probes in the vicinity of the impurity .
such results will apply to nmr and tunneling experiments . in considering such response function
it is useful to use the language of boundary conformal field theory @xcite : we are considering here a bulk @xmath184-dimensional conformal field theory , @xmath32 at @xmath126 , which contains a one - dimensional ` boundary ' degree of freedom at @xmath164 .
correlation functions near the boundary will be controlled by _
boundary scaling dimensions _ which are distinct from the scaling dimensions of the bulk theory we have considered so far .
the most important of these controls the long - time decay of the impurity spin : @xmath185 ( here , and in the remainder of the paper we are indulging in a slight abuse of terminology by identifying these as the correlations of the `` impurity spin '' ; as in fig [ stagg ] it is often the case that there is no spin on the impurity site in such cases we are referring to spin correlations at sites very close to the impurity . )
the quantity @xmath186 is an anomalous boundary exponent : we will compute its value in the expansion in @xmath101 in section [ rg ] .
clearly , ( [ nn ] ) implies the scaling dimension @xmath187 = \eta^{\prime}/2 , \label{dimn}\ ] ] which corrects the tree - level result in ( [ dimphin ] ) .
( this is a good point to mention that loop corrections also modify the scaling dimension of the bulk field @xmath20 in ( [ dimphin ] ) to @xmath90 = ( d-1+\eta)/2 , \label{dimphi}\ ] ] where @xmath188 ( like @xmath47 ) is a known exponent which is a property of @xmath32 alone @xcite . ) for @xmath189 , there is a finite remnant moment in the boundary spin correlations . for @xmath113 ,
this is simply a consequence of the bulk nel order also breaking the spin rotation symmetry on the boundary too .
however , for @xmath48 , this is a somewhat more non - trivial quantum effect : the presence of the spin gap in the bulk , means that the boundary excitations , associated with the berry phase term in @xmath76 are confined to the impurity and maintain a permanent static moment .
so we may generalize ( [ nn ] ) to @xmath190 the impurity moment , @xmath191 , ( the remarks below ( [ nn ] ) on abuse of terminology apply to the `` impurity moment '' too ) behaves like @xmath192 a consequence of the scaling dimension ( [ dimn ] ) .
we have computed a large number of correlation functions describing the spatial and temporal evolution of the spin correlations in the vicinity of the impurity .
we will leave a detailed discussion of these to the body of the paper , but note here some simple arguments which allow deduction of important qualitative features .
the experimental probes mentioned earlier can follow the spatial evolution of the correlators of the bulk antiferromagnetic order parameter field @xmath20 , and also that of the uniform ( ferromagnetic ) magnetization density @xmath102 . as the spatial arguments of @xmath19 and @xmath193 approach the impurity , their critical correlations must mutate onto those of the boundary degrees of freedom .
this transformation is encoded in the statements of the operator product expansion @xmath194 where the powers of @xmath195 follow immediately from the scaling dimensions ( [ dimn],[dimphi],[diml ] ) .
the results ( [ nn ] ) and ( [ ope ] ) can be combined with scaling arguments to determine the important qualitative features of the spatial , temporal , and temperature dependencies of most observables in the vicinity of the impurity : details appear in sections [ rg1 ] and [ rg2 ] .
the problem of a finite density of impurities is one of considerable complexity , and we will only address a particular aspect for which we have new , significant , and experimentally testable predictions .
we will not settle the issue of whether fig [ fig0]a or b , or some other possibility , is the correct phase diagram .
we will be interested in dynamical properties of the region , @xmath48 , as characterized by the susceptibility at the antiferromagnetic wavevector @xmath196 where @xmath197 is the volume of the system , and @xmath198 is a matsubara imaginary frequency . in the absence of impurities in an insulating antiferromagnet , this susceptibility has a pole at the spin gap energy @xcite ( at @xmath83 ) : @xmath199 where @xmath200 is now a real frequency , @xmath201 represents a positive infinitesimal , and @xmath61 is a residue determining the spectral weight of the spin-1 collective mode . the form in ( [ pole ] ) is valid only in the vicinity of the pole , and quite complicated structures appear well away from the pole frequency .
such a pole will also be present in a d - wave superconductor if momentum conservation prohibits decay of @xmath20 into low - energy fermionic @xmath140 quasiparitcles .
the case for which such a decay is allowed is considered in appendix [ quasi ] : we show there that ( [ pole ] ) is replaced by the more general scaling form @xmath202 where @xmath203 is a universal scaling function .
the form ( [ pole2 ] ) predicts that in such d - wave superconductors the pole in ( [ pole ] ) will be universally broadened on an energy scale @xmath0 . such a pole ( or broadened pole )
has immediate experimental consequences : it leads to a ` resonance peak ' in the neutron scattering cross - section , as is seen in @xmath5 @xcite . at current experimental resolution ,
no intrinsic broadening has been observed at low temperatures , and so it is reasonable in a first theory to work with the sharp pole in ( [ pole ] ) , as we do in the body of this paper .
our primary interest will be in the fate of the pole ( or of ( [ pole2 ] ) ) in the presence of a finite density of impurities , @xmath2 .
this is a property at the energy scale of order @xmath0 , and useful results can be obtained without resolving the very low energy properties which distinguish the phase diagrams of fig [ fig0 ] . at small @xmath2 , the typical interaction @xmath131 in ( [ mattis ] ) is exponentially small in @xmath204 , and we will neglect such effects . instead , at the energy scales we are interested in
, we use mean - field approach to account for a finite @xmath2 , but include the full dynamics of the interaction between the @xmath20 and a single impurity : the finite @xmath2 will only modify the density of states of the @xmath20 excitations , and this will be determined self - consistently .
let us first review the exact scaling arguments .
the expression ( [ defchiq ] ) defines a bulk observable which can not acquire any of the anomalous boundary dimensions .
consequently its deformation due to the impurities is fully determined only by the energy scale @xmath115 defined in ( [ defgamma ] ) , and we have the scaling prediction that ( [ pole ] ) is modified to @xmath205 where @xmath206 is a fully universal function of its arguments ( we have restricted attention to @xmath83 here for @xmath207 there is universal dependence on an additional argument , @xmath208 ) . for the case
where there is no bulk coupling to fermionic quasiparticles at the lowest energies , we have @xmath209 , and so ( [ polescale ] ) reduces to ( [ pole ] ) for zero impurity density , @xmath210 ; with a bulk coupling of @xmath20 to fermionic quasiparticles , as in appendix [ quasi ] , ( [ polescale ] ) reduces to ( [ pole2 ] ) at @xmath211 . the scaling form ( [ polescale ] ) applies to the dynamic susceptibility for both possibilities of the phase diagram in fig [ fig0 ] ; if fig [ fig0]b is the correct phase diagram , then the scaling function @xmath206 will have a singularity at the critical value of @xmath117 .
notice that the impurities induce broadening at an energy scale @xmath212 . on the other hand , intrinsic broadening
is suppressed by momentum conservation ; in the special case that @xmath213 , intrinsic broadening exists at the lowest energies , and then it is at most of order @xmath0 .
so for small @xmath0 , the impurity broadening dominates in both cases . in experimental comparisons it may be useful to use fits with a linewidth written as a sum of intrinsic and extrinsic contributions
. further discussion of these issues is in appendix [ quasi ] .
we will obtain results for the form of @xmath206 for non - zero @xmath117 in section [ sec : ncamany ] in a self - consistent non - crossing approximation , for the case where the coupling to fermionic quasiparticles is irrelevant . as we mentioned earlier
, one of our important results will be that the @xmath214 lineshape has an asymmetric shape , with a tail at high frequencies .
this is a significant prediction of our theory , which should be testable in future experiments .
the experiment of ref . has @xmath56 , @xmath215 mev , @xmath216 and we display our predicted lineshape for this value of @xmath117 in fig [ fig2 ] ( results for other values of @xmath117 appear in section [ sec : ncamany ] ) .
the half - width of the line is approximately @xmath115 , and this is in excellent accord with the measured linewidth of 4.25 mev . the sharp lower threshold in the spectral weight
is an artifact of the non - crossing approximation : consideration of griffths effects ( which , in principle , are contained in the exact scaling function in ( [ polescale ] ) ) , or the contribution of the `` irrelevant '' fermionic excitations ( which are not part of the scaling limit ( [ polescale ] ) , will always lead to a small rounding of the lower edge . as a function of @xmath217 for @xmath43 , @xmath218 , and @xmath216 , the values corresponding to the experiment of ref . .
this result has been obtained in a self - consistent ` non - crossing approximation ' analysis of @xmath129 described in section [ sec : ncamany ] .
, width=326 ] we now sketch the contents of the remainder of the paper .
readers not interested in details of the derivation of the results may wish to skip ahead to section [ conc ] .
section [ sec : eps ] will obtain renormalization group results in an expansion in @xmath219 for the single impurity problem .
similar physical results will be obtained in section [ sec : ncasingle ] , but in a different large-@xmath220 or ` non - crossing ' approximation ( nca ) .
the advantage of the latter approach is that it can be readily extended to obtain a self - consistent theory of the magnon lineshape broadening ; this shall be presented in section [ sec : ncamany ] .
we review our main results and discuss some further experimental issues in section [ conc ] .
the appendices contain extensions and further calculational details ; in particular , appendix [ quasi ] considers the case where momentum conservation allows a bulk coupling to the fermionic quasiparticles of a d - wave superconductor at the lowest energies , while appendix [ incomm ] generalizes our results to antiferromagnets and d - wave superconductors with incommensurate spin correlations _ i.e. _ @xmath221 .
the central ingredient behind the universal structure of all of our results is the analysis of the renormalization group equations presented in section [ rg ] .
this is performed order - by - order in an expansion in @xmath219 .
the results will be used to obtain expressions for a number of physical observables in the subsequent subsections : we will consider the paramagnetic region , @xmath223 , in section [ rg1 ] , and the magnetically ordered phase , @xmath113 , in section [ rg2 ] .
all of the computations in this section will consider only the single impurity problem .
we will therefore use the notation in ( [ single1 ] ) and also denote @xmath224 also in the remainder of this paper , we will use units of time , temperature , and length in which @xmath225 we will follow the orthodox field theoretic approach @xcite in obtaining the renormalization group equations of @xmath85 : generate perturbative expansions for various observables correlators , multiply them with renormalization factors to cancel poles in @xmath101 , and finally use the independence of the bare theory on the renormalization scale to obtain the scaling equations .
this approach is rather abstract , but has the advantage of allowing explicit calculations to two - loop order and establishing important scaling relations to all orders . a more physically transparent ` momentum shell
' formalism can be used to obtain equivalent results at one loop , but several key properties are not delineated at this order .
it is clearly advantageous to have a diagrammatic method which allows expansion of the correlators in powers of the couplings @xmath52 and @xmath226 .
while the standard time - ordered perturbation theory can be used for @xmath52 , the non - linear berry phase term in @xmath76 ( and the associated unit length constraint on @xmath89 ) makes the expansion in @xmath226 more intricate .
building on earlier work in the context of the kondo problem@xcite , we have developed a diagrammatic representation of the perturbative expansion in @xmath226this is described in appendix [ diag ] .
the representation works to all orders in @xmath226 , but does not have the benefit of a dyson theorem or the cancellation of disconnected diagrams ; nevertheless , all terms at a given order in perturbation theory can be rapidly written down .
for instance , the diagrams shown in fig [ diag1 ] lead to the following lowest order representation of the two - point correlator of @xmath89 in the paramagnetic phase : ) in the diagrammatic approach of appendix [ diag ] .
the open circles represent the external sources , the dashed lines are @xmath20 propagators , and the filled circles contribute factors of @xmath226 .
the impurity spin is represented by the single full line along which imaginary time runs periodically from 0 to @xmath227 .
the correlator ( [ eps1 ] ) is the ratio of the diagrams in ( a ) to those in ( b ) , with the latter representing the normalizing partition function @xmath80 .
, width=307 ] @xmath228 \label{eps1}\end{aligned}\ ] ] where @xmath229 and @xmath230 is the two - point @xmath20 propagator at @xmath164 @xmath231 here the 0 subscript in the correlator indicates that it is evaluated to zeroth order in @xmath52 .
however , we have included hartree - fock renormalizations in the determination of the ` mass ' @xmath232 ; these have been computed in earlier work in the @xmath101 expansion @xcite , and we quote some limiting cases : @xmath233 recall that @xmath0 was defined earlier as the exact @xmath83 spin gap of the host antiferromagnet .
the crossover function between the two limiting results in ( [ eps3 ] ) is also known for small @xmath101 , but we refer the reader to the original paper @xcite for explicit details . the @xmath83 limit of ( [ eps1 ] ) must be taken with some care : the function @xmath230 in ( [ eps2 ] ) is periodic as a function of @xmath16 with period @xmath227 , and so any significant contributions for small @xmath234 have periodic images in the region with small @xmath235 . indeed , at @xmath83 , and at the critical coupling @xmath126 ( where @xmath236 ) , ( [ eps1 ] ) reduces to @xmath237 , \label{eps4}\end{aligned}\ ] ] where @xmath238 with @xmath239 simple power counting of the integrals in ( [ eps4 ] ) shows that they lead to poles in @xmath101this is as expected from the tree - level scaling dimension of @xmath226 in ( [ dimgamma ] ) .
in the field theoretic rg these poles have to be cancelled by appropriate renormalization factors which we will describe shortly . evaluating the integrals in ( [ eps4 ] ) , and also those in the two - loop corrections to ( [ eps4 ] ) described in appendix [ diag ]
, we obtain @xmath240 .
\label{eps7}\end{aligned}\ ] ] we have retained only poles in @xmath101 in the co - efficient of the @xmath241 term , as that is all that shall be necessary for our subsequent analysis .
we now describe the structure of the renormalization constants that are needed to cancel the poles in ( [ eps7 ] ) and in other observable correlation functions .
first , the renormalization factors of the host antiferromagnet can only depend upon the bulk theory @xmath32 as a single impurity can not make a thermodynamically significant contributions .
these are well known and described in numerous text books and review articles ; we will use here the conventions of brezin _ _ et al.__@xcite they define the renormalized field @xmath242 and the dimensionless coupling constant , @xmath62 by @xmath243 where @xmath244 is a renormalization momentum scale , @xmath80 is the wave - function renormalization factor of the field @xmath20 , @xmath245 is a coupling constant renormalization , and @xmath246 the explicit expressions for @xmath80 and @xmath245 , obtained by brezin _
@xcite , to the order we shall need them are @xmath247 now let us turn to the boundary renormalization factors associated with the presence of the impurity spin . as in the bulk , we have wavefunction ( @xmath248 ) and coupling constant ( @xmath249 ) renormalization factors : @xmath250 where @xmath251 is the renormalized , dimensionless boundary coupling . in both ( [ eps8 ] ) and ( [ eps11 ] ) we have inserted judicious factors of the wavefunction renormalizations in the redefinitions of the coupling constants @xmath52 and @xmath226these are determined simply by the powers of the fields multiplying the couplings in the action .
the action has a term @xmath252 , and so the renormalization of @xmath226 picks up one power each of the wavefunction renormalizations of @xmath20 and @xmath89 ; similar considerations hold for @xmath52 .
the expression ( [ eps11 ] ) contains two boundary renormalization factors , but so far we have evaluated only one correlation function , ( [ eps7 ] ) . to obtain an independent determination of @xmath249 we consider the correlation function @xmath253 which will involve the coupling @xmath226 at leading order , as shown in fig [ diag4]a . ) in the diagrammatic approach of appendix [ diag ] ; the diagram in ( a ) is the zeroth order result , while ( b ) contains diagrams of order @xmath254 .
the wavy line is the interaction @xmath52 and other conventions are as described in fig [ diag1 ] .
, width=345 ] this leading term evaluates at @xmath83 and @xmath126 to @xmath255 let us first consider corrections to ( [ eps13p ] ) from interactions involving only the boundary coupling @xmath226 .
it is easy to see that the diagrams for these have interactions along the impurity spin loop that are in one - to - one correspondence with those appearing in the evaluation of the two - point @xmath89 correlator ; these are shown in figs [ diag1 ] , [ diag2 ] , [ diag3 ] and lead to ( [ eps7 ] ) .
however , such diagrams are obviously associated with the wavefunction renormalization , @xmath248 , of @xmath89 .
so there are no contributions to the coupling constant renormalization , @xmath249 , from the boundary interactions alone ; this leads to the important result @xmath256 next , we consider interference between the bulk interaction , @xmath52 , and the boundary interaction , @xmath226 .
now there are contributions which are specifically associated with the renormalization of @xmath226 and these are shown in fig [ diag4]b . evaluating the graphs of fig [ diag4]b at @xmath83 and @xmath257 as described in appendix [ diag ]
, we obtain the following contribution to the correlator in ( [ eps12p ] ) @xmath258 after determining , by standard methods , the residue of the pole in @xmath101 associated with the integral in ( [ eps14 ] ) , we can write down the coupling constant renormalization @xmath259}{6 \epsilon } g \gamma^2 + \cdots . \label{eps15}\ ] ] the above result for @xmath260 is smaller by a factor of 3 from that quoted earlier by us @xcite , and this corrects a numerical error in our computation .
this correction also leads to corresponding minor changes in ( [ eps21 ] ) , ( [ eps22 ] ) , ( [ eps24 ] ) and ( [ n22 ] ) below .
we now have all the ingredients to determine the final renormalization factor @xmath248 .
using the definitions ( [ eps11 ] ) in ( [ eps7 ] ) , along with the result ( [ eps15 ] ) , and demanding that all poles in @xmath101 cancel , we obtain @xmath261 now the rg beta - functions can be obtained by standard manipulations @xcite .
for the bulk coupling @xmath62 we obtain from the definition ( [ eps8 ] ) and the values ( [ eps10 ] ) , the known result @xmath262 similarly , we can obtain the beta - function for the boundary coupling @xmath251 defined as @xmath263 taking the @xmath244 derivative of the second equation in ( [ eps11 ] ) at fixed bare couplings @xmath52 , @xmath226 , we get @xmath264 solving ( [ eps20 ] ) using ( [ eps10],[eps15],[eps17],[eps18 ] ) we obtain to the order we are working @xmath265 ; \label{eps21}\ ] ] the first two terms in this beta - function agree with the earlier results of smith and si @xcite and sengupta @xcite
. the beta - functions ( [ eps18 ] ) and ( [ eps21 ] ) have infra - red stable fixed points at @xmath266 \right ) \label{eps22}\end{aligned}\ ] ] these fixed point values control all of the universal physical quantities computed in this paper .
in particular , we can now immediately obtain the exponents @xmath188 and @xmath186 .
the value of @xmath188 is , of course , known previously , and is given by @xmath267 for the new boundary anomalous dimension @xmath186 we have @xmath268 \right ) .
\label{eps24}\end{aligned}\ ] ] this exponent @xmath186 controls the decay of the two - point @xmath89 correlations at the @xmath83 critical point at @xmath126 , as in ( [ nn ] ) . for @xmath58 in the @xmath83 quantum paramagnet , @xmath186 determines the impurity static moment , as was asserted in ( [ defm ] ) and ( [ defm1 ] ) .
these facts are demonstrated explicitly in perturbation theory in appendix [ mimp ] .
we close this section by mentioning an interesting bi - product of our analysis : we will obtain an exact exponent for some simpler models considered earlier in the literature . consider the fixed point of the beta - functions above with @xmath269 and @xmath270 , so that the bulk fluctuations of the @xmath20 are described by a gaussian theory .
such a fixed point is unstable in the infra - red to the fixed point we have already considered , and so does not generically describe impurities in any antiferromagnet .
nevertheless , it is an instructive model to study , and closely related models appear in mean - field theories of quantum spin glasses @xcite .
the fixed point value of @xmath271 can of course be easily obtained from ( [ eps21 ] ) to second order in @xmath101this value will have corrections at all orders in @xmath101 , and an exact determination does not seem to be possible .
even so , we can obtain an exact result for @xmath186 .
this is a consequence of the identity ( [ eps16 ] ) . using this result , and the fact that all @xmath62 derivatives can be neglected at @xmath269 , the expressions ( [ eps20 ] ) and ( [ eps24 ] ) simplify to @xmath272 it now follows immediately from the fact that @xmath273 , that @xmath274 exactly at the @xmath269 , @xmath275 fixed point .
the results in this section will be limited to @xmath223 , _
i.e. _ , the right half of fig [ fig1 ] . we will be interested in numerous different linear response functions of @xmath85 . to discuss these in some generality , we extend the actions by coupling them to a number of different external fields : in @xmath32 we make the transformation @xmath276 while we add to @xmath76 the coupling @xmath277 the field @xmath278 is an external magnetic field which varies as a function of @xmath195 only on a scale much larger than the lattice spacing of the underlying antiferromagnet . in contrast
, @xmath279 is a _ staggered _ magnetic field which couples to the antiferromagnetic order parameter
so it oscillates rapidly on the lattice scale , but has an envelope which is slowly varying ; such fields can not be imposed directly in the laboratory , but associated response function can be measured in nmr and neutron scattering measurements . finally @xmath280 is the magnetic field at the location of the impurity .
so a space - independent , uniform magnetic field @xmath281 applied to the antiferromagnet corresponds to @xmath282 , @xmath283 , and @xmath284 .
we have also taken all fields to be time independent for simplicity - it is not difficult to extend our approach to dynamic response functions with time - dependent fields .
we can now consider a total of six different response function to these external fields .
we tabulate below results for these response functions obtained in bare perturbation theory , to lowest non - trivial order .
the computation is performed by the methods of appendix [ diag ] and is quite straightforward
we do not explicitly show the feynman diagrams associated with these results .
the results will subsequently be interpreted using the rg formulated in section [ rg ] .
@xmath285 \nonumber \\ \chi_{\text{u},\text{s } } ( x , x^{\prime } ) & \equiv & \frac{t}{3 } \frac{\delta^2 \ln z}{\delta h_{\text{u}\alpha } ( x ) \delta h_{\text{s } \alpha } ( x^{\prime } ) } \nonumber \\ * & = & \gamma_0 \chi_{\text{u},\text{imp } } ( x ) g(x^{\prime } ) \nonumber \\ \chi_{\text{imp},\text{imp } } & \equiv & \frac{t}{3 } \frac{\delta^2 \ln z}{\delta h_{\text{imp},\alpha } \delta h_{\text{imp } , \alpha } } \nonumber \\ * & = & \frac{s(s+1)}{3 t } \left ( 1 - 2 \gamma_0 ^ 2 \mathcal{f } \right ) \nonumber \\ \chi_{\text{s},\text{imp } } ( x ) & \equiv & \frac{t}{3 } \frac{\delta^2 \ln z}{\delta h_{\text{s}\alpha}(x ) \delta h_{\text{imp } , \alpha } } \nonumber \\ * & = & -2 \frac{s(s+1)}{3 t } \gamma_0 ^ 3 g(x ) \mathcal{f } \nonumber \\ \chi_{\text{s},\text{s } } ( x , x^{\prime } ) & \equiv & \frac{t}{3 } \frac{\delta^2 \ln z}{\delta h_{\text{s}\alpha } ( x ) \delta h_{\text{s } \alpha } ( x^{\prime } ) } \nonumber \\
* & = & \frac{s(s+1)}{3 t } \gamma_0 ^ 2 g(x ) g(x^{\prime } ) \label{par3}\end{aligned}\ ] ] where @xmath286 the prefactors of @xmath287 in ( [ par3 ] ) merely perform the rotational averaging over the directions in spin space .
it is important to note that we have only included impurity related contributions in the above results : the portions of the susceptibilities which are a property of the bulk theory , @xmath32 , alone have been dropped as they were considered in earlier work .
we will now combine these results into various observable correlators and discuss their scaling properties .
first , let us consider the impurity susceptibility , @xmath68 , defined in ( [ defchi ] ) .
this is related to the expressions in ( [ par3 ] ) by @xmath288 . \label{par5}\end{aligned}\ ] ] an important property of the above expression is that as @xmath65 for @xmath48 , the term of order @xmath289 is exponentially suppressed and the response reduces to that of a free spin @xmath3 : this indicates , as expected , that the total magnetic moment associated with the impurity is precisely @xmath3 in paramagnetic phase .
this last fact can also be established directly by computing the magnetization in the presence of a finite field @xmath281 at @xmath83 ; this can be done to all orders in @xmath281 by the methods to be developed in section [ rg2 ] , and the result ( [ n17 ] ) holds we will not present the details here .
next , we examine the behavior of @xmath68 as @xmath33 approaches @xmath34 from above .
the momentum integral in ( [ par5 ] ) is exponentially convergent , and so there are no poles in @xmath101 .
this implies , as expected from the conservation of total spin , that @xmath68 acquires no anomalous dimensions , and ( [ par5 ] ) can be written in the scaling form @xmath290 with @xmath291 a universal scaling function .
the latter can be evaluated from ( [ par5 ] ) by setting @xmath292 , and by using the crossover function for @xmath232 in ref . .
let us consider such an evaluation explicitly at @xmath126 .
then , from ( [ eps3 ] ) , @xmath232 is proportional to @xmath18 , but with the proportionality constant of order @xmath293 .
it is therefore tempting , to leading order in @xmath101 , to simply set @xmath294 in the integrand in ( [ par5 ] ) .
however , this leads to trouble the integrand in ( [ par5 ] ) is infrared singular , and behaves like @xmath295 at low momentum .
so we have to keep a finite value of @xmath232 , and the correction of order in @xmath289 in ( [ par5 ] ) , which is superficially of order @xmath101 , turns out to be of order @xmath293 .
by such a calculation , we find that at @xmath126 , @xmath68 takes the form ( [ i1 ] ) , with @xmath296 . \label{par7}\ ] ] actually , it is possible to also obtain the @xmath297 contribution above without too much additional difficulty . to do this , we have to identify only the contributions to @xmath68 at order @xmath241 and @xmath298 , which are infrared singular enough to reduce the expression from superficial order @xmath299 to order @xmath101 .
it turns out that there is only a single graph for @xmath300 which can accomplish this , and it contributes @xmath301 ^ 2 \label{par8}\ ] ] to @xmath68 . evaluating ( [ par8 ] ) , and also identifying the @xmath297 contribution from ( [ par5 ] ) , we correct ( [ par7 ] ) to @xmath302 .
\label{par9}\ ] ] second , consider the response , at or close to the impurity site , to a local magnetic field applied near the impurity site .
this is usually denoted by @xmath303 and could be measured in a muon spin resonance experiment ; we define it by @xmath304 we have inserted a prefactor of @xmath305 because @xmath306 is the two - point correlator of the @xmath89 , and this acquires a field renormalization factor @xmath307 in ( [ eps11 ] ) .
it is now easy to verify from ( [ par3 ] ) and ( [ eps17 ] ) that the poles in @xmath101 do cancel , and that ( [ par10 ] ) can be written in the form @xmath308 again @xmath309 is a universal scaling function ; it reaches a constant value at zero argument and so @xmath303 diverges as @xmath310 at @xmath126 . for @xmath48 , we see from ( [ par3 ] ) that @xmath311 as @xmath312 .
we can consider this as arising from the overlap of the impurity moment with the total , freely fluctuating , moment of @xmath3 that was considered below ( [ par5 ] ) , and so write @xmath313 from the expressions in ( [ par3 ] ) , we deduce @xmath314 .
\label{par11b}\ ] ] it can now be checked that this value for @xmath191 agrees precisely with that computed from the definition ( [ defm ] ) this is shown in appendix [ mimp ] , where @xmath315 evaluates to ( [ mimpsq ] ) .
also this last result , or ( [ par11 ] ) , show that ( [ defm1 ] ) is obeyed with @xmath316 .
alternatively , we can use the methods of section [ rg2 ] to compute the @xmath83 impurity magnetization in the presence of a finite field @xmath281 , to all orders in @xmath281 , and the result agrees with ( [ par11b ] ) .
finally , we measure the space - dependent response to a uniform magnetic field , @xmath281 .
this can be measured in a nmr experiment as a knight shift , as was done in refs . , with the results indicated in fig [ stagg ] .
we are considering linear response in @xmath281 , and so such results are implicitly valid for @xmath317 .
however , experiments @xcite are often in a regime where @xmath281 is of order @xmath18 , and so it is useful to go beyond linear response , and obtain the full @xmath318 dependence of the knight shift .
we will show that this can be done using some simple arguments in the spin - gap regime ( @xmath36 ) with @xmath319 , but @xmath318 arbitrary : from the knowledge of the linear response in @xmath281 , the entire @xmath318 dependence can be reconstructed . in the quantum critical region ( @xmath320 , @xmath223 )
we will be satisfied by exploring @xmath321 in linear response ; in the opposite limit , the host antiferromagnet undergoes a phase transition to a canted state@xcite induced by the applied field at @xmath322 , and we wish to avoid such complications here .
we can identify three important components of a knight shift : ( _ i _ )
@xmath323 , the knight shift of a nucleus at , or very close to , the impurity site ; ( _ ii _ ) @xmath324 , the envelope of a knight shift which oscillates rapidly with the orientation of the antiferromagnetic order parameter , ( _ i.e. _ , as @xmath325 ; and ( _ iii _ ) @xmath326 , the uniform component of the knight shift away from the impurity site .
we will consider these three in turn .
we measure the knight shift as simply the mean electronic moment at a particular location , and will drop the factor of the electron - nucleus hyperfine coupling .
( _ a _ ) @xmath323 we define @xmath323 by @xmath327 h \nonumber \\ & & \quad = z^{\prime -1/2 } \frac{s(s+1)}{3 t } \bigg [ 1 - \gamma_0 ^ 2 \int \!\ ! \frac{d^d
k}{(2 \pi)^d } \bigg ( \frac{\coth ( \varepsilon_k /2t)}{2 \varepsilon_k^3 } \nonumber \\ & & \qquad \qquad \qquad - \frac{1}{4 t \varepsilon_k^2 \sinh^2 ( \varepsilon_k / 2 t ) } \bigg ) \bigg ] h. \label{par12}\end{aligned}\ ] ] this has a renormalization factor of only @xmath307 because it is the correlator of @xmath89 with the total magnetization , and the latter conserved quantity requires no renormalization
. there is an overall factor of @xmath281 because the magnetization is induced by the external field , and we are considering linear response . as for ( [ par11 ] ) , it can be verified from ( [ par3 ] ) and ( [ eps17 ] ) that the poles in @xmath101 cancel , and the result is of the form @xmath328 the scaling function @xmath329 reaches a constant value at zero argument , and so @xmath323 diverges as @xmath330 at @xmath126 . for @xmath48 , @xmath331 as @xmath65 , and by the analog of the arguments associated with ( [ par11a ] ) we can now write @xmath332 the resulting value for @xmath191 agrees with ( [ par11b ] ) .
we can extend ( [ par13a ] ) to beyond linear response in @xmath281 for @xmath333 by realizing that the important thermal excitations in such a regime are simply those of the free spin in the external field ; so the prefactor of the free spin susceptibility in ( [ par13a ] ) is simply the component of the free spin wavefunction near the impurity .
we can expect that the same prefactor applies for arbitrary @xmath318 , and so the general response is the same prefactor times the free spin magnetization in a field @xmath281 at temperature @xmath18 ; this generalizes ( [ par13a ] ) to @xmath334 where @xmath335 is the familiar brillouin function for spin @xmath3 : @xmath336 note that @xmath337 for small @xmath338 , and @xmath339 .
naturally , ( [ par13b ] ) reduces to ( [ par13a ] ) for @xmath340 . for large @xmath318 ,
the impurity knight shift is @xmath3 times the impurity moment @xmath191 .
( _ b _ ) @xmath341 for the staggered knight shift , we have the expression @xmath342 h. \label{par14}\ ] ] notice that now we only have the field scale renormalization of the bulk theory , as we are considering a correlator of @xmath343 with the total conserved spin .
to the order we are working , we can simply set @xmath344 , and then the expressions in ( [ par3 ] ) yield @xmath345 h. \label{par14a}\end{aligned}\ ] ] upon using ( [ eps11 ] ) and ( [ eps17 ] ) it can be verified that the above is free of poles in @xmath101 , and the result is of the form @xmath346 with @xmath347 a universal function , as is expected from general scaling arguments . at the approximation we have computed things , @xmath348 , and @xmath349^{\epsilon/2}$ ] . as for @xmath323 , we can generalize ( [ par15 ] ) to obtain the non - linear response in @xmath281 in the spin gap regime ( @xmath350 ) .
we first identify a local staggered moment as in ( [ par13a ] ) @xmath351 we have identified the staggered moment as proportional to the expectation value of the antiferromagnetic order in an applied field in the @xmath88 direction , and the factor of @xmath352 follows because we have absorbed a factor of @xmath3 in defining the free moment of which @xmath353 is an envelope . from ( [ par14a ] )
we obtain immediately @xmath354 ; \label{par15b}\ ] ] this expression for @xmath355 can also be obtained by the methods of section [ rg2 ] by computing the staggered magnetization in the presence of a finite field at @xmath83 .
the factor in the square brackets , after using ( [ eps11 ] ) and ( [ eps17 ] ) , is free of poles in @xmath101 , and evaluates to order @xmath356 at the fixed point value for @xmath251 .
because of the @xmath357 factor , @xmath355 decays exponentially for large @xmath195 , and we will consider its small @xmath195 behavior shortly . to obtain the non - linear response we now have the analog of ( [ par13b ] ) : @xmath358 so at large @xmath318 , the staggered knight shift simply measures the space - dependent antiferromagnetic moment induced by the impurity spin . we examine the behaviors of @xmath324 and @xmath359 for small @xmath195 ; these are specified by the operator product expansion in ( [ ope ] ) : @xmath360 where @xmath191 was obtained in ( [ par13a ] ) ; a similar result holds for the knight shift in linear response in @xmath281 but all @xmath18 : @xmath361 provided the value of @xmath186 is such that @xmath362 ( which we definitely expect ) , the staggered knight shift will increase as one approaches the impurity , as seen in refs . .
indeed , all of the results of this subsection are qualitatively consistent with the trends in fig [ stagg ] .
( _ c _ ) @xmath363 the remaining uniform knight shift is given by @xmath364 h .
\label{par17}\ ] ] now no renormalization factors are necessary because this is a correlator of the conserved spin of the bulk theory with the conserved spin of the total theory , and neither of them acquire any anomalous dimensions .
evaluation of ( [ par17 ] ) using ( [ par3 ] ) gives @xmath365 at the fixed point of the beta - functions , this is of the universal scaling form @xmath366 as in the staggered case , we can go beyond linear response in @xmath281 for @xmath367 . then , following ( [ par15a ] ) , we define a uniform magnetization @xmath368 by @xmath369 and an expression for @xmath370 follows from ( [ par18a ] ) .
the non - linear response , generalizing ( [ par15c ] ) is @xmath371 as before , the small @xmath195 behavior of the scaling function is controlled by the operator product expansion ( [ ope ] ) : @xmath372 and @xmath373 we consider impurity properties when spin rotation invariance has been broken in the host antiferromagnet for @xmath113 .
we will restrict our attention to @xmath83 .
the field @xmath20 has an average orientation which we take in the @xmath88 direction .
to lowest order in @xmath52 this expectation value is @xmath374 there is no dependence on @xmath195 at this order .
we will consider corrections to this result in renormalized perturbation theory . in preparation , we quote some additional properties of the host antiferromagnet .
we will need the value of the critical coupling to leading order in @xmath34 @xcite @xmath375 we also define a parameter @xmath376 measuring the deviation from the critical point @xmath377 associated with this is a renormalized @xmath378 and a bulk renormalization factor @xmath379 @xmath380 and to lowest order in @xmath62 we have@xcite @xmath381 the ordering in the host antiferromagnet produces an effective field on the impurity spin , and so its fluctuations are anisotropic . in principle
, these can be computed order - by - order in @xmath226 by the perturbative methods discussed in appendix [ diag ] .
however , spin anisotropy leads to a proliferation in the number of diagrams , and we found it more convenient to use in alternative approach .
indeed , the broken spin rotation invariance implies that traditional methods developed for ordered magnets can be applied here : we used the dyson - maleev representation for the impurity spin @xcite .
a potential problem with the dyson - maleev representation is that it is designed to obtain results order - by - order by an expansion in @xmath352 , while all our results so far have been exact as a function of @xmath3 .
however , it turns out that , at @xmath83 , the feynman graph expansion of the dyson - maleev representation of our problem remains a perturbation theory in @xmath52 and @xmath226 : to each order in these parameters , the results are exact as a function of @xmath3 .
this is a consequence of there being only a small number of dsyon - maleev bosons in every intermediate state in the perturbation theory , and these can couple only through a limited number of non - linear interactions .
in contrast , at @xmath207 , the dyson - maleev perturbation theory sums over an infinite number of intermediate states with an arbitrary number of bosons , and the results are then no longer exact as a function of @xmath3 ; in this case the methods of appendix [ diag ] should be used , as they remain exact even for @xmath207 .
our interest here is only in @xmath83 , and so we will use the more convenient dyson - maleev method it has the advantage of using canonical bosons and so permits use of standard time - ordered diagrams , the dyson theorem , and automatic cancellation of disconnected diagrams .
let us introduce the dyson - maleev formulations .
we label the states of the impurity spin by the occupation number of a canonical boson , @xmath382 .
the spin - operator is given by @xmath383 notice that the representation does not appear to respect the hermiticity of the spin operators ; this is because a similarity transformation has been performed on the hilbert space
we refer the reader to the literature for more discussion on this point .
it is also convenient to define a ` circularly ' polarized combination of the bulk field @xmath20 , and to shift the longitudinal component from the mean value in ( [ n1 ] ) : @xmath384 now @xmath76 in ( [ simp ] ) takes the form ( we only have a single impurity at @xmath385 and have dropped the sum over @xmath70 ) @xmath386 , \label{n8}\end{aligned}\ ] ] where it is understood that @xmath387 and @xmath388 are evaluated at @xmath164 in the above .
notice that , at zeroth order , the @xmath382 bosons have finite energy gap of @xmath389it is this gap which stabilizes the perturbation theory and permits evaluation of results exact in @xmath3 at each order in @xmath226 and @xmath52 . for completeness
, we also quote the bulk action @xmath32 in ( [ sb ] ) in this representation : @xmath390 , \label{n9}\end{aligned}\ ] ] the partition function is now an unrestricted functional integral over @xmath391 , @xmath392 and @xmath393 with weight @xmath394 .
computation of the perturbation theory in @xmath226 and @xmath52 in above representation is a completely straightforward application of standard methods .
there are a fair number of non - linear couplings , and so tabulation of all the graphs can be tedious .
we will be satisfied here by simply quoting the results of perturbation theory and then providing a scaling interpretation .
we consider a few different observables in the following subsections .
we consider the static magnetization of the impurity site , and also the staggered and uniform moments in the host antiferromagnet .
first , the static magnetization of the impurity , @xmath191 , defined in ( [ defm ] ) . in bare perturbation theory
we obtain @xmath395 \label{n10}\end{aligned}\ ] ] we evaluate the integral , perform the substitutions to the renormalized couplings and fields in ( [ eps8 ] ) , ( [ eps11 ] ) , ( [ n3 ] ) and ( [ n4 ] ) , use the renormalization constants in ( [ eps10 ] ) , ( [ eps15 ] ) , ( [ eps17 ] ) and ( [ n5 ] ) , and then expand to the appropriate order in @xmath101 . all poles in @xmath101 cancel and we obtain @xmath396 where @xmath397 is the euler - mascheroni constant . substituting the fixed point values @xmath398 , @xmath399 , and exponentiating we see that @xmath400 by ( [ defm1 ] ) this defines the exponent @xmath401 , and is consistent with the leading order values @xmath402 , @xmath316 .
next we turn to the spatial dependence of the static moment in the host antiferromagnet .
there will be a staggered contribution to this which oscillates with the local orientation of the antiferromagnetic order , given by @xmath355 .
evaluating the latter in perturbation theory we obtain @xmath403 .
\label{n12}\end{aligned}\ ] ] this is evaluated by the method described below ( [ n10 ] ) ; again poles in @xmath101 cancel and we obtain @xmath404 .
\label{bulkmag}\end{aligned}\ ] ] the function @xmath405 has the properties @xmath406 where @xmath407 is a modified bessel function . after substituting fixed points values for @xmath251 and @xmath62 in ( [ bulkmag ] ) , we see that for @xmath408 , the moment has the bulk behavior@xcite @xmath409 with exponent @xmath410 , as expected . in the opposing limit ,
the result is consistent with that expected from the operator product expansion ( [ ope ] ) , given in ( [ n14 ] ) , to leading order in @xmath101 . a second contribution to the host static magnetization is uniform on the lattice scale this is given by the expectation value of the magnetization , @xmath370 , defined in ( [ defl ] ) .
perturbation theory yields @xmath411 we will leave this integral in the unevaluated form above : suffice to say that @xmath370 decays to 0 at large @xmath412 , while ( [ n16 ] ) holds for small @xmath195 . in the notation introduced at the beginning of section [ rg1 ]
, a uniform magnetic field corresponds to @xmath413 , @xmath414 , and @xmath283 .
we have to differentiate between the responses parallel and orthogonal to the bulk order parameter . in the direction parallel to the bulk order ( the @xmath88 direction )
, the symmetry of rotations about the @xmath88 axis is preserved , and this means that total spin is a good quantum number .
consequently , the total magnetic moment is quantized@xcite precisely at @xmath3 @xmath415 it can be verified that this holds order - by - order in perturbation theory in @xmath226 and @xmath52 ; the sensitive cancellations required to make this happen are consequences of gauge invariance .
this magnetic moment is pinned to the direction of the bulk antiferromagnetic order and is not free to rotate
so unlike the situation in the paramagnet , it does not contribute a curie susceptibility .
indeed , the direction of the bulk antiferromagnetic order is invariably pinned by very small anisotropies which are always present ; in such a situation , the total impurity moment is also static , and longitudinal susceptibility is zero .
now consider the response to a field in a direction ( say @xmath195 ) transverse to the antiferromagnetic order , @xmath182 .
such a field induces numerous additional terms in the action @xmath416 which can be deduced from ( [ par1 ] ) , ( [ par2 ] ) , ( [ n6 ] ) and ( [ n7 ] ) .
we then expanded the partition function to second order in @xmath417 in a perturbation theory in @xmath226 and @xmath52 .
this required a total of 22 one - loop feynman diagrams .
we will refrain from listing them here as the computations are completely standard details of these diagrams are available from the authors .
these diagrams were evaluated and expressed in terms of renormalized couplings as described below ( [ n10 ] ) using the mathematica computer program .
the final expressions were free of poles in @xmath101 , and this was a very strong check on the correctness of the computations . at the fixed point value for @xmath251 and @xmath62 ,
the final result was @xmath418 .
\label{n18}\end{aligned}\ ] ] let us express this result in terms of the spin stiffness of the ordered state , @xmath51 .
we expect @xmath419 to scale as an inverse energy , and the parameter which sets the energy scale for general @xmath31 is@xcite @xmath420^{1/(d-1)}. \label{n19}\ ] ] the numerical factors have been chosen for convenience ; note that in @xmath43 @xmath421 is simply proportional to @xmath51 .
the @xmath101 expansion for @xmath422 is available in ref . :
@xmath423 . \label{n20}\ ] ] if we eliminate @xmath378 between ( [ n18 ] ) and ( [ n20 ] ) we see that the @xmath244 dependence also disappears : this verifies that @xmath419 and @xmath421 are universally proportional to each other
. we may generalize ( [ i2 ] ) to @xmath424 by @xmath425 our results yield the @xmath101 expansion for the universal constant @xmath426 : @xmath427 . \label{n22}\end{aligned}\ ] ]
in this section we complement the rg analysis of section [ sec : eps ] by a self - consistent diagrammatic approach . this new approach will allow us to obtain more detailed dynamic information for the single impurity problem .
it can also be easily extended to treat the problem with a finite density of impurities , and this will be considered in section [ sec : ncamany ] .
one way of motivating the analysis is the large-@xmath220 approximation : the symmetry group of the impurity spin is extended from su(2 ) to su(@xmath220 ) and the @xmath428 limit is taken by a saddle - point of the functional integral . alternatively , the resulting saddle - point equations can also be interpreted as the summation of all ` non - crossing ' feynman diagrams this is the so - called non - crossing approximation @xcite ( nca ) . while the nca and large-@xmath220 approaches are equivalent for the single impurity problem , this will no longer be the case in the many - impurity analysis of section [ sec : ncamany]there we will use the nca , but will not have @xmath429 as a control parameter . it is worth pointing out that @xmath429 is the only small parameter in the present single impurity analysis ; we are not restricted to small @xmath219 or to a perturbation theory in the coupling between bulk and impurity .
indeed , we will sum diagrams to all orders in the latter coupling , and can work directly @xmath43 .
however , we will formulate results in arbitrary @xmath31 to allow for a comparison with the expressions obtained in the @xmath101 expansion of section [ sec : eps ] .
it is convenient to present the nca analysis in a hamiltonian formulation of @xmath85 .
the bulk system shall be represented by a heisenberg model of spins @xmath430 on a regular two - dimensional lattice ; a concrete example is an array of coupled ladders @xcite . at zero temperature
the bulk system can be driven from a paramagnetic state with energy gap @xmath0 to a nel state by varying a coupling constant @xmath431 .
for an explicit derivation we assume that the paramagnetic phase of the bulk is dimerized .
its excitations shall be described using the bond - operator formalism @xcite where the spins of each pair forming a dimer are represented by bosonic singlet ( @xmath432 ) and triplet ( @xmath433 , @xmath434 ) bond operators : @xmath435 note that the index @xmath40 here labels bonds , _
i.e. _ , pairs of lattice sites , while the subscripts @xmath436 identify the two spins in each pair .
the bulk hamiltonian can now be expressed terms of the bond operators . using standard mean - field - type approximations ( _ i.e. _ , condensation of singlet operators and decoupling of quartic triplet terms ) one arrives at @xmath437 which contains only bilinear terms in @xmath438 .
@xmath439 denotes a characteristic bulk energy scale ( _ i.e. _ , the nearest - neighbor coupling constant ) .
the dimensionless functions @xmath440 and @xmath441 contain the geometry of the system and depend upon the coupling constant @xmath431 and the mean - field parameters .
@xmath442 can be easily diagonalized by a bogoliubov transformation leading to @xmath443 where the new triplet bosons are defined through @xmath444 , @xmath445 and @xmath446 are the bogoliubov coefficients with @xmath447 , and @xmath448 denotes the dispersion of the triplet modes , @xmath449 .
note that @xmath440 and @xmath441 are finite , smooth functions of @xmath450 .
for a detailed discussion of appropriate mean - field calculations see e.g. refs .. the spin-1 excitations @xmath451 in ( [ hbosdiag ] ) appear to be non - interacting bosons . however , this is somewhat deceptive : interactions between these bosons where necessary to obtain the self - consistent dispersion @xmath452 , and these interactions also lead to a crucial temperature dependence in @xmath448 .
the dispersion @xmath448 in the paramagnetic state is given by @xmath453 at small @xmath454 ( the velocity @xmath455 in our conventions ) , where @xmath456 denotes the renormalized mass introduced in ( [ eps2 ] ) , ( [ eps3 ] ) .
at zero temperature we have @xmath457 , the spin gap . for @xmath207 ,
the interaction @xmath52 in @xmath32 led to the temperature dependence in ( [ eps3 ] ) ; similarly here we find that the solution of the mean - field equations leads to the scaling form @xmath458 where @xmath459 is a universal scaling function .
expressions for @xmath459 in general @xmath31 are available @xcite , and in @xmath43 we have the simple explicit result @xcite : @xmath460 note that in the quantum - critical region , @xmath320 , @xmath461 , a dependence similar to ( [ eps3 ] ) in the @xmath101 expansion . in analogy to section [ qimp ] we now introduce magnetic impurities into the bulk system . here
we restrict ourselves to a single impurity spin @xmath430 coupled to the bulk at ( bond ) site 0 with a coupling constant @xmath151 : @xmath462 here @xmath463 are the components of the impurity spin , and @xmath464 is the number of dimer sites of the bulk lattice .
note that we have omitted terms of the form @xmath465 ; such terms correspond to the irrelevant coupling in ( [ impl ] ) .
the second identity in eq .
( [ himp ] ) holds for @xmath466 . at this point
it is useful to establish the correspondence between the operators in @xmath467 and the fields in @xmath85 .
the triplet bosons @xmath468 near the antiferromagnetic wavevector @xmath469 represent the fluctutations of the antiferromagnetic order parameter ; they are related to the @xmath470 fields by @xmath471 the impurity spin operator @xmath472 directly corresponds to @xmath473 appearing in eq .
( [ simp ] ) . to allow for a controlled approximation
we generalize the impurity spin symmetry to su(@xmath220 ) and consider the large-@xmath220 limit .
all results of this and the following subsections are limited to @xmath474 , _
i.e. _ , to a bulk phase without spontaneously broken symmetry .
the impurity spin is represented by auxiliary fermions @xmath475 ( @xmath476 ) , so the @xmath477 traceless components of the spin @xmath478 can be written as @xmath479 . a chemical potential @xmath480 is introduced to enforce the constraint @xmath481 ; for @xmath482 this corresponds to @xmath82 , and all results in this and the next section will be restricted to this value of the spin .
the generalization of the bulk system to su(@xmath220 ) symmetry leads to @xmath477 triplet operators @xmath483 .
the hamiltonian for the impurity takes the form @xmath484 we are interested in the dynamics of the impurity spin arising from the coupling to the bulk .
feedback effects are not present for the case of one impurity ; they will be discussed in section [ sec : ncamany ] .
the imaginary time green s function for the auxiliary fermions @xmath475 is introduced as @xmath485 here , @xmath486 is the time ordering operator in imaginary time .
if we set @xmath487 , the unperturbed green s function is simply given by @xmath488 where @xmath489 denotes a fermionic matsubara frequency and @xmath490 .
furthermore we introduce the green s functions for the bulk bosons : @xmath491 from the bulk hamiltonian @xmath492 ( [ hbosdiag ] ) the fourier transform of @xmath493 is found to be @xmath494 with @xmath495 representing a bosonic matsubara frequency . in order to calculate the self - energy @xmath496 of the @xmath497 particles arising from the interaction with the bulk bosons we employ a non - crossing approximation ( nca ) @xcite .
this amounts to the summation of all self - energy diagrams with non - crossing boson lines .
the nca approach can be derived from a saddle - point principle in the large-@xmath220 limit , see e.g. refs . . within nca
we obtain the following equations for the green s functions : @xmath498 where the self - energies @xmath496 are defined by : @xmath499 the quantity @xmath500 represents the bulk fluctuations at the impurity site , it is given by the bulk green s function of the operator sum @xmath501 together with the momentum dependence of the vertex in ( [ himpn ] ) : @xmath502 here we have suppressed all su(@xmath220 ) indices since the bulk is assumed to be isotropic ( no broken symmetry ) .
we remind the reader that @xmath503 and @xmath500 are related by fourier transforms to the propagators of the @xmath20 field in ( [ mattis ] ) and ( [ eps2 ] ) .
the fourier transform of @xmath504 follows from ( [ hbosdiag ] ) : @xmath505 finally , @xmath480 is determined by the equation : @xmath506 the symmetry of the nca equations permits solutions for the green s functions which are independent of @xmath47 ; therefore we drop the su(@xmath220 ) index from now on .
( note that this symmetry would not hold in a magnetically ordered phase of the bulk . )
we introduce spectral densities corresponding to @xmath507 and @xmath504 , @xmath508 where @xmath509 follows from the definition of @xmath500 . from this symmetry property of
@xmath510 it can be easily shown that for @xmath511 the chemical potential @xmath480 is zero for all temperatures and values of @xmath456 ( see appendix [ appnca ] ) .
the system of nca equations can be analyzed in the low - temperature , low - energy regime where all energies are well below a high - energy cut - off set by the lattice ( e.g. the nearest - neighbor exchange @xmath439 ) . in the scaling limit
the momentum sums can be transformed to integrals with an upper cut - off @xmath512 , so we consider @xmath513 , @xmath514 , @xmath515 . for @xmath424
there are no ultraviolet divergences , _
i.e. _ , the results are independent of @xmath512 for large @xmath512 .
we start with the bulk spin fluctuations .
the triplet modes are gapped with an effective mass @xmath516 , their dispersion near @xmath469 is given by @xmath517 .
( note that the momentum @xmath454 is now measured relative to the antiferromagnetic wavevector @xmath146 , and we will explicitly display the velocity @xmath1 here and in the following . ) the scaling limit of the local bulk green s function can be found from ( [ gttlattice ] ) .
note that the dominant contributions to the momentum integral arise near @xmath518 , so the smooth function @xmath440 can be replaced by its @xmath518 value @xmath519 .
@xmath520 with @xmath521 defined in ( [ eps9 ] ) .
we have dropped an additive contribution to @xmath504 , dependent on @xmath512 , but non - singular as a function of frequency .
this is made clearer by a fourier transform to the time domain , where the @xmath512-dependent contributions are negligible at long times : at @xmath83 , the fourier transform can be written as @xmath522 where the scaling function @xmath523 depends on the product @xmath524 only .
it has the following asymptotic behavior : @xmath525 from this it is clear that the contribution arising from the upper cut - off @xmath512 falls off as @xmath526 , and becomes @xmath527 for @xmath515 , which will drop out of the scaling equation ( see below ) for @xmath511 .
for finite temperature , the scaling form ( [ gttscaltau ] ) has to be replaced by @xmath528 and so as @xmath65 we must have @xmath529 if we insert the expression for @xmath504 into the nca equations it turns out that the dimensionful factors can be combined into a single energy scale @xmath530 which plays a role similar to the renormalization scale @xmath244 in the rg analysis : @xmath531 for convenience we have absorbed the dimensionless quantities @xmath519 , @xmath521 into @xmath530 , too .
the solution of the nca equations can be written in terms of two - parameter scaling functions : @xmath532 here , @xmath533 and @xmath534 are universal scaling functions which do not depend on the microscopic details of the bulk or the cutoff .
note that the scaling form for @xmath456 in ( [ mscale ] ) , ( [ phim ] ) can be used to write the scaling functions in terms of the arguments @xmath535 and @xmath536 .
the fourier transform of the scaling function @xmath533 is given by the solution of the nca equations ( [ ncasigma ] ) , ( [ defsigma ] ) in the scaling limit : @xmath537 where @xmath538 , and we have used that @xmath539 . for shortness we have dropped the argument @xmath540 in the scaling functions .
( for general values of @xmath541 one finds @xmath542 for @xmath543 , _
i.e. _ , only @xmath544 obeys a scaling form , and any dependence on the cut - off @xmath512 is absorbed in the chemical potential @xmath480 . ) in general , the solution of ( [ ncascalet ] ) must be found numerically .
however , some special cases which permit an analytical solution will be discussed in turn .
we first consider the bulk critical point ( @xmath545 ) at zero temperature .
the triplet modes are gapless and follow a linear dispersion @xmath546 .
it can be easily shown that the long - time behavior of the local bath green s function ( [ gttscaltau ] ) is given by @xmath547 by matching the @xmath200 powers in the nca equations it is found that the green s function @xmath507 has the following scaling behavior in the zero - temperature limit : @xmath548 more generally , these relations are valid provided that @xmath549 and @xmath550 .
it is possible to obtain a complete solution @xcite of the nca equations in the scaling limit for @xmath294 at low temperature which means that we take @xmath551 while keeping @xmath552 finite .
we note that in general these solutions do not correspond to a physical situation since , at non - zero temperatures and @xmath553 , bulk interactions always lead to a finite effective mass @xmath554 .
it is , however , instructive to display the complete solutions for @xmath294 . by explicitly performing the fourier transformations and using the scaling form @xmath555^{d-1 } } \ ] ] it can be shown @xcite that the scaling functions have the form @xmath556 with @xmath557 , @xmath558 .
the amplitude @xmath559 is a constant depending on @xmath31 only .
periodicity requires @xmath560 .
note that the above scaling function has a conformal - invariant form , _
i.e. _ , the finite - temperature green s function follows from the @xmath83 green s function by applying the conformal transformation @xmath561 @xcite .
this holds for the value @xmath562 ; there the system obeys particle - hole symmetry among the pseudo - fermions @xmath497 under @xmath563 , which follows from the fact that the average number of fermions per `` flavor '' is @xmath430 .
the connection between the nca - type approximation and conformal field theory has been discussed in ref . ; see also ref . .
for the case of non - zero @xmath456 we have not succeeded in finding a complete analytical solution of the nca equations . here
we shall briefly discuss the general behavior of @xmath565 at zero temperature and @xmath566 .
the asymptotic expressions for @xmath504 ( [ gtlimits ] ) allow to determine the asymptotic behavior of @xmath567 : @xmath568 where now @xmath569 .
in the long - time limit the green s function @xmath567 decays to a finite value which scales as @xmath570 ( see also next subsection ) .
the spectral density of the bulk fluctuations , @xmath571 , has a gap of size @xmath572 . at small frequencies @xmath573
the impurity spin behaves like a free spin which implies that @xmath574 has a @xmath575-peak at @xmath576 with a weight of order @xmath577 . at small , but finite @xmath200
, @xmath574 inherits the gap from @xmath571 , _
i.e. _ , @xmath578 for @xmath579 . for @xmath580
, @xmath574 is non - zero ; the singularity at @xmath581 has the form @xmath582 [ @xmath583 for @xmath43 ] .
furthermore , @xmath574 has singularities at all integer multiple frequencies of @xmath0 . for large frequencies , @xmath585
, it decays as @xmath586 . a numerical result for the scaling function @xmath587 associated with the fermion spectral density at @xmath43
is shown in fig .
[ figrhof ] . associated with the fermion spectral density @xmath588 as a function of @xmath589 , calculated for @xmath43 .
the arrow denotes the @xmath575-peak at @xmath576 with an amplitude of @xmath590 .
, width=326 ] in this section we consider the response of the system to an external magnetic field @xmath281 in analogy to section [ rg1 ] .
we restrict our attention to a homogeneous field , _ i.e. _ , we set the staggered component @xmath591 and neglect any spatial dependence of the uniform component @xmath592 . within our large-@xmath220 generalization the field @xmath281 couples to the @xmath477 components of the impurity spin as well as to the bulk bosons .
the field will be described by the hermitean matrix @xmath593 , the hamiltonians are extended according to @xmath594 which can be deduced requiring that a homogeneous field @xmath595 couples to a conserved quantity .
it is easily seen that at large-@xmath220 the susceptibility of the bulk alone is of order @xmath220 whereas all impurity corrections which will be considered below are of order unity . without loss of generality
we assume in the following @xmath596 , _
i.e. _ , the field couples to a traceless combination of two ` colors ' which reduces to the @xmath195 component of the spin for the usual su(2 ) case
. the susceptibility of a free spin in the large-@xmath220 limit is easily found as @xmath597 with @xmath598 ; the exact answer for su(2 ) is , of course , @xmath180 . as in section [ rg1 ]
we consider response functions @xmath599 , @xmath600 , and @xmath601 . the feynman diagrams being relevant in the large-@xmath220 limit are shown in fig .
[ figchidgr ] . , the full lines are propagators of the auxiliary fermions @xmath497 representing the impurity spin , and the dashed lines are @xmath602 propagators . , width=326 ]
note that there are two contributions to @xmath603 where @xmath604 has four interaction vertices . at the critical point ( @xmath553 )
the four diagrams have the following low - temperature behavior : @xmath605 in the gapped phase at low temperatures , @xmath606 , the susceptibilities are @xmath607 we now proceed by combining the results into observable quantities .
first , we have the impurity susceptibility , @xmath68 , as defined in ( [ defchi ] ) .
it is given by @xmath608 from ( [ chidgr1 ] ) , ( [ chidgr2 ] ) and @xmath609 in the scaling limit it follows that the low - temperature behavior of @xmath68 is dominated by @xmath599 , whereas the other contributions can be considered as corrections to scaling .
the impurity susceptibility takes the scaling form quoted in ( [ par6 ] ) , with the universal scaling function @xmath291 given by @xmath610 consistent with our predictions , there are no anomalous dimensions involved in @xmath68 , the reduced coupling constant @xmath530 and the bulk energy scale @xmath439 have dropped out .
the contributions to the susceptibility can be evaluated numerically . at the critical coupling , @xmath553
, we find @xmath68 to have a curie form , @xmath611 , as predicted in the introduction ( [ i1 ] ) , with a universal prefactor @xmath612 defining an effective spin , as discussed in the caption to fig [ fig1 ] . in @xmath43 ,
the large @xmath220 value of @xmath612 is @xmath613 , \ , , \label{valc1}\ ] ] where @xmath614 is defined in ( [ chifree ] ) . in the gapped phase at @xmath615
, it is seen from ( [ chidgr2 ] ) that the low-@xmath18 behavior of @xmath68 is determined by @xmath604 alone . as shown in appendix [ appnca ]
, the curie prefactor of @xmath604 equals that of a free spin , _
i.e. _ , @xmath616 .
this implies that in the gapped phase the impurity contributes to the total uniform susceptibility in the same manner as a free spin , _
i.e. _ , we have shown the confinement of the impurity spin . as a second observable , we again consider the local susceptibility which is defined as the response of the system to a magnetic field applied to the impurity spin only .
it is given by @xmath617 , and is proportional to the autocorrelation function of the of the impurity spin at zero frequency .
in the large-@xmath220 limit it is simply given by @xmath618 if we evaluate this integral using the scaling solution for @xmath507 we arrive at @xmath619 with the universal scaling function @xmath620 determined by @xmath621 from eq .
( [ gfscalgapped ] ) : @xmath622 the asymptotic behavior of @xmath620 is found to be @xmath623 the local impurity susceptibility diverges at @xmath553 as @xmath624 , similarly the impurity spin autocorrelation function decays as @xmath625 at the critical coupling .
this implies that we have @xmath626 in the large-@xmath220 limit . for @xmath615
, @xmath303 follows a curie law with a prefactor of @xmath627 which corresponds to a remnant impurity moment of @xmath628 consistent with the prediction ( [ defm1 ] ) .
in this section we will discuss the case of a finite density of magnetic impurities using the formalism presented in section [ sec : ncasingle ] .
a finite impurity density can lead to variety of phenomena which were discussed in sections [ qimp ] and [ many ] ; we will restrict our attention here to the scattering of the bulk spin excitations from the impurities .
interactions between the impurities which eventually may lead to an ordered state at very low temperatures will be neglected .
we shall assume @xmath629 and work directly in @xmath43 and real frequency ; all green s functions are retarded quantities , @xmath200 implicitly carries a positive imaginary part everywhere .
all the computations in this section will be for the case in which all impurities have spin @xmath56 . the important new effect from a finite @xmath2
is that bulk spin excitations described by the green s function @xmath630 acquire a self - energy from the scattering at the impurities this changes the local density of states of the bulk excitations .
we will determine the modified density of states self - consistently in this subsection .
the lowest - order scattering process for the @xmath602 bosons is described by the @xmath18-matrix @xmath631 given by the bubble diagram of fermion green s functions : @xmath632 the interaction @xmath633 gives rise a momentum - dependent self - energy for the @xmath602 bosons .
if we neglect higher - order scattering contributions ( born approximation ) the self - energy is given by @xmath634 where @xmath635 is the impurity concentration , and we have already performed an averaging over the ( randomly distributed ) impurities in the sample .
the green s function @xmath636 in the presence of scattering can be written as [ cf .
( [ gttlattice ] ) ] @xmath637 ^ 2 } \,.\end{aligned}\ ] ] in the scaling limit we can replace @xmath440 by @xmath519 as above leading to @xmath638 where the self - energy @xmath639 ( @xmath640 is now momentum - independent , it is given by @xmath641 the momentum integral can be performed in the scaling limit @xmath642 which replaces the expression ( [ gttscalom ] ) found at zero impurity concentration .
we have used @xmath517 .
it is important to note that the effective mass @xmath456 acquires a renormalization from the scattering processes since the hartree contribution to @xmath456 contains a boson propagator which is affected by the impurities ; @xmath456 has to be determined by @xcite @xmath643 where @xmath644 denotes the green s function ( [ gttscalom ] ) in the absence of impurities .
equations ( [ ttdef ] ) , ( [ sigmatt ] ) , ( [ gttscatt2d ] ) determine @xmath510 from @xmath507 , on the other hand @xmath507 is connected with @xmath510 through the nca equations ( [ ncasigma ] ) , ( [ defsigma ] ) .
solutions have to be found self - consistently , this procedure is similar to a self - consistent born approximation .
it is possible to write the solutions in terms of scaling functions , where the only additional parameter introduced by the impurities is the energy scale @xmath115 ( [ defgamma ] ) defined in section [ intro ] .
we restrict the presentation to zero temperature , and introduce scaling functions as follows : @xmath645 with the notations @xmath646 , @xmath647 , and @xmath648 we can re - write the equations ( [ ttdef ] ) , ( [ sigmatt ] ) , ( [ gttscatt2d ] ) : @xmath649 \label{scba1 } \ , , \\ {
\rm im}\:\phi_\sigma'(\bar\omega,\bar\gamma ) & = & \frac{2 \bar \gamma}{s_2 } \int \frac{d\bar\omega_1}{\pi } { \rm im}\:\phi_g'(\bar\omega-\bar\omega_1,\bar\gamma ) \nonumber\\ \times \ : { \rm im}\:&\phi_g'&(\bar\omega_1,\bar\gamma ) \left [ n_f(\bar\omega_1 ) - n_f(\bar\omega-\bar\omega_1 ) \right ] \ , , \nonumber\end{aligned}\ ] ] and the nca equations ( [ ncasigma ] ) , ( [ defsigma ] ) : @xmath650 \ , . \nonumber\end{aligned}\ ] ] the functions @xmath651 and @xmath652 are fermi and bose functions in the zero - temperature limit , @xmath653 , @xmath654 , where @xmath655 is the step function .
@xmath656 and @xmath657 are analytic functions of @xmath658 in the upper - half complex plane , so their real parts are given by kramers - kronig relations .
finally , the renormalized mass @xmath659 is determined by the constraint equation ( [ constraint ] ) written in terms of @xmath523 .
the system of equations ( [ scba1 ] ) , ( [ scba2 ] ) can be solved self - consistently for a given value of @xmath660 , and universally describes the spin dynamics of both the host antiferromagnet and the impurities . for @xmath211
the equations of course reduce to the zero - temperature limit of the nca equations ( [ ncasigma ] ) , ( [ defsigma ] ) quoted in section [ sec : ncasingle ] .
we now turn to the broadening of the boson spin-1 collective mode peak seen e.g. in a neutron scattering experiment .
therefore we have to calculate the dynamic spin susceptibility which is identical to the green s function @xmath636 .
we consider the dynamic susceptibility at the antiferromagnet wavevector @xmath661 the scaling form for @xmath639 in ( [ scalforms ] ) implies that @xmath662 indeed obeys the scaling form ( [ polescale ] ) with the scaling function @xmath206 given by @xmath663 the results show that @xmath664 has a hard gap at small frequencies , with a gap size of @xmath665 .
the lineshape is strongly asymmetric with a tail at high frequencies .
the half - width of the resulting line is approximately given by @xmath115 , see figs .
[ fig2 ] and [ fig3 ] . as a function of @xmath217 similar to fig .
[ fig2 ] , but for @xmath666 , 0.1 , 0.15 , and 0.2 . ,
width=345 ] the asymmetry of the line is a consequence of the interplay between the gap , @xmath0 , in the bulk antiferromagnet and the inelastic scattering from the impurity spin .
the bulk magnons self - consistently form an ` impurity band ' whose lower edge extends to an energy of order @xmath115 below @xmath0 .
there is no phase space for scattering below this band edge , while frequencies above have scattering states readily available
the result is an asymmetric line .
we now apply the present analysis to the broadening of the ` resonance peak ' observed recently in @xmath667-doped @xmath668 @xcite . using the values @xmath669 ev , @xmath670 mev , and @xmath671 , we obtain @xmath215 mev and @xmath672 .
a linewidth of roughly 5 mev is in excellent accord with the measured value of 4.25 mev @xcite .
the lineshape observed in ref .
appears to be asymmetric in accordance with our analysis ; however , future experiments should be able to test this prediction in more detail .
the object of attention in this paper has been a ` nearly - critical ' , two - dimensional , host antiferromagnet with a spin - gap @xmath0 .
we described this antiferromagnet by a universal , continuum , quantum field theory , @xmath673 : this is obtained obtained by deforming a o(3)-symmetric , 2 + 1 dimensional , conformally - invariant fixed point by its single relevant operator the ` mass ' term , @xmath674 , in ( [ sb ] ) .
then we ` punctured ' this continuum field theory by quantum impurities at a collection of spatial point @xmath71 .
we argued that each such impurity was completely characterized by a single integer / half - odd - integer valued spin , @xmath55 , which determined the berry phase of spin precession , see ( [ simp ] ) .
in particular , for a non - extensive number of impurities , there were no new relevant operators , and so the energy scale , @xmath0 , continued to characterize the dynamics of the antiferromagnet in the vicinity of the impurities .
a non - zero density of impurities , @xmath2 , is a relevant perturbation , but its impact was universally determined entirely by the dimensionless number @xmath675 , and was _ independent _ of the magnitude of the bare exchange between the impurities and the antiferromagnet .
results for many physical properties of this antiferromagnet were obtained .
we also showed how the above results could be extended to apply to d - wave superconductors .
for the case where momentum conservation prohibits a coupling of bosonic @xmath20 field to low energy fermionic bogoliubov quasiparticles , no changes are necessary : the theory for the insulating antiferromagnet applies quantitatively to the collective spin excitations of the superconductor .
the situation where there is an efficient coupling of @xmath20 to the bogoliubov quasiparticles is considered in appendix [ quasi ] : we show there that the structure of all the scaling forms remains unchanged , but there are quantitative changes to the numerical values of the scaling functions . as a guide to the reader who read section [ intro ] and skipped the body of the paper , we itemize our main results . for the case of a single impurity ,
our main results were as follows .
* the response to a uniform magnetic field is summarized in fig [ fig1 ] .
the @xmath101 expansion for the constant @xmath612 is in ( [ par7 ] ) : it is evident that the series can not be used directly for a quantitative estimate , and we have not attempted any resummation analysis .
the large-@xmath220 prediction for @xmath612 is in ( [ valc1 ] ) with @xmath180 . on the ordered side ,
the @xmath101 expansion for the constant @xmath426 is in ( [ n22 ] ) .
* local response functions are governed by the known bulk anomalous dimension , @xmath188 , and the new boundary anomalous dimension @xmath186 .
the @xmath101 expansion for the latter is in ( [ eps24 ] ) . * the local susceptibility , @xmath303 , measures the impurity site response to a field applied at the impurity site only .
the @xmath101 expansion for this is in ( [ par10],[par11 ] ) , while the large-@xmath220 result is in ( [ locn ] ) . *
the knight shift measures the localized response to a uniform magnetic field applied everywhere .
the knight shift very close to the impurity site is in ( [ par12],[par13 ] ) .
away from the impurity site , there is a response which oscillates with the orientation of the nel order parameter , as in fig .
[ stagg] the envelope of this oscillating component is given in ( [ par14],[par15 ] ) .
there is also a smaller uniform ( ferromagnetic ) component to the knight shift this is in ( [ par18a],[par18 ] ) .
* in the ordered nel state , the spontaneous moment on the impurity site obeys ( [ defm1],[n11 ] ) .
away from the impurity site , there is a staggered component obeying ( [ bulkmag ] ) , and a uniform component obeying ( [ n15 ] ) . for the system with a finite density of impurities , we obtained results for the magnon lineshape in the spin gap phase .
this line obeys the exact scaling form ( [ polescale ] ) , and nca results for the scaling function are in figs [ fig2 ] and [ fig3 ] , and in ( [ scba1],[scba2],[xyz ] ) .
the rest of this section discusses some remaining experimental issues .
one interesting observation of the neutron scattering experiments of ref was that zn doping caused the resonance peak to appear at larger temperatures significantly above the superconducting transition temperature .
we do not have a theory of the superconducting order in this paper , and so can not directly address this question .
however , this experimental result supports the interpretation that the resonance is very much like the gapped , spin-1 excitation in the paramagnetic phase of an antiferromagnet @xcite . in the `` swiss - cheese '' model of zn doping developed by uemura and collaborators @xcite ,
such antiferromagnetic correlations are enhanced in the vicinity of the impurity ( as in fig [ stagg ] and refs . ) , and should lead to increased stability of the resonance mode at higher temperatures .
recent measurements of the effective moment on the zn site in @xmath676 at fairly large temperatures yield a value @xmath677 which is smaller than the value expected for @xmath678 , @xmath82 , @xmath679 . while some of this reduction is surely due to inter - impurity interactions , we speculate that the crossover to the quantum - critical region is playing a role . in the latter regime , we expect @xmath680 ; using the value in ( [ valc1 ] ) we obtain @xmath681 . next , we discuss the relationship of our work to analyses of impurity effects in d - wave superconductors @xcite and a recent tunneling experiment@xcite .
the former have all been carried out using actions closely related to @xmath682 ( in ( [ spsi],[spsiimp ] ) .
however , we have argued in this paper that the coupling of the impurity spin to the collective , spin-1 , mode @xmath20 is a more relevant perturbation .
one consequence of such a coupling is that the impurity spin has structure in its spectral function at frequencies which are integer multiples of @xmath0 , as we found in section [ sec : n1 ] .
it is therefore natural to expect that some of this structure will appear also in the quasiparticle tunneling spectrum : it is interesting that a sideband at a frequency of order @xmath0 does appear to be present in the tunneling spectrum at the impurity site in ref . .
as we move away from the impurity site , the coupling between the quasiparticles and the @xmath20 mode becomes negligible , and the frequency @xmath0 should not be visible in tunneling this also appears to be consistent with ref . .
we note that this sideband phenomenon is similar in spirit to arguments made by abanov and chubukov @xcite for photoemission experiments @xcite . also , some related computations on the relationship of tunneling spectra to antiferromagnetic collective modes appear in a recent work by chubukov _
@xcite in closing , we mention some consequences of our theory to ni doping of two - dimensional spin - gap compounds of spin-1/2 cu ions
. each ni ion has spin-1 , and so the net uncompensated moment in its vicinity will still have @xmath683 .
our theory , therefore implies that the damping of the bulk @xmath20 quasiparticle mode due to the ni impurities will be the same as that due to non - magnetic zn impurities , because both depend only the bulk parameters and the values of @xmath55 ; all differences arise only from irrelevant couplings and these are suppressed by factors of @xmath118
. however , the atomic scale , staggered spin polarization will be very different in the two cases ; in particular , a large contribution to the net moment of @xmath112 comes from the magnetic moment on the ni ion itself , while no magnetic moment is present on the zn ions : this implies strong differences in the stm spectra .
we thank i. affleck , h. alloul , r. bulla , n. bulut , a. castro neto , a. chubukov , seamus davis , e. demler , h. fukuyama , a. georges , m. imada , m .- h .
julien , a. kapitulnik , b. keimer , c. ppin , a. sengupta , t. senthil , q. si , o. starykh , m. troyer , and j. zaanen for useful discussions .
we are grateful to m .- h .
julien for providing fig .
[ stagg ] .
this research was supported by us nsf grant no dmr 9623181 and by the dfg ( vo 794/1 - 1 ) .
this appendix will continue the discussion of section [ dwave ] on the consequences of the gapless , fermionic , bogoliubov excitations in a d - wave superconductor . as was noted in section [ dwave ] , for quasiparticles with vanishing energy at wavevectors @xmath139 , the key issue was whether the antiferromagnetic wavevector , @xmath146 , equaled @xmath684 or not . for @xmath148 , the bulk coupling between the antiferromagnetic order parameter , @xmath20 and the fermionic quasiparticles can be neglected at low enough energies , and this was assumed in the body of the paper . in this appendix
, we describe the changes necessary for @xmath685 .
the critical theory of an antiferromagnetic ordering transition in a d - wave superconductor with @xmath685 was described recently by balents , fisher and nayak @xcite . here
we will review their results in a field - theoretic approach designed for efficient treatment of the impurity problem .
first , we specify the fermionic action ( [ spsi ] ) more precisely . let the components of electron annihilation operator in the vicinity of the wavevectors @xmath686 , @xmath687 , @xmath688 , @xmath689 be @xmath690 , @xmath691 , @xmath692 , @xmath693 respectively , where @xmath694 is the electron spin component .
now introduce the 4-component nambu spinons @xmath695 and @xmath696 where @xmath697 is an antisymmetric tensor with @xmath698
. then the full form of ( [ spsi ] ) is @xmath699 here @xmath700 are pauli matrices which act in the fermionic particle - hole space , the wavevectors @xmath701 have been rotated by 45 degrees from the axes of the square lattice , and we have scaled the fermionic velocities to unity ( following ref . , we assume that the velocities in the two directions are equal near the critical point ) . with @xmath702 ,
the allowed bulk coupling between the antiferromagnetic order parameter @xmath20 and @xmath703 is @xcite @xmath704 where @xmath705 are pauli matrices in spin space .
the antiferromagnetic ordering transition in the d - wave superconductor is described by the action @xmath706 specified in ( [ sb ] ) , ( [ spsi2 ] ) , ( [ spsiphi ] ) .
we perform its renormalization group analysis in @xmath707 dimensions by the renormalizations ( [ eps8 ] ) and @xmath708 at one loop , the results of ref translate to the following renormalization constants in the minimal subtraction scheme ( ( [ eps10 ] ) no longer applies ) : @xmath709 we have taken the liberty of adding a single additional two - loop contribution above : the order @xmath710 term for @xmath80 which was also present in ( [ eps10 ] ) .
these translate into the beta functions ( replacing ( [ eps18 ] ) ) @xmath711 the stable fixed point of these equations is @xmath712 and the value of @xmath713 replaces that in ( [ eps22 ] ) .
the anomalous dimension in ( [ eps23 ] ) is replaced by @xmath714 where @xmath715 is anomalous dimension of the fermions @xmath703 .
these fixed point couplings can be used to compute the universal correlations of the antiferromagnetic order parameter . in particular , in the paramagnetic phase , the dynamic susceptibility at the antiferromagnetic wavevector obeys ( [ pole2 ] ) , where @xmath0 is an energy scale measuring the distance from the critical point to long - range antiferromagnetic order . to leading order in @xmath101
we find @xmath716 with a self - energy , @xmath717 , with imaginary part @xmath718 this is the main difference between the antiferromagnetic transition in an insulator and a d - wave superconductor with @xmath702 : the scaling structure in the two cases is identical , but the latter has damping in the @xmath20 collective mode of the paramagnetic phase even at @xmath83 .
as discussed in section [ qimp ] , we couple quantum impurities to the bulk theory of appendix [ hostd ] .
the fermionic couplings in @xmath160 ( eqn .
( [ spsiimp ] ) ) continue to be irrelevant , and so we discuss the theory with action @xmath719 specified in ( [ sb ] ) , ( [ simp ] ) , ( [ spsi2 ] ) , ( [ spsiphi ] ) .
the analysis is essentially identical to that in section [ rg ] .
at the one loop level , there are no new feynman diagrams , and only those in fig [ diag1 ] appear . moving to the two loop level
, we find the diagrams in fig [ diag4 ] , [ diag2 ] , [ diag3 ] , and also the single new category of diagrams in fig [ diag5 ] . in the theory of appendix [ quasi ] .
each propagator of @xmath20 ( represented by the dashed line ) in fig [ diag1 ] acquires a fermion loop ( represented by the dash - dot line ) self energy correction ; the filled square is the @xmath720 coupling . only the correction to one diagram of fig [ diag1 ]
is shown here ; corresponding corrections to the others are implied .
, width=240 ] evaluating these new diagrams , we find that at this order , the value of the boundary renormalization constant @xmath249 remains unchanged to that in ( [ eps15 ] ) , while the constant @xmath248 is modified from ( [ eps17 ] ) to @xmath721 where @xmath722 is euler s constant .
the renormalization constants in ( [ eps15],[hostd2],[hostd6 ] ) modify the beta function for the boundary coupling @xmath251 from ( [ eps21 ] ) to @xmath723 - 2 \lambda^2 \gamma^3 ( 1 + 2c - \ln 4 ) \nonumber \\ & ~ & + \vartheta \lambda^4 \gamma . \label{hostd7}\end{aligned}\ ] ] here @xmath724 is a numerical coefficient whose determination requires the order @xmath725 term in the bulk renormalization factor @xmath80 in ( [ hostd2 ] ) ; the latter has not yet been computed .
eqn ( [ hostd7 ] ) modifies the fixed point value @xmath94 from ( [ eps22 ] ) to @xmath726 armed with the fixed point values in ( [ hostd9],[hostd8 ] ) , we can now extend all the computations of section [ sec : eps ] to the theory of this appendix .
there are no changes in the structure of the exact scaling forms , but the scaling functions and exponents have to be recomputed in the present theory . in particular , the value of the boundary anomalous dimension @xmath186 changes from ( [ eps24 ] ) to @xmath727
the body of the paper considered antiferromagnets with a collinear spin polarization commensurate with the underlying lattice .
here we discuss the case of incommensurate spin correlations .
several classes of incommensurate magnets need to be distinguished .
if the incommensurate order is _ spiral _ , then the spin polarization is non - collinear , and a very different theory is necessary @xcite the quantum paramagnetic phases of such magnets possibly have deconfined excitations , and their coupling to the impurity spin will be quite different .
we will not consider such a situation here .
restricting attention to incommensurate collinear antiferromagnets , we have to further distinguish two cases .
let us describe the spin polarization in the ordered phase of such a magnet by @xmath728 where @xmath146 is the incommensurate ordering wavevector , and @xmath729 is a _
complex _ three - component field . associated with any such spin polarization , simple symmetry considerations @xcite
imply that there must be a corresponding charge density modulation @xmath730 where @xmath731 is an arbitrary phase offset .
the final situation depends upon the status of the charge density modulation at the quantum transition at which the magnetic order disappears .
if the magnetic order ( [ ic1 ] ) vanishes while the charge modulation in ( [ ic2 ] ) is preserved on both sides of the transition , then the theory in the body of the paper continues to apply to the spin excitations .
this is because we may view the spin modulation as ( in a sense ) commensurate with the underlying charge density wave in other words , the value of @xmath731 pins the phase of @xmath729 and the order parameter becomes effectively real . only in the case where the modulations in ( [ ic1 ] ) and ( [ ic2 ] ) vanish at the same quantum critical point
is a new theory necessary ( the remaining case , in which the order ( [ ic2 ] ) vanishes , while ( [ ic1 ] ) is preserved is not possible on general symmetry grounds ) .
we have to extend the theory @xmath85 from a real field @xmath89 to a complex field @xmath729
. we will not enter into this in any detail , but mention the changes necessary . in @xmath32 ,
in addition to the usual quartic term , @xmath732 , a second quartic term , @xmath733 is permitted ; in @xmath76 , the coupling @xmath79 becomes complex .
the phase of @xmath79 is a renormalization group invariant ( rather than just its sign ) , and its initial value plays a role similar to that of @xmath97 .
we describe a diagrammatic approach to expanding arbitrary correlators @xmath85 in powers of @xmath226 and @xmath52 . we will restrict ourselves to a single impurity here , with the notations ( [ single1 ] ) and ( [ single2 ] ) .
the generalization to the many impurity case is straightforward . the simplest way to generate
this is to use a hamiltonian representation of the quantum spin in terms of heisenberg spin operators ; so we replace @xmath734 where the @xmath735 satisfy the standard spin commutation relations @xmath736 = i \epsilon_{\alpha\beta\gamma } \hat{s}_{\gamma}. \label{aa2}\ ] ] now we represent the partition function as a time - ordered exponential of the interaction terms , expand the exponential in powers of @xmath226 and @xmath52 , and take the trace over the hilbert space of the spin operators .
closely related methods have been developed for the traditional kondo problem @xcite .
we can reduce the results of such an expansion to a simple set of diagrammatic rules .
it is important to remember that there is no automatic cancellation of disconnected diagrams in such an approach .
so when evaluating the expectation value @xmath737 of an arbitrary observable @xmath738 , @xmath739 we must separately expand the numerator and denominator consistently to each order in @xmath226 and @xmath52 .
the diagrammatic rules can be applied to both these expansions .
we represent the time evolution of the spin operator by a single , directed loop along which imaginary time runs periodically from 0 to @xmath227 .
this loop is shown as a full line , and there is one , and only one , loop in each diagram .
each factor of @xmath226 is associated with a filled circle on the loop at some time @xmath740 .
associated with each such filled circle we have : * a propagator for the @xmath20 field emerges from the filled circle ( represented by a dashed line ) , and these are contracted together in powers of @xmath52 by standard @xmath741 perturbation theory .
the @xmath52 interaction is represented by a wavy line . * a factor of the spin operator @xmath735these
are placed in order of the increasing @xmath740 and then the trace is taken over the spin hilbert space .
there are also factors of @xmath735 associated with external sources of @xmath742 in the operator @xmath738these are represented by open circles . *
all the times @xmath740 are integrated over the maximum possible interval between @xmath743 to @xmath227 subject to the constraint that they maintain their time ordering along the directed loop .
all of the diagrams contributing at order @xmath289 to the numerator and denominator of the two - point correlator of the @xmath89 are shown in fig [ diag1 ] . as an application of the rules above
, we see that the contribution of the last diagram in the numerator ( the one with the time labels shown ) is @xmath744 the trace over spin operators evaluates to @xmath745 \label{aa5}\ ] ] other diagrams can be evaluated similarly , and we obtain ( [ eps1 ] ) after the quotient has been taken . proceeding to order @xmath241 for the same correlator ,
the graphs shown in fig [ diag2 ] are added to the numerator , ) at order @xmath241 .
these have to be added to the diagrams in fig [ diag1]a .
they have been collected in groups of 5 for reasons described in appendix [ diag ] . ,
width=307 ] ) at order @xmath241 .
these have to be added to the diagrams in fig [ diag1]b .
, width=307 ] while those shown in fig [ diag3 ] add to the the denominator .
the new non - trivial spin traces which arise are @xmath746 \nonumber \\ & & \text{tr } \left ( \hat{s}_{\alpha } \hat{s}_{\beta } \hat{s}_{\gamma } \hat{s}_{\alpha } \hat{s}_{\beta } \hat{s}_{\gamma } \right ) = ( 2s+1)\nonumber \\ * & & \qquad \times s(s+1 ) [ s^2 ( s+1)^2 -3 s(s+1 ) + 2 ] .
\label{aa6}\end{aligned}\ ] ] we now take the quotient of the additional contributions from fig [ diag2 ] and [ diag3 ] . although this can be a laborious process , the work can be shortened by observing many cancellations diagrammatically .
we have collected the diagrams in fig [ diag2 ] in groups of 5 to aid this process ; all the diagrams within each group differ only by a time displacement on the @xmath226 interactions through the external vertices .
indeed , the union of the domains of the time integrals in each group precisely equals the domain of the time integral of a single diagram in fig [ diag3 ] . were it not for the differences between the traces of the spin operators in ( [ aa6 ] ) , these contributions to the numerator and the denominator would precisely cancel .
consequently , we need only keep track of such differences , and can then quickly write down the final quotient .
the result is the following contribution to the two - point @xmath89 correlator at order @xmath241 : @xmath747 d(\tau_3 - \tau_4 ) \label{aa7}\end{aligned}\ ] ] after carefully taking the @xmath83 limit of the above expression as described above ( [ eps4 ] ) , and evaluating the integrals using ( [ eps5 ] ) , we obtain the two - loop contribution in ( [ eps7 ] ) .
at @xmath83 , and @xmath223 , the two - point correlations of the renormalized @xmath748 field can be written down from ( [ eps1 ] ) : @xmath749 /z^{\prime } , \label{mimp1}\end{aligned}\ ] ] where have retained terms accurate only to order @xmath750 , and at @xmath83 , @xmath230 in ( [ eps2 ] ) evaluates to @xmath751 with @xmath752 a modified bessel function .
the @xmath753 limit of ( [ mimp2 ] ) is ( [ eps5 ] ) .
we first evaluate ( [ mimp1 ] ) at the critical point @xmath753 . performing the integral and using ( [ eps17 ] )
we have @xmath754\right ) \label{nren}\ ] ] to leading order in @xmath101- the poles in @xmath101 have cancelled . for large @xmath16 we can exponentiate this at the fixed point coupling value @xmath755 and find that @xmath756 which predicts @xmath757 to leading order in @xmath101 .
we now consider the @xmath758 limit of the correlations for @xmath759 : @xmath760\end{aligned}\ ] ] introducing new dimensionless coordinates @xmath761 ) , we write @xmath762\end{aligned}\ ] ] the relevant integral to be evaluated is ( with @xmath763 ) @xmath764 in the second line above we have used the asymptotic forms of the function @xmath765 for very small and large values of its argument @xcite , and @xmath766 is the incomplete gamma function .
now putting it all back together @xmath767\end{aligned}\ ] ] recalling the definition of @xmath768 ( [ eps17 ] ) , we conclude that @xmath769 \label{mimpsq}\end{aligned}\ ] ] near the fixed point we use @xmath755 and exponentiate to obtain from ( [ defm ] ) that @xmath770 , which agrees with ( [ defm1 ] ) and @xmath316 .
this appendix provides some details of the large-@xmath220 calculations presented in section [ sec : ncasingle ] .
first we prove that the chemical potential @xmath539 in the particle - hole symmetric case @xmath562 . from @xmath771 and @xmath772 ( for @xmath562 )
follows that @xmath773 is purely imaginary , @xmath774 , and the self - energy obeys @xmath775 .
therefore the chemical potential @xmath480 in the scaling limit is zero for any temperature @xmath18 and effective mass @xmath456 .
second we consider the impurity contributions @xmath599 to the uniform susceptibility .
the diagrams shown in fig .
[ figchidgr ] evaluate to : @xmath776 the prefactor @xmath430 arises from the summation over the su(@xmath220 ) indices together with the form of the field matrix @xmath593 defined below ( [ hfield ] ) .
@xmath777 and @xmath778 are abbreviations for the product of two and three boson green s functions respectively , together with the momentum dependence of the vertex ; note that the external frequency is zero here . due to the antisymmetric structure of the field coupling ( [ hfield ] ) both @xmath779 and @xmath780 enter the diagrams , with the result being @xmath781 \,,\nonumber
\\ g_{{{t}}}^{(3)}(i \nu_n ) & = & \frac{1}{n_s } \sum_{\bf k } { 2 { { j}}a_{\bf k } \over { { \epsilon}}_{\bf k } } \left [ g_{{{t}}}({\bf k},i \nu_n)^3 + g_{{{t}}}({\bf k},-i \nu_n)^3 \right ] \nonumber \:.\end{aligned}\ ] ] evaluation of the momentum integrals using the expression for @xmath493 in the scaling limit and performing the matsubara summations in ( [ chiimp1 ] ) one obtains the behavior quoted in ( [ chidgr1 ] ) , ( [ chidgr2 ] ) . in the gapped phase , @xmath615 , we expect from the confinement of the impurity spin that @xmath782 ( with the field conventions used in sec . [ ncasusc ] ) . in the low - temperature limit @xmath783 is dominated by @xmath604 , since @xmath784 remains finite for @xmath615 , and both @xmath600 and @xmath601 show a weaker divergence with @xmath543 .
we prove @xmath785 for @xmath615 .
we start with the observation that @xmath786 which holds at any temperature and independent of the dispersion of the bulk bosons .
it follows @xmath787 \:.\end{aligned}\ ] ] from the nca equations ( [ ncasigma ] ) , ( [ defsigma ] ) it is easily seen that the last term in the @xmath788 $ ] brackets equals @xmath789 in the scaling limit . with this
we can express @xmath604 in the scaling limit as @xmath790 the dominant contributions to the frequency sum ( for @xmath543 ) arise from low frequencies @xmath791 .
therefore it is safe to use @xmath792 ( which holds for small @xmath198 provided that @xmath615 ) .
in turn we have @xmath793 which implies @xmath794 j. bobroff , w. a. macfarlane , h. alloul , p. mendels , n. blanchard , g. collin , and j .- f .
marucco , phys .
lett . * 83 * , 4381 ( 1999 ) ; w. a. macfarlane , j. bobroff , h. alloul , p. mendels , n. blanchard , g. collin , and j .- f .
marucco , http://arxiv.org/abs/cond-mat/9912165[cond-mat/9912165 ] .
p. bourges in _ the gap symmetry and fluctuations in high temperature superconductors _ ed .
j. bok , g. deutscher , d. pavuna , and s. a. wolf ( plenum , new york , 1998 ) ; http://arxiv.org/abs/cond-mat/9901333[cond-mat/9901333 ] .
s. sachdev , _
http://pantheon.yale.edu/~subir/qptweb/toc.html[quantum phase transitions ] _ , cambridge university press , cambridge ( 1999 ) ; see also s. sachdev , http://pantheon.yale.edu/~subir/physworld/index.html[_physics world _ , * 12 * , no .
4 , pg . 33 ( 1999 ) ] .
m. p. gelfand , phys .
b * 53 * , 11309 ( 1996 ) ; y. matsushita , m. p. gelfand , and c. ishii , j. phys .
jpn * 66 * , 3648 ( 1997 ) ; m. p.
gelfand , zheng weihong , c. j. hamer , and j. oitmaa , phys . rev .
b * 57 * , 392 ( 1998 ) . however ,
even if @xmath795 , it is possible for @xmath20 to couple to ` massive ' fermions , @xmath796 , which are away from the nodes and have a finite excitation energy ; such a coupling can lead to a small irrelevant damping of the @xmath20 mode , when the latter also has a finite excitation energy ( we thank e. demler for pointing this out to us ) .
schematically , we have the cubic coupling @xmath797 ; after integrating out the massive fermions , this generates the quartic couplings like @xmath798 .
such quartic couplings are also permitted on general symmetry grounds and by momentum conservation however , it is easily shown by power - counting that they are irrelevant . c. ppin and p. a. lee , phys .
* 81 * , 2779 ( 1998 ) ; http://arxiv.org/abs/cond-mat/0002227[cond-mat/0002227 ] .
these papers consider @xmath31-wave superconductors with purely potential scattering and no independent spin degrees of freedom on the impurities ; however they succeeded in obtaining non - perturbative results for a finite density of impurities .
they show that in the unitarity limit of potential scattering , quasi - bound states form at the impurity sites at the fermi level .
we believe that after accounting for the strong local coulomb repulsion ( neglected by ppin and lee ) , each bound state will capture only a single electron , and the low energy physics will then be described by impurity spin models similar to those considered in our paper ; the residual impurity potential scattering in the resulting models will not be in the unitarity limit . | we present the universal theory of arbitrary , localized impurities in a confining paramagnetic state of two - dimensional antiferromagnets with global su(2 ) spin symmetry .
the energy gap of the host antiferromagnet to spin-1 excitations , @xmath0 , is assumed to be significantly smaller than a typical nearest neighbor exchange . in the absence of impurities
, it was argued in earlier work ( chubukov _ et al .
_ phys .
rev .
b * 49 * , 11919 ( 1994 ) ) that the low - temperature quantum dynamics is universally and completely determined by the values of @xmath0 and a spin - wave velocity @xmath1 .
here we establish the remarkable fact that _ no additional parameters _ are necessary for an antiferromagnet with a dilute concentration of impurities , @xmath2each impurity is completely characterized by a integer / half - odd - integer valued spin , @xmath3 , which measures the net uncompensated berry phase due to spin precession in its vicinity .
we compute the impurity - induced damping of the spin-1 collective mode of the antiferromagnet : the damping occurs on an energy scale @xmath4 , and we predict a universal , asymmetric lineshape for the collective mode peak .
we argue that , under suitable conditions , our results apply unchanged ( or in some cases , with minor modifications ) to d - wave superconductors , and compare them to recent neutron scattering experiments on @xmath5 by fong _
et al .
_ ( phys
. rev .
lett . *
82 * , 1939 ( 1999 ) ) .
we also describe the universal evolution of numerous measurable correlations as the host antiferromagnet undergoes a quantum phase transition to a nel ordered state .
= 1 |
ramsey s method of atom interferometry with separated oscillating fields @xcite provides the basis of present primary time standards , as cesium - beam or cesium - fountain setups @xcite .
basically , one aims to lock an oscillator exactly to a given atomic transition frequency to achieve stability and accuracy of the oscillator and thus of the clock .
the physical quantity that indicates possible deviations from the reference transition as a function of detuning is the excitation probability of the ground state atom after the interaction with two separated field pulses , and this function shows the well - known ramsey fringes .
note that as long as quantum reflections of the atom at the fields can be neglected , the operation of the interferometer in time domain ( temporally separated pulses and fixed atom ) or in space domain ( spatially separated fields and moving atom ) is equivalent for atoms moving along classical trajectories , as we shall exploit hereafter . a central requirement for frequency standards is a narrow interference pattern with respect to detuning , to allow for a precise lock of the oscillating fields to the atomic clock transition .
it has been shown by ramsey @xcite that the width of the central peak of the pattern is inversely proportional to the intermediate time between the two pulses @xmath0 , in contrast to single - field methods where it is inversely proportional to the field - crossing time . in the case of cesium - beam standards , for example , this implicates the desire of using tall `` fountain '' configurations , which of course have a limit in practice because of space constraints , magnetic field control , temperature homogeneity and other technical aspects @xcite ; or very slow ( ultracold ) atoms in reduced gravity @xcite . even though the dependence on the free flight time is far better than on field - crossing time because of the difficulty to implement a homogeneous and stable field
, the free flight occurs for atomic states with an excited component , which should be as stable as possible against radiative decay @xcite . in this paper
we report an effect that leads to a considerable narrowing of the ramsey interference fringes which is stable with respect to the atomic velocity or field mode functions , and relies on the free flight of ground state atoms before entering the first field region .
this effect is based on the use of two field pulses with _
different _ detuning . for simplicity of the presentation
, we neglect in our calculations the transverse momentum transfer on the atom , which is reasonable for moving atoms and microwave frequencies or for trapped ions or atoms in the lamb - dicke regime and optical fields @xcite . for a detailed study of these recoil effects in connection with ramsey interferometry we refer to refs . @xcite . for freely moving atoms interacting with optical fields ,
one is led normally to consider the multi - beam schemes proposed by kasevich and chu or by bord to compensate for the wavepacket separation due to recoil effects , see @xcite for reviews . however , to suppress transversal momentum transfer one may think on moving atoms in a narrow waveguide , interacting with optical fields @xcite .
in fact , for a waveguide width of @xmath1 and for cesium , the energy gap to the first transversely excited state is @xmath2 .
now , a minor modification of ref .
@xcite to incorporate detuning shows that excitation mainly occurs at the rabi resonances " @xmath3 , where @xmath4 is the rabi frequency and @xmath5 denotes the detuning between laser frequency and atomic transition frequency . for rabi frequencies of the order of @xmath6 one therefore would have a detuning range @xmath7 for which transversal excitation can be neglected .
we consider the basic ramsey setup where either a two - level atom in the ground state moves along the @xmath8 axis and crosses two separated oscillating fields localized between @xmath9 and @xmath10 and between @xmath11 and @xmath12 ( see fig .
[ fig : setup ] ) , or the atom is trapped and the two interaction pulses are separated in time .
in contrast to the standard setting , we allow for different detuning of the two fields with respect to the atomic transition frequency @xmath13 , respectively . the measured quantity is the transmission probability of excited atoms , @xmath14 , as a function of the detuning @xmath15 , @xmath16 , where @xmath17 is the frequency of the @xmath18th field .
the hamiltonian describing the moving atom reads in the dipole and rotating - wave approximation and in the schrdinger picture @xmath19 where @xmath20 , @xmath16 , are the rabi frequencies of the two spatially localized fields and the hat @xmath21 is used to distinguish operators from the corresponding c - numbers .
it is important in the following that the time origin has been chosen in a way that the two fields are in phase at @xmath22 . for trapped ions or atoms ,
the kinetic energy term would be absent and the rabi frequencies would be time - dependent , which will not alter the results within the following treatment . the crucial point to note is that for @xmath23 there is no interaction picture for which this time - dependent hamiltonian can be made time - independent , as it is the case for @xmath24 .
thus , the quantum mechanical probabilities , and in particular @xmath14 , will also be time - dependent . in the atom - adapted interaction picture
with @xmath25 one has @xmath26 in this interaction picture , the atom - field interaction is zero in between the fields and therefore we will favor it for the calculation of @xmath14 . [ [ remark ] ] remark : + + + + + + + one may also choose a field - adapted interaction picture with respect to the first or the second field by setting @xmath27 , which leads for @xmath28 to @xmath29 and for @xmath30 to @xmath31 in these interaction pictures the hamiltonian is either constant or periodic with period @xmath32 . for equal detuning , @xmath33 , these pictures would be time - independent and the standard result for @xmath14 would arise .
for fast enough particles , i.e. for kinetic energies @xmath34 much larger than @xmath35 and @xmath36 , the center - of - mass motion of the atom can be treated classically and independently of the internal dynamics . in that case
the two - component wavefunction @xmath37 which accounts for the internal dynamics in the interaction picture i1 is a solution of the internal schrdinger equation @xmath38 where @xmath39 and @xmath40 is the position of the atom at time @xmath22 . in the following , we consider the internal dynamics for a given kinetic energy @xmath41 , i.e. for a single atom whose center - of - mass follows the classical trajectory @xmath42 .
we denote by @xmath43 the time of the first interaction with the leftmost field , @xmath44 .
this treatment neglects initial uncertainties in position and momentum and is well justified for atoms in a trap and a pulsed experiment .
the case of a bunch of atoms with initial position and momentum distribution crossing to separated fields leads to a distribution of entrance times @xmath43 and it is considered later in sec .
[ sec : packet ] . for an initial internal state
@xmath45 the solution of eq .
( [ eq : se_internal ] ) is given in terms of the evolution operator @xmath46 , @xmath47 where @xmath46 fullfills @xmath48 to obtain analytical results , we first consider the case of mesa mode functions for the two fields , @xmath49 for @xmath50 and zero elsewhere and @xmath51 for @xmath52 and zero elsewhere . in sec .
[ sec : gaussmode ] we will show the stability of our results with respect to more realistic field modes . in the field free regions
the hamiltonian in this interaction picture is zero and thus the evolution operator is unity . within the @xmath18th field , the solution of eq .
( [ eq : udgl ] ) is given by @xmath53 - \frac{i\delta_j}{\omega'_j } \sin\left [ \frac{\omega'_j(t - t_0)}{2 } \right ] \right\ } & -\frac{i\omega}{\omega'_j } { \mathrm{e}}^{i\delta_j ( t+t_0)/2 } \sin\left [ \frac{\omega'_j(t - t_0)}{2 } \right ] \\ & \\ -\frac{i\omega}{\omega'_j } { \mathrm{e}}^{-i\delta_j ( t+t_0)/2 } \sin\left [ \frac{\omega'_j(t - t_0)}{2 } \right ] & { \mathrm{e}}^{-i\delta_j(t - t_0)/2 } \left\{\cos\left [ \frac{\omega'_j(t - t_0)}{2 } \right ] + \frac{i\delta_j}{\omega'_j } \sin\left [ \frac{\omega'_j(t - t_0)}{2 } \right ] \right\ } \end{pmatrix},\ ] ] where the effective rabi frequencies @xmath54 , @xmath16 , have been defined and @xmath55 , @xmath56 .
now assume that at time @xmath57 the ground state atom interacts with the first field for a time @xmath58 , evolves freely a time @xmath59 and finally interacts with the second field for another time @xmath60 .
thus , the final internal state is @xmath61 this yields for the probability of a transmitted excited state @xmath62 \frac{\omega}{\omega'_2 } \nonumber\\ & & + ~ { \mathrm{e}}^{-i(\delta_1 - \delta_2)(t_0+\tau)/2 } { \mathrm{e}}^{i\delta_2 t/2 } \sin\left(\frac{\omega'_1
\tau}{2}\right ) \nonumber\\ & & \times \left[\cos\left(\frac{\omega'_2 \tau}{2}\right ) + \frac{i \delta_2}{\omega'_2 } \sin\left(\frac{\omega'_2 \tau}{2}\right ) \right ] \frac{\omega}{\omega'_1 } \biggr|^2.\nonumber\\&&\end{aligned}\ ] ] note that @xmath63 is periodic in the initial time @xmath43 with period @xmath64 . as a check we consider the limiting case of equal detuning , @xmath33 , leading to @xmath65 ^ 2,\end{gathered}\ ] ] which is independent of @xmath43 and coincides with the well - known result obtained by ramsey @xcite .
another interesting case is that of equal modulus of detuning but opposite sign in both fields , @xmath66 \\ \times \left [ \cos^2\left(\frac{\omega ' \tau}{2 } \right ) + \frac{\delta^2}{\omega'^2 } \sin^2\left(\frac{\omega ' \tau}{2 } \right)\right],\end{gathered}\ ] ] which maximizes the effect of different field frequencies .
we plot @xmath67 in fig .
[ fig : p12contour ] as a function of @xmath68 and @xmath43 . remarkably , the interference pattern with respect to @xmath68 becomes narrower if the initial entrance time @xmath43 is increased .
this may appear astonishing at first sight since we expect periodicity in @xmath43 . as it is shown in fig .
[ fig : p12cuts]a , @xmath14 is indeed periodic in @xmath43 for fixed detuning , but with a detuning - dependent period @xmath69 , leading for increasing @xmath43 to a narrower ramsey pattern as a function of @xmath68 , see fig . [
fig : p12cuts]b .
an estimation for the width of the central fringe is obtained if one expands @xmath67 in a series around @xmath70 . assuming a @xmath71-pulse for the fields , @xmath72
, this gives @xmath73 ^ 2 \delta^2/4 + \mathcal{o}(\delta^3)$ ] , such that the first zeros of the pattern are approximately given by @xmath74 the central width is inversely proportional to the sum of the intermediate crossing time @xmath0 and the entrance time @xmath43 .
of course , we could change the time @xmath43 by changing the origin of time .
the important fact is that any such time shift would leave invariant the time
_ interval _ @xmath43 between the instant in which the fields are in phase ( @xmath22 in our time reference system so far ) and the entrance of the atom .
note the roles of @xmath43 and @xmath0 in eq .
( [ width ] ) : first of all , @xmath43 is twice as efficient as @xmath0 to produce a desired width ; moreover @xmath43 it is a time for free flight of ground state atoms , whereas @xmath0 is a free flight time for atoms with excited components , which are amenable of decay .
= 8.8 cm = 8.8 cm an alternative viewpoint of the effect can be given by analyzing the phase mismatch between the two oscillating fields .
this is better suited to understand the experiment with trapped ions and pulsed fields .
let us consider a more general version of the hamiltonian ( [ eq : hscl ] ) where the phases of the fields at @xmath22 are given by @xmath75 and @xmath76 , respectively , @xmath77 an equivalent derivation as above , but for an entrance time @xmath78 , yields for this hamiltonian @xmath79 \\
\times \left [ \cos^2\left(\frac{\omega ' \tau}{2 } \right ) + \frac{\delta^2}{\omega'^2 } \sin^2\left(\frac{\omega ' \tau}{2 } \right)\right].\end{gathered}\ ] ] this agrees with eq .
( [ eq : p12_minpl ] ) if one sets @xmath80 thus , the effect of different entrance times @xmath43 is equivalent to a fixed entrance time @xmath78 but with an initial phase difference @xmath81 of the fields .
in particular , this means that in order to obtain one of the curves shown in fig .
[ fig : p12cuts]b one has to change for every value of @xmath68 this initial phase difference to fullfill eq .
( [ eq : phase_condition ] ) . in practice
, it might be difficult to hold @xmath43 stable due to phase fluctuations of the fields .
however , this will not affect the central fringe as we will show in the section [ sec : packet ] .
up to now , the monochromatic case with a fixed entrance time @xmath43 has been considered .
this holds true for the pulsed experiment and if the micromotion of the atom within the trap can be neglected .
however , if the incoming ensemble of moving atoms is described by an initial position and momentum distribution , the entrance and crossing times for the individual atoms will be different . in this case one
has to integrate @xmath14 over all possible classical trajectories , weighted by a phase space distribution @xmath82 .
as before we assume free motion for the center - of - mass , unperturbed by the fields . in the following ,
we restrict our analysis to the case @xmath83 and we assume @xmath82 to describe a minimum uncertainty packet when its center impinges the origin @xmath84 at time @xmath85 .
thus , @xmath82 is given by @xmath86 \\
\times \exp\left\{-\frac{[x-\hbar k ( t - t^c_0)/m]^2}{2(\delta x)^2}\right\},\end{gathered}\ ] ] where @xmath87 and @xmath88 are the uncertainties of position and momentum , connected by @xmath89 and @xmath90 is the mean momentum . now in eq .
( [ eq : p12_general ] ) one has to replace @xmath60 by @xmath91 , @xmath0 by @xmath92 and @xmath43 by a varying entrance time @xmath93 , where @xmath8 and @xmath94 are distributed according to eq .
( [ eq : wigner ] ) . finally , one obtains @xmath95\right\ } \nonumber\\ & & \quad\quad\quad\quad \times \left [ \cos^2 \left(\frac{m\omega ' l}{2\hbar k}\right ) + \frac{\delta^2}{\omega'^2}\sin^2 \left(\frac{m\omega ' l}{2\hbar k}\right ) \right],\end{aligned}\ ] ] where @xmath96 .
if one translates for every fixed velocity @xmath97 the position width @xmath87 into an uncertainty of the entrance times , @xmath98 , one sees from the exponential in the integrand of eq .
( [ eq : p12int2 ] ) that for @xmath99 much larger than the period @xmath100 the dependence on @xmath44 and thus the effect of narrowing disappears , whereas for a sudden entrance , @xmath101 , the integrand coincides with eq .
( [ eq : p12_minpl ] ) .
the @xmath94-integration has to be performed numerically and the result is shown in fig .
[ fig : p12_int ] as a function of @xmath68 for fixed values of @xmath102 .
one clearly sees the ramsey pedestal due to the averaging of the outer fringes , but the central fringe is not affected , it remains stable and becomes narrower for increasing @xmath102 .
to investigate the stability of the narrowing effect with respect to the field mode function we withdraw the assumption of a mesa - shaped field within this section . as a more realistic example we consider a shape of the two field pulses which is very much reminiscent of a gaussian but limited to a finite duration , @xmath103}(t)\\ \omega_2(t ) & = & \displaystyle \omega \sin^4\left(\frac{\pi ( t - t_0-\tau - t)}{\tau}\right)\nonumber\\ & & \qquad\qquad\qquad\times \chi_{[t_0+\tau + t , t_0 + 2\tau+t]}(t),\end{aligned}\ ] ] where @xmath104}(t)=1 $ ] for @xmath105 and zero elsewhere . these field shapes
have also been used for stirap calculations @xcite . again , the evolution operator between the two pulses is unity , whereas it has to be determined numerically by means of eq .
( [ eq : udgl ] ) in the field regions , where now the constant @xmath4 has to be replaced by @xmath106 or @xmath107 , respectively .
the probability of excitation is calculated by means of eqns .
( [ eq : final_internal ] ) and ( [ eq : p12_general ] ) and the result is shown in fig .
[ fig : p12gaussmode ] .
irrespective of the sinusoidal mode , the effect of narrowing with respect to an increasing @xmath43 remains stable .
since the central peak is located at @xmath70 for all values of @xmath43 , a phase - space integration as in section [ sec : packet ] will not change the central fringes and the result would be qualitatively similar as it has been shown in fig . [
fig : p12_int ] for the case of mesa mode functions .
in this paper we have studied the interference fringes in a ramsey interferometer , for the case that the separated fields have different detuning .
the excitation probability @xmath63 , derived within a semiclassical picture neglecting quantum reflections at the fields , depends on the entrance time @xmath43 of atom at the first field , or , more precisely , on the phase difference of the two fields at this time .
our main result is that the width of the ramsey fringes decreases for increasing @xmath43 .
moreover , we have shown that this effect remains stable for atomic clouds with an initial momentum and position distribution and for sinusoidal field mode functions .
a similar effect is achieved in a time domain configuration of atoms at rest by controlling the laser field phases .
it would be interesting to investigate the ramsey fringes for differently detuned fields beyond the semiclassical approximation , i.e. taking into account quantum reflections at the fields for very slow ( ultracold ) atoms .
this has been shown to yield interesting effects in the case of equal detuning @xcite .
however , the quantum treatment of the time - dependent problem leads to difficulties in the calculation of the matching conditions between regions of constant potential and time - periodic potential .
in fact , one has to introduce energy sidebands and to match them individually , as it has been done in the case of a one - channel oscillating barrier to model quantum traversal times @xcite .
the solution of the two - channel case with two separated oscillating fields in an open problem so far .
the authors thank d. guery - odelin for helpful discussion . this work has been supported by ministerio de educacin y ciencia ( bfm2003 - 01003 ) and upv - ehu ( 00039.310 - 15968/2004 ) .
d.s . acknowledges a fellowship within the postdoc - programme of the german academic exchange service ( daad ) .
bord , c. r. acad .
paris srie iv * 2 * , 509 ( 2001 ) ; p. wolf _
et al . _ , recoil effects in microwave atomic frequency standards : an update , proc .
6th symp . on frequency standards and metrology , gill p. ed . , p. 593
, ( world scientific , 2002 ) ; c.j .
bord , metrologia * 39 * , 435 ( 2002 ) . | we report a narrowing of the interference pattern obtained in an atomic ramsey interferometer if the two separated fields have different frequency and their phase difference is controlled .
the width of the ramsey fringes depends inversely on the free flight time of ground state atoms before entering the first field region in addition to the time between the fields .
the effect is stable also for atomic wavepackets with initial position and momentum distributions and for realistic mode functions . |
in recent years the instanton liquid model has been very successful in describing the low energy properties of light hadrons @xcite . on the other hand
, it seems quite unlikely that instantons can account for confinement @xcite . this is certainly the case if the instanton liquid is sufficiently dilute and the instanton size distribution falls off rapidly enough for large instantons . if the fall - off is slow , e.g. as @xmath0 or @xmath1 , a linear term in the heavy quark potential due to instantons has been claimed @xcite , although these results are controversial . still , a proper understanding of the tail of the instanton size distribution may have physical significance .
unfortunately very little is known about it .
lattice simulations are typically done on too small volumes and have too poor statistics to contain any precise information on the tail of the distribution . moreover , even the question of what non - perturbative mechanism suppresses large instantons is still unanswered , although some recent attempts have been made @xcite . in this paper
we would like to make some remarks on these issues by studying the most general charge two instanton solution .
our observations will be about the ( possible mis)interpretations of the lattice data . in lattice determinations
@xcite of the instanton liquid parameters one usually starts with the assumption that the liquid is dilute enough , i.e. the individual pseudoparticles are far enough apart that they do not distort one another considerably .
only then does the `` instanton size '' have an unambiguous meaning . treating the @xmath2 case
exactly can be thought of as the next order approximation when one takes into account the distorting effect of like charge nearest neighbour pairs . to fix the notation
, we start with briefly summarising the adhm construction @xcite of the general @xmath2 solution @xcite .
then we discuss the ambiguities arising in the identification between the `` physical '' parameters and the ones appearing in the adhm construction , giving rise to two dual descriptions of the same physical configuration .
this duality maps large separation between the constituents to small separation @xcite and depends non - trivially on the relative gauge orientation of the two instantons .
it has two important consequences for the identification of instantons from the charge density profile , most clearly seen in the two extreme cases .
namely parallel and perpendicular relative gauge orientation ( the su(2 ) invariant angle being 0 and @xmath3 respectively ) .
if the orientation is perpendicular , the two instantons can not be closer to each other than a minimal distance @xmath4 set by their sizes @xmath5 .
if the relative orientation is parallel and the two constituents have the same size @xmath6 , when they get close to each other , the charge density looks like the super - imposition of a small instanton of size proportional to the separation , plus another instanton of size @xmath7 right under the small one .
if the relative orientation and sizes are not fine tuned , some combination of the above effects will take place .
the first effect controls how close two instantons can get to each other
. the second one can potentially hide large instantons from instanton finding algorithms based on the diluteness assumption and thereby it can significantly distort the instanton size distribution . in the last part of the paper we present quantitative data on how these effects are manifested in realistic lattice situations .
this is done by generating the charge density of instanton pairs , as given by the exact @xmath2 solutions , and comparing the size distribution thus prescribed by the original data with that found by lattice instanton finder algorithms .
we observe that the results depend very strongly on the relative orientation in group space , yielding a strong suppression of large instantons for parallel orientation .
the fundamental objects describing the most general su(2 ) charge @xmath8 instanton solution in the adhm construction are the @xmath8-dimensional row vector @xmath9 and the @xmath10 symmetric matrix @xmath11 , both having quaternionic elements @xcite .
( a quaternion @xmath12 can be parametrised with 4 real numbers as @xmath13 , where @xmath14 are the pauli matrices . )
the adhm data can be conveniently summarised in a single @xmath15 quaternionic matrix @xmath16 which in the @xmath2 case can be parametrised as @xmath17 @xmath16 describes a charge two self dual solution of the yang - mills equations if and only if the matrix @xmath16 satisfies the adhm constraint that @xmath18 be real quaternionic ( i.e. proportional to @xmath19 ) and invertible .
here @xmath20 , @xmath21 , @xmath22 , and @xmath23 are quaternionic parameters , and @xmath24 denotes a space - time position . in terms of our parametrisation of the charge two case , the adhm constraint reads as @xmath25 where the symbol @xmath26 is introduced to simplify the notation .
this equation does not specify @xmath23 unambiguously ; the most general solution can be written as @xcite @xmath27 where @xmath28 is an arbitrary real constant .
at this point it is instructive to compare this most general adhm ansatz with the special case of the t hooft ansatz @xcite in order to identify the physical parameters .
this special case is obtained for parallel gauge orientations , @xmath29 and @xmath30 real and @xmath31 , which solves the adhm constraint , eq .
( [ eq : lambda ] ) .
we can identify @xmath32 and @xmath33 as the scale parameters ( sizes ) , and @xmath34 as the locations of the two instantons .
the action density of the solution , valid for all choices of parameters @xcite , @xmath35 indeed agrees in this case with the action density of the t hooft solution . the most general charge two adhm solution can be described by the following set of free parameters : @xmath36 , the scale parameters ; latexmath:[$\lambda_1^\dagger \lambda_2/(|\lambda_1| @xmath34 the location of the constituents .
this gives a total number of 13 real parameters , in agreement with the general result that the charge @xmath8 solution has @xmath38 parameters .
for @xmath2 the conformal generalisation of the t hooft ansatz @xcite also has 13 parameters , which can be related to the adhm parametrisation @xcite .
the charge @xmath8 adhm ansatz is well known to have an @xmath39 symmetry acting on the parameters as @xmath40 in the @xmath2 case , using our parametrisation , the @xmath41 symmetry amounts to @xmath42 while @xmath21 is left unchanged .
the @xmath43 transformation that extends this to the full @xmath44 symmetry is generated by @xmath45 , which interchanges @xmath46 with @xmath47 and changes the sign of @xmath22 .
different sets of parameters related by this @xmath44 symmetry describe the same physical solution @xcite @xmath48}f(x ) \lambda^\dagger\right),\quad f^{-1}(x)\equiv\delta^\dagger(x)\delta(x),\ ] ] where @xmath49 . at first sight
this seems to imply that we have one less free parameter .
this is , however , compensated by the fact that , in general , solutions to the adhm constraint with different values of @xmath28 ( eq . ( [ eq : u ] ) )
result in physically different gauge field configurations , as one can easily be convinced of by computing @xmath50 .
ideally one would like to fix this ambiguity together with the @xmath44 symmetry to have a one - to - one correspondence between the gauge inequivalent solutions and the 13 parameters describing them .
our choice of these parameters , which we call `` physical '' , should be as close as possible to a superposition of two instantons .
looking at the situation of large separations ( @xmath51 large ) , where the relative gauge orientation does not play a role , the action density should be the sum of two instantons of sizes @xmath52 located at @xmath34 .
this imposes @xmath53 , or equivalently @xmath54 .
( the equivalence of the two conditions can be easily proved using eq .
( [ eq : lambda ] ) and @xmath55 which holds for any two quaternions @xmath56 and @xmath57 . ) choosing this particular solution to the constraint ensures that when the orientation of the constituents is parallel , the adhm solution coincides with the t hooft ansatz . with this convention
the identification of the `` physical '' parameters becomes almost unique .
the prescription is that from each @xmath44 orbit we have to choose the point that satisfies @xmath58 . generally or @xmath59 .
] there are 16 such points on an @xmath44 orbit describing the same gauge field configuration , as has been noted before @xcite .
they are generated by rotations ( see eq .
( [ eq : o(2 ) ] ) ) over multiples of @xmath60 and the @xmath43 transformation , @xmath61 , @xmath62 .
most of these do not affect the `` physical '' interpretation .
but one ambiguity remains for which we choose the representative described by the @xmath43 symmetry ( @xmath45 ) combined with the rotation over @xmath63 , providing the relation @xmath64 we note that if the distance of the two instantons is @xmath65 in one description , it is proportional to @xmath66 in the `` dual '' description , as long as the relative gauge orientation is not parallel ( @xmath67 ) .
going beyond the issue of finding a unique parametrisation , the question now arises which of these two descriptions is the `` physical '' one . to answer this
it is instructive to look at the charge density profile of a set of solutions with varying separations , keeping the other parameters fixed . in fig .
[ fig : scatter ] we show such a sequence .
the parametrisation listed is in terms of the l.h.s . of eq .
( [ eq : duality ] ) .
the scale parameters and relative orientation are described by @xmath68 , @xmath69 , and the separation between the constituents is _ always _ along the 0-axis .
when they are far apart ( a ) , corresponding to @xmath70 , the solution indeed looks like a pair of instantons along the 0-axis . as the separation decreases
the two lumps merge together into an asymmetric ring ( b - c ) ( @xmath71 and 10.5 respectively ) .
for even smaller separation : @xmath72 ,
( d ) in the figure , the two lumps separate again but now displaced along the 1-axis . clearly , our parametrisation is not `` physical '' any more at this stage ; instead of two very close lumps separated along the 0-axis , we see two lumps farther apart but along the perpendicular 1-axis . on the other hand
, we also have the dual parametrisation ( r.h.s .
( [ eq : duality ] ) ) at our disposal .
indeed , a short computation shows that in the dual description of ( d ) we have two instantons of the same scale parameter @xmath73 separated along the 1-axis at a distance of 22 , evidently the correct `` physical '' description .
the general picture emerging from this exercise is quite clear .
when @xmath51 is large , more precisely @xmath74 , the original description is `` physical '' , i.e. describing two superposed instantons separated by a distance @xmath65 .
when @xmath51 is small , however , the dual description is the more `` physical '' one .
there is an intermediate region where the solution can not be approximated by a combination of two instantons , here the question which parameter set to use for the physical description is ill defined . in the rest of this section
we discuss two simple consequences of this dual choice of parametrisation .
the first one is that two instantons , as identified from maxima in the action density , can never get closer to each other than a minimal distance @xmath75 where @xmath76 is the invariant angle of the relative group orientation .
@xmath77 is defined by the property that the two dual descriptions have the same separation @xmath78 between the constituents .
if @xmath51 is chosen to be smaller than this , the two instantons `` scatter off '' in the @xmath79 direction and one has to switch to the dual description .
it is also interesting to observe the very special case when @xmath80 and @xmath81 . in this case
, @xmath82 gives a self - dual point where the two parametrisations completely coincide and the charge density has an axial symmetry in the @xmath83 plane .
for these parameters the density profile is ring - like , not unlike the case of monopoles @xcite , but apparently not noted before for instantons . in particular , the solution does not degenerate to one with topological charge @xmath84 , as conjectured in ref .
this _ only _ happens in the case of parallel gauge orientation , for @xmath85 and @xmath86 ( see eq .
( [ eq : duality1 ] ) below ) .
the discussion so far applies only to the @xmath67 case .
moreover , if @xmath87 , then the switching between the two parametrisations occurs at very small @xmath51 ( see eq.([eq : zmin ] ) ) and other interesting effects might come into play when the two constituents are very close .
let us now look at the extreme situation when @xmath88 , the relative orientation is parallel and consequently @xmath89 . in this case
the two parametrisations are @xmath90 in the limit @xmath91 the two descriptions become equivalent , there is no way to choose between them .
one can see that two instantons of scale parameters @xmath52 ( @xmath92 ) on top of each other is equivalent to a small instanton of size @xmath93 on top of a larger one of size @xmath94 .
in this extreme case any lattice instanton finder based on the diluteness assumption would find only the former smaller instanton and nothing else .
moreover due to the strongly non - linear character of the yang - mills equations , when two instantons come close with parallel orientation , both peaks become narrower and sharper and are thus identified by any lattice algorithm as `` small '' instantons .
at best one misses one large instanton in the background ; at worst one misinterprets two large instantons as two small ones .
the effect is similar at @xmath87 and can provide a mechanism to hide large instantons as some smooth background that remains unnoticed by lattice instanton finder algorithms .
we have seen how the identification of single instanton parameters becomes ambiguous when instantons overlap .
we discussed what happens in the two extreme cases of parallel and perpendicular orientation in group space . in between , some combination of the two above described effects takes place .
we expect that the exact way this affects the instanton size distributions measured on the lattice will depend on the relative orientation of nearest neighbour pairs . in the remainder of the paper we study numerically how the relative orientation can affect the lattice instanton size distribution .
in order to see the trends as clearly as possible , instead of trying to mimic the `` real '' distribution of relative orientations , we use two simple orientation distributions : the haar measure , and all instantons taken parallel . due to the @xmath95 factor , the haar measure very strongly favours ( close to ) perpendicular orientation , thus our two distributions almost represent the two possible extremes . we generated the charge density of a set of instanton pairs with the adhm construction using eq .
( [ eq : charge ] ) .
the parameters of the adhm ansatz were taken as follows .
the instanton scale parameters @xmath96 were distributed independently and qualitatively similar to that found on the lattice , except for an enhanced tail ( for @xmath6 large ) .
we artificially enhanced the tail of the distribution in order to test whether such a tail can remain undetected by the lattice instanton finders .
the separation @xmath65 was gaussian distributed with mean 7.0 , and variance of 1.0
. the resulting charge densities each pair resolved on a @xmath97 grid were then analysed using two different instanton finder algorithms @xcite .
the details of these algorithms are not relevant in the present context .
however their most important common feature is that they are both based on the dilute gas assumption .
they identify the highest peaks in the charge density and estimate the instanton sizes from the fall - off of the density in the vicinity of the maximum . in fig .
[ fig : haar_all ] we show the instanton size distributions found by the algorithms of refs .
@xcite ( dotted line ) and @xcite ( dashed line ) along with the distribution of the adhm size parameters used to construct the charge densities ( solid line ) , representing the physical choice of parameters .
the haar measure was used for the gauge orientation .
one of the instanton finders seems to somewhat suppress the enhanced tail while this effect is not significant with the other . in fig .
[ fig : par_all ] we plotted the size distributions obtained when all the pairs were taken parallely oriented in group space .
all the conventions are the same as in fig .
[ fig : haar_all ] . here
the two instanton finders both yield a significantly suppressed tail .
we also considered ensembles ( not shown ) which in terms of the relative orientation are in between the two shown .
we can draw the following general conclusion .
the two instanton finders display the same trend ; the more parallel the pair is in group space the more suppressed the large instantons become .
this is partly due to the effect that large instantons can `` hide '' under small ones that produce sharper peaks in the charge density .
the other effect is that , when two instantons come close with approximately parallel orientation , they look narrower and are thus interpreted by the lattice algorithm as `` small '' instantons
to summarise , we studied the question of what happens when the instanton liquid is not dilute enough to be considered as a collection of individual pseudoparticles .
the next approximation is to treat nearest pairs of like charge exactly .
we established a correspondence between the parameters of individual instantons and the parameters of the charge two adhm construction .
after fixing most of the ambiguities of this correspondence we still have two dual sets of parameters describing the same charge two instanton solution .
this duality turned out to imply the existence of a minimal distance between the two instantons , which is maximal in the case of perpendicular orientation . in the other extreme case of ( nearly ) parallel orientation , we found that instanton finders based on the diluteness assumption can have a tendency to miss large instantons or to underestimate instanton sizes .
we numerically confirmed that this indeed happens for realistic parameters of the instanton liquid and the effect becomes larger for orientations closer to being parallel .
in fact , a recent lattice study indicates that like charge pairs strongly favour parallel orientation compared to the haar measure @xcite , therefore the effect we found can be potentially important .
we thank jan de boer and thomas kraan for discussions .
this work was supported in part by a grant from `` stichting nationale computer faciliteiten ( ncf ) '' for use of the cray y - mp c90 at sara .
t. kovcs was supported by fom and m. garca prez by cicyt under grant aen97 - 1678 .
d. diakonov and v. petrov , `` confinement from instantons ? '' , in : _ non - perturbative approaches to quantum chromodynamics _ , proceedings of the ect@xmath98 workshop , trento , july 10 - 29 , 1995 , ed .
d. diakonov , gatchina , 1995 , p. 239 ; m. fukushima , h. suganuma , and h. toki , phys .
* d60 * ( 1999 ) 094504 ; r.c .
brower , d. chen , j.w .
negele , and e. shuryak , nucl .
b ( proc . suppl . )
* 73 * ( 1999 ) 512 . | by studying the non - linear effects of overlapping instanton pairs we address difficulties in the identification of instanton distributions when the average instanton size is comparable to the average distance .
for the exact charge two solution , we study how its parametrisation relates to a description in terms of individual instantons .
there exist two dual sets of parameters describing the same charge two instanton solution .
this duality implies the existence of a minimal separation between two instantons .
conventionally used lattice instanton finder algorithms based on the assumption of diluteness tend to underestimate instanton sizes .
finally we numerically confirm this for realistic parameters of the instanton liquid .
the effect is enhanced by parallel orientation in group space . |
in a heavy fermion ( hf ) kondo lattice system , the ground state is a fermi - liquid ( fl ) state constituting the landau quasi - particles . in ce- , yb- , and u - based intermetallic systems the conduction electrons compensate or screen the localized moments of the @xmath8-electrons where localized electrons together with their screening cloud form quasi - particles .
these quasi - particles have heavy masses , reflected in an enhanced value of the sommerfeld coefficient , @xmath9 = @xmath10 , at low temperatures .
when @xmath8-electrons so strongly couple with conduction band there is an increased overlap of the electronic state which enhances the hybridization and band widths @xcite .
the fl state in hf kondo lattice systems shows strong correlations among physical quantities .
one such correlation is the kadowaki - woods ( k - w ) ratio , a relation between the electrical resistivity ( @xmath11=@xmath12 ) and specific heat ( @xmath13=@xmath14 ) @xcite , given by the universal ratio @xmath15 = 1.0@xmath1610@xmath17 @xmath18cm/(mj / mol@xmath19k)@xmath20 . recently ,
systematic deviations of the k - w ratio in many hf systems ( especially for yb - based compounds ) have been explained by tsujii @xmath21 @xmath22 . , taking into account the ground state degeneracy ( @xmath23=2@xmath24 + 1 ) @xcite .
a fl state is also characterized by the wilson ratio ( @xmath25 ) which links @xmath9 to the pauli susceptibility @xmath26 @xcite , which is given by @xmath25 = @xmath27 , where @xmath28 , @xmath29 , and @xmath30 are the boltzman constant , lande s factor , and bohr magneton , respectively @xcite .
in addition to the @xmath25 and k - w ratio , the zero temperature limit of the thermoelectric power ( tep ) , @xmath31=@xmath6 , for several correlated materials has shown a strong correlation with @xmath9 via the dimensionless ratio @xmath32=@xmath33 = @xmath34 @xcite , where @xmath35 is the avogadro number and @xmath36 is the carrier charge . for yb - based hf systems ,
the electrical resistivity and tep reveal complex temperature dependencies with a local extrema . in general
, these extrema are related to kondo scattering associated with the ground state and excited states of the cef energy levels @xcite .
the characteristic temperature of the local maximum shown in @xmath37 and the local minimum developed in @xmath0 allow for an estimate of the kondo temperature , @xmath38 , and the crystalline electric field ( cef ) splitting , @xmath39 , as relevant energy scales in yb - based hf systems . in this paper ,
tep measurements on ybt@xmath1zn@xmath2 ( t = fe , ru , os , ir , rh , and co ) are presented as functions of temperature and magnetic field to evaluate the correlation between specific heat and tep in the zero temperature limit .
these compounds crystallize in the cubic cecr@xmath1al@xmath2-type structure ( @xmath40 , no.227 ) @xcite and have been reported to be hf metals with no long range order down to 20mk @xcite . in the fl regime it has been shown that the @xmath25 and k - w ratios in this family follow the theoretical predictions with different ground state degeneracies .
the tep data of yt@xmath1zn@xmath2 ( t = fe , co ) are also presented for comparison .
yfe@xmath1zn@xmath2 is one of the examples of a nearly ferromagnetic fermi liquids ( nffl ) with a highly enhanced magnetic susceptibility value at low temperatures @xcite , whereas yco@xmath1zn@xmath2 shows normal metallic behavior .
single crystals of ybt@xmath1zn@xmath2 ( t = fe , ru , os , ir , rh , and co ) and isostructural yt@xmath1zn@xmath2 ( t = fe and co ) were grown out of excess zn @xcite using standard solution growth techniques @xcite .
the tep was measured using a dc , alternating heating , technique that utilizes two heaters and two thermometers @xcite .
a quantum design physical property measurement system provided the temperature ( from 2 to 300k ) and magnetic field ( up to 140koe ) environment . for t
= fe , rh , and co , zero - field tep measurements were extended down to 0.4k , measured using the same technique @xcite , in a cryo industries of america @xmath41he system .
the heat current was generated in the ( 111)-plane of the samples ( @xmath42@xmath3@xmath43(111 ) ) and the magnetic field was applied along the [ 111]-direction maintaining a transverse configuration , @xmath44@xmath45*h*.
figure [ zn20yst ] shows the tep data for yfe@xmath1zn@xmath2 and yco@xmath1zn@xmath2 .
the temperature - dependent tep , @xmath0 , of these compounds is similar to normal metallic systems . at 300k , @xmath0 of both compounds is positive and has an absolute value of @xmath469@xmath4v / k for yfe@xmath1zn@xmath2 and @xmath465@xmath4v / k for yco@xmath1zn@xmath2 , and then decrease monotonically to below 25k with decreasing temperature . with further cooling ,
@xmath0 of yco@xmath1zn@xmath2 passes through a broad peak ( @xmath47/12 @xcite , where @xmath48 is the debye temperature ) expected to be due to phonon - drag @xcite . on the other hand , @xmath0 of yfe@xmath1zn@xmath2 shows a local minimum around 14k ( @xmath47/23 @xcite ) that is not currently understood .
the absolute value of the tep for yfe@xmath1zn@xmath2 is much smaller than other nffl systems .
a signature of the spin fluctuation temperature , @xmath49 , has been inferred from a shoulder in @xmath50fe@xmath51sb@xmath52 ( @xmath50 = ca , sr , and ba ) data @xcite and as a minimum developed in @xmath53co@xmath1 ( @xmath53 = y , sc , and lu ) data @xcite .
the minimum developed near 14k may be related to the signature of spin fluctuation , combined with phonon - drag in the yfe@xmath1zn@xmath2 system . in the @xmath540k limit , @xmath31 of yfe@xmath1zn@xmath2 is larger than that of yco@xmath1zn@xmath2 as shown in the inset of fig .
[ zn20yst ] . in zero field ,
the @xmath0 data of the ybt@xmath1zn@xmath2 ( t = fe , ru , os , ir , rh , and co ) compounds are plotted in fig .
[ zn20st1 ] .
in contrast to the isostructural y - based compounds , @xmath0 of the yb - based compounds exhibits a large , negative minimum ( between -70 and -40 @xmath4v / k ) and the sign of @xmath0 changing above 150k from negative to positive ( not observed in this temperature range for t = ir ) .
the absolute tep values of yb - based compounds are much larger than y - based compounds at low temperatures , while they have a similar order of magnitude compared to y - based compounds around 300k . a negative
, highly enhanced value of the tep , over the temperature region measured , is typical of those found in other yb - based kondo lattice systems @xcite .
figure [ zn20st2 ] shows the low temperature @xmath0 of ybt@xmath1zn@xmath2 . for t = fe and ru
, a broad minimum of @xmath55 -70@xmath4v / k is shown at the temperature @xmath56@xmath5522k . for t
= os , ir and rh , a similar broad minimum develops at a temperature of @xmath56@xmath5516 - 23k , where the width of the peak is wider than that for t = fe and ru . for t
= co , @xmath0 shows a similar temperature dependence but with the minimum shifted to @xmath56@xmath554k and it also shows slope changes around @xmath551k and @xmath558k .
the width of the minimum for t = co is narrower than that for the other members of this family . above 10k ,
the absolute value of the tep for t = co reduces rapidly as the temperature increases and the sign of the tep changes from negative to positive close to 150k . for comparison ,
@xmath0 curves for t = co together with t = fe and rh are plotted on a semi - logarithmic scale in the inset of fig .
[ zn20st1 ] .
a smaller local minimum ( @xmath55 - 48@xmath4v / k ) is observed for ybos@xmath1zn@xmath2 .
it is not clear at present if this is related to the electrical resistivity measurement that showed a larger residual resistivity in ybos@xmath1zn@xmath2 compared to other members ( t = fe , ru , ir , and rh ) @xcite .
@xmath0 of ybir@xmath1zn@xmath2 is negative over the whole temperature range measured , the sign change from negative to positive being expected around @xmath55400k , based on a linear extrapolation of @xmath0 above 250k . below 10k ( or 3k for t = co ) , @xmath0 data for all compounds show a tendency of approaching zero and reveal linear temperature dependencies , @xmath0=@xmath5 .
since the tep was measured down to 0.4k for t = fe , rh , and co , the linear temperature dependence of @xmath0 is even more clearly revealed for these compounds as shown in the inset of fig .
[ zn20st2 ] .
figure [ zn20st3 ] presents @xmath31 of ybt@xmath1zn@xmath2 below 10k .
since no signature of any kind of long range order down to 20mk has been observed in the resistivity measurements @xcite , the zero temperature limit of @xmath31 can be reliably estimated by extrapolating @xmath0 from 2k ( or 0.4k ) to @xmath3 = 0 , where the inferred @xmath57 values for t = fe , ru , os , ir , and rh range between -3.8 @xmath55 -10@xmath4v / k@xmath20 . for t
= co , @xmath31 value at 0.4k reaches -42@xmath4v / k@xmath20 and is still decreasing ( see inset of fig .
[ zn20st3 ] ) . by using a linear extrapolation , @xmath57 value for t = co
is found to be @xmath55 - 57@xmath4v / k@xmath20 . in fig .
[ zn20sh1 ] , the results of @xmath0 measurements at @xmath58 = 0 and 140koe are shown for t = fe , ru and ir . for clarity ,
the absolute value of the tep is shifted by -20@xmath4v / k for t = ru and -40@xmath4v / k for t = ir .
a slight change of @xmath56 and a reduction of absolute value are seen for the @xmath58 = 140 koe data . above 100k , @xmath0 for @xmath58 = 140koe remains essentially the same as @xmath0 for @xmath58 = 0 .
in the zero temperature limit for @xmath58 = 140koe data , whereas @xmath31 for t = ru remain essentially the same , @xmath31 at 140koe for t = fe and ir is decreased from @xmath55 - 3.8 to @xmath55 - 6.4@xmath4v / k@xmath20 and from @xmath55 - 4 to @xmath55 - 6.6@xmath4v / k@xmath20 , respectively . in the inset ,
the tep measured at @xmath3 = 2.2k is plotted as a function of magnetic field for t = fe , ru , and ir , where @xmath59 .
an interesting point of this result is the appearance of a maximum around @xmath5570koe for t = fe and ru and a minimum around @xmath55100koe for t = ir . for t = ir the local minimum field shown in tep
is roughly matched with the metamagnetic - like anomaly seen around @xmath58 = 120koe in magnetization isotherms , @xmath60 , @xcite for * * h**@xmath43[110 ] . for t
= fe and ru the @xmath60 data at @xmath3 = 2k do not show any signature of metamagnetic - like behavior up to 70koe @xcite , with @xmath60 being linear in magnetic field for both compounds . in order to clarify this point ,
it is necessary to measure @xmath60 for magnetic fields higher than 70koe , to see whether the anomaly in @xmath61 is related to features in magnetization or electronic data .
based on earlier thermodynamic and transport measurements of this family @xcite , @xmath0 data for ybt@xmath1zn@xmath2 ( t = fe , ru , os , ir , rh , and co ) can be understood qualitatively by considering the kondo ( @xmath38 ) and cef ( @xmath39 ) effects .
the compounds in this series appear to be a set of model kondo lattice systems with varying energy scales : @xmath38 and @xmath39 . in fig .
[ zn20parameters ] ( a ) , the kondo temperature , @xmath38 , determined from @xmath9 @xcite and local minimum temperature , @xmath56 , observed in the zero field @xmath0 data are plotted as a function of the transition metal , t. the value of @xmath56 correlates strongly with the value of @xmath38 for t = os , ir , rh , and co. a similar trend can be found in the previously published electrical resistivity , @xmath37 , results @xcite . for t
= co , @xmath37 manifests a clear local maximum , @xmath62 , around 2.4k followed by a logarithmic temperature dependence as temperature decreases . whereas @xmath62 is clear in the @xmath37 data for t = co
, @xmath37 data from the other members of this family only show a clear local maximum after subtracting the resistivity data of the isostructural lut@xmath1zn@xmath2 ( t = fe , ru , os , ir , and rh ) compounds .
the local maximum temperatures , @xmath62 , taken from ref .
@xcite are plotted in fig .
[ zn20parameters ] ( a ) .
the variation of @xmath62 follows the same trend as @xmath56 with @xmath62 @xmath55 2@xmath56 even for t = fe and ru . in a kondo lattice system
, a single minimum developed in @xmath0 is expected when @xmath38 is either close to or higher than @xmath39 . typically , an intermediate valence system such as ybal@xmath63 @xcite and ybcu@xmath1si@xmath1 @xcite and a fully degenerate kondo lattice system such as yb@xmath1pt@xmath64al@xmath65 @xcite exhibit a single minimum in the tep , developing below @xmath38 . when @xmath38@xmath66@xmath39 , more than one peak has been frequently observed in the tep @xcite .
the low temperature extremum is usually located close to @xmath38 , and the high temperature extremum located at 0.4 - 0.6@xmath39 is attributed to kondo scattering off of the thermally populated cef levels , which is in agreement with theoretical predictions @xcite . therefore , the peak position can represent @xmath38 and @xmath39 as relevant energy scales in kondo lattice systems . for the ybt@xmath1zn@xmath2 family , @xmath38 and the ground state degeneracy play important roles in the thermodynamic and transport properties . by considering the ground state degeneracy ( @xmath23=8 for t = fe and ru , and @xmath23=4 for t = os , rh , ir , and co @xcite ) it is expected that @xmath38@xmath67@xmath39 for t = fe and ru and @xmath68 for t = os , ir , rh , and co. based on this , for t = fe and ru , it is reasonable to assume that @xmath56 and @xmath62 simply reflect @xmath38 ; with the fully degenerate case corresponding to @xmath23=8 . for t
= os , ir , rh , and co , the two extrema in the @xmath0 data associated with kondo scattering on the ground state and thermally populated cef levels could be expected , however , only one broad peak structure is developed for t = os , ir , rh , and co. we thus expect that a single broad minimum is produced by merging more than one peak structure due to the relatively small cef level splitting ( @xmath38 @xmath55 @xmath39 ) . to reiterate : a strong correlation between the two local extrema @xmath62 and @xmath56 develops and remains robust even when dependence on @xmath38 appears to break down ( fig .
[ zn20parameters ] ( b ) ) .
@xmath62 @xmath55 2@xmath56 for t = fe , ru , os , ir , and rh , and for t = os , ir , rh , and co @xmath56 @xmath55 @xmath38 and @xmath62 @xmath55 2@xmath38 . as shown in fig
. [ zn20sh1 ] the magnetic field dependence of the tep observed in ybt@xmath1zn@xmath2 ( t = fe , ru , and ir ) is anomalous . in the simplest case of a two band model
, the carrier density of electrons , @xmath69 , and holes , @xmath70 , can be taken as @xmath71 = @xmath69 = @xmath70 .
the diffusion tep in magnetic field with several assumptions @xcite can be expressed as @xmath72 where @xmath73 , and @xmath74 with @xmath75 and @xmath76 ( lorentz number ) ; @xmath77 = 1/@xmath78 and @xmath79 are the electrical conductivity and thermal conductivity , respectively , in zero magnetic field . at low temperatures
@xmath80 = @xmath81 , @xmath82 ( the residual resistivity ) , and the diffusion tep in zero magnetic field is proportional to the temperature , @xmath83 @xmath84 @xmath3 . therefore , for simple metals @xmath85 = 0
when @xmath3 = 0 , and for very low temperatures @xmath85 @xmath84 @xmath3 . at high temperatures @xmath80 = @xmath81 , and @xmath83 and @xmath78 are both proportional to temperature , so that @xmath85 tends to zero like 1/@xmath3 as @xmath86 .
in general , the change in the tep will be too small to be detected at room temperature .
since the magnetoresistance ( mr ) for t = fe and ru is positive and increases monotonically at 2k for * * h**@xmath43[111 ] up to 140koe @xcite , the change of the tep ( @xmath85 ) should increase or saturate with increasing magnetic field .
the field dependence of the tep is not consistent with the mr results .
generally , the phonon - drag itself is not sensitive to the applied magnetic field @xcite , so it is clear that neither conventional phonon - drag nor diffusion tep of conduction electrons can account for the magnetic field dependence of the tep in these compounds .
thus , multiple factors , such as the kondo effect and cef contributions , have to be considered . in order to understand
the observed behavior in more detail , a theoretical analysis of the tep as a function of field for this systems will be necessary .
earlier thermodynamic and transport measurements @xcite showed that the @xmath25 and k - w ratios of ybt@xmath1zn@xmath2 agree well with the fl picture of the hf ground state .
a clear dependence of the @xmath15 ratio on the degeneracy @xmath23 is shown in the inset of fig .
[ zn20fl ] , where the @xmath50 and @xmath9 values are taken from ref . @xcite and
lines for degeneracies @xmath23 are based on ref .
a fermi liquid state can also be characterized by the ratio between @xmath9 and the zero temperature limit of @xmath31 @xcite : a quasi universal " ratio @xmath32 = @xmath87 remains close to @xmath32 = @xmath881 for metals and the sign of @xmath32 depends on the type of carriers . although for strongly correlated electronic materials like hf systems , a single band and single scattering process is not generally thought to be sufficient for explaining the strong correlation effects , given that @xmath7 and @xmath31 are most sensitive to the position of the heavy band , a quasi universal ratio is expected to hold at low temperatures @xcite .
the experimental correlation between the zero temperature limit of @xmath31 and @xmath9 for ybt@xmath1zn@xmath2 ( t = fe , ru , os , ir , rh , and co ) is presented in fig .
[ zn20fl ] , where the phonon - drag effect is ignored since it is small . for comparison , data for several other yb - based hf compounds are also plotted in the same figure @xcite .
the calculated @xmath32 values of yb - based compounds vary from -0.77 for t = fe to -1.4 for t = rh , which are close to the value @xmath32 = -1 , expected for hole - like charge carriers . as shown in the figure , each data point is close to a line represented by @xmath32 = -1 which means that the zero temperature limit of @xmath31 is strongly correlated to @xmath9 due to the enhanced density of state at the fermi level ; the larger density of states at the fermi level results in a larger @xmath9 and @xmath57 .
the thermoelectric power measurements on the ybt@xmath1zn@xmath2 ( t = fe , ru , os , ir , rh , and co ) compounds are in agreement with the behavior observed in many heavy fermion kondo lattice systems .
the evolution of the local minimum in @xmath0 and the local maximum ( coherence temperature ) in @xmath37 with variation of the transition metals can be understood based on the energy scale of kondo temperature in conjunction with the influence of the crystalline electric field splitting .
the large value of @xmath31 in the zero temperature limit can be scaled with the electronic specific heat coefficient , @xmath9 , which is reflected by a strong correlation via the universal ratio @xmath89 and confirms the validity of fermi - liquid descriptions .
s. jia , n. ni , g. d. samolyuk , a. safa - sefat , k. dennis , hyunjin ko , g. j. miller , s. l. budko , and p. c. canfield , phys .
b * 77 * , 104408 ( 2008 ) .
note that the debye temperatures used in this paper are recalculated ; @xmath48 ( yfe@xmath1zn@xmath2 ) = 350 k and @xmath48 ( yco@xmath1zn@xmath2 ) = 344 k. f. j. blatt , p. a. schroeder , c. l. foiles , and d. greig , _ thermoelectric power of metals _ ( plenum , new york , 1976 ) . f. j. blatt , a. d. caplin , c. k. chiang , and p. a. schroeder
. solid state commun . * 15 * , 411 ( 1974 ) .
s. yoshiuchi , m. toda , m. matsushita , s. yasui , y. hirose , m. ohya , k. katayama , f. honda , k. sugiyama , m. hagiwara , k. kindo , t. takeuchi , e. yamamoto , y. haga , r. settai , t. tanaka , y. kubo , and y. onuki , j. phys .
78 * , 123711 ( 2009 ) .
eundeok mun , ph .
d. thesis , iowa state university . in zero field ,
the resistivity data for ybt@xmath1zn@xmath2 compounds were taken from ref .
the resistivity data for lufe@xmath1zn@xmath2 and luco@xmath1zn@xmath2 compounds were taken from : s. jia , n. ni , s. l. budko , and p. c. canfield , phys .
b * 80 * , 104403 ( 2009 ) . the resistivity data for lut@xmath1zn@xmath2 ( t = ru , os , ir , and rh ) are unpublished . the zero temperature limit of @xmath90 and @xmath9 for ybcual , ybinau@xmath1 , ybal@xmath63 , ybcu@xmath1si@xmath1 , ybagcu@xmath51 , and ybcu@xmath91 are taken from the table of ref . @xcite . @xmath90 and @xmath9 of yb@xmath1pt@xmath64al@xmath65 are taken from ref .
@xmath9 value of ybni@xmath1b@xmath1c and ybni@xmath1ge@xmath1 are taken from ref .
@xcite and ref .
@xcite , respectively .
the tep data of ybni@xmath1b@xmath1c and ybni@xmath1ge@xmath1 are taken from ref .
@xcite .
s. li , m. c. de andrade , e. j. freeman , c. sirvent , r. p. dickey , a. amann , n. a. frederick , k. d. d. rathnayaka , d. g. naugle , s. l. budko , p. c. canfield , w. p. beyermann , m. b. maple , philos . mag .
b * 86 * , 3021 ( 2006 )
. vs. @xmath3 for ybt@xmath1zn@xmath2 below 10k in zero applied magnetic field .
inset : @xmath31 vs. @xmath3 for ybco@xmath1zn@xmath2 .
the solid line represents the linear extrapolation curve to @xmath3 = 0 . ] of ybt@xmath1zn@xmath2 ( t = fe , ru , and ir ) at @xmath58=0 ( closed symbols ) and 140koe ( open symbols ) . for clarity
, the data for t = ru and t = ir are shifted by -20 @xmath4v / k and -40 @xmath4v / k , respectively .
inset : @xmath42@xmath92 = @xmath93 at 2.2 k for t = fe , ru , and ir . ]
vs. @xmath9 ( log - log ) plot of ybt@xmath1zn@xmath2 ( t = fe , ru , os , ir , rh , and co ) . the zero temperature limit of @xmath31 and @xmath9 for ( 1 ) ybcu@xmath91 , ( 2 ) ybcual , ( 3 ) ybagcu@xmath51 , ( 4 ) ybcu@xmath1si@xmath1 , ( 5 ) ybal@xmath63 , and ( 6 ) ybinau@xmath1 are taken from the table of ref . @xcite .
@xmath31 and @xmath9 of ( 7 ) yb@xmath1pt@xmath64al@xmath65 are taken from ref .
@xcite . @xmath31 and @xmath9 of ( 8) ybni@xmath1b@xmath1c and ( 9 ) ybni@xmath1ge@xmath1 are taken from refs .
@xcite , respectively .
the solid line represents @xmath9/@xmath94 .
inset : kadowaki - woods plot ( log - log plot of @xmath50 vs. @xmath9 ) of ybt@xmath1zn@xmath2 .
symbols are taken from ref .
@xcite and solid lines correspond to @xmath23 = 2 , 4 , 6 , and 8 based on ref .
@xcite , respectively . ] | the thermoelectric power , @xmath0 , of the heavy fermions ybt@xmath1zn@xmath2 ( @xmath3 = fe , ru , os , ir , rh , and co ) has been measured to characterize their strong electronic correlations . a large , negative , local minimum in @xmath0 with approximately -70@xmath4v / k is found for all compounds . from the observed local minimum , the energy scales associated with both the kondo temperature and the crystalline electric field splitting are deduced and compared to previous specific heat measurements . at low temperatures , @xmath0 follows a linear temperature dependence , @xmath0=@xmath5 , with an enhanced value of @xmath6 which is characteristic for a heavy fermion state . in the zero temperature limit
, the enhanced @xmath6 value strongly correlates with the electronic specific heat coefficient , @xmath7 , due to the large density of states at the fermi level . |
the results presented in the main text are obtained by first - principles calculations as implemented in vienna _ ab initio _ simulation package ( vasp ) @xcite with the projector augmented wave ( paw ) method @xcite . the generalized gradient approximation ( gga ) with perdew - burke - ernzerhof ( pbe ) @xcite realization
were adopted for the exchange - correlation potential .
the plane - wave cutoff energy was taken as 500 ev .
the monkhorst - pack @xmath38-point mesh @xcite of size @xmath53 was used for brillouin zone sampling .
the crystal structures were optimized until the forces on the ions were less than 0.01 ev / . as the transition metal
@xmath0 orbitals may have notable correlation effects , we also validate our results by gga+u method@xcite . on - site hubbard u parameters ranging from 1 ev to 7 ev were tested .
we find that the key features are qualitatively the same as the gga results , which also agree with previous studies @xcite .
so in the main text , we focus on the gga results . from the dft results , we construct the maximally localized wannier functions ( mlwf )
@xcite for cu ( ag ) @xmath4 , @xmath0 and s ( se ) @xmath5 orbitals , and effective model hamiltonian for bulk and semi - infinite layer are built to investigate the surface states . in this work ,
we focus on the antifluorite structure ( @xmath11-phase , space group @xmath14 ( no . 225 ) ) of the cu@xmath1s - family materials . for the band structure calculations ,
we take the experimental lattice parameters @xmath54 for cu@xmath1s @xcite , @xmath55 for cu@xmath1se @xcite , and @xmath56 for cuags @xcite . for cuagse , since we also need to analyze its result under strain , we used its optimized lattice parameter @xmath57 in the calculation .
considering the possible underestimation of band gap by gga , we further check the band structure by the hybrid functional approach with modified becke - johnson ( mbj ) potential @xcite as implemented in the wien2k package @xcite .
we find that the band inversion features are maintained .
the band inversion energy between @xmath2 and cu-@xmath10 bands are of about 0.1 ev , 0.24 ev , 0.53 ev and 0.26 ev for cu@xmath1s , cu@xmath1se , cuagse and cuags , respectively .
the topological phases are still maintained .
as shown in fig .
[ fig : mbj ] , cu@xmath1s is still a negative soc ti , with direct ( indirect ) band gap of 65 mev ( 30 mev ) ; cuags also keeps the negative soc band structure with direct ( indirect ) band gap of 71 mev ( 38 mev ) ; cu@xmath1se and cuagse are still topological semimetals .
band inversion is a necessary condition for a topological nontrivial band structure , but it is not sufficient to identify a ti .
topological invariant is a global character of the electronic structure in the whole brillouin zone .
the definition of parity production for occupied bands at trim points is a convenient method to distinguish tis with inversion symmetry , but it can not be used for inversion - asymmetric systems such as cuagse .
the wilson loop method can be employed for such cases .
it traces the evolution of the wannier function centers @xcite . the wannier center evolution for four representative planes ( in the brillouin zone ) of cuagse are shown in fig .
[ fig : wfc ] , from which we can find the @xmath3 indices are respectively 1 and 0 for @xmath48 plane and @xmath58 plane , indicating that there must be a pair of gapless states in edge spectrum of the @xmath48 plane . as discussed in the main text , this dictates the existence of a dirac - cone on the @xmath59 surface .
for the two mirror planes ( 110 ) and @xmath51 , we find @xmath49 . however , this does not lead to an additional dirac - cone for @xmath52 surface because the pattern of the fermi arcs already satisfies the requirement . | we reveal a class of three - dimensional @xmath0-orbital topological materials in the antifluorite cu@xmath1s family . derived from the unique properties of low - energy @xmath2 states ,
their phases are solely determined by the sign of spin - orbit coupling ( soc ) : topological insulator for negative soc , whereas topological semimetal for positive soc ; both having dirac - cone surface states but with contrasting helicities . with broken inversion symmetry ,
the semimetal becomes one with a nodal box consisting of butterfly - shaped nodal lines that are robust against soc . further breaking the tetrahedral symmetry by strain leads to an ideal weyl semimetal with four pairs of weyl points .
interestingly , the fermi arcs coexist with a surface dirac cone on the ( 010 ) surface , as required by a @xmath3-invariant . when individual atoms are brought together to form crystalline solids , the atomic orbitals overlap and form extended bloch states . in the energy - momentum space ,
discrete atomic levels evolve into dispersive electronic bands .
the interaction between orbitals and with further coupling to spin may generate inverted band ordering and lead to topological states of matter , which is a focus of recent physics research @xcite .
it is now established that nontrivial topology can occur for both gapped ( insultor ) and gapless ( semimetal ) systems .
for topological insulators ( tis ) , an invariant is defined for the bulk valence bands below the gap @xcite ; whereas for topological semimetals ( tsms ) , the characterization is on the topology of band - crossings near the fermi level @xcite , leading to a variety of tsms , among which the weyl semimetal and nodal - line semimetal states with respective 0d and 1d band - crossings are attracting great interest and actively searched for @xcite .
the nontrivial bulk topology manifests on the sample surface as the existence of protected surface states : tis have dirac - cone like surface states with spin - momentum - locking @xcite ; whereas weyl semimetals possess open fermi arcs connecting pairs of projected weyl points on the surface @xcite .
in forming topological band structures , the orbital character of bands plays an important role , e.g. , it determines the band inversion , the low - energy quasiparticle dispersion , and furthermore , the type and strength of the effective spin - orbit coupling ( soc ) . for almost all
the tis identified so far , the low - energy bands are of @xmath4- and/or @xmath5-orbital character @xcite .
meanwhile , it is known that @xmath0-orbitals could host interesting soc physics .
for instance , with tetrahedral coordination , the five @xmath0-orbitals will split into an @xmath6 doublet and a @xmath2 triplet .
with soc , the @xmath2 states can exhibit a unique property that its effective soc is negative , with @xmath7 doublet energetically higher than the @xmath8 quartet .
this mechanism has inspired works in exploring tis with negative soc .
however , in the few examples predicted to date @xcite , the low - energy bands are still dominated by @xmath4 and @xmath5 characters , while the @xmath0-bands are away from the fermi level and only act indirectly .
thus , one may wonder whether we can find a genuine @xmath0-orbital ti , and how would the @xmath0 character produce any new physics ?
s - family materials .
the low - energy bands are dominated by 3@xmath0 @xmath2-orbitals , which are splitted by soc into @xmath7 and @xmath8 states ( each line here is assumed doubly spin degenerate ) .
the sign of soc @xmath9 determines the phase : ti for negative soc , while tsm for positive soc .
the band inversion is between the @xmath2 band and the @xmath10 band.,width=340 ] here , we answer the above questions by revealing intriguing @xmath0-orbital topological phases in the cu@xmath1s material family with antifluorite structure . by first - principles calculations ,
we show that the low - energy bands in these materials are dominated by the @xmath2-orbitals , of which the sign of effective soc can be made either negative or positive by tuning the orbital interaction .
remarkably , this sign completely determines the phase . as illustrated in fig .
[ fig1 ] , enforced by band filling , the system must be an insulator ( semimetal ) when the sign of soc is negative ( positive ) .
moreover , in both cases , there is band inversion between the cation @xmath4-band and the @xmath2-band , hence both phases are topological .
we explicitly show that the dirac - cone surface states in the two phases have opposite helicities in spin - momentum - locking , consistent with their sign of soc .
furthremore , novel features are observed for the tsm phase . in ternary compounds with intrinsic inversion asymmetry ,
the system becomes a novel semimetal with a nodal box comprising butterfly - shaped nodal lines . in known examples of nodal - line materials , the nodal lines are unstable when soc is included ; in contrast , the butterfly nodal line here is stabilized with soc . with further lowering of cubic symmetry by strain ,
an ideal weyl semimetal emerges with four pairs of weyl points lying exactly at the fermi level .
we find the phenomenon of hybrid surface states , i.e. , coexisting fermi - arc and dirac - cone surface states on certain surfaces , as required by a bulk @xmath3 invariant .
our predicted features in the bulk band structure and the surface states ( including the spin texture ) can be readily probed by the arpes experiment .
the cu@xmath1s - family materials typically have three structures ( denoted as @xmath11 , @xmath12 , and @xmath13 ) . here
, we focus on the @xmath11-phase structure @xcite , also known as the antifluorite structure , having the space group @xmath14 ( no .
225 ) .
the conventional unit cell has a cubic shape and the structure can be regarded as a double nested zinc - blende lattice . as shown in fig .
[ fig2](a ) , for binary compounds like cu@xmath1s and cu@xmath1se , an inversion center is preserved .
however , inversion symmetry is broken for ternary compounds like cuagse ( see fig .
[ fig4](a ) ) where the cu sites in one of the nested zinc - blende lattices are occupied by ag .
we perform first - principles calculations based on the density functional theory ( dft ) .
the calculation details and the structural parameters are in the supplemental material @xcite . in the following discussion , we shall mainly focus on the representatives cu@xmath1s , cu@xmath1se , and cuagse . s lattice . ( b ) band structure of cu@xmath1s .
the colored circles indicate the weight of cu-@xmath2 ( red ) , cu-@xmath10 ( green ) , s-@xmath15 ( blue ) orbital characters .
( c ) enlarged view of ( b ) near the fermi level , showing the splitting between @xmath7 and @xmath16 states due to negative soc .
( d ) projected spectrum on ( 001 ) surface , and ( e ) the corresponding fermi circle , exhibiting a right - handed spin - momentum - locking ( spin - polarization marked by the green arrows ) .
, width=347 ] in cu@xmath1s , each cu is surrounded by a tetrahedron of s atoms .
the tetrahedral crystal field splits the cu 3@xmath0-orbitals into @xmath6 and @xmath2 states , with @xmath2 having a higher energy .
focusing on the @xmath2 states , the interaction between the two cu atoms in a primitive cell leads to bonding and antibonding states , with the latter energetically higher than the former . in cu@xmath1s ,
the cu @xmath0-orbitals have higher energy than that of s-3@xmath5 orbitals , so that the @xmath5-@xmath0 hybridization further pushes the @xmath2 antibonding states up to the fermi level .
the @xmath2 triplet ( containing @xmath17 , @xmath18 , and @xmath19 orbitals ) have an effective orbital moment @xmath20 , which are then split by the soc @xmath21 into a @xmath7 doublet and a @xmath8 quartet . as mentioned , a unique feature for @xmath2 is that the soc can be negative with @xmath22 , opposite to the soc splitting of @xmath5-orbitals @xcite . indeed , as shown in fig . [ fig2 ]
, our dft result confirms the above picture .
the low - energy bands near the fermi level are mainly from the @xmath2 states , and the s-3@xmath5 bands are below @xmath23 ev . due to combined inversion symmetry and time reversal symmetry , each band is spin degenerate .
around @xmath24-point , one observes that : ( i ) the originally degenerate @xmath2 states are split by soc , and the @xmath7 doublet is higher than the @xmath8 quartet in energy , showing a negative soc ; ( ii ) the cu-4@xmath4 states dive below the @xmath2 states by about 1.1 ev , indicating an inverted band ordering .
band filling dictates that the fermi level lies exactly in the gap ( @xmath25 mev ) between the @xmath7 and @xmath8 states ( fig .
[ fig2](c ) ) .
the band inversion at a single time reversal invariant momentum ( trim ) point directly indicates that the system is a strong ti . to further confirm the nontrivial band topology
, we calculate the @xmath3 invariant for the bulk band structure . with inversion symmetry ,
the task is simplified by analyzing the product of parity eigenvalues at the eight trim points @xcite .
we find that this product is positive at @xmath24 and negative at other trims , in accordance with our analysis of the band inversion , which leads to a strong ti with @xmath3 indices @xmath26 .
the hallmark of ti is the existence of protected dirac - cone surface states with spin - momentum - locking . in fig .
[ fig2](d ) , we plot the calculated surface energy spectra for ( 001 ) surface , clearly showing a single surface dirac - cone . here the dirac point is buried in the bulk valence bands .
notice that for a constant energy above the dirac point , the surface states show a right - handed spin - momentum - locking pattern ( see fig .
[ fig2](e ) ) .
this is consistent with the negative soc , and is in contrast with almost all other tis ( which have left - handed helicity due to positive soc ) @xcite .
the same feature is also observed for other surfaces .
all these evidences confirm that cu@xmath1s is a @xmath0-orbital ti . se .
( a ) low - energy band structure .
( b ) projected spectrum on ( 001 ) surface showing dirac - cone surface states buried in the bulk valence band .
( c ) surface states at @xmath27 ev ( marked in ( b ) ) , exhibiting a left - handed spin - momentum - locking.,width=347 ] the effective soc strength @xmath9 could be tuned by varying the crystal environment , especially through the interaction with the anion @xmath5-orbitals .
the effect depends on both the interaction strength and the soc of @xmath5-orbitals .
consider cu@xmath1se which has the same structure as cu@xmath1s .
compared with s-3@xmath5 , the se-4@xmath5 orbitals are closer to the cu-@xmath2 orbitals in energy , and they also possess a stronger soc ( which is positive ) . consequently , one expects that the interaction with se-4@xmath5 would decrease the negativeness of @xmath9 for cu-@xmath2 states .
this speculation is confirmed by our dft result .
we find that the composition of the low - energy bands in cu@xmath1se is similar to that of cu@xmath1s , however , its @xmath8 quartet is above the @xmath7 doublet ( see fig . [ fig3](a ) ) , indicating that soc has changed from negative to positive .
band filling dictates that the quartet is half - filled , so that the fermi level intersects with the degenerate states and the system must be a semimetal .
note that the band inversion near @xmath24-point between the cu-4@xmath4 states and the @xmath2 states is preserved , not affected by the sign change of @xmath9 .
hence the surface states still exist for cu@xmath1se , as shown in fig .
[ fig3](b ) , although they are submerged in the bulk states .
importantly , the spin - momentum - locking pattern here becomes left - handed ( fig .
[ fig3](c ) ) , which is consistent with the positive soc . point .
( c ) dispersion of the ( @xmath8 ) low - energy bands in [ 110 ] mirror plane , showing ( d ) a butterfly - shaped nodal line , where @xmath28 and @xmath29 correspond to the crossing points marked in ( b ) .
( e ) combining all the butterfly nodal lines leads to a nodal box .
, width=347 ] breaking inversion symmetry would split the spin - degenerate bands with soc , adding new ingredients to the physics .
consider cuagse in which the inversion symmetry is naturally broken ( see fig .
[ fig4](a ) ) .
similar to cu@xmath1se , its @xmath2 states have an effectively positive soc , hence the system is also a semimetal , as shown in fig .
[ fig4](b ) . in cu@xmath1se , the conduction band and the valence band touch at a single point , but we shall see that in cuagse , the splitting of the @xmath8 quartet near @xmath24-point leads to nodal - line band - crossings . the symmetry group contains six mirror planes which may be collectively denoted as @xmath30 . in fig .
[ fig4](b ) , one observes that the four states in the quartet fully split along the @xmath24-@xmath31 line , on which the two middle bands cross each other at a point @xmath29 near the fermi level .
meanwhile , along the @xmath24-@xmath32 direction , two middle bands become degenerate , which then cross the upper band at a triply - degenerate point ( labeled as @xmath28 ) . a careful scan of the band structure around @xmath24-point shows that these crossing points are not isolated ( see fig . [ fig4](c ) ) .
remarkably , in each @xmath30 mirror plane , the crossings near fermi level form a butterfly - shaped nodal line , which is possible since it is formed by the pair - wise crossings of three bands .
this is shown in fig . [ fig4](d ) : along the diagonal ( i.e. @xmath24-@xmath32 ) direction , the crossing is protected and pinned by the @xmath33 symmetry such that the two states at each point on this line form a 2d irreducible representation @xmath34 ( @xmath35 ) ; whereas for the arcs connecting the @xmath28 points , they are protected by the mirror plane because the two crossing bands have opposite mirror eigenvalues @xmath36 ( see fig .
[ fig4](b ) ) .
note that unlike most previously identified nodal lines which are unstable under soc @xcite , the butterfly nodal line here is robust against soc , and actually it appears only when soc is included . combining the ` butterflies ' from all the mirror planes leads to the interesting nodal - box pattern shown in fig .
[ fig4](e ) .
it has been theoretically argued that a weyl semimetal phase must occur during the transition of an inversion - asymmetric system from a ti phase to a normal insulator phase @xcite . for cuagse
, we find that it is a ti under a small uniaxial compression , and it becomes a normal insulator under a tensile strain .
hence there must exist a weyl semimetal phase in - between .
we carefully monitor the band structure change under uniaxial strains and find that the butterfly nodal lines quickly disappear upon applying a tensile strain .
the reduction of the cubic symmetry leads to a dramatic change in the low - energy bands .
figure [ fig5](a ) shows the band structure at 3% strain , in which the gap almost closes along the @xmath24-@xmath37 line .
we perform a scan of the possible band - crossing points using a dense @xmath38-mesh and reveal that there exist four pairs of weyl points at ( @xmath390.001235 , 0 , @xmath390.1012 ) and ( 0 , @xmath390.001235 , @xmath390.1012 ) in unit of reciprocal lattice vectors . as schematically shown in fig .
[ fig5](b ) , the points in each time reversal pair ( i.e. , at opposite @xmath38-points ) have the same chirality , and the four pairs are further connected by the two remaining mirror planes @xmath40 and @xmath41 ( two points connected by mirror have opposite chiralities ) , so all the eight points are tied by symmetry and must locate at the fermi level .
moreover , there is no other coexisting electron / hole pockets .
such kind of weyl semimetals is referred to as ideal " @xcite , regarded as good platforms for studying the interesting weyl physics .
point .
a small gap is opened along @xmath24-@xmath37 .
( b ) schematic view of 4 pairs of weyl points in the brillouin zone .
( c ) fermi arcs on the ( 001 ) surface form a closed loop .
( d ) for ( 010 ) surface , fermi arcs coexist with dirac - cone surface states .
( e ) shows states at energy of @xmath42 mev , where the dirac - cone states can be cearly distinguished .
the green dots mark the surface - projections of weyl points.,width=347 ] the surface of a weyl semimetal features fermi arcs connecting the surface - projected weyl points . on ( 001 )
surface , there are four projected points , each having a chirality of @xmath43 ( see fig . [ fig5](b ) ) , hence there must be two arcs connected to each point .
this is verified by the surface spectrum in fig .
[ fig5](c ) , where the fermi arcs are connected into a loop .
meanwhile , for ( 010 ) surface , there are six projected weyl points , of which the middle ones on the @xmath44-axis have chirality @xmath45 and other four have chirality @xmath46 ( fig . [ fig5](b ) ) . in the slice at fermi energy ( fig . [ fig5](d ) ) , one observes the nearly straight fermi arcs connecting each @xmath47 triplet .
but surprisingly , there appears an additional fermi circle , which looks like the ti dirac - cone surface states . by checking the surface state variation at constant energies around fermi level
, one confirms that the circle is indeed separated from the fermi arcs ( see fig .
[ fig5](e ) ) . to explain its existence
, we note that @xmath48 time reversal invariant plane is fully gapped in the bulk hence carries a @xmath3 invariant .
using the wilson loop method @xcite , we find that its @xmath49 , indicating that there must be a pair of gapless states in edge spectrum of the @xmath48 plane .
this means that , in fig .
[ fig5](d ) and [ fig5](e ) , there must exist a time reversal pair on the @xmath48 line .
therefore , the existence of the dirac - cone is in fact required by the @xmath3 invariant .
in retrospect , one notes that planes such as @xmath50 and @xmath51 also have nontrivial @xmath3 , but this does not lead to an additional dirac - cone for @xmath52 surface because the pattern of fermi arcs in fig .
[ fig5](c ) already satisfies the requirement .
in this work , we have revealed an interesting family of topological materials in which the sign of soc controls topological phase . on the negative soc side , these materials are the first @xmath0-orbital ti discovered to date .
we note that a few materials like tln and hgs have been proposed as tis with negative soc @xcite , but their low - energy bands are still dominated by @xmath4 and @xmath5 orbitals .
the sign of soc determines the helicity of dirac - cone surface states like in fig .
[ fig2](e ) and [ fig3](c ) , which can be detected by spin - resolved arpes experiment . the ability to control the helicity gives us additional freedom in utilizing these states for spintronics applications .
interestingly , it has been proposed that gapless interface states will emerge when two surfaces with opposite helicities are contacted @xcite . the effect may be explored in our identified materials .
as mentioned , almost all the identified nodal lines are unstable under soc .
certain kind of robust nodal lines requires nonsymmorphic symmetries @xcite .
the butterfly nodal line discovered here is distinct in that : ( i ) it is formed under soc ; ( ii ) it does not require nonsymmorphic symmetries ; and ( iii ) it is from the pair - wise crossings of three bands , not just two bands .
moreover , the multiple ` butterflies ' constitute a novel nodal box under tetragonal symmetry .
finally , the coexistence of fermi arcs and dirac - cone surface states was proposed at an interface between a weyl semimetal and a ti @xcite , and is also proposed in a model study @xcite .
our work predicts the first material for its realization .
the intriguing surface - state patterns can be directly probed via arpes experiment .
furthermore , the interference between the fermi arcs and the dirac - cone may also lead to signatures in the quasiparticle interference pattern detectable via scanning tunneling microscopy @xcite . |
the photodisintegration of @xmath1h and @xmath2he and the inverse reaction , the radiative capture of protons or neutrons by deuterons , have been intensively investigated in the past . due to the fact that the corresponding matrix elements contain both the three - body bound and continuum states
, these reactions are expected to be a good testing ground for the underlying two - body potential . in early calculations by barbour and phillips
@xcite it has been shown that an exact treatment of the continuum states results in a considerable enhancement of the cross section in the peak region . in that work and in the following by gibson and lehman
@xcite only simple @xmath3-wave interactions of yamaguchi type have been used , while @xmath4-components @xmath5 were incorporated in @xcite . the role of @xmath6-wave contributions in the two - body input with respect to the three - body cross section
was investigated by fonseca and lehman @xcite , using again yamaguchi terms in the interaction .
more recently , calculations based on separable representations of the bonn _ a _ and paris potentials , including their higher partial wave contributions , were performed for polarization observables at some specific energies @xcite . in the present work
we calculate the differential cross sections at @xmath7 for the yamaguchi , malfliet - tjon , paris , and bonn _ b _ potentials over an energy region from threshold up to @xmath8 mev . the potential dependence , the effect of higher partial waves , and the role of meson exchange currents , taken into account via siegert s theorem , are investigated .
technically the calculations are based on the faddeev - type alt - grassberger - sandhas ( ags ) formalism @xcite adjusted to photonuclear processes @xcite , as done already in @xcite .
the separable representation of the subsystem _
matrices , relevant in this context , is chosen according to the _
w_-matrix approach @xcite .
this approximation combines high accuracy with considerable simplicity @xcite . a more reliable , still fairly simple representation of the two - body input is provided by the ernst - shakin - thaler method @xcite .
calculations employing the corresponding pest , best , and nest potentials @xcite will be presented in a subsequent publication .
the alt - grassberger - sandhas ( ags ) equations are well known to go over into effective two - body lippmann - schwinger equations @xcite when representing the input two - body
_ t_-operators in separable form .
the neutron - deuteron off - shell scattering amplitude @xmath9 , thus , is determined by @xmath10 applying the same technique to the photodisintegration of the triton into neutron and deuteron , i.e. to the process @xmath11 , an integral equation of rather similar structure is obtained @xcite , @xmath12 where @xmath13 represents an off - shell extension of the full photodisintegration amplitude @xmath14 in both these equations the kernel is given by the effective neutron - deuteron potential @xmath15 and the corresponding effective free green function @xmath16 defined in ref .
however , in eq .
( [ pampl ] ) the inhomogeneity of eq .
( [ scampl ] ) is replaced by an off - shell extension @xmath17 of the plane - wave ( born ) amplitude @xmath18 here @xmath19 and @xmath20 denote the triton and deuteron states , @xmath21 the momentum state of the neutron relative to the center of mass of the deuteron , and @xmath22 the electromagnetic operator .
thus , by this replacement any working program for @xmath23@xmath4 scattering can immediately be applied to the above photoprocess .
the results presented in this paper are , in fact , obtained by extending recent @xmath23@xmath4 calculations @xcite in this manner . as in these references , the separable part of the _ w_-matrix representation @xcite of the two - body _ t _
matrices is employed , @xmath24 here @xmath25 and @xmath26 are the orbital angular momenta , and @xmath27 stands for the spin , the angular momentum j [ with the coupling sequence @xmath28 and the isospin @xmath29 of the two - body subsystem . with the form factors @xmath30 and the propagators @xmath31 of this representation eq .
( [ pampl ] ) reads , after partial wave decomposition and antisymmetrization , @xmath32 and its inhomogeneity is given by @xmath33 here , @xmath34 and @xmath35 represent the free three - body green function and the partial wave projection of the three - body plane wave state @xmath36 .
the detailed structure of the effective potential @xmath37 can be found in ref .
the label @xmath38 denotes the set @xmath39 of quantum numbers , where @xmath40 and @xmath41 are the channel spin of the three nucleons [ with the coupling sequence @xmath42 and the relative angular momentum between the two - body subsystem and the third particle , respectively .
@xmath43 is the total angular momentum following from the coupling sequence @xmath44 , and @xmath45 is the total isospin .
the born amplitude contains the triton bound state @xmath46 , which may be calculated by any of the various bound - state methods . consistently with the present approach we employ for this purpose the partial wave projected homogeneous version of eq .
( [ scampl ] ) , @xmath47 its solutions , the three - body `` form - factors '' @xmath48 , are related to @xmath46 according to @xcite by @xmath49 here , the summation runs over all two - fragment partitions @xmath50 , the variables and quantum numbers being understood in the corresponding set of jacobi coordinates .
the electromagnetic operator entering eq .
( [ bampl ] ) is , at the low energies considered , essentially a dipole operator . ignoring meson - exchange currents it is given by @xmath51 with exchange currents it takes , according to siegert s theorem @xcite , the form @xcite
@xmath52 here @xmath53 and @xmath54 denote the final and the triton energies , @xmath55 the nucleon center - of - mass coordinates , @xmath56 the corresponding momenta , @xmath57 the electric charges , and @xmath58 the polarization of the photon .
the on - shell restricted solutions @xmath59 of the integral eq .
( [ phampl ] ) yield the photodisintegration amplitude via the partial - wave summation @xmath60 where now , in contrast to eq .
( [ mampl ] ) , the spin and angular momentum quantum numbers of the neutron ( @xmath61 ) , deuteron ( @xmath62 ) , triton ( @xmath63 ) , and the polarization ( @xmath58 ) of the photon are explicitly given . denoting this amplitude by @xmath64 ,
the cross section is obtained in the standard way by @xmath65 here we have averaged over the initial states and summed over the final states .
as a first test of our numerical program we performed calculations for the yamaguchi @xcite potential in order to compare them with the corresponding results by gibson and lehman @xcite .
unfortunately , employing the same sets of parameters as in this reference , we read off from the analytically given relations somewhat different values for the two - body scattering length , effective range , and deuteron binding energy ( table i ) .
the triton binding energies , obtained by solving the homogeneous three - body equations , differ in one case , the symmetric tabakin potential , where we found @xmath66mev instead of @xmath67mev ( table iii ) .
these discrepancies mean that our photodisintegration results can not be expected to fully agree with the ones of ref .
@xcite .
figure 1 compares our cross sections for triton - photodisintegration ( solid line ) with the calculations by gibson and lehman @xcite ( dashed line ) .
the disagreement , in particular in the peak region , is reduced when replacing in the siegert operator our triton energy by the gibson - lehman value ( short - dashed line ) .
figure 2 shows the same comparison , but now for the @xmath68he photodisintegration . since there is no disagreement between the @xmath68he binding energies
, a correspondingly modified curve does not exist .
the remaining discrepancies , therefore , have to be attributed to numerical uncertainties , which are not unexpected in view of the level of accuracy reached in early calculations . within these limits
we consider our results for the yamaguchi potential consistent with the ones of ref .
@xcite .
the differential cross sections obtained for the paris @xcite , bonn _ b _
@xcite , malfliet - tjon ( mt i - iii ) @xcite , and yamaguchi potentials are shown in figure 3 .
most remarkable is the strong potential dependence in the peak region , which vanishes at higher energies .
a further observation , which should be relevant in model calculations or when going over to higher particle numbers @xcite , is the proximity of the paris and malfliet - tjon results on the one hand , and less closely the bonn _
b _ and yamaguchi potentials on the other hand . in figure 4
we contrast , for the paris potential , the solution of the integral equation ( solid line ) with the corresponding plane - wave ( born ) approximation ( dashed line ) .
it is seen that the full solution is enhanced by about 24% at the peak .
a similar enhancement was observed for simpler interactions already in ref .
the upper curves are the ones based on the siegert - operator ( [ sieg2 ] ) , the lower ones correspond to the non - siegert operator ( [ sieg1 ] ) .
there is a factor of two between the siegert and non - siegert results , which demonstrates the relevance of meson - exchange currents . for a detailed discussion of the same phenomenon in case of deuteron and @xmath69he photodisintegration
we refer to refs . @xcite and @xcite , respectively .
figures 5 and 6 compare our cross sections for the s- and d - wave projected paris and bonn _ b _ potentials with the data from refs .
figures 7 and 8 show the same with inclusion of the subsystem @xmath6 waves , which leads to a reduction of the peak by 810% . up to 25 mev the best agreement with the data is achieved for the bonn _
b _ potential . in view of the experimental errors
the relevance of this observation is , of course , somewhat questionable . at higher energies the potential dependence vanishes .
it is , however , seen that the incorporation of the @xmath6 waves is essential for the remarkable agreement with most recent experimental data @xcite .
note that for the inverse reaction the relevance of the @xmath6 wave contributions has been pointed out also in @xcite .
two fragment photodisintegration of the three - nucleon bound states , thus , provides a sensitive tool for testing the underlying two - nucleon potentials . a repetition of the corresponding low - energy measurements with much higher accuracy is strongly suggested by this observation . | the process @xmath0 is treated by means of three - body integral equations employing in their kernel the _ w_-matrix representation of the subsystem amplitudes . as compared to the plane wave ( born ) approximation the full solution of the integral equations , which takes into account the final state interaction , shows at low energies a 24% enhancement .
the calculations are based on the semirealistic malfliet - tjon and the realistic paris and bonn _ b _ potentials . for comparison with earlier calculations we also present results for the yamaguchi potential . in the low - energy region a remarkable potential dependence is observed , which vanishes at higher energies . |
colloidal suspensions of monodisperse spherical particles , with a diameter between 1 nm and 1 @xmath0 m , in liquids or gases self assemble onto three - dimensional simple lattices whose lattice parameter can be easily tailored , providing new opportunities for fundamental as well as for applied research ( see , e.g. , ref . and references therein ) .
the optical properties of these colloidal crystals are being investigated by many research groups because they seem to be good photonic band - gap materials ; but they have interesting acoustic properties as well .
@xcite moreover , experiments relating to the reflection , transmission and absorption of ultrasonic waves by colloidal crystals can be very useful for the characterization of such systems , provided the means exist for a proper theoretical analysis of the experimental data .
finally , theoretical results for colloidal crystals offer a starting point for the understanding , at a semiquantitative level , of the propagation of acoustic waves in corresponding random media consisting of monodispersed spherical particles in a fluid . @xcite the vibration modes ( normal modes of the elastic field ) of a phononic crystal , by which we mean a composite material whose density @xmath1 and lam coefficients @xmath2 and @xmath3 vary periodically in space , are bloch waves with a corresponding frequency band structure which is analogous to that of electrons in ordinary crystals and electromagnetic waves in photonic crystals .
@xcite with an appropriate choice of the parameters involved one may obtain phononic crystals with absolute ( omnidirectional ) frequency gaps ( phononic gaps ) in selected regions of frequency .
an elastic wave , whose frequency lies within an absolute frequency gap of a phononic crystal , incident on a slab of the crystal of certain thickness will be reflected by it , the slab operating as a perfect nonabsorbing mirror of elastic waves in the frequency region of the gap . @xcite it may be possible , for example , to construct in this way vibration - free cavities which might be very useful in high - precision mechanical systems operating in a given frequency range . in this paper
we investigate in detail the acoustic properties of fcc colloidal crystals of polystyrene spheres in water and of close - packed silica spheres in air ( opals ) . given that these systems have a characteristic length scale of the order of 1 @xmath0 m , they should exhibit interesting acoustic properties at ultrasonic frequencies of a few hundred mhz to a few ghz .
it should be noted , however , that our results apply to different regions of frequency of the acoustic field provided that the size of the spheres and of the unit cell are scaled accordingly , and provided the elastic coefficients can be taken as constants independent of frequency over the said regions .
the calculations of the frequency band structure and of the transmission coefficient of acoustic waves through a slab of the material were done using the layer - multiple - scattering formalism we have developed for this purpose .
@xcite a formalism along the same lines has also been published by liu _
et al._. @xcite we have already demonstrated the efficiency of this method in relation to solid - solid composites .
@xcite the present paper shows that the method applies equally well to phononic crystals consisting of nonoverlapping solid spheres in a fluid host ; it appears that the plane - wave method for calculating the frequency band structure of such systems has convergence problems .
@xcite besides , we shall pay particular attention to the physical origin of the different modes of the acoustic field in the systems under consideration .
we consider a model colloidal crystal consisting of polystyrene spheres in water . the mass density and the sound velocities for polystyrene are : @xmath4 , @xmath5 , @xmath6 . for water
we have : @xmath7 , @xmath8 .
we begin by considering the scattering of a harmonic plane acoustic wave ( a longitudinal wave ) of angular frequency @xmath9 , by a single polystyrene sphere of radius @xmath10 in water ; the water extends over all space .
the incident plane wave can be written as a sum of spherical waves associated with the spherical harmonics @xmath11 , where @xmath12 and @xmath13 , as usual .
@xcite .
the solid / dotted / dashed lines refer to bands of @xmath14/@xmath15/@xmath16 symmetry , respectively .
( b ) transmittance of an acoustic wave incident normally on a slab of the above crystal consisting of 32 planes of spheres parallel to the ( 111 ) surface .
the positions of the first two resonant modes of a single polystyrene sphere in water are indicated in the margin.,height=226 ] a spherical wave of given @xmath17 scatters independently of the others , of different ( @xmath18 ) , because of the spherical symmetry of the scatterer ; therefore the total scattering cross section is the sum of partial @xmath17 cross sections , with @xmath19 up to a maximum @xmath20 depending on the size parameter , @xmath21 , of the sphere .
we assume that waves with @xmath22 do not scatter from the sphere and do not contribute to the total scattering cross section .
of course partial cross sections of different @xmath23 ( but of the same @xmath19 ) are equal because of the spherical symmetry of the system . in fig .
[ fig1 ] we show the total scattering cross section for a plane wave scattered by a polystyrene sphere in water as a function of @xmath21 .
the peaks correspond to resonant modes of the acoustic field about the sphere with an angular distribution of the displacement field at the surface of the sphere determined by the spherical waves which contribute to these modes .
we find that the resonance at @xmath24 is an @xmath25 resonance : the displacement field associated with it is made up almost entirely ( 98% ) from @xmath25 spherical waves ; the one at @xmath26 is an @xmath27 resonance : the corresponding displacement field is made up mostly ( 90% ) from @xmath27 spherical waves ; finally the field associated with the resonance at @xmath28 is made up mostly from @xmath29 waves ( 65% ) and @xmath30 waves ( 22% ) . and
we remember that there will be ( @xmath31 ) resonant modes of the displacement field corresponding to @xmath13 of the same frequency @xmath32 , i.e. @xmath32 is ( @xmath31)-degenerate .
we now consider a fcc crystal , with lattice constant @xmath33 , of nonoverlapping polystyrene spheres in water .
( only bands of @xmath14 symmetry are shown ) , calculated ( a ) : with @xmath34 and ( b ) : with @xmath35.,height=226 ] we view the crystal as a sequence of ( 111 ) planes of spheres .
the spheres of a plane are arranged on a hexagonal lattice of lattice constant @xmath36 , defined by the primitive vectors @xmath37 and @xmath38 , in the @xmath39 plane ( plane of the spheres ) .
the ( @xmath40)th plane along the @xmath41 axis is obtained from the @xmath42th plane by a primitive translation , @xmath43 , of the crystal .
the planes are separated along the @xmath41 direction by a distance @xmath44 .
[ fig2](a ) shows the frequency band structure of acoustic waves in this crystal normal to the ( 111 ) surface , obtained by the method of ref .
, when the fractional volume occupied by the spheres is @xmath45 . in this case
, the component of the reduced wavevector within the surface brillouin zone ( sbz ) of the fcc ( 111 ) surface , @xmath46 , equals zero .
the symmetry of the bands @xmath47 is determined by the symmetry group for this direction ( the @xmath48 group ) and the corresponding propagating modes ( bloch waves ) of the acoustic field belong to the different irreducible representations of this group , denoted by @xmath14 , @xmath15 and @xmath16 ( see , e.g. , ref . ) ; @xmath14 and @xmath15 are one - dimensional : the corresponding frequency bands are non - degenerate , and @xmath16 is two - dimensional : the corresponding frequency bands are doubly degenerate . at the @xmath49 point
( @xmath50 , @xmath51 ) the modes of the acoustic field belong to the irreducible representations of the @xmath52 group associated with this point , as shown in fig .
[ fig2](a ) .
we see that the frequency bands of fig .
[ fig2](a ) appear in pairs : @xmath53 , @xmath54 .
however , this symmetry property is not valid for an arbitrary value of @xmath46 , because the crystal under consideration , described by the @xmath52 group , does not have a plane of mirror symmetry parallel to the ( 111 ) surface which would transform @xmath55 to @xmath56 .
the above symmetry property holds only for @xmath46 on the @xmath57 axis [ @xmath58 , because a rotation through @xmath59 about the @xmath57 axis is a symmetry operation of @xmath52 and it transforms @xmath60 to @xmath61 .
the ( longitudinal ) wavefield , of given @xmath9 and @xmath62 , in the host region between the @xmath42th and the ( @xmath40)th planes of spheres , can be expanded into plane waves propagating ( or decaying ) to the left and to the right as follows @xmath63+\ { \bf u}_{{\mathbf{g}}n } ^{- } \exp\left[i\,{\mathbf{k}}_{{\mathbf{g}}l}^{- } \cdot ( { \mathbf{r}}-{\mathbf{a}}_n)\right]\biggr\}\ ; , \label{eq : unn+1}\end{aligned}\ ] ] with @xmath64^{1/2}\right)\ ; , \label{eq : kg}\ ] ] where @xmath65 are the two - dimensional reciprocal vectors corresponding to the lattice defined by @xmath66 and @xmath67 above , and @xmath68 is a point between the @xmath42th and @xmath69th planes .
a generalized bloch wave satisfies the equation @xmath70 where @xmath71 and @xmath72 .
we note that , although both longitudinal and transverse modes of the elastic field are considered within the ( solid ) spheres , only longitudinal modes exist in the ( fluid ) host region in the absence of viscosity .
therefore , in a binary composite of nonoverlapping solid spheres in a fluid host , where the solid component does not form a continuous network , there can not be propagating transverse waves .
consequently , at the @xmath49 point ( @xmath73 @xmath51 ) the dispersion curves [ fig .
[ fig2](a ) and also figs . [ fig3 ] , [ fig5 ] , and [ fig7](a ) below ] , show only one ( longitudinal ) branch starting from zero frequency , instead of the three ( corresponding to both longitudinal and transverse modes ) appearing in solid - solid composites .
@xcite one can easily show that when @xmath50 , the @xmath74 component of the wavefield , described by eq .
( [ eq : unn+1 ] ) , vanishes ( @xmath75 ) for the modes of the @xmath15 or the @xmath16 symmetry ; only the @xmath14 symmetry allows a non - vanishing @xmath76 component of the wavefield .
now assume that we have a slab of the crystal of finite thickness , i.e. @xmath77 fcc ( 111 ) planes of polystyrene spheres ; between the spheres and to the left and right of the slab , extending to infinity , we have water . there , and for the frequency range which interests us here , the acoustic wavefield has only the one component @xmath76 . and therefore the external field couples with the field inside the slab essentially only through the @xmath76 component of the latter .
therefore an acoustic wave of given frequency , incident normally on the ( 111 ) slab of the crystal , will excite essentially only @xmath14 modes of the crystal .
if such a band does not exist at the given frequency the wave will be totally reflected by the slab .
there will be no transmitted wave .
the reader will see that this is indeed the case in the present instance : we show in fig .
[ fig2](b ) , opposite the frequency band structure , the transmission coefficient of an acoustic plane wave incident normally on a slab of the crystal consisting of 32 planes of polystyrene spheres in water .
the transmission coefficient opposite the @xmath14 bands exhibits the well - known fabry - perot - like oscillations due to multiple scattering between the surfaces of the slab ; elsewhere it practically vanishes , and this includes regions of frequency where only bands of the @xmath15 and/or @xmath16 symmetry exist .
of course the above argument holds for normal incidence ; off this direction the symmetry argument does not apply , and there will be some transmission if there are any propagating modes of the acoustic field at the given frequency ( see below , fig . [ fig4 ] ) . from a practical point of view , once the frequency band structure and , if required , the transmission coefficient have been calculated , one has all that is necessary for a comparison with relevant experimental data . however , it is worthwhile to look at the physics behind the band structure of fig .
we note that the five narrow bands about @xmath78 ( we remember that the @xmath16 band is doubly degenerate ) derive from the @xmath25 resonances on the individual spheres , which interact weakly between them .
the five - fold degeneracy of the resonance of the single sphere is split by this interaction in accordance with the lower ( cubic ) symmetry of the crystal field .
similarly the seven narrow bands about @xmath79 derive from the @xmath27 resonances on the individual spheres .
we observe the small hybridization gaps opening up about @xmath80 and about @xmath79 , between bands of the same symmetry . , for @xmath81 .
( b ) transmittance of an acoustic wave incident with @xmath82 on a slab of 32 fcc ( 111 ) planes of polystyrene spheres in water.,height=188 ] there is also a mode of propagation of acoustic waves corresponding to almost free propagation in an effective homogeneous medium . in the absence of the resonances associated with the spheres ,
this mode of propagation , which has the @xmath14 symmetry , would dominate the frequency band structure of the acoustic field as shown in fig .
[ fig3](a ) .
it was calculated by suppressing the @xmath83 resonances , i.e. cutting off the @xmath83 spherical waves , which give rise to these resonances , from the spherical - wave expansion of the wavefield ( see ref . ) . in the long - wavelength limit ( @xmath84 )
we obtain a linear dispersion curve , the slope of which gives an effective velocity of sound for the composite medium , @xmath85 , which is in very good agreement with the result , @xmath86 , of the effective - medium approximation .
@xcite the small gap about @xmath78 in fig .
[ fig3](a ) , is a bragg gap ; it is analogous to the small gaps one obtains at the edges of the brillouin zone in the electronic band structure of free - electron - like metals .
when the resonances on the spheres are allowed in , one obtains the band structure shown in fig .
[ fig3](b ) , with apparent hybridization gaps between the continuum band [ of fig .
[ fig3](a ) ] and resonance bands of the same symmetry ( @xmath14 ) .
we believe that the disorder that is naturally there in a colloidal solution , does not eradicate the essential characteristics of the acoustic modes as calculated here , which are determined by the local environment about a sphere and less so by long range order .
we would expect the fine features associated with narrow bands to be smoothed out by disorder , but bands of modes separated by gaps could remain
. this would be in accordance with the results of brillouin - scattering experiments on disordered colloidal suspensions of monodispersed polymethylmethacrylate ( pmma ) spheres in transparent oil .
it has been shown that in these colloidal suspensions different longitudinal modes of propagation of the acoustic field exist which are separated by frequency gaps about the resonance frequencies of an individual pmma sphere , @xcite which agrees , at a semiquantitative level , with the hybridization - induced gaps discussed above . in fig .
[ fig4](a ) we show an example of the band structure for a @xmath87 , and next to it , in fig .
[ fig4](b ) , the transmission coefficient of an acoustic wave , with the same @xmath46 , incident on a slab of the material consisting of 32 fcc ( 111 ) planes of polystyrene spheres .
the thing to note is that all bands are active in this case , although the transmission coefficient is not always unity [ compare with fig .
[ fig2](a ) ] .
we remember that the incident wave can not have a frequency smaller than @xmath88 .
finally , it can be seen that now the frequency bands do not have the reflection symmetry found in the case of @xmath50 [ fig .
[ fig2](a ) ] , for the reasons given above .
we calculated the frequency band structure for a sufficient number of @xmath46 points within the sbz .
there is a near - omnidirectional gap about @xmath89 ; we tried to turn this gap into a proper omnidirectional gap by changing some of the parameters , but we did not succeed . and ( b ) : @xmath90 .
the solid / dotted / dashed lines refer to bands of @xmath14/@xmath15/@xmath16 symmetry , respectively .
the positions of the first two resonant modes of a single polystyrene sphere in water are indicated in the margin.,height=188 ] in fig .
[ fig5 ] we look at the dependence of the frequency band structure on the fractional volume occupied by the spheres .
we see in particular that the width of the resonance bands increases with @xmath91 , apparently because the spatial overlap of the wavefield associated with resonances on neighboring spheres increases with @xmath91 .
in this section we consider a phononic crystal , an artificial opal , consisting of two media ( silica spheres in air ) with density and velocity contrasts much higher than was the case in the crystal ( polystyrene spheres in water ) studied in the preceding section .
opalescent structures ( usually referred to as synthetic or artificial opals ) can be obtained from monodispersed silica colloids , e.g. by sedimentation in the gravitational field . in this way , fcc arrays of silica microspheres in air , near the close - packing density ( @xmath92 ) , similar to a natural opal , have been obtained .
@xcite here we present some results on the acoustic properties of such an opal .
the mass density and the sound velocities for silica are : @xmath93 , @xmath94 , @xmath95 .
for air we have : @xmath96 , @xmath97 .
proceeding as in section [ poly ] , we first consider the scattering of a harmonic plane acoustic wave , of angular frequency @xmath9 , by a single silica sphere of radius @xmath10 in air . in this case
the total scattering cross section is a slowly increasing function of @xmath98 ( see fig .
[ fig6 ] ) . however , the partial cross sections , corresponding to different values of @xmath19 , exhibit well defined resonance peaks , of some width , as can be seen from fig . [ fig6 ] . at @xmath99
we obtain an @xmath100 resonance , at @xmath101 an @xmath30 resonance , and at @xmath102 an @xmath25 resonance .
we consider then an fcc crystal of nonoverlapping silica spheres in air with @xmath103 and view it as a sequence of ( 111 ) planes of spheres .
[ fig7](a ) shows the frequency band structure of the infinite crystal , for @xmath50 .
the physical origin of the bands shown in fig .
[ fig7](a ) , other than the flat band about @xmath104 , can be understood in much the same way as for the bands of fig .
[ fig2](a ) . in the absence of hybridization
we would have , extending practically over the entire frequency range , a mode corresponding to almost free propagation in an effective homogeneous medium , with bragg gaps opening up about @xmath105 at the edges of the brillouin zone and about @xmath106 at the center of the brillouin zone ; and we would have also resonance bands developing about the @xmath100 and @xmath30 resonances of the individual spheres . from fig .
[ fig6 ] , we expect a single @xmath100 resonance band of @xmath14 symmetry about @xmath107 ; and three @xmath30 bands about @xmath108 , of which one should be non - degenerate ( @xmath14 symmetry ) and one doubly degenerate ( @xmath16 symmetry ) .
we expect a degree of hybridization between bands of the same ( @xmath14 ) symmetry , and this naturally leads to the bands shown in fig .
[ fig7](a ) , except for the one flat band , a doubly degenerate band of @xmath16 symmetry , about @xmath104 , which appears to have a very different physical origin ( see below ) . in the long - wavelength limit ( @xmath84 )
we obtain a linear dispersion curve , the slope of which gives an effective velocity of sound in the composite medium , @xmath109 m / s , which is in very good agreement with result , @xmath110 m / s , of the effective - medium approximation .
@xcite we note that this agreement is greatly improved at lower values of the volume - filling fraction @xmath91 .
it is worth noting , also , that the effective velocity of sound in a system of silica spheres in air is smaller than in air because of the high density contrast between silica and air .
indeed , in the limit @xmath111 , the effective - medium approximation gives practically identical results with the simple expression @xmath112^{1/2}$ ] , which gives @xmath113 at any value of @xmath91 . .
the solid and dashed lines refer to bands of @xmath14 and @xmath16 symmetry , respectively .
( b ) transmittance of an acoustic wave incident normally on a slab of the above crystal consisting of 32 planes of spheres parallel to the ( 111 ) surface .
the positions of the first two resonant modes of a single silica sphere in air are indicated in the margin.,height=226 ] it seems that the modes corresponding to the flat band about @xmath104 , which are deaf modes , are those of a wavefield highly concentrated between consecutive planes of spheres , with very little interaction between neighbor regions of high concentration . in order to demonstrate the above
, we looked for the eigenmodes of the acoustic field , for @xmath114 , in a slab of @xmath115 planes of spheres , in the manner described for the corresponding problem of the electromagnetic field , @xcite over a frequency range about the said flat band of fig .
[ fig7](a ) .
because of the two - dimensional translation symmetry of the slab , the wavefield , in the air regions , can be expanded [ as in eq . ( [ eq : unn+1 ] ) ] in a series of plane waves with wavevectors @xmath116^{1/2}\right)$ ]
; now , as it turns out , for the eigenmodes in question the term corresponding to @xmath76 in the above plane - wave expansion vanishes , which implies that the wavefield decreases exponentially on both sides out of the slab , since latexmath:[$\omega / c_a < for a single pair of planes ( @xmath118 ) we obtain a pair of such modes corresponding to a doubly degenerate eigenfrequency . for three pairs of planes ( @xmath119 ) we obtain three pairs of doubly degenerate eigenfrequencies , and similarly in each case we obtain a pair of doubly degenerate eigenfrequencies for every pair of planes in the slab .
the eigenfrequencies corresponding to the bound states of the different slabs in fig .
[ fig8 ] , have been plotted against values of the reduced wavenumber @xmath120 , @xmath121 , for the slab consisting of @xmath77 planes , to show how this band of states develops with increasing @xmath77 . in every case
, @xmath122 equals the separation of two consecutive fcc ( 111 ) planes in the infinite crystal , so that @xmath123 equals the thickness of the slab .
we note that the eigenfrequencies of the finite slabs coincide with the corresponding eigenfrequencies on the dispersion curves ( solid lines in fig .
[ fig8 ] ) of the infinite crystal .
the question now arises as to whether one can see experimentally this band .
the corresponding modes do not couple to an incident wave .
this is demonstrated quite clearly in fig .
[ fig7](b ) , which shows the transmission coefficient of a wave incident normally on a slab of the material consisting of 32 planes of spheres .
we can see that the transmission coefficient vanishes at and about @xmath104 where the band under consideration exists .
however , we have verified that the said band survives for @xmath124 ( at least in the neighborhood of @xmath50 ) where it couples with an incident wave of the same @xmath46 leading to measurable transmittance . in fig .
[ fig9 ] , we present the projection of the frequency band structure of the acoustic field of the phononic crystal under consideration , on the symmetry lines of the sbz of the fcc ( 111 ) surface .
the shaded regions extend over the frequency bands of the acoustic field : at any one frequency within a shaded region , for a given @xmath46 , there exists at least one propagating acoustic mode in the infinite crystal .
the blank areas correspond to frequency gaps .
we note that knowing the modes with @xmath46 in the shaded area ( @xmath125 ) of the sbz and @xmath126 is sufficient for a complete description of all the modes in the infinite crystal .
the modes in the remaining of the reduced @xmath127 space are obtained through symmetry .
one clearly sees that for the opal under consideration , and unlike the colloidal crystal studied in section [ poly ] , one obtains an omnidirectional frequency gap extending from @xmath128 to @xmath129 .
we verified that this is indeed so by calculating the band structure at a sufficient number of @xmath46 points in the sbz .
finally , we should note the existence of a very narrow omnidirectional gap at higher frequencies , extending from @xmath130 to @xmath131 .
, on the sbz of the fcc ( 111 ) surface , along the symmetry lines shown in the inset . propagating waves in the air about a slab of the crystal exist for frequencies above a threshold value ( a function of @xmath62 ) @xmath132 denoted by the dashed line.,height=188 ] this work has been supported by the institute of communication and computer systems ( iccs ) of the national technical university of athens . partial support from the university of athens is also acknowledged .
r. sainidou is supported by the state foundation ( i.k.y . ) of greece . | we present a systematic study of the frequency band structure of acoustic waves in crystals consisting of nonoverlapping solid spheres in a fluid .
we consider colloidal crystals consisting of polystyrene spheres in water , and an opal consisting of close - packed silica spheres in air .
the opal exhibits an omnidirectional frequency gap of considerable width ; the colloidal crystals do not .
the physical origin of the bands are discussed for each case in some detail .
we present also results on the transmittance of finite slabs of the above crystals . |
to introduce and place the ideas of the present contribution in context , consider the paradigmatic problem of estimating the volume @xmath0 of a given region as shown in figure [ fig : illustration ] . in its most rudimentary form ,
mc estimates of @xmath0 proceed using the `` hit - or - miss '' idea , whereby the user designs a reference region of known volume @xmath1 that fully overlaps with @xmath0 , and draws random points uniformly distributed in @xmath1 ( cf .
[ fig : illustration ] , top panel ) .
an estimate of @xmath0 can then be obtained from @xmath2 , where @xmath3 is the fraction of such points that fall inside @xmath0 . as is immediately apparent
, the efficiency of this simple method hinges upon the amount of overlap between the volumes @xmath1 and @xmath0 : the tighter @xmath1 bounds @xmath0 , the more the hits and hence the better the quality of the estimate .
conversely , bounding volumes that have little overlap with @xmath0 give rise to more misses than hits and hence to large errors in the estimate .
this poses a major obstacle to the implementation of such methods , especially in cases where it is difficult to guess the precise location of the volume and its boundaries , often leading to the design of an unnecessarily conservative ( large ) reference region , and hence to very inefficient estimates of @xmath0 . despite its simplicity , the above example captures the central issues pertinent to integral estimation problems in general , for which more sophisticated methods exist @xcite .
notably , in most efficient monte carlo techniques for normalizing constant or free energy estimation , a suitable family of intermediate integrands is required to interpolate between the two desired integrands ( or between the available reference system and the integrand of interest , in absolute integral estimation ) , the extent of overlap between the integrands dictating the efficiency of the methods much like in the example above @xcite .
such requirements become particularly daunting when the integrands in question are highly concentrated in unknown and disparate regions of configuration space , as is typically the case in most interesting physical problems , thereby demanding substantial user insight prior to the applicability of such methods . in the present contribution
, a new family of mc integration strategies that can greatly alleviate the demand for such types of insights will be introduced . as expressed by the central results underpinning these methods , eqs .
( [ eq : z_1st ] ) and ( [ eq : z_2nd ] ) , the integrals of interest ( @xmath4 , eq . ( [ eq : z ] ) ) are computed individually , thus fundamentally departing from the aforementioned alternatives that rely on ratios of similar integrals . at the core of these new strategies
is the use of integration spaces of enlarged dimensionality ( `` replicas '' ) , a concept already widely invoked to speed up markov chain simulations ( parallel tempering @xcite , evolutionary monte carlo @xcite , etc . ) , combined with the replacement of reference integrals by normalized transition matrices @xcite ( also known as transition functions @xcite or kernels @xcite in the markov chain literature ) .
two main variants will be presented : the first , more physically intuitive , requires a fluctuating number of replicas ( fig .
[ fig : illustration ] , bottom panel ) ; the second , more abstract but also more easily adaptable to existing replica simulations , uses a fixed number of replicas along with `` virtual '' insertions / deletions of replicas .
these two variants are complementary to each other much like grand - canonical monte carlo ( gcmc ) and widom s test particle insertion method in simulations involving chemical potentials @xcite .
to see how in principle the simultaneous use of multiple configurations ( replicas ) allows for the absolute computation of integrals , let us go back to the volume estimation problem of fig .
[ fig : illustration ] . as illustrated in the bottom panel of that figure , the volume @xmath0 of interest can be estimated by equilibrating it with an infinite reservoir of ideal gas particles ( replicas ) of density @xmath5 , and measuring the average number @xmath6 of particles inside @xmath0 , to obtain @xmath7 .
this physical idea can be implemented computationally by means of a generalized form of the usual grand - canonical monte carlo method @xcite , where the particle reservoir is at density @xmath5 ( or , equivalently , at the corresponding chemical potential ) , and attempted particle insertions / deletions take place in the neighborhood of an existing particle rather than inside a fixed volume that bounds @xmath0 , that neighborhood being defined by a transition matrix @xmath8 .
these are the central ideas that motivated the development of the two versions described below . in general
, we would like to estimate integrals of the type @xmath9 where @xmath10 is the support of the integral , and @xmath11 is a positive - definite function ( see below , however ) of the @xmath12-dimensional vector @xmath13 ; e.g. @xmath14 for most physical problems with energy function @xmath15 .
partition functions of discrete systems , i.e. @xmath16 , can be dealt with in an analogous fashion .
we are given a ( normalized ) transition matrix @xmath8 , such as those routinely used in the trial part of the metropolis algorithm @xcite ; for example , @xmath8 could be a uniform probability distribution up to some distance @xmath17 from @xmath13 ( see dashed circle in fig .
[ fig : illustration ] ) , or a gaussian function centered about @xmath13 . as with metropolis and other markov chain simulations ,
the width of these distributions can be chosen during a preliminary run of the simulation ( see below ) . for non - positive definite integrands
, one can invoke the identity @xmath18 where @xmath4 is defined by eq .
( [ eq : z ] ) , @xmath19 , and the average of the sign function of @xmath11 is with respect to points @xmath13 sampled from @xmath20 .
provided they exist , both quantities on the right hand side of this identity are immediately available from the methods below .
for the first version of the method , we would like to simulate a system of replicas in contact with a reservoir of ideal ( non - interacting ) replicas of specified density @xmath5 , such that each replica @xmath21 in @xmath10 independently samples the distribution @xmath22 .
the corresponding grand - canonical partition function is thus @xmath23 note that , unlike the traditional case where @xmath24 , here the sum starts at @xmath25 , as in the present method ( see below ) replica insertions / deletions take place in the neighborhood of at least one existing replica .
provided we can simulate according to this partition function , the desired integral @xmath4 can be found from the equality @xmath26 which follows straightforwardly from eq .
( [ eq : q ] ) by computing each moment of @xmath27 separately .
( alternatively , one can use @xmath28 or any other relationship between moments of @xmath27 and @xmath4 , and numerically solve the transcendental equation for @xmath4 ; the question of which @xmath4 estimator is more efficient will be left for future studies ) . an algorithm that samples according to
the above grand - canonical partition function goes as follows . at the beginning of the algorithm
we are given at least one point @xmath29 that belongs to @xmath10 ; let us assume the general case where we already have @xmath27 replicas in @xmath10 , and let us denote the vector of coordinates of the replicas by @xmath30 .
we then decide whether to insert or remove a replica , typically but not necessarily ( see methods section)with equal probability . if the decision was an insertion , we sample a new replica coordinate @xmath31 from @xmath32 , where @xmath21 is a randomly chosen coordinate from the existing @xmath27 , and accept its insertion with probability @xmath33 except when @xmath31 lies outside @xmath10 , in which case an immediate rejection takes place . similarly ,
if the decision was to try a deletion , we randomly pick a replica @xmath34 from the existing @xmath27 , and accept its deletion with probability @xmath35 where @xmath36 is @xmath37 excluding @xmath34 , and the sum over @xmath38 excludes @xmath39 .
an exception is the case where only one replica remains , which is always rejected .
a proof that this algorithm samples according to eq .
( [ eq : q ] ) follows by detailed balance ( see methods section ) .
when the replicas correspond to particles inserted uniformly in a fixed volume @xmath0 , i.e. @xmath40 , the above acceptance probabilities reduce to those of ref .
@xcite ( with the due mappings between @xmath5 and chemical potential , and between @xmath11 and the boltzmann factor ) . of course , it is also possible to perform ordinary @xmath41-preserving monte carlo moves on each replica before attempted insertion / deletions @xcite ( fig .
[ fig : sinx ] uses this idea ) . by repeating the above procedure a number of times
, the simulation will eventually equilibrate , and the number of replicas will fluctuate about its mean value @xmath6 . if the equilibration is too slow , i.e. too few replica insertions / deletions are accepted , the width of the distribution @xmath8 about @xmath13 can be adjusted accordingly during a preliminary run , in a fashion analogous to what is done in metropolis monte carlo to keep the rate of accepted trial moves within a reasonable range @xcite . likewise ,
if the number of replicas starts to grow beyond one s computational capabilities , or diminish until it hardly departs from unity , @xmath5 can be adjusted so that @xmath6 is a reasonable number consistent with one s computing power . in practice , for many - particle systems with extensive free energies ( i.e. @xmath42 ) , it is best to write @xmath43 and adjust @xmath44 instead .
note that the present algorithm generalizes standard grand - canonical simulation @xcite in two crucial ways .
first , it inserts and removes entire replicas of the system of interest as opposed to individual particles of a many - body system .
this conceptual difference is essentially what allows one to relate moments of @xmath27 to the partition function of the system of interest ( eq .
( [ eq : z_1st ] ) ) .
second , the replicas are introduced in the neighborhood of an existing replica instead of inside a fixed region , as prescribed by the transition matrix .
this allows for efficient simulation when the integrand of interest is sharply peaked about unknown regions of configuration space ( cf .
[ fig : illustration2 ] ) .
although the use of arbitrary transition matrices in grand - canonical simulations is known in the mathematical literature @xcite , to our knowledge the use of such ideas for integral / partition function estimation is new .
the second version of the replica gas method introduces two important advantages .
first , the number of replicas is constant as opposed to fluctuating , making it more convenient for parallel computing architectures , and second the replicas can be simulated at different temperatures .
these features also make the method easily implemented in existing parallel tempering ( replica exchange ) simulations , thereby benefiting from the greatly enhanced equilibration rates of these simulations @xcite .
the integrals of interest can then be estimated by computing two separate averages involving both the integrand @xmath11 and the transition matrix @xmath8 , as described below . to introduce the method in its simplest form ,
let us first assume that only two replicas exist ( @xmath45 ) , each of them independently sampling the distributions @xmath41 and @xmath46 , via e.g. metropolis .
the integral of interest is @xmath4 , as in eq .
( [ eq : z ] ) , and @xmath47 is an auxiliary integral ; the auxiliary distribution @xmath46 is arbitrary ( for example , it could be @xmath41 itself ) , but in typical applications it corresponds to @xmath41 at a higher temperature , i.e. @xmath48 , where @xmath49 .
then the following identity holds : @xmath50 \int_\omega dx ' \ , t(x'|x ) \cdot \pi(x ' ) } { \int_\omega dx [ \tilde{\pi}(x)/\tilde{z } ] \int_\omega dx ' [ \pi(x')/z ] \cdot t(x'|x ) } \nonumber \\ & \equiv
\frac{\left\langle \pi(x ' ) \right\rangle_{\tilde{\pi},t } } { \left\langle t(x'|x ) \right\rangle_{\tilde{\pi},\pi } } , \label{eq : z_2nd}\end{aligned}\ ] ] where the average @xmath51 of an observable @xmath52 means that configurations @xmath13 ( @xmath53 ) are sampled from the distribution @xmath3 ( @xmath54 ) .
thus , the numerator in the above result requires @xmath13 to be sampled from @xmath55 while @xmath53 is sampled from @xmath8 ( `` virtual replica insertion , '' in analogy with widom s method @xcite ) , and for each such pair of configurations one computes the value of @xmath56 .
likewise , for the average in the denominator , @xmath13 is sampled from @xmath55 while @xmath53 is sampled from @xmath56 , and for each pair one evaluates @xmath8 ( `` virtual replica deletion '' ) . in the limit of infinite samples
, the ratio of these two averages converges to @xmath4 as expressed by eq .
( [ eq : z_2nd ] ) . for simulations with multiple replicas at different temperatures ( @xmath57 ) , one can simply combine ( i.e. sum ) the above result for each pair of replicas .
thus , for the ising model results in fig .
[ fig : ising2d ] , the equation @xmath58 was used .
the sums run over each replica @xmath59 at temperature @xmath60 , except @xmath61 .
the energy function @xmath15 for a spin configuration @xmath13 is the usual ising model function @xmath62 with periodic boundary conditions @xcite .
the transition matrix adopted @xmath8 generates a new spin configuration @xmath53 by flipping each spin of @xmath13 with probability @xmath63 .
thus , @xmath64 , where @xmath65 is the hamming distance ( number of spin mismatches ) between the configurations @xmath13 and @xmath53 , and @xmath27 is the total number of spins .
for illustrative purposes , the results of a volume estimation problem in two spatial dimensions using the version with varying number of replicas are reported in figure [ fig : vol2d ] .
this example was chosen due to its infinite support , a property that would render the use of importance sampling methods difficult , as they would require the design of a non - trivial reference volume @xmath1 with similar support and known quadrature ( recall that in principle we do not know where the integrand is concentrated , or where its boundaries are ) .
the present method performs well in such circumstances without any prior information concerning the support of the integrand , by using a simple uniform transition matrix @xmath8 ( fig .
[ fig : vol2d ] , dashed square ) . as an application to integrals more general than simple volumes , in fig .
[ fig : sinx ] a representative estimate of @xmath66 is shown .
note that this integrand is non - positive definite , so eq . ( [ eq : trick ] ) was used .
the quantities on the right hand side of that equation were estimated using the varying number of replicas version of the method , with the positive - definite integrand @xmath67 . in figure [ fig : ising2d ] the partition function of the two - dimensional ising model is computed to demonstrate the version with fixed number of replicas ( similar results are obtained with the non - fixed version ) . at low temperatures , i.e. ordered states , the replicas are densely localized about the spin - up and spin - down states , and hence a local transition matrix @xmath8 is sufficient to ensure efficient convergence of the averages .
conversely , for higher temperatures close to the disordered state and above , the relevant configuration space and hence the spread of the replicas grows beyond the reach of the local transition matrix adopted , thus causing the averages to converge more slowly ( cf .
right side of fig .
[ fig : illustration2 ] ) . to understand such convergence issues in more detail ,
consider for simplicity eq .
( [ eq : z_2nd ] ) when @xmath68 ( see below for the version with varying number of replicas ) . in order for the averages in eq .
( [ eq : z_2nd ] ) to converge efficiently , the transition matrix @xmath8 has to be such that : * most configurations @xmath53 sampled from @xmath8 fall in typical regions of @xmath41 for any typical configuration @xmath13 sampled from @xmath41 ( so that the numerator is not dominated by those rare configurations with high values of @xmath56 ) ; * most configurations @xmath13,@xmath53 sampled from @xmath41 fall in typical regions of @xmath8 ( so that the denominator is not dominated by those rare events that cause @xmath8 to be of appreciable value ) . in the ising model example of fig .
[ fig : ising2d ] , where @xmath8 typically flips only a few spins of @xmath13 , the low temperature estimates converge faster as typical spin configurations only differ by a few spins , thereby satisfying both requirements , while at higher temperatures close to @xmath69 and above , any two typical spin configurations differ by a substantial number of spins , and hence the requirement ( b ) is violated . adding more replicas , increasing the value of @xmath63 for higher temperatures , or using non - local transition matrices ( such as those of cluster algorithms @xcite )
can alleviate the problem , but such issues will be left for future studies . note
that analogous convergence issues arise in the version with varying number of replicas .
indeed , as can be seen by inspection of eqs .
( [ pacc1 ] ) and ( [ pacc2 ] ) , the acceptance probabilities for insertion and deletion are affected by the choice of @xmath8 much like the averages in eq .
( [ eq : z_2nd ] ) are affected by criteria ( a ) and ( b ) above .
( a separate issue is how well the replicas explore the energy landscape . in the fixed number of replicas version ,
different temperatures are used to overcome energy barriers .
although replica exchange operations can be combined with the varying number of replicas method , as a proof of concept for the convergence issues above , it suffices to start a population of replicas that populate spin - up and spin - down states equally ) .
as illustrated by the above example , the most attractive use of the present methods lies in problems where the integrand is sufficiently localized , so that a general - purpose , local transition matrix can be used to yield efficient results with moderate numbers of replicas .
these are precisely the problems for which existing importance sampling - based methods are most demanding , and thus these methods can be seen as complementary to each other ( see fig .
[ fig : illustration2 ] ) .
it should be noted that the so - called `` flat histogram '' monte carlo methods @xcite are also able to bypass some of the difficulties with importance sampling strategies in some cases , especially for discrete systems .
however , the required human input and scope of such integration methods are rather different : they require the existence and knowledge of suitable order parameters and their ranges ( this being particularly difficult for entropic problems , such as that of fig .
[ fig : vol2d ] ) , knowledge of ground state degeneracies , and for continuum systems suffer from systematic errors due to discretization schemes , although attempts to alleviate some of these problems have been put forward @xcite . in summary ,
the present contribution has introduced two variants of a novel monte carlo strategy for estimating integrals and partition functions .
both versions can be seen as complementary to existing importance sampling or free energy methods @xcite , in that their utility is generally best when the integrands are concentrated in relatively small and unknown regions of configuration space . both continuum and discrete systems are equally amenable to their use . by shifting focus from importance sampling functions to transition matrices ,
it is expected that these methods will encourage a change of paradigm in monte carlo integration .
in this section it will be shown that eqs .
( [ pacc1 ] ) and ( [ pacc2 ] ) satisfy the detailed balance condition @xmath70 according to the grand - canonical partition function eq .
( [ eq : q ] ) , the probability of observing the microstate @xmath37 is given by @xmath71 where the proportionality constant does not depend on the replica coordinates or @xmath27 .
note carefully the difference between the distribution of the labeled microstate @xmath37 , corresponding to replica with label `` 1 '' being at @xmath29 , `` 2 '' at @xmath72 , etc , and that of the unordered set of coordinates @xmath73 , corresponding any replica being at @xmath29 , another arbitrary replica at @xmath72 , etc .
this probability is given by @xmath74 where @xmath75 is the sum over all possible permutations of @xmath76 .
since the replicas are indistinguishable , we have @xmath77 , as in above . of course , it is possible to use either description ( labeled or unlabeled ) , provided the correct probability distributions are used ( eq . ( [ labeled ] ) or eq .
( [ unlabeled ] ) , respectively ) . in this section ,
following @xcite , we will only show the proof using labeled states , i.e. eq . ( [ labeled ] ) .
it is a simple exercise to modify the development below for the unlabeled case ; the acceptance probabilities , of course , are unchanged .
our acceptance probabilities are of the metropolis - hastings type , which by construction satisfy detailed balance . in the present notation ,
the formulas are @xmath78 and analogously for @xmath79 . according to the insertion / deletion algorithm described before eq .
( [ pacc2 ] ) , the trial probabilities for going between the states @xmath80 and @xmath81 are given by @xmath82{q \cdot \frac{1}{n+1 } \cdot \sum_{i=1}^n \frac{1}{n } t(\xi|x_i ) } ( x_1 , \ldots , x_n , \xi ) , \ ] ] where @xmath83 is given by the expression above the arrows , and @xmath84 by the one below them . in the insertion trial probability
, @xmath85 is the probability to try an insertion as opposed to a deletion , @xmath86 is the probability that the new coordinate @xmath87 is inserted in a given slot of the vector @xmath88 ( in the above case , the last slot ) , @xmath89 is the probability to pick coordinate @xmath21 as reference , and @xmath90 is the probability to sample the candidate position @xmath87 given the chosen reference . in the deletion trial probability , @xmath91 is the probability to try a deletion , and @xmath86 is the probability that the replica at @xmath87 will be chosen for attempted removal among the existing ones . plugging these trial probabilities in eq .
( [ methastings ] ) , we obtain eq .
( [ pacc1 ] ) ( an analogous procedure gives eq .
( [ pacc2 ] ) ) .
the case where @xmath92 can be easily taken care of by modifying the acceptance probabilities accordingly .
alternatively , detailed balance can be directly proven by plugging the trial probabilities in eq .
( [ trial ] ) and the acceptance probabilities given by eqs .
( [ pacc1 ] ) and ( [ pacc2 ] ) into eq .
( [ detbalance ] ) .
monte carlo estimation of a volume @xmath0 ( shaded region ) by means of sampling from a reference volume @xmath1 ( top ) , and equilibration with a hypothetical , infinite reservoir of ideal gas particles at density @xmath5 ( bottom ) . in the former ,
one draws random points uniformly from @xmath1 and counts the fraction @xmath3 that lands in @xmath0 , to obtain @xmath93 . in the latter , one performs a grand - canonical monte carlo simulation @xcite at reservoir density @xmath5 , and monitors the average number @xmath6 of ideal gas particles in @xmath0 ; upon equilibration , the density of particles in @xmath0 equals that of the reservoir , and thus @xmath94 .
( in practice , due to the constraint @xmath95 , this formula for @xmath0 needs to be modified slightly ; see eq .
( [ eq : z_1st ] ) ) .
each particle corresponds to a point ( `` replica '' ) residing in @xmath0 , and attempted replica insertions / removals take place in the neighborhood ( dashed circle ) of an existing replica @xmath13 , defined by the transition matrix @xmath8 of the method .
a version of the algorithm with fixed number of replicas possibly at different temperatures is also described in the text . ]
estimation of a two - dimensional volume @xmath4 ( yellow region ) using the replica gas method with varying number of replicas , eq .
( [ eq : z_1st ] ) . the volume @xmath4 is defined by the region @xmath96 for @xmath97 , and @xmath98 for @xmath99 .
the points correspond to the replica configurations at the end of one simulation , and the dashed square defines the boundaries of the adopted transition matrix @xmath8 ( uniform distribution , each side of length unity ) for the particular configuration @xmath13 shown .
_ inset : _ histogram of @xmath100 independent estimates of @xmath4 using the replica gas method with @xmath101 attempted insertions / deletions , and @xmath102 .
the exact value of @xmath4 , obtained by analytic quadrature , is @xmath103 . ]
an illustrative non - positive definite integrand , @xmath104 .
the scale in the center of the graph corresponds to the parameter @xmath105 in the uniform transition matrix @xmath106 for @xmath107 ( zero otherwise ) adopted for the results shown in the inset .
_ inset : _ running estimate of @xmath108 using eq .
( [ eq : trick ] ) , where @xmath109 is estimated via the replica gas method with varying number of replicas , eq .
( [ eq : z_1st ] ) .
the mean sign function of @xmath41 required by this last equality , @xmath110 , is also obtained from this run , by averaging @xmath111 over all replicas @xmath13 during the simulation .
the dashed red line is the exact result @xmath112 .
for this example , @xmath113 , and @xmath114 ordinary monte carlo moves per replica are performed between every attempted insertion / deletion ( gcmc step ) .
the ensuing number of replicas fluctuated about @xmath115 . ]
natural logarithm of the partition function @xmath4 of the two - dimensional ising model with @xmath116 spins , according to the replica gas method ( circles ) , and the exact kaufman formula @xcite ( dashed curve ) .
the replica gas results were obtained using the fixed number of replicas version , eq .
( [ eq : z_2nd_manybeta ] ) , with @xmath100 replicas at the temperatures corresponding to the data points shown ( similar results are obtained with varying number of replicas , see text ) , with error bars indicating the standard deviation of 8 independent runs .
importance sampling results were obtained using the ideal ( non - interacting ) spin partition function @xmath117 and @xmath118 , with @xmath119 independent configurations @xmath13 sampled from the ideal reference system ( increasing this number to @xmath101 does not lead to appreciable changes in the results ) . for disordered states ( i.e. @xmath120 close to or higher than the critical temperature @xmath121 ) ,
the partition function is no longer dominated by a small fraction of the configuration space , and the replica gas method converges more slowly with the adopted ( local ) transition matrix ( see also fig .
[ fig : illustration2 ] ) .
the replica exchange simulation took @xmath119 mc steps , with an attempted exchange every @xmath122 steps , where each mc step is a simple spin flip .
for the transition matrix sampling , @xmath123 ( cf .
discussion after eq .
( [ eq : z_2nd_manybeta ] ) ) . ]
comparison of the merits of importance sampling ( left ) and replica gas ( right ) methods for sparse ( top ) and localized ( bottom ) integrands .
integrands are represented by their densest regions ( shaded curvy shapes ) . for physical systems ,
sparse integrands correspond to boltzmann factors at high temperatures , while at low temperatures the integrands tend to be localized in a small fraction of configuration space ( e.g. magnetized spin systems , crystals , proteins in their native state , etc ) . in importance sampling ,
one typically has at their disposal a general - purpose sparse reference system ( polygon on the left ) such as an ideal gas , which is generally sufficient to ensure proper sampling at high temperatures , but not at lower ones .
conversely , in replica gas methods one typically has at their disposal a local transition matrix ( small boxes on the right ) that is generally sufficient to sparingly `` cover '' the integrand of interest ( cf .
efficiency criteria ( a ) and ( b ) discussed in the text ) at low temperatures , but not at higher temperatures . ] | owing to their favorable scaling with dimensionality , monte carlo ( mc ) methods have become the tool of choice for numerical integration across the quantitative sciences .
almost invariably , efficient mc integration schemes are strictly designed to compute ratios of integrals , their efficiency being intimately tied to the degree of overlap between the given integrands .
consequently , substantial user insight is required prior to the use of such methods , either to mitigate the oft - encountered lack of overlap in ratio computations , or to find closely related integrands of known quadrature in absolute integral estimation . here a simple physical idea measuring the volume of a container by filling it up with an ideal gas is exploited to design a new class of mc integration schemes that can yield efficient , absolute integral estimates for a broad class of integrands with simple transition matrices as input .
the methods are particularly useful in cases where existing ( importance sampling ) strategies are most demanding , namely when the integrands are concentrated in relatively small and unknown regions of configuration space ( e.g. physical systems in ordered / low - temperature phases ) .
examples ranging from a volume with infinite support to the partition function of the 2d ising model are provided to illustrate the application and scope of the methods . |
thompson s groups @xmath0 and @xmath3 have been studied from many perspectives .
both groups can be understood as groups of locally orientation - preserving piecewise - linear maps of the unit interval . in the case of @xmath0 ,
these maps are homeomorphisms , and in the case of @xmath3 the maps are right - continuous bijections . in both cases the breakpoints and discontinuities
are restricted to be dyadic rational numbers , and the slopes , when defined , are powers of 2 .
both groups can also be understood by means of rooted binary tree pair diagrams order - preserving in the case of @xmath0 .
cannon , floyd and parry @xcite give an excellent introduction to these groups and several approaches to understanding their properties .
a rooted binary tree is a finite tree where every node has valence 3 except the root , which has valence two , and the leaves , which have valence one .
we draw such trees with the root on top and the nodes descending from it to the leaves along the bottom .
the two nodes immediately below a node are its _
children_. a node and its two children form a _ caret_. a caret whose two children are leaves is called an _ exposed caret_. we number the leaves of a rooted binary tree with @xmath4 carets and @xmath5 leaves from @xmath6 to @xmath4 in any order , although we frequently choose to number the leaves of a rooted tree in order from left to right . a _ tree pair diagram _ is a triple @xmath7 , where @xmath8 and @xmath9 are two binary trees with the same number of leaves @xmath4 , and @xmath10 is a permutation in @xmath11 . a reduction can be performed in a diagram if both leaf numbers of an exposed caret in @xmath8 are mapped by @xmath10 to the two leaf numbers of an exposed caret in @xmath9 . in cases where a reduction is possible , we can replace the exposed caret with a leaf and renumber leaves to give an equivalent representative with a new permutation in @xmath12 .
a tree pair diagram is _ reduced _ if no reductions are possible .
the set of binary tree pair diagrams thus admits an equivalence relation , whose classes consist of those diagrams which have a common reduced representative , with such reduced representatives being unique . in figure [ fig : velem ] a reduced diagram is depicted , where the leaf numbering describes the permutation .
the elements of @xmath3 which are actually in @xmath0 are precisely those elements for which the permutation is the identity .
composition in @xmath3 can be understood by means of these binary tree diagrams . if two elements in @xmath3 are given by their representative diagrams @xmath7 and @xmath13 ,
their composition can be found by finding two tree diagrams in the corresponding equivalence classes , @xmath14 and @xmath15 , such that @xmath16 .
when this is achieved , the product element is represented by the diagram @xmath17 .
[ fig : velem],title="fig:",width=288 ] + brin @xcite and dehornoy @xcite describe braided thompson s group @xmath18 , incorporating braids into tree pair diagrams .
the permutation @xmath10 in a triple for an element of @xmath3 is replaced by a braid , giving the notion of a _ braided tree diagram_. a braided tree diagram is then a triple @xmath19 , where the two trees have the same number of leaves @xmath4 , and @xmath20 is a braid in @xmath4 strands .
reductions can still be performed if two exposed carets are joined by two parallel strands . in figure
[ fig : bvelem ] an element of @xmath18 is depicted .
we now draw the rooted tree @xmath8 with the root at the top , and the tree @xmath9 below with the root at the bottom , and then draw the braid between the leaves of the two trees as indicated .
, one of the preimages of the element drawn in figure [ fig : velem ] under the map @xmath21.[fig : bvelem],title="fig:",width=144 ] + in @xcite , brin also describes a larger " group @xmath22 as a group of braided forest diagrams having all but finitely many of the forests trivial .
not only does @xmath22 naturally contain @xmath18 as the subgroup of forest diagrams where all trees are trivial except the first pair , but @xmath22 also sits inside @xmath18 as the subgroup of braided tree diagrams where the rightmost strand is always unbraided .
we begin this paper by providing finite and infinite presentations of @xmath18 which contain the presentations provided by brin in @xcite as subpresentations .
next , we describe subgroups @xmath1 of @xmath18 and @xmath2 of @xmath22 . just as @xmath0 is the subgroup of @xmath3 of order - preserving right continuous bijections of @xmath3 , the groups @xmath1 and @xmath2 are the subgroups of order - preserving elements of the braided versions of @xmath3 .
the order is preserved by using generators which come from @xmath0 and generators which involve pure braids .
we describe normal forms for elements in these subgroups and obtain infinite and finite presentations for these groups .
dehornoy @xcite calls this pure braid subgroup @xmath23 , the group of pure parenthesized braids .
in @xcite , an infinite presentation for @xmath22 is given .
the generators in this presentation are the generators in the standard infinite presentation for thompson s group @xmath0 , as well as the generators for @xmath24 . here
@xmath24 is considered as a direct limit of the groups @xmath25 , where @xmath25 is included in @xmath26 via adding one strand at the right .
now @xmath22 sits naturally as a subgroup of @xmath18 ; it is isomorphic to the subgroup of all elements represented by braided tree diagrams in which the rightmost strand is unbraided . although presentations for @xmath18 , both finite and infinite , are given in @xcite , they are not related in a simple way to the presentation for @xmath22 . instead
, we give a presentation for @xmath18 which contains brin s presentation for @xmath22 as a subpresentation .
first , we define the set of generators .
any element of @xmath18 can be represented by a braided tree diagram @xmath27 where both @xmath8 and @xmath9 have @xmath4 leaves and @xmath28 is a braid in @xmath25 .
a single tree can be thought of as a _ positive _ element of thompson s group @xmath0 , when we take it as being paired with an _ all - right _ tree , which is a tree whose carets are all right children of their parent carets .
these positive elements correspond to elements which are positive words in @xmath0 with respect to the infinite generating set @xmath29 the correspondence between tree pair diagrams and normal forms with respect to the infinite generating set is given by the process of exponents of leaves , as described by cannon , floyd and parry @xcite and fordham @xcite .
all - right trees have all leaf exponents zero , and thus the normal forms for tree pair diagrams which involve one all - right tree will be purely negative or purely positive .
we will denote by @xmath30 the all - right tree which has @xmath4 leaves .
we can factor an element @xmath19 into three pieces , using all - right trees of the appropriate number of leaves , in a manner similar to that done for elements of thompson s group @xmath31 by burillo , cleary , stein and taback @xcite .
the resulting three elements in this factorization are @xmath32 and the product of the elements represented by these three diagrams yields the original group element . in general , these three tree pair diagrams will not be reduced ; in order for each of them to have the same number of carets , we may need to take unreduced representatives for as many as two of the three terms . by enlarging trees in this manner ,
it is clear that every element of @xmath18 can be factored this way .
hence , we can always think of an element of @xmath18 as if it were composed of two elements of thompson s group @xmath0 , one positive and one negative , and one braid .
it makes sense then to consider , as a set of generators of @xmath18 , the set of generators for @xmath0 and the set of generators for the braid groups , interpreted as braided tree pair diagrams between all - right trees .
the infinite set of generators for @xmath0 consists of the elements @xmath33 with @xmath34 .
figure [ fig : x2 ] shows @xmath35 in both tree pair diagram form and in braided tree form .
these generators from @xmath0 are enough to produce the two elements @xmath36 and @xmath37 in @xmath18 . of @xmath0 , in standard and in braided form.[fig :
x2],title="fig:",width=288 ] + we can consider the element @xmath38 as an element in the appropriate braid group @xmath25 . now
this copy of @xmath25 is generated by @xmath39 transpositions , the @xmath40 of which braids strand @xmath41 over strand @xmath42 .
we do not , however , need to include all of these as generators .
the generators which do not involve braiding the last two strands can be obtained from the generators of @xmath43 by splitting the last strand .
for this reason we must consider two sets of generating braids , one which leaves the rightmost strand unbraided and one which does not .
we define @xmath44 to be the element represented by the braided tree diagram @xmath45 , where @xmath46 is the braid on @xmath47 strands which crosses strand @xmath41 over strand @xmath42 .
similarly , @xmath48 is the element represented by the diagram @xmath49 , where @xmath50 is the braid on @xmath42 strands which crosses strand @xmath41 over stand @xmath42 .
then the set @xmath51 generates the copy of @xmath25 containing all elements of @xmath18 represented by diagrams of the form @xmath52 .
notice that the @xmath33 together with the @xmath44 generate the copy of @xmath22 inside @xmath18 , and they correspond to brin s generators .
we have shown : [ bvgens ] the elements @xmath33 , for @xmath53 , @xmath44 for @xmath54 , and @xmath48 for @xmath54 form a set of generators for @xmath18 .
there are three types of natural relations among these generators .
first , there are the generators involving only the generators of @xmath0 , namely @xmath55 for @xmath56 .
these are the generators for the standard presentation for @xmath0 .
next , we expect to need generators for each copy of @xmath25 . these yield four types of relations : * @xmath57 , for @xmath58 * @xmath59 * @xmath60 , for @xmath61 * @xmath62 . finally , there are relations governing the interactions between the generators for @xmath0 and the generators for the braid groups @xmath25 .
* @xmath63 , for @xmath64 * @xmath65 * @xmath66,for @xmath67 * @xmath68 * @xmath69 , for @xmath70 * @xmath71 * @xmath72 . in preparation for showing that the relations above give a presentation , we first introduce a special class of words in the generators .
we would like to identify those words in the generators which could be identified easily with a triple of diagrams in @xmath18 .
as noted earlier , any element of @xmath18 can be represented by a triple of braided tree diagrams of the form @xmath73 .
such a triple leads easily to a word in the generators as follows .
the group element represented by the first diagram is a positive element @xmath74 of @xmath75 , and may be expressed uniquely as a word of the form @xmath76 , where @xmath77 and @xmath78 for all @xmath79 .
similarly , the group element @xmath80 represented by the third diagram can be uniquely expressed as a word of the form @xmath81 , where @xmath82 and @xmath83 for all @xmath79 .
now the group element represented by the middle diagram may be represented as some word in the generators @xmath84,@xmath85 , and their inverses . for convenience , we will call any word in this set of generators and their inverses _ a word in the @xmath25 generators_. note that if such a word contains no @xmath86 generators , it can be considered a word in the @xmath25 generators for many values of @xmath4 .
notice that the minimum number of carets required in the trees for tree pair diagrams representing @xmath74 and @xmath87 respectively , is at most @xmath39 .
the concatenation of the three words described above yield a word which can not serve as a normal form , since we have not specified preferred arrangements of the @xmath88 s and @xmath86 and furthermore , there are many different triples of tree pair diagrams representing any element . however , these words are nice in that any word of the above special form can be easily translated into a triple of diagrams , and we find them to be useful tools . given a word @xmath89 , denote by @xmath90 the number of carets in the reduced binary tree diagram representing it . here is the algebraic description of blocks , which are these words which come from a single triple of diagrams .
a word in the generators @xmath91 is called a block if it is of the form @xmath92 where * @xmath93 is of the form @xmath76 , where @xmath77 and @xmath78 for all @xmath79 .
* @xmath94 is of the form @xmath95 , where @xmath96 and @xmath83 for all @xmath79 , and by @xmath97 we mean the word @xmath98 .
* let @xmath99 .
then there exists an integer @xmath4 , @xmath100 , such that @xmath101 is a word in the @xmath25 generators .
then we have the following lemma : [ blockid ] a block @xmath92 is the identity in @xmath18 if and only if @xmath93 and @xmath94 are the same word , and @xmath101 is the identity in the copy of the braid group @xmath25 generated by @xmath102 and @xmath103 where @xmath104 .
the lemma follows directly from the fact that any word which is a block @xmath92 can be represented by a braided tree diagram @xmath27 where @xmath93 is represented by @xmath105 , @xmath101 is represented by @xmath38 , and @xmath97 is represented by @xmath106 , and from any such triple of diagrams a block can be read off , unique up to the choice of the word in @xmath25 expressing @xmath28 . since the identity in @xmath18
can be represented by the diagram consisting of the tree with only one vertex , and the trivial braid on one strand , all other diagrams representing the identity result from splitting strands , and will always have two identical trees with the trivial braid .
but such diagrams translate into blocks of the form described in the lemma .
we will use these blocks to prove : [ bvinfpres ] the group @xmath18 admits a presentation with generators : * @xmath33 , for @xmath53 , * @xmath44 , for @xmath54 , * @xmath48 , for @xmath54 . and relators * @xmath107 * @xmath108 * @xmath109 * @xmath110 * @xmath111 .
* @xmath112 * @xmath65 * @xmath113 * @xmath68 * @xmath114 * @xmath115 * @xmath72 .
this presentation appears without proof in j. belk s thesis @xcite .
let @xmath116 be the abstract group given by the presentation above .
we map @xmath116 to @xmath18 via @xmath117 by sending each generator to the element of @xmath18 with the same name .
all relations in the presentation hold in @xmath18 , so @xmath117 is a well - defined homomorphism .
proposition [ bvgens ] shows that the map is surjective , so it remains only to show that @xmath117 is injective . to show this
, we must show that any word in the generators which maps to the identity in @xmath18 is already the identity in @xmath116 .
now by lemma [ blockid ] , this is true if the word in question happens to be a block .
so we are done once we show that the relations in @xmath116 are sufficient to transform any word into a block .
but since any generator is itself a block , an arbitrary word of length @xmath118 is trivially the product of @xmath118 blocks .
so to show @xmath117 is injective it is sufficient to prove that a word in @xmath116 which is the product of two blocks can be rewritten , using the relations in @xmath116 , as a single block .
we first prove a series of 3 preliminary lemmas , from which we will deduce this fact in lemma [ 2blockstoone ] , which will complete the proof of the theorem .
our first lemma permits us to push the @xmath119 generators to the left or right of the braid generators , which helps move a word toward block form .
[ push ] if @xmath120 is a word in the @xmath25 generators , and @xmath121 , then @xmath122 is equivalent in @xmath116 to either @xmath123 or @xmath124 , where @xmath124 is a word in the @xmath26 generators , and @xmath125 .
similarly , under the same conditions on all indices , a word @xmath126 may be replaced by either @xmath127 or @xmath124 .
we describe first how to push @xmath128 past the @xmath88 and @xmath86 generators . pushing past @xmath88 type generators
is always possible , but in order to push past @xmath86 s we must carefully keep track of the index of the @xmath129 as it moves along . using the relations of type c , we may replace @xmath130 by @xmath131 , where @xmath132 is a word in the @xmath88 generators and their inverses of length one or two .
furthermore , the maximum index appearing in @xmath132 is @xmath133 . now
the index @xmath134 can , in general , be either @xmath135 , or @xmath136 .
however , it only increases to @xmath136 in the case where @xmath137 also .
so since the initial index @xmath41 satisfies @xmath121 and @xmath138 , even a series of such replacements results in the presence of @xmath139 with @xmath125 .
this is important , since relations @xmath140 and @xmath141 allow replacement of @xmath142 by @xmath143 , and relations @xmath144 allow us to replace @xmath145 by @xmath146 if @xmath147 .
hence , @xmath148 can be replaced by either @xmath149 or simply @xmath124 as claimed .
the argument for @xmath126 is similar .
next we prove a lemma showing that a word in the @xmath25 generators can always be pumped up to a word in the @xmath26 generators at the expense of tacking on an @xmath150 generator . [ pumping ]
let @xmath120 be a word in the @xmath25 generators . then using the relators in @xmath116 ,
@xmath120 may be replaced by either @xmath124 or @xmath151 where @xmath121 and @xmath124 is a word in the @xmath26 generators .
similarly , @xmath120 may also be replaced by either @xmath124 or @xmath152 where @xmath124 is a word in the @xmath26 generators .
consider the leftmost occurrence of @xmath85 in the word @xmath120 , that is , @xmath153 , where @xmath93 has only @xmath88 generators .
using relations @xmath140 or @xmath141 depending on the exponent of @xmath85 , replace @xmath120 by @xmath154 , and then apply lemma [ push ] to @xmath155 to replace it with either @xmath156 or @xmath157 with @xmath121 and where @xmath157 is a word in the @xmath26 generators .
then the desired @xmath124 is @xmath158 .
similarly , working from the right , @xmath120 can be replaced by either @xmath159 or @xmath124 .
the two previous lemmas will now be used to show that the relators allow us to transform the product of two blocks to a new product of two blocks where the combined length of the middle two of the 6 subwords involved is reduced .
[ squish ] let @xmath160 and @xmath161 be two blocks .
then the relations in @xmath116 allow us to replace the word @xmath162 by @xmath163 , the product of two blocks @xmath164 and @xmath165 , where @xmath166 .
let @xmath167 and @xmath168 be the first letters in @xmath94 and @xmath169 . if they are the same , we can delete the pair @xmath170 and we are done .
if not , suppose @xmath171 ( if @xmath172 a similar argument works , truncating @xmath173 and absorbing @xmath168 into @xmath120 ) .
let @xmath174 , @xmath175 , and let @xmath176 be @xmath94 with @xmath167 deleted .
now we use the relations ( a ) to replace @xmath177 by @xmath178 . note that @xmath179 and @xmath169 have the same length , but @xmath180 , since each index in @xmath169 is increased by @xmath181 as @xmath182 moves past it ( see theorem 3 of @xcite ) .
next , suppose @xmath183 is a word in @xmath184 and @xmath185 with @xmath186 .
then @xmath187 and @xmath188 , so @xmath189 .
we must replace @xmath190 by @xmath191 so that @xmath192 is a block .
we will consider two cases .
_ case 1 : _ if @xmath193 , we use lemma [ push ] to replace @xmath194 by @xmath195 with @xmath121 and @xmath196 a word in @xmath197 and @xmath185 .
then @xmath198 can be rewritten using relations ( a ) as a word @xmath199 so that @xmath200 is a positive word in the generators of @xmath0 with increasing indices from left to right .
then it again follows from @xcite that @xmath201 .
hence @xmath202 , since @xmath203 and @xmath204 , and this implies that @xmath205 is a block as desired .
_ case 2 : _ if @xmath206 , it is necessary to first use lemma [ pumping ] to replace @xmath183 by either @xmath207 or @xmath208 where @xmath121 . if @xmath128 is present , we use relations ( a ) to replace @xmath198 by @xmath209 , and we see that @xmath210 . we continue applying lemma [ pumping ] and absorbing any resulting @xmath128 letters into the @xmath211 part of the word in this manner , and after @xmath212 repetitions we have replaced @xmath213 by @xmath214 , where @xmath207 a word in the @xmath215 generators , and @xmath216 is a word in the @xmath33 with indices increasing from left to right with @xmath217 . now just as before we can apply lemma [ push ] to replace @xmath214 by @xmath218 where @xmath219 , and @xmath196 is a word in the @xmath220 generators . when we use relations ( a ) to replace @xmath221 by @xmath222 , @xmath223 .
but since @xmath224 and @xmath225 , @xmath226 , and hence @xmath227 is a block .
now we are in a position to prove the final lemma which completes the proof of theorem [ bvinfpres ] .
[ 2blockstoone ] the product of two blocks may be rewritten , using the relations of @xmath116 , as a single block .
let @xmath160 and @xmath161 be two blocks .
we apply lemma [ squish ] , at most @xmath228 times , to replace @xmath162 by @xmath229 where @xmath230 is a word in the @xmath25 generators , @xmath196 is a word in the @xmath231 generators , @xmath232 , and @xmath233 . if @xmath234 , declaring @xmath235 , @xmath236 , and @xmath237 shows that @xmath238 is a block .
if not , say @xmath239 , we apply lemma [ pumping ] @xmath240 times to replace @xmath196 by a word @xmath207 , a word in the @xmath25 generators , followed by some new @xmath129 generators , so that @xmath191 has been replaced by @xmath241 where @xmath242 for @xmath243 . now
so inductively , we have that @xmath245 , and hence declaring @xmath235 , @xmath246 , and @xmath247 , shows that @xmath248 is a block .
of course , if @xmath249 , a similar argument works , pumping up indices in @xmath230 instead .
we remark that an infinite presentation of thompson s group @xmath3 is easily obtained from the presentation for @xmath18 by adding two more infinite families of relators , @xmath250 and @xmath251 for all @xmath41 .
it is common for these type of infinite presentations for thompson - type groups to reduce to finite presentations . for example , in @xcite an inductive argument is spelled out which obtains the standard two generator - two relator presentation for @xmath0 from the standard infinite presentation .
brin uses similar arguments to obtain finite presentations for @xmath18 and @xmath22 from his infinite ones . in a similar manner ,
the infinite presentation for @xmath18 in the previous section reduces to a finite presentation with 4 generators and only 18 relators , an improvement over the presentation in @xcite , which has 4 generators and 26 relators .
[ finitebv ] the group @xmath18 admits a finite presentation with generators @xmath252 and relators * @xmath253 , @xmath254 * @xmath255 , @xmath256 , @xmath257 , @xmath258 * @xmath259 , @xmath260 * @xmath261 , @xmath262 * @xmath263 , @xmath264 * @xmath265 , @xmath266 * @xmath267 * @xmath268 * @xmath269 * @xmath270 where the letters in the relators not in the set of 4 generators are defined inductively by @xmath271 for @xmath272 , @xmath273 for @xmath54 , and @xmath274 for @xmath54 . that the two ( a ) relators yield inductively all ( a ) relators in the infinite presentation is a standard argument , given in @xcite .
notice that the relators ( c2 ) and ( d3 ) in the infinite presentation are precisely the relations used to inductively define the higher index generators in the infinite presentation .
now a straightforward induction yields the ( c1 ) relators in the infinite presentation from the ( c1 ) relators , then the ( c3 ) relators from the ( c3 ) relators , and so on , in the order the groups of relators are listed in the finite presentation above . as an example , we spell out the induction for the ( b3 ) relators .
so suppose we have ( b3 ) , or @xmath269 .
then suppose inductively that we have established @xmath275 for @xmath276 , where @xmath277 .
then @xmath278 using the relator defining @xmath279 .
now we can move the @xmath280 to the right , first using the ( c1 ) relators , then the inductive hypothesis , and finally the ( b1 ) relators , and then use the defining relation for @xmath279 in the other direction , to obtain the relator @xmath281 . therefore , by induction , @xmath281 for all @xmath282 .
now suppose that we have @xmath283 for @xmath284 , @xmath285 , and @xmath286 .
then it follows that @xmath287 for @xmath288 .
we replace @xmath289 in the word @xmath290 by @xmath291 , and then moves the @xmath292 to the left , first using the inductive hypothesis to obtain @xmath293 , then using the ( d1 ) relators to obtain @xmath294 , and finally using the inductive hypothesis again to obtain @xmath295 . now use ( d1 ) relators
to move @xmath296 left , to obtain @xmath297 .
but now the rightmost four letters can be replaced by @xmath289 using the defining relation for @xmath289 in reverse , showing that @xmath287 .
hence , by induction , all ( b3 ) relators hold . for the corresponding finite presentation for @xmath3
, we note that the relations @xmath251 and @xmath250 for @xmath298 can be deduced inductively from the two relations @xmath299 , using the ( c2 ) and ( c3 ) relators in the case of @xmath44 , and the ( d2 ) and ( d3 ) relators for @xmath48 .
this yields a presentation for @xmath3 with 4 generators and 20 relators , not quite as efficient as the presentation in @xcite with 14 relators .
the pure braid groups @xmath301 are the groups of braids where the @xmath41th strand is braided with the other strands but returns to the @xmath41th position .
there are several possible ways to construct analogous subgroups of @xmath18 .
one way is by considering the standard short exact sequences for the braid groups , involving the pure braid groups and the permutation groups . for each @xmath4
, we have : @xmath302 which maps a braid to its permutation , and whose kernel is the pure braid group @xmath301 .
this family of maps @xmath303 collectively induces a map @xmath304 defined by @xmath305 , where we use the appropriate @xmath303 for the number of leaves in either tree .
let @xmath306 .
by definition , a diagram @xmath19 represents an element in @xmath300 if it maps to the identity in @xmath3 , that is , if @xmath307 , and if @xmath308 .
hence , @xmath300 is the subgroup of @xmath18 which consists of those elements which admit a representative @xmath309 on which the two trees are the same and the braid is pure .
if an element admits one representative where the two trees are equal , then every representative will have the trees being equal .
so @xmath300 is a subgroup , because the product of such two elements also has representatives where the two trees are equal .
note that it is crucial in this construction that the braid is a pure braid .
the main result concerning the group @xmath300 is the following .
[ pbvnotfp ] the group @xmath300 is not finitely generated .
_ given two elements of @xmath300 by their diagrams @xmath309 and @xmath310 , their product always admits a representative diagram @xmath311 , where @xmath312 is the _ least common multiple _ of @xmath313 and @xmath31 that is , the minimal tree which contains both @xmath313 and @xmath31 as subtrees .
hence , if @xmath300 were to be generated by a finite set @xmath314 , for @xmath315 , every tree in @xmath300 would admit a representative whose tree would be the least common multiple of the @xmath316 .
there are elements whose smallest representatives are of increasing size and thus @xmath300 can not be finitely generated .
from the map @xmath21 defined above , since @xmath0 is the subgroup of @xmath3 of those elements whose permutation is the identity , we can define the group @xmath317 .
the group @xmath1 is the subgroup of @xmath18 of those elements which admit a representative @xmath318 , where @xmath319 is a pure braid .
note the contrast with @xmath300 , since here the two trees are not necessarily equal .
in fact , @xmath300 is a subgroup of @xmath1 , and observing the restriction to @xmath1 of the map @xmath21 above , it is easy to see that @xmath300 is also the kernel of @xmath320 .
thus the diagram below is commutative : @xmath321 the main goal of the remainder of this paper is to prove that @xmath1 is finitely presented and to find both finite and infinite presentations .
finding generators for @xmath1 is not difficult . just as for @xmath18 , an element of @xmath1
is given by a triple @xmath318 , where this time @xmath319 is a pure braid .
again , we factor the element into three pieces @xmath322 where the individual diagrams may not be reduced . hence , we can always think of an element of @xmath1 as if it were composed of two elements of thompson s group @xmath0 , one positive and one negative , and one pure braid .
again , just as for @xmath18 , we take as a set of generators of @xmath1 , the set of generators for @xmath0 and the set of generators for the pure braid groups , interpreted as braided tree pair diagrams between all - right trees .
we consider the element @xmath323 as an element of the appropriate group @xmath301 of pure braids . to generate these groups
@xmath301 we would like to use the braids @xmath324 , for @xmath325 , which wrap the @xmath41-th strand around the @xmath326-th one .
see hansen @xcite for details of these generating sets .
the process of obtaining the generators of @xmath1 from the generators @xmath324 of @xmath301 is the same than the process specified above for @xmath18 from the standard braid generators .
we will denote by @xmath327 the element @xmath328 , and by @xmath329 the element @xmath330 . as in @xmath18 ,
the differences between these two families of generators are whether or not the last strand is involved in the braiding .
figure [ fig : a13 ] shows an example of the two generators of @xmath1 corresponding to a generator @xmath324 . in @xmath331 , with its corresponding generators @xmath332 and @xmath333 in @xmath1 .
[ fig : a13],title="fig:",width=288 ] + the proof of the following proposition is analogous to the proof of proposition [ bvgens ] .
[ gens ] the elements @xmath33 , for @xmath34 , @xmath327 , for @xmath334 , and @xmath329 , for @xmath334 , form a set of generators of @xmath1 . in the next theorem
we will give a presentation for the group @xmath1 .
the relators are going to be divided into four families .
the family ( a ) is obtained from the relators of @xmath0 .
the family ( b ) is obtained from the presentation of the pure braid group : * @xmath335 * @xmath336 * @xmath337 * @xmath338 obtained from @xcite .
the family ( d ) reflects the interactions between generators of @xmath0 and pure braids . in @xcite
, brin constructs these relators using the structure of zappa - szp product of the monoid associated to @xmath22 .
this construction is not possible here because the presentations for the groups @xmath339 are not monoid presentations .
in fact , the monoid of pure positive braids is not finitely generated as shown by burillo , gutierrez , krsti and nitecki .
@xcite .
the family ( c ) of relators is given by the special way that the pure braid groups are embedded into each other inside @xmath1 .
to embed @xmath301 into @xmath340 we split the last strand in two .
if the last strand is not braided , this does not affect the element , but if the last strand takes part in the actual braiding , then these elements in @xmath301 change when embedded in @xmath340 . when a generator @xmath329 has its last strand split , now the @xmath41-th strand wraps around two strands ( the @xmath326-th and the @xmath341-th ) , and the element is now a product of two generators .
the group @xmath1 admits a presentation with generators : * @xmath33 , for @xmath34 , * @xmath327 , for @xmath334 , * @xmath329 , for @xmath334 and relators : * @xmath342 * @xmath343 * @xmath344 * @xmath345 * @xmath346 * @xmath347 * @xmath348 * @xmath349 * @xmath350 * @xmath351 * @xmath352 * @xmath353 * @xmath354 * @xmath355 * @xmath356 * @xmath357 * @xmath358 * @xmath359 * @xmath360 _ proof .
_ as in the proof of theorem [ bvinfpres ] , we consider the algebraic and the geometric group , establish a homomorphism between them , which is well - defined and surjective .
for instance , one needs to check geometrically the relators to see it is well defined .
see figure [ fig : relator ] for an example .
after this , it only remains to check the injectivity of the map . and
the following steps are the multiplication processes which transform it into @xmath361[fig : relator],title="fig:",width=480 ] + to prove that the homomorphism is injective , we consider an element of @xmath116 , given as a word @xmath362 , and imagine that it is mapped to the identity in @xmath116 .
we need to prove that it is consequence of the relators listed above . as a first step
, we can see that the relators ( d1 ) to ( d9 ) and ( a ) can be used to transform any such word into the product of three words , @xmath363 we can arrange this in such a way that the words @xmath93 and @xmath94 contain only generators @xmath33 and not their inverses . this special expression is then particularly useful for studying the element , because it corresponds easily to the representative @xmath318 .
we will use now the following lemma whose proof is analogous to the proof of lemma [ blockid ] : a triple @xmath318 in @xmath1 represents the identity element if and only if the braid @xmath319 is the trivial braid and the element @xmath364 represents the identity in @xmath0 . with this lemma , we can assume now that the two words @xmath365 and @xmath101 map to the identity , and we must prove that they are consequence of the relators .
the word @xmath366 lies in the subgroup isomorphic to @xmath0 .
so if it is the identity , it is consequence of the relators ( a ) .
the word @xmath101 is a product of some @xmath367 and @xmath368 generators .
we would like to consider this word inside some @xmath301 , for a fixed @xmath4 .
the image in @xmath1 of @xmath329 has @xmath326 strands , and the image of @xmath327 has @xmath133 strands .
so the appropriate @xmath4 to use is the maximum of the following set : @xmath369 if we have a @xmath329 with @xmath370 , we use the relators ( c ) to increase the second index of that @xmath329 to @xmath4 .
this way , the only generators involved are @xmath327 , for @xmath371 , and @xmath372 , for @xmath373 , which generate a copy of @xmath301 inside @xmath1 .
so , we can use the relators ( c ) to have our word expressed in this small set of generators and assume that it is a word in @xmath301 .
so if the word is the identity , it is consequence of the relators of @xmath301 . but these relators correspond to the relators ( b1 ) to ( b8 ) .
as is common in the groups of the thompson family , the infinite presentations are interesting and useful because of their symmetry and associated normal forms , but often it turns out that there are finite presentations from which the infinitely many generators and relations can be constructed and deduced . in this section , we construct a finite presentation for @xmath1 .
thompson s group @xmath0 admits a finite presentation which is merely the first two generators @xmath374 and @xmath375 and the first two non - trivial relations .
we construct @xmath376 from @xmath374 and @xmath375 as @xmath377 and from the two first two non - trivial relations we can deduce all of the relations in ( a ) above .
this is the first building block for our finite presentation . in a similar way we can construct all generators @xmath327 from a few ones .
the generators needed are @xmath378 , @xmath332 , @xmath379 , @xmath380 .
the idea is that conjugating a braid with @xmath381 has the effect of splitting the @xmath382 strand , so from these four generators and the generators for @xmath0 , we can construct any generator @xmath327 by the process of splitting as many strands as necessary to produce the strands before the @xmath41-th and between the @xmath41-th and @xmath326-th .
this process is as follows : * given @xmath383 , with @xmath384 , we use the relators ( d3 ) to decrease the distance between @xmath41 and @xmath326 until 2 : @xmath385 if @xmath386 then this process brings any generator @xmath387 down to @xmath332 . for any other value of @xmath41 it reduces to @xmath388 .
* we reduce the generators @xmath389 with @xmath390 to @xmath380 with relators of type ( d1 ) : @xmath391 * and finally , we reduce generators of type @xmath392 to @xmath379 by again using ( d1 ) : @xmath393 the generators @xmath368 are constructed in exactly the same way , where we replace @xmath367 by @xmath368 , and with the same constraints on indices . to see which relators to include in the finite presentation ,
we see which can be used to get the full families .
we consider the relators ( d ) first and we will use those to help with the other families . in each of the relators of ( d1 ) , there are three strands which are important : the strands labelled @xmath41 , @xmath326 and @xmath118 .
the idea is that between those , we only need to have one strand , because by splitting it , we can get to any number of strands in that position .
thus , we can get all of the relators ( d1 ) from the following * @xmath394 * @xmath395 * @xmath396 * @xmath397 * @xmath398 * @xmath399 * @xmath400 * @xmath401 every relator of the type ( d1 ) is a consequence of the definitions above and of these eight relators . as an example
, we will show the relator @xmath402 . if @xmath403 , the relator is ( d1.1 ) above .
if @xmath404 , then we use the definitions : @xmath405 and using ( d1.3 ) we get to @xmath406 which is @xmath407 finally equal to @xmath408 all the other relators of type ( d ) are very similar to this case , and we leave the details to the reader as they are straightforward but tedious .
the families of relators ( d1 ) and ( d3 ) are especially important , because they are used to split or combine adjacent strands which are not involved in the braiding .
any two adjacent strands which are not braided can be joined using a relator from one of these families .
this is useful for the families ( b1 ) to ( b9 ) .
for instance , the relators ( b1 ) show that two braids @xmath327 and @xmath409 commute if @xmath410 .
this relation only involves the strands @xmath411 , @xmath412 , @xmath41 and @xmath326 .
if there are strands in between , they are uninvolved in the braiding .
if there is more than one strand , we can apply the conjugating relations ( d1 ) or ( d3 ) to bring the relators down to a simple one where there is just one strand between .
for instance , we show the relator @xmath413 in figure [ fig : a1456 ] . .
the commutativity is apparent .
[ fig : a1456],title="fig:",width=144 ] + we see that the two strands between the first and the fourth are straight .
it is clear they can be obtained by splitting a single strand from an analogous relator whose braided strands are 1 , 3 , 4 and 5 .
a conjugation then by @xmath375 brings it down , according to the definitions above .
now we have that @xmath414 which is of the type ( d3 ) , and @xmath415 which is of the type ( d1 ) .
so the relator is a consequence of the relator @xmath416 using only ( d1 ) and ( d3 ) relators . in this way
we see that the only relators that we need to construct all the relators in ( b1 ) are those that have either zero or one strand at the beginning , or between the @xmath41 , @xmath326 , @xmath411 , and @xmath412-th strands .
these are : * @xmath417 * @xmath418 * @xmath419 * @xmath420 * @xmath421 * @xmath422 * @xmath423 * @xmath424 * @xmath425 * @xmath426 * @xmath427 * @xmath428 * @xmath429 * @xmath430 * @xmath431 * @xmath432 the other families , including the ones which involve @xmath368 generators , are completely analogous .
we are left only with the family ( c ) .
the relators are all of the type @xmath433 , with @xmath64 . as before ,
it is clear that if @xmath434 , we can conjugate by @xmath374 to get a relator with lower indices .
so only ones with @xmath386 or @xmath403 are needed .
again we see that all are consequences of a fundamental finite set of relations : * @xmath435 * @xmath436 * @xmath437 * @xmath438 * @xmath439 * @xmath440 for example , we see that @xmath441 if @xmath442 .
and the cases for @xmath403 are exactly similar .
hence , we have proved the following theorem : [ bffp]the group @xmath1 is finitely presented the finite presentation for @xmath1 is the following : generators : @xmath374 , @xmath375 , @xmath378 , @xmath332 , @xmath379 , @xmath380 , @xmath443 , @xmath333 , @xmath444 , @xmath445 relators : * @xmath446 * @xmath447 , for @xmath448 * @xmath449 , for @xmath450 * @xmath451 , for @xmath452 * @xmath453 , for @xmath454 * @xmath455 , for @xmath448 * @xmath456 , for @xmath450 * @xmath457 , for @xmath452 * @xmath458 , for @xmath454 * @xmath433 , for @xmath459 * @xmath460 , for @xmath461 * @xmath462 , for @xmath463 * @xmath464 , for @xmath465 * @xmath466 , for @xmath467 * @xmath468 , for @xmath469 * @xmath470 , for @xmath461 * @xmath471 , for @xmath463 * @xmath472 , for @xmath465 * @xmath473 , for @xmath467 this gives a total of 10 generators and 192 relators .
in the previous section we constructed a presentation of @xmath1 . brin @xcite described both @xmath18 and @xmath22 .
brin describes the group @xmath22 via a zappa - szp product of @xmath0 and @xmath474 , and describes @xmath18 as a subgroup of @xmath22 .
the group @xmath22 is the group of braided forest diagrams , where all but finitely many of the forests are trivial , and the group @xmath18 is the subgroup of @xmath22 where all of the trees in the forest pairs are trivial except the first pair , which have the same number of leaves . not only is @xmath18 a subgroup of @xmath22 , but also @xmath22 is a subgroup of @xmath18 , as described by brin @xcite . we take the standard identification of the real line with the unit interval which is compatible with the relevant dyadic subdivisions which sends the interval @xmath475 $ ] of @xmath476 with the interval @xmath477 $ ] and then we see that @xmath22 is the subgroup of @xmath18 in which the last strand is not braided with any other strands .
similarly , we have the group @xmath2 which can either be regarded as the supergroup of @xmath1 of pure braided forest diagrams , or as a subgroup of @xmath1 where the braiding does not involve the last strand . here
, we easily describe the subgroup @xmath2 of @xmath1 by omitting the generators and relations from @xmath1 which involve braiding the last strand .
so we obtain presentations for @xmath2 which are sub - presentations of the infinite and finite presentations for @xmath1 given in the earlier sections above .
_ the proof is similar to the case for @xmath1 .
here we note that @xmath2 is exactly the subgroup where the last strand is not braided with any previous strands , and by omitting the @xmath368 generators we guarantee that the last strand is not braided .
an argument similar to the earlier one for @xmath1 shows that these generate @xmath2 . to find relators for a presentation of @xmath2 we use the same sets of generators and relators as for @xmath1 , deleting the generators in the @xmath368 family and deleting all relations which include any of the @xmath329 .
_ proof._using the interpretation of the geometric group @xmath2 as the subgroup of @xmath1 where braiding never involves the last strand , the same proof used in the section 5 to establish the presentation for @xmath1 goes through in this situation .
the only difference is that since we have no @xmath368 generators , once we rearrange the word to have all @xmath367 generators in the middle , they are already generators for one copy of the pure braids on @xmath4 strands with the rightmost strand unbraided , so there is no need for the step using relators of type ( c ) .
note that the group @xmath2 will also be finitely presented .
the arguments needed to see this are similar to those for @xmath1 .
the finite presentation for @xmath2 can be easily obtained from the finite presentation for @xmath1 by deleting all generators @xmath329 and the relations where they appear . | we describe pure braided versions of thompson s group @xmath0 . these groups , @xmath1 and @xmath2 , are subgroups of the braided versions of thompson s group @xmath3 , introduced by brin and dehornoy . unlike @xmath3 , elements of @xmath0 are order - preserving self - maps of the interval and we use pure braids together with elements of @xmath0 thus preserving order .
we define these groups and give normal forms for elements and describe infinite and finite presentations of these groups . |
the _ swift _ gamma - ray burst explorer @xcite has been returning unprecedented data about gamma ray bursts ( grbs ) for the past 18 months . of particular interest in the bursts followed by _
swift _ have been the early time lightcurves of the afterglows which have shown much greater structure and different temporal decay properties than expected , leading to much discussion in the literature regarding the nature of the transition between the prompt emission , thought to be due to synchrotron radiation from internal collisions @xcite , and the afterglow , also thought to be due to synchrotron radiation , though it remains somewhat unclear whether the emission arises in internal or external shocks . _ swift _ observations have shown that grb lightcurves can be described by a canonical 3-segment shape .
this shape consists of 1 - an early steep decay phase ( f@xmath2 where 3@xmath3 @xmath4 @xmath3 5 ; t @xmath5 500 s ) 2 - a very shallow decay phase ( 0.5 @xmath3 @xmath6 @xmath3 1.0 ; 500 s @xmath5 t @xmath5 @xmath7 s ) and 3 - a normal decay phase ( 1 @xmath3 @xmath8 @xmath3 1.5 ; @xmath7 s @xmath5 t ) @xcite .
superimposed on this smooth decay profile , _ swift _ has shown that bright x - ray flares , 2 to 100 times as bright as the underlying afterglow , are common during the early ( t @xmath3 10 ks ) stages of grbs @xcite .
the observation of grb050117 @xcite marked the first time that _ swift _ slewed to and settled on a grb while the prompt gamma ray emission was still in progress , arriving 192 seconds after the burst alert telescope ( bat ) triggered on the burst which had a t@xmath9 of 168 seconds .
due to irregularities in the observing mode of the xrt and the proximity of the south atlantic anomaly to _ swift _ during the observation of grb050117 , however , only very sparse data was collected by the xrt , totaling 11.4 seconds in the first orbit and 946 seconds overall .
this left large gaps in the lightcurve coverage and severely limited the quality of the spectral analysis that could be performed .
we report here on the observation of grb 050713a @xcite , a burst of t@xmath9 = 70 seconds to which _ swift _ slewed and began collecting data with the narrow field instruments ( nfis ) in just 72.6 seconds , while the prompt gamma ray emission was still detectable by the bat .
this burst marks just the second time that the bat and xrt have collected simultaneous data on a burst and it marks the first time that both instruments have produced a well sampled , simultaneous dataset covering multiple flares in the prompt emission .
in addition to the _ swift _ coverage of grb 050713a , we report also on prompt and followup observations carried out by konus - wind , magic , _ xmm - newton _ and ground based optical observatories . in section 2 we describe the observations and data analysis from all instruments including ground follow - up . in section 3
we discuss the implications of the observations in light of the new theoretical understanding emerging from _ swift _ observations of grbs . in section 4
we summarize and present our conclusions .
quoted uncertainties are at the 90% confidence level for one interesting parameter ( i.e. , @xmath10=2.71 ) unless otherwise noted .
many different observatories and instruments have observed grb 050713a .
we devote the following section to a description of the observations and analysis carried out by each instrument team .
all spectral fits were performed using xspec v11.3 .
the _ swift _ bat @xcite triggered on grb 050713a at 04:29:02.39 ut , measuring a peak 1-second flux of @xmath11 photons @xmath12 s@xmath13 .
t@xmath9 measured in the 15350 kev energy range is @xmath14 s @xcite .
the onset of the burst as defined by the bat trigger is preceded by a weak , hard ( photon index = 1.26 ) precursor at t@xmath060 s. bat data were processed using the bat ground software build 11 and bat calibration database files build 11 . at the time of the bat trigger
, the flux rose rapidly and remained elevated during a 12 s long , multipeaked burst ( fig 1 ) . at t@xmath0 + 12
s , the bat flux rapidly decayed with a powerlaw decay rate of @xmath15 for 5 seconds before breaking to a more shallow decay of @xmath16 at t@xmath0 + 17 s. this decay continued until t@xmath0 + 40 s at which point the bat flux had decayed to near background levels . at t@xmath0 + 50
s , a flare is seen with peak flux @xmath17 ergs @xmath12 s@xmath13 , extrapolated into the xrt 0.210.0 kev bandpass , followed by a flare with peak flux @xmath18 ergs @xmath12 s@xmath13 at t@xmath0 + 65 s , another at t@xmath0 + 105 s with peak flux @xmath19 ergs @xmath12 s@xmath13 and some hint of further emission at the onset of a flare seen in the xrt at t@xmath0 + 160 s. a weak but statistically significant precursor is seen at t@xmath070 s to t@xmath050 s followed by a period of no significant emission from t@xmath050 s to the burst trigger .
the spectrum of the entire bat dataset is well fit by a power - law spectrum with photon index = @xmath20 , though there is evidence for a slightly harder index of 1.45 during the plateau and a softening to @xmath21 during the rapid decay , and further softening to @xmath22 during the weak flares . using the global fit of @xmath23 ,
the fluence is @xmath24 ergs @xmath12 in the 15350 kev energy range .
grb 050713a triggered konus - wind ( k - w ) @xcite at t@xmath0(k - w)= 04:29:01.745 ut .
it was detected by the s2 detector , which observes the north ecliptic hemisphere , with an incident angle of 18@xmath251 .
the k - w lightcurve in 3 bands is shown in figure 2 . the propagation delay from wind to _
swift _ is 1.387 s for this grb .
correcting for this factor , one sees that the k - w trigger time corresponds to t@xmath0 + 0.742 s. prior to t@xmath0(k - w)-0.512 s data were collected by k - w in a survey mode with lower time resolution of 2.944 s and only 3 broad spectral channels , 1870 kev , 70300 kev and 3001160 kev . from t@xmath0(k - w ) to t@xmath0(k - w)+491.776 s , 64 spectra in 101 channels were accumulated on time scales varying from 64 ms near the trigger to 8.19 s by the time the signal became undetectable .
the multichannel spectra cover the 18 kev14 mev energy range but no statistically significant emission is seen above 2 mev .
data were processed using standard konus - wind analysis tools .
joint spectral analysis was carried out using the bat data between 15 and 150 kev and the konus data from 20 to 2000 kev .
the spectra were fit by a power law model with an exponential cut off : dn / de @xmath26 e@xmath27 e@xmath28 where e@xmath29 is the peak energy of the @xmath30 f@xmath31 spectrum and @xmath32 is the photon index .
the spectrum of the main pulse is well fit ( fig 3 ) with photon index @xmath33 and e@xmath34 kev ( @xmath35=138/119 dof ) .
joint fits between bat and konus were also made for other time intervals , including one which shows the faint precursor detected by both instruments at t@xmath3660 s , and will be addressed in greater detail in section 3.3 .
the main pulse fluence in the 20 kev to 2 mev range is @xmath37 erg @xmath12 .
the 256-ms peak flux measured from t@xmath0 + 1.2 s in the 20 kev to 2 mev band is @xmath38 erg @xmath12 s@xmath13 and the t@xmath9 durations of the burst in the g1 , g2 , g3 energy bands are @xmath39 s , @xmath40 s and @xmath41 s , respectively .
the xrt @xcite performs an automated sequence of observations @xcite after _ swift _ slews to a grb detected by the bat .
when the spacecraft first settles on the target , a short image ( 0.1 s followed by a longer 2.5 s image if a position is not determined in 0.1 s ) is taken to determine an accurate position .
following the image , the xrt switches into either windowed timing ( wt ) mode ( a high timing accuracy mode with 1 dimensional position information ) if the source count rate is above 2 counts s@xmath13 , or photon counting ( pc ) mode ( the more traditional operating mode of x - ray ccds in which full 2 dimensional position information is retained but with only 2.5 s timing resolution ) if the count rate is below 2 counts s@xmath13 .
xrt collected a 0.1 s image mode frame upon settling on grb 050713a 73 seconds after the bat trigger , which yields a count rate of 314 counts s@xmath13 . following the image mode frame ,
xrt cascaded down through its automated mode sequence and collected its first wt frame 4.5 seconds later . at the onset of the wt data , the xrt count rate was about 100 counts s@xmath13 and decaying as a powerlaw .
this initial powerlaw decay in the xrt wt data together with the image mode data point measured at a flux level @xmath13 times higher just 4.6 s earlier clearly indicates that the xrt settled and began taking data during the latter portion of the flare detected in the bat at t@xmath0 + 65 seconds ( see fig 4 ) .
xrt remained in wt mode throughout the entire first orbit of data collection on grb 050713a , also observing the flare detected by the bat at t@xmath0 + 105 and a lower level flare not clearly detected by the bat at t@xmath0 + 155 s. following a 65 minute period of occultation by the earth , xrt began observations again at t@xmath0 + 4300 s , now observing in pc mode since the countrate of the source had decayed below 2 counts s@xmath13 . a small flare at t@xmath0 + 10 ks and the indication of another flare at t@xmath0 + 45
ks are seen in the late time xrt lightcurve data , superimposed on an otherwise steady powerlaw decay .
xrt observations continue to monitor the source until t@xmath0 + 1.8@xmath42 s , a total exposure time of 178 ks , at which time the source had decayed below the xrt detection threshold .
xrt data are processed using the _ xrtpipeline _ ground software version 0.9.9 , the redistribution response matrices swxwt0to2_20010101v007.rmf ( wt data ) and swxpc0to12_20010101v007.rmf ( pc data ) , and ancillary response files generated with the _ xrtpipeline _ task _ xrtmkarf_. the x - ray afterglow position determined from ground processing of the data is with an uncertainty of 3.2 arcseconds .
this is 10.5 arcseconds from the reported bat position , 0.5 arcseconds from the optical counterpart reported by malesani et al . ,
@xcite and 1.5 arcseconds from the initial xrt position calculated onboard the satellite and automatically distributed via the gcn network @xcite .
an x - ray image compiled from the first segment of xrt pc data is shown as figure 5 with the bat , xrt and optical counterpart error circles displayed .
a faint background source is detected 30 arcseconds due south of the grb afterglow at a constant flux level of @xmath43 counts s@xmath13 .
the contribution of this steady source has been removed from the calculation of the afterglow lightcurve .
a timeline of the xrt ( as well as other ) observations of grb 050713a is shown in table 1 .
the lightcurve will be broadly treated in two parts .
the first part is the initial orbit of data , during which the lightcurve is characterized by bright flares which are simultaneously observed by the bat as well as the k - w instrument at higher energies . due to the extreme variability in this portion of the lightcurve , a global decay index
can not be determined from the xrt data .
the second part is the remainder of the xrt data from the second orbit onward , which is characterized primarily by a broken powerlaw decay , though at least one small flare is seen superimposed atop this global decay .
_ swift _ finished slewing to grb 050713a at t@xmath0 + 73 s , during the flare which began at t@xmath0 + 65 s. the xrt short image frame is collected just after the peak of this flare , at a flux of @xmath44ergs @xmath12 s@xmath13 , and the first 20 frames of wt data record the decay of the flare . fitting a simple powerlaw to this decay from t@xmath0 + 79 s to t@xmath0 + 100 s setting t@xmath0 to be the bat burst trigger time , we find a powerlaw index of @xmath45 ( 1 @xmath46 ) . at t@xmath0 + 105
s a new flare begins , which rises with a powerlaw index of @xmath47 for 5 - 10 s , flattens at the peak of @xmath48 ergs @xmath12 s@xmath13 for 5 - 10 s , then decays with a more shallow powerlaw index of @xmath49 for about 30 s. at t@xmath0 + 165 s a third flare is detected , which rises with a powerlaw slope of @xmath50 for 5 - 10 s , flattens at the peak of @xmath51 ergs @xmath12 s@xmath13 for 5 - 10 s , then decays with a slope of @xmath52 for 70 s before the end of the observing window due to earth occultation .
the second orbit of data in the xrt is the only single orbit of data in which the afterglow is characterized by a well sampled ( greater than 100 events total ) lightcurve devoid of any obvious flaring activity . during the 1600 seconds of data in this orbit , from t@xmath0 + 4360
s to t@xmath0 + 5952 s , the lightcurve decays steadily as a powerlaw with decay index of about 1.0 .
the third orbit of data is characterized by another flare , beginning at t@xmath0 + 10 ks , lasting throughout the entire orbit ( about 2 ks ) and reaching a peak flux of @xmath53 ergs @xmath12 s@xmath13 .
a powerlaw fit to the rising portion of the flare yields a slope of @xmath54 while the decaying portion yields a slope of @xmath55 .
this flare seems to be superposed atop the underlying afterglow decay of decay index @xmath56 .
observations were interrupted after 150 s during the fourth orbit due to the occurrence of grb 050713b , and observations of grb 050713a remained suspended until t@xmath0 + 40 ks .
some suggestion of another flare is seen in the orbit of data beginning at t@xmath0 + 45 ks , though the statistics are poor . while afterglow data from the xrt alone do not clearly require a break in the afterglow powerlaw , _ xmm - newton _ data ( see @xmath572.4 ) from t@xmath0 + 21 ks to t@xmath0 + 50 ks provide an accurate measure of the late - time decay slope ( @xmath58 ) which can not fit the xrt data from orbits 2 and 3 without a break in the powerlaw .
the joint xrt-_xmm - newton _ lightcurve will be further discussed in @xmath572.4 .
table 2 summarizes the flares and their temporal fits . the xrt spectral analysis of this burst is somewhat complicated by the high degree of flaring activity seen . in all cases , spectra are binned to a minimum of 20 counts per bin in order to use @xmath35 statistics .
fitting the entire first orbit of data , the spectrum is well fit by a highly absorbed powerlaw with photon index @xmath59 and @xmath60 @xmath12 , which is significantly above the galactic column ( @xmath61 @xmath12 ) in the direction of grb 050713a .
we are also able , due to the large number of counts in each of the early flares in the dataset , to fit a spectrum to both the rising and decaying portions of the flares . in doing
so we see the typical hard to soft evolution of the flares @xcite .
the second orbit of data shows a significantly different spectrum from the first , with a harder spectrum with photon index @xmath62 and a lower value of @xmath63 , possibly indicating a period of energy injection @xcite .
the third orbit is well fit by a softer powerlaw similar to that which fit the first orbit with photon index @xmath64 and @xmath65 . during the period of overlapping coverage between _ swift _ and _ xmm - newton _ , xrt has 3.5 ks of exposure time at a mean countrate of 0.04 counts s@xmath13 for a total of about 150 events during the simultaneous observing period .
fitting a spectrum to this overlapping coverage yields a photon index @xmath66 and @xmath67 .
the corresponding mean unabsorbed 0.210.0 kev flux during the overlap period as measured by xrt is @xmath68 ergs @xmath12 s@xmath13 . the data collected after the third orbit ( i.e. , after the temporal break in the lightcurve at t@xmath0+@xmath69 ks ) are too sparse to justify fitting with higher order models , but a simple absorbed powerlaw fit yields a spectrum of photon index @xmath70 with @xmath71 .
this is consistent with the x - ray photon index found in orbits 1 and 3 and is marginally softer than the photon index found during orbit 2 which , as noted above , suggests a period of energy injection .
_ xmm - newton _ follow - up observations of grb 050713a commenced t@xmath0 + 23.6 ks ( for the epic - pn ) and t@xmath0 + 20.9 ks ( for the two epic - mos cameras ) .
the _ xmm - newton _ data were processed with the epproc and emproc pipeline scripts , using the _ xmm - newton _ sas analysis package , version 6.5 .
a bright rapidly decaying source is detected near the aimpoint of all three epic detectors , localized at .
the net exposures after screening and deadtime correction are 24.1 ks ( pn ) and 27.0 ks ( mos ) .
all three epic cameras ( pn and 2 mos ) were used in full window mode with the medium filter in place .
source spectra and lightcurves for all 3 epic cameras were extracted from circular regions of 20 arcseconds radius centered on the afterglow .
background data were taken from a 60 arcseconds circle on the same chip as the afterglow , but free of any x - ray sources .
fitting the afterglow lightcurve with a simple power - law decay results in a decay index of @xmath72 .
several flares are present in the background lightcurve , so as a conservative check , we also excluded times where the background rate is @xmath73countss@xmath13 .
the afterglow decay rate is then @xmath74 , consistent with the above value .
the decay rate from the mos lightcurve ( for the two detectors combined ) is also consistent at @xmath75 .
afterglow and background spectra were extracted with the same regions used for the lightcurves , while ancillary and redistribution response files were generated with the sas tasks arfgen and rmfgen respectively . as with xrt data ,
source spectra were binned to a minimum of 20 counts per bin in order to use @xmath76 statistics .
the pn and mos spectra were fitted jointly , allowing only the cross normalization to vary between the detectors , which is consistent within @xmath77 .
the two mos spectra and responses were combined to maximize the signal to noise , after first checking that they were consistent with each other .
the average net source count rates obtained over the whole observation are @xmath78countss@xmath13 for the pn and @xmath79countss@xmath13 per mos module . allowing the absorption column to vary in the spectral fit results in a formally acceptable fit ( @xmath80 ) .
the @xmath81 obtained is @xmath82@xmath12 , while the continuum photon index @xmath83 . the time - averaged , unabsorbed , 0.210.0 kev flux obtained for the afterglow is @xmath84erg@xmath12s@xmath13 .
these values are consistent with the _ swift _ xrt measurement obtained at the time of the _ xmm - newton _ observation .
the _ xmm - newton _ afterglow spectra were also sliced into three segments of approximately 8 ks in length , in order to search for any spectral evolution within the _ xmm - newton _ observation .
no change in the continuum parameters was found , all three spectral segments being consistent with photon index @xmath85 and @xmath86@xmath12 .
the spectrum obtained from the pn detector and residuals to an absorbed power - law model ( with @xmath87 and @xmath88@xmath12 ) are shown in figure 6 .
the power - law decay index obtained from the _ xmm - newton _ observation ( @xmath89 ) appears to be steeper than that obtained from the _ swift _ xrt in orbit 2 ( @xmath90 ) . in order to compare between the _ xmm - newton _ and
_ swift _ afterglow lightcurves , a combined lightcurve from the _ xmm - newton _ and _ swift _ observations was produced , scaling to the absorbed continuum fluxes measured in the 0.510 kev band .
the joint _ swift _ and _ xmm - newton _ lightcurve is shown in figure 7 , zoomed to better display the region at which the lightcurve break occurs .
a single power - law decay slope of @xmath91 is an extremely poor fit to the lightcurve in this region , with a fit statistic of @xmath92 . indeed
the lightcurve from t@xmath0 + 4 ks until t@xmath0 + 1000 ks can be better fitted with a broken power - law .
there is a flat decay index of @xmath93 at early times and a steeper decay index of @xmath94 at later times , with the break in the decay occurring at t@xmath95 ks .
the fit statistic is then @xmath96 .
the remaining contribution towards the @xmath76 originates from two small possible flares present near t@xmath97 ks and t@xmath98 ks .
the magic telescope @xcite was able to observe part of the prompt emission phase of grb 050713a as a response to the alert provided by _
swift_. the observation , at energies above @xmath99 , started at t@xmath0 + 40 s , 20 s after reception of the alert .
it overlapped with the prompt emission phase measured by _
swift _ and k - w , and lasted for 37 min , until twilight .
the observation window covered by magic did not , however , contain the burst onset peak detected at kev - mev energies , where the _ swift _ and k - w spectra were taken .
the same region of the sky was observed 48 hours after the burst onset , collecting an additional 49 minutes of data , which was used to determine the background contamination .
the magic ( _ major atmospheric gamma imaging cherenkov _ ) telescope is currently the largest single - dish imaging air cherenkov telescope ( iact ) in operation , with the lowest energy threshold ( @xmath100 at zenith , increasing with zenith angle ) . in its fast slewing mode
, the telescope can be repositioned within @xmath130 s. in case of an alert by gcn , an automated procedure takes only a few seconds to terminate any pending observation , validate the incoming signal and start slewing toward the grb position , as was the case for grb 050713a . using the standard analysis ,
no significant excess of @xmath101-like air showers from the position of grb 050713a above @xmath99 was detected @xcite .
this holds both for the prompt emission and during the subsequent observation periods .
figure 8 shows the number of excess events during the first 37 minutes after the burst , in intervals of 20 s. for comparison , the number of expected background events in the signal region , stable and compatible with statistical fluctuations , is shown .
upper limits to the gamma - ray flux are given in table 3 .
this is the first observation of the grb prompt emission phase performed by an iact .
optical followup observations of grb 050713a performed by the uvot and by ground based observatories are summarized in table 4 .
the earliest optical afterglow measurement comes from the raptor - s robotic telescope at the los alamos national laboratory in los alamos , new mexico at r=18.4 @xmath102 0.18 in a coadded series of 8x10 second images with a midpoint observation time of t@xmath0 + 99.3 s @xcite .
a nearly simultaneous measurement was made by the robotic liverpool telescope in a coadded series of 3 @xmath12 minute exposures in the r@xmath103 band with a midpoint of observation of t@xmath0 + 3 minutes @xcite .
later detections below the digitized sky survey limits were reported within the first 60 minutes after the burst trigger in the r band by the nordic optical telescope ( t@xmath0 + 47 m ) , in the i band by the galileo italian national telescope in the canary islands and in the infrared j , h , and k bands by the astronomical research consortium telescope at apache point observatory ( t@xmath0 + 53 m ) . due to the bright ( v=6.56 ) star hd204408 which is located just 68 arcseconds from the position of the burst ,
the uvot background level at the position of the afterglow is significantly higher than usual , resulting in abnormally poor sensitivity of the instrument in detecting the afterglow of grb 050713a . considering this high background ,
the non - detection of the afterglow by the uvot is not surprising .
all other reported optical observations of the afterglow position have yielded only upper limits .
most of the upper limits are near in time to the actual detections but at brighter magnitudes and thus do not produce strong constraints on the decay rate of the optical afterglow .
the r - band measurement made at t@xmath0 + 10.3 hours by the lulin telescope in taiwan , however , is at a sufficiently late epoch to place a useful constraint on the optical decay rate .
fitting a simple powerlaw to the two well defined measurements at t@xmath0 + 99.3 s and t@xmath0 + 180 s and the upper limit at t@xmath0 + 10.3 hours yields an upper limit on the power law decay slope of @xmath104 , as is shown in figure 4 . a radio followup observation made with the vla reports no detection at t@xmath0 + 4.3 days .
the k - w light curve in the 181160 kev energy range is similar to the _
swift_-bat light curve ( fig 1 ) .
the small precursor peak detected by bat at t@xmath070 to t@xmath050 s is seen by k - w at statistically significant levels in all three broad , pre - trigger bands : g1 ( 1870 kev ) , g2 ( 70300 ) , and g3 ( 3001160 kev ) .
the other smaller peaks detected by the bat after the burst trigger are not seen at statistically significant levels in the k - w data , despite the fact that the peaks at t@xmath0 + 50 s and at t@xmath0 + 65 s are more intense in the bat energy range than the precursor is .
the detection by k - w of the precursor but not the later flares is indicative of the harder spectral index seen in the precursor as compared to the later flares ( see section 3.3 for discussion of separate spectral fits to individual flares ) .
the xrt lightcurve with bat data overplotted is shown in figure 4 .
both the x - ray and gamma - ray data in the first orbit are dominated by flaring activity , making it difficult to draw a conclusion regarding the underlying powerlaw decay index from this orbit alone .
the xrt data beginning at t@xmath0 + 4 ks ( orbit 2 ) and extending until t@xmath0 + 40 ks show a significantly flatter powerlaw decay slope of @xmath105 , implying that a break in the powerlaw decay has occurred near the end of the first orbit of xrt coverage at t@xmath0+@xmath1300 s and that a period of energy injection occurs from t@xmath0+@xmath1300 s to t@xmath0+@xmath115 ks .
another break in the lightcurve then occurs near t@xmath0 + 25 ks to a steeper , `` normal '' , pre - jetbreak decay slope , as shown by the _ xmm - newton _ data ( @xmath106 ) .
support for this notion of the presence of an energy injection phase may be drawn from the harder x - ray spectral slope of the second orbit of xrt data ( photon index = 1.9 @xmath102 0.13 ) compared to the first orbit ( photon index = 2.28 @xmath102 0.04 ) , the third orbit ( photon index = 2.25 @xmath102 0.23 ) , and the later data ( photon index = 2.8 @xmath102 0.6 ) ( table 5 ) .
_ xmm - newton _ data coverage nicely fills much of the data gap in the xrt coverage between t@xmath0 + 15 ks and t@xmath0 + 40 ks and provides high signal to noise data in this regime , producing a confident determination of the lightcurve break .
the global picture of the lightcurve of this burst is one in which the early data ( prior to t@xmath0 + 12 s ) shows a bright plateau in the 15 kev to 1 mev energy range , consisting of multiple overlapping peaks . at t@xmath0 + 12
s the emission drops rapidly , consistent with a curvature radiation falloff @xcite until subsequent flaring activity begins to be seen in the 0.3150 kev region with some indication of flux at higher energies from k - w . due to the rapid rise and decay of the flares , internal shocks from continued central engine activity appears to be the most likely explanation for these flares @xcite . the earliest ground based optical detections are reported at this time also , suggesting that the flares may also be optically bright .
the lack of higher resolution timing information in the optical data , though , admits the possibility that the optical emission may be unassociated with the emission mechanism responsible for the x - ray flares .
it is possible that the optical emission is due to synchrotron emission from the reverse shock ( rs ) , though the much higher flux level of the x - ray flare peaks compared to the optical measurements suggests that the x - ray flares themselves are not due to inverse compton scattering of the optical synchrotron emission of the rs @xcite .
following this prompt emission phase , an energy injection phase begins which dominates the lightcurve until at least t@xmath0 + 16 ks . during the energy injection phase ,
continued activity of the central engine adds energy to the afterglow of the burst , either through additional ejection events or through the realization of energy contained in previously ejected outward moving relativistic shells which only collide at later times , producing late time internal shock emission which is then added to the overall decay @xcite .
it may be expected , if the energy injection phase is due to continued central engine activity , that flaring behavior would continue to be observed during this period and , indeed , some evidence for small scale flaring activity during both the second and third orbit of xrt data can be seen , though at a much reduced significance in comparison to the flaring activity of the first orbit . near t@xmath0 + 25 ks , the energy injection phase ends , giving way to a steeper decay slope similar to what is often seen in grbs after the prompt emission phase and prior to the possible onset of a traditional jet - break @xcite .
many flares superimposed on top of the overall decay of grb 050713a show the typical properties seen in other bursts : that @xmath107t / t @xmath1 0.1 and that the peak flux level is negatively correlated with the time of the flare @xcite .
these two properties of flares seen in _ swift _ grb afterglows have been cited as evidence for flares being produced through accretion processes onto the central compact object @xcite , but we offer here that the constancy of the @xmath107t / t value of flares may partly be a by - product of the overall decay of the afterglow in so much as the sensitivity of the xrt to flares is naturally degraded as the overall flux level of the afterglow decays , thus _ requiring _ flares at later times ( and hence , lower flux levels ) to be longer in duration for enough counts to be collected to produce a significant flare seen above the background . such a case can be seen in comparing the early time flares in the first orbit of grb 050713a to the flare seen in the the third orbit . during the first orbit , the underlying flux level beneath the flares
is poorly determined , but can be assumed to be 10 - 100 counts s@xmath13 .
we are dominated in this portion of the lightcurve by the poissonian error in the flux , which in a 10 second integration will be 10 - 32 counts , or 3 - 10@xmath108 .
thus , for a flare to appear at the 6 sigma level above the background during this portion of the lightcurve , at most a 60@xmath108 increase in fluence above the normal powerlaw decay is needed , which can be acquired in a few seconds by the introduction of a flare with twice the flux of the underlying afterglow . during the third orbit , however , the underlying afterglow flux level has dropped to @xmath1 0.1 counts s@xmath13 . during a 10 second integration at 0.1 counts
s@xmath13 the poisson error alone is 1 count , so for a flare to be detectable at 6 sigma above background at these count levels , the total fluence must be 6 counts , implying an increase in the rate from 0.1 counts s@xmath13 to 0.6 counts s@xmath13 during the 10 s interval , a 6 fold increase , which has been seen only in the brighter flares . in order to be sensitive to the same 60@xmath108 increase in flux level as during the first orbit , the flare which occurs at a flux level of 0.1 counts s@xmath13 needs to have a poission error which is 1/6 of the total counts in the observation ,
i.e. , 36 counts must be collected , which implies an exposure time of at least 180 s if produced by the introduction of a flare with twice the flux of the underlying afterglow . in other words , because the afterglow flux level decays as t@xmath27 , the exposure time needed to acquire the same fluence level increases as t@xmath109 .
thus , we see that in moving from the first orbit at t@xmath0 + 100 s to the third orbit at t@xmath0 + 10000 s , assuming a typical underlying powerlaw decay of the afterglow of @xmath110 1 , we have greatly decreased the temporal resolution of xrt to detect flares ( from a few seconds to a few hundred seconds ) .
this is not to imply that there is not another more physical cause for the constancy of the @xmath107t / t ratio seen in flares , but rather to note that the typical grb seen by the _
swift _ xrt does not provide sufficient flux at times typically greater than a few ks to detect the shorter timescale flares that are so often seen during the first orbit .
in grb 050713a , a hint of emission above the afterglow powerlaw decay appears in the xrt data at t@xmath0 + 45 ks , though the statistics are , predictably , poor .
this time is overlapped by _
xmm - newton _ data , though , so we can look for evidence of a short flare in the _ xmm - newton _ data at this time . in figure 9
we show the _ xmm - newton _ lightcurve , plotted linearly and zoomed near t@xmath0 + 45 ks . though a 1 - 2 sigma deviation above the background decay is seen at t@xmath0 + 45 ks , the _ xmm - newton _ data appear consistent with a statistical fluctuation rather than a true flare similar to those seen earlier during the burst .
the presence of multiple flares in grb050713a argues against `` one - shot '' emission mechanisms such as synchrotron self - compton emission in a reverse shock or deceleration of the blastwave @xcite and it argues in favor of a mechanism which can produce repeated flares , such as late time central engine activity . while it may remain possible that one of the several flares in grb050713a is due to the rs or the onset of the afterglow due to external shocks , the steep temporal decays of all the temporally fitted flares coupled with the photon indices of the flares ( 1.25 @xmath1 2.5 ; table 6 ) do not satisfy the closure relations of sari et al . , (
1998 ) , chevalier & li ( 1999 ) and sari et al . , ( 1999 ) for propagation of the blast wave into either a wind or constant density ism .
together these points seem to argue in favor of an internal shock origin for the flares seen in this burst . due to the relatively narrow spectral response function of the bat ( 15150 kev for mask - tagged events ) and the xrt ( 0.310 kev ) , a spectral fit to data from only one of the two high energy instruments on _ swift _ is usually not able to discriminate between higher order spectral models .
analysts and authors are usually limited to choosing between a power - law or band function . in grb
050713a we have a rare case of data coverage overlap between bat and xrt ( 0.3150 kev ) and also between bat and k - w ( 15 kev14 mev ) .
taking advantage of this where appropriate , considering the relative flux levels in the three instruments , we have jointly fitted spectral datasets between the two pairs of instruments . during the precursor and from t@xmath0 + 0 to t@xmath0 + 16.5 s , we perform joint fitting between bat and k - w data . from t@xmath0
+ 16.5 to t@xmath0 + 78 s we have only bat data . from t@xmath0 + 78 to t@xmath0 + 116 s and during the onset of the flare at t@xmath0 + 160 we perform joint fitting between xrt and bat .
we have grouped the data into segments ( as shown in table 6 ) in order to temporally separate data which we expect may show significantly different spectral parameters .
segments 1 - 4 contain bat and k - w data and are segmented to separate the precursor from the prompt emission and the prompt emission from the rapid decay phase .
segment 5 contains bat data only and segments 6 - 10 contain xrt and bat data .
these are segmented to distinguish the 3 flares which have overlapping data and also to separate the rise of each flare from the decay of each flare .
we attempt fits to each of these segments using 4 different spectral models : 1 ) an absorbed powerlaw 2 ) an absorbed cutoff or broken powerlaw ( cutoff for data extending beyond 150 kev , broken otherwise ) 3 ) an absorbed band function and 4 ) an absorbed blackbody plus powerlaw .
the precursor is the most poorly sampled of all the regions . despite the low number of counts in the region , a cutoff powerlaw is favored over a single powerlaw at @xmath111 confidence according to the f - test .
of all the segments fit , the precursor has the hardest photon index , regardless of the model which is used to perform the fits .
the plateau of the prompt emission is best fit by an exponentially cutoff powerlaw model with photon index @xmath112 and e@xmath113 kev .
next to the precursor , the prompt plateau has the second hardest photon index of all segments fit , regardless of the model used . as with the other data segments which contain k - w data , the rapid decay segment is poorly fit by a simple powerlaw and is best fit by a cutoff powerlaw or band function .
the photon index of the cutoff powerlaw in segment 3 is quite similar to that in the prompt plateau , but the cutoff energy is somewhat lower ( 312 kev compared to 421 kev in the plateau ) , suggesting that the highest energy flux is `` shutting off '' during the rapid decay phase .
this segment is an extension of the prompt segment to slightly later times , encompassing slightly more data .
the cutoff powerlaw or band function is the best fit , with photon indices similar to segment 2 and e@xmath114 between that in segments 2 and 3 .
this segment contains only bat data and is included for completeness , though the narrowness of the bat spectral response limits the ability to distinguish between models .
a simple powerlaw is a good fit with photon index of 1.83 .
is unconstrained .
the powerlaw plus blackbody model produces a good fit to this segment but only with a very minimal blackbody component , effectively reproducing the fit of the simple absorbed powerlaw .
therefore we consider the powerlaw plus blackbody model inapplicable to this segment . only in this segment
, the data time ranges are mismatched between xrt and bat ( due to xrt observations beginning towards the end of the flare decay ) . rather than
ignore this flare or consider only the later part of the flare decay where xrt and bat data coverage overlap , we have chosen to fit the entire bat time range from t@xmath0 + 68 to t@xmath0 + 95 s together with the t@xmath0 + 79 to t@xmath0 + 95 s xrt data ( note that the image mode data taken at t@xmath0 + 73 s are highly piled up and can not be used spectrally ) for consistency with our treatment of the other flares . a simple powerlaw is a good fit to this segment , yielding @xmath115@xmath12 and a photon index of 2.47 , significantly softer than the rise of the flare , as expected . in the rise of the brightest flare seen in xrt ,
both an absorbed powerlaw plus blackbody model and an absorbed broken powerlaw model are significantly better fits ( f - test probability @xmath116 ) than a simple absorbed powerlaw .
the powerlaw plus blackbody indicates @xmath117@xmath12 and a relatively soft photon index of 2.0 with a blackbody temperature of kt@xmath118 kev .
we note that this value of kt is below the xrt energy band and may therefore indicate a non - physical spectral solution .
the absorbed broken powerlaw indicates @xmath119@xmath12 and photon indices of @xmath120 and @xmath121 , broken at 3.4 kev .
these two models are somewhat degenerate in this dataset , with both models producing a roll over in flux at low ( below 0.5 kev ) and high ( above 50 kev ) energies .
the decay portion of this flare is well fit by a simple absorbed powerlaw with @xmath122@xmath12 and photon index of 2.68 .
we note , however , that both an absorbed broken powerlaw and absorbed powerlaw plus blackbody are equally good fits to the data .
the rise of the last flare with overlapping data is well fit by a simple absorbed powerlaw with @xmath123@xmath12 and photon index @xmath124 , however the absorbed powerlaw plus blackbody is , strictly , a better fit according to the f - test , though only at about the @xmath125 confidence level ( f - test probability = 0.219 ) , with @xmath126@xmath12 , kt=200 kev and photon index of 2.43 .
we note that this value of kt is above the xrt - bat energy band and may therefore indicate a non - physical spectral solution .
the decay of this flare is well fit by an absorbed powerlaw with @xmath127@xmath12 and photon index of 2.55 , though as with segment 9 , the absorbed powerlaw plus blackbody is also an acceptable fit with @xmath128@xmath12 , kt=5.2 kev and photon index of 2.83 .
it should be noted that the bat flux is very near the noise level in this segment and really provides only an upper limit on the spectral fitting process in the higher energy region .
we have produced the broadband sed ( spectral energy distribution ) of the afterglow of grb 050713a over the time range from t@xmath0 + 20 s to t@xmath0 + 1000 s ( fig 10 ) .
this timerange includes detections of the burst afterglow in the optical from the raptor - s and liverpool telescopes ( corrected for the galactic extinction in this direction of @xmath129=1.04 @xcite ) and in the x - ray from _ swift _ bat and xrt .
it also includes upper limits in the gamma - ray energy range from k - w ( whose detectable emission ends at t@xmath0+@xmath110 s ) and in the gev energy range from magic .
a similar sed has been addressed by the magic collaboration in their paper regarding the magic flux upper limit @xcite in which they note that the sed composed of data from _ swift _ and magic ( 0.2 kev to 400gev ) is fit by a band function at low energy and that the magic data are consistent with a single unbroken powerlaw extending from e@xmath114 ( at @xmath1 400 kev ) to the magic limits up to 500 gev .
we confirm this result , citing a best fit photon index for a single powerlaw fit from 400 kev to 500 gev of @xmath130 = 2.1 @xmath102 0.1 and a reduced @xmath131 = 1.66 for 63 dof .
we further note that in performing our fit to the magic data , we have treated the magic upper limits as data points during our fit , thus our photon index or 2.1 is only a lower limit on the true photon index of a powerlaw which would fit the true flux level at gev energies .
our results here are , therefore , consistent with the analysis of the albert et .
, in which they show that their data are consistent with a powerlaw photon index of 2.5 from 400 kev to 500 gev .
we add that a band function fit is not , however , consistent with the data when we also consider the contemporaneous optical detections .
the relative faintness of the optical detections compared to the x - ray detections combined with the upper limits from k - w and magic requires an absorbed broken powerlaw fit .
figure 10 shows the best fit to the entire sed using an absorbed powerlaw ( dotted ) , absorbed broken powerlaw ( solid ) and absorbed band function ( dashed ) .
the spectral parameters and fit characteristics for each of these fits are shown in table 7 .
we have not corrected for the attenuation of flux above 10gev due to photon - photon interactions with the infrared background @xcite , however , our spectral fit results will remain valid independent of this effect due to the constraints placed by the k - w limiting flux measurement from 20 kev to 14 mev .
grb 050713a is one of the rare bursts observed simultaneously in soft x - rays ( xrt ) , hard x - rays ( bat ) and gamma - rays ( k - w ) .
the broad spectral coverage of these simultaneous measurements has allowed us to fit the early prompt emission , rapid decay , and several flares in the early emission with several different spectral models . in general
we find a cutoff powerlaw model to be a good fit to segments with data extending into the mev range , thus able to constrain the high energy component of the model .
for data segments with 0.3150 kev coverage ( bat and xrt data ) we find that a simple absorbed powerlaw is often an adequate fit to the data , though an absorbed powerlaw plus blackbody or absorbed broken powerlaw model seems to sometimes be a marginally better fit during periods of flaring activity .
the lightcurve structure of grb 050713a is quite typical of many grbs that have been observed by _
it has an early section showing steep decay slopes of @xmath132 and bright flares extending until t@xmath0+@xmath11 ks , followed by a break to a flatter section with decay slope @xmath133 lasting until t@xmath0+@xmath125 ks , followed by a break to a steeper slope of @xmath58 .
we have temporally separated the early , flaring portion of the burst into 10 segments and attempted to fit each segment using 4 different spectral models : 1 ) an absorbed powerlaw 2 ) an absorbed cutoff or broken powerlaw ( cutoff for data extending beyond 150 kev , broken otherwise )
3 ) an absorbed band function and 4 ) an absorbed blackbody plus powerlaw . in all segments where at least two instruments provide significant , simultaneous levels of emission , and
hence the spectral data span more than 2 decades in energy , we find that at least one of the higher order spectral models is acceptable and , in several cases , is a better fit to the data than a simple absorbed powerlaw .
this suggests that the spectral shape of grb flares , while consistent with a simple absorbed powerlaw when viewed through any particular narrow spectral window , is intrinsically fit in the broadband by a model with attenuated flux above ( and possibly below ) some threshold energy .
it has long been known that grb prompt emission is better fit by spectral models with a high ( and sometimes low ) energy cutoff than by a simple absorbed powerlaw @xcite , and thus the indication that grb flares are fit by a similar spectral model suggests that similar emission mechanisms may be responsible for the production of flares and for the prompt emission itself , namely internal shocks produced as a result of central engine activity . since the discovery of x - ray flares in grbs by _ swift _ ,
relatively few of the flares have been observed simultaneously across a broad enough energy range to determine whether such higher order models are necessary to fit their spectra , making the multi - instrument observations of grb 050713a unique and valuable .
we have also examined the temporal properties of the flares seen in grb 050713a as early as t@xmath0 + 80 s and as late as t@xmath0 + 10 ks . in all cases
we find the flares to have steep powerlaw rise and decay slopes and 0.1 @xmath5 @xmath107t / t @xmath5 1 , which also suggests internal rather than external shocks as the production mechanism for the flares @xcite .
we have noted that the presence of multiple flares and the failure of those flares to fit the closure relations associated with the external shock in a wind or constant density ism further argues in favor of the internal shock origin for the flares .
we have furthermore discussed the difficulty that _ swift _ xrt will face in detecting late time flaring activity .
we have noted that the xrt will have difficulty resolving late , short - timescale flares due to the low xrt count rates typically involved .
data from higher throughput instruments such as _ xmm - newton _ _ epic _ will be important for constraining flares at these times
. grb 050713a has simultaneous coverage at moderately late times with _ xmm - newton _ _ epic _ , but no conclusive evidence of flaring in the _ xmm - newton _ data has been found in this case . finally , we have created a broadband sed of the flaring region of grb 050713a from 0.002 kev to 500gev at times from t@xmath0 + 20 s to t@xmath0 + 1000 s. we find that the sed is inconsistent with a single absorbed powerlaw or an absorbed band function and is best fit by an absorbed broken powerlaw .
this overall sed again implies that grb flares are best fit by a spectral model similar to that of the prompt emission itself and thus suggests a common mechanism for the emission from the prompt phase and from flares .
this work is supported at penn state by nasa contract nas5 - 00136 .
we gratefully acknowledge the contributions of dozens of members of the _ swift _ team at psu , university of leicester , oab , gsfc , asdc , mpe and our subcontractors , who helped make this observatory possible and to the flight operations team for their support above and beyond the call of duty .
this work contains observations obtained with _ xmm - newton _ , an esa science mission with instruments and contributions directly funded by esa member states and the usa ( nasa ) .
the k - w experiment is supported by russian space agency contract and rfbr grant 06 - 02 - 16070 .
albert , j. , et al .
, 2006 , apj , 641 , l9-l12 aptekar , r.l . , et al . , 1995 , , 71 , 265 band , d. , et al . , 1993 , , 413 , 231 barthelmy , s. , et al . , 2005 , , 120 , 143 barthelmy , s. , et al . , 2005 ,
nature 438 , 994 burrows , d.n .
, et al . , 2005a , , 120 , 165 burrows , d.n . , et al
, 2005b , science , 309 , 1833 chevalier , r.a . , and li , z.y . , 1999 , , 520 , l29 dickey , j.m . and
lockman , f.j .
28 , p215 falcone , a.d . ,
et al . , 2005 , gcn 3581 falcone , a.d . , 2006 , , 641 , 1010 gallant , y.a . , et al .
, 1999 , a&as , 138 , 549 g gehrels , n. , et al . , 2004 , , 611 , 1005 gendre , b. , et al . , 2005 ,
a&as , accepted , astro - ph/0603431 hill , j.e . , et al . , 2004 , , vol .
5165 , 217 hill , j.e . , et al . , 2005 , , 639 , 303 ioka , k. , et al
, 2005 , apj , 631 , 429 de jager , o.c . and stecker , f.w .
2002 , apj , 566 , 738 kneiske , t. m. , et al . , 2004 ,
a&as , 413 , 807 kobayashi , , s. , et al . , 2005 , astro - ph/0506157 kraft , r.p . ,
burrows , d.n . , nousek , j.a . , 1991 , , 344 , 355 kouveliotou , c. , et al . , 1993 , , 413:l101 , l104 malesani , d. , et al . , 2005 ,
gcn 3582 mirzoyan , r. et al . , 2005 ,
ip procs 29@xmath134 icrc nousek , j.a . , et al . , 2005 , , 642 , 389 palmer , d.m . , et al . , 2005 ,
gcn 3597 perna , r. , armitage , p.j . ,
zhang , b. , 2006 , apj , 636 , 29 piro , l. , 2005 , , 623 , 314 piro , l. , 2005 , astro - ph/0506020 primack , j.r . ,
bullock , j.s . , somerville , r.s .
2005 , aip conf .
proceedings # 745 , p23 romano , p. , et al . , 2006 , a&as , 450 , 59 roming , p. , et al . , 2005 , , 120 , 95 ryde , f. , 2005 , , 625 , 95 sari , r. , et al . , 1998 , , 497 , l17 sari , r. , and piran , t. , 1999 , , 520 , 641 schlegel , d.j . , finkbeiner , d.p . , and davies , m. , 1998 , , 500 , 525 wren , j. , et al . , 2005 , gcn 3604 zhang , b. and meszaros , p. , 2004
, ijomp a , v19 , 15 , 2385 zhang , b. , et al .
, 2006 , , 642 , 354 cllcll swift - bat & 05 - 07
- 13 - 04:29:02.4 & 05 - 07 - 13 - 04:32:00 & 178 & 0 + konus - wind & 05 - 07 - 13 - 04:29:03.1 & 05 - 07 - 13 - 04:37:14.8 & 491.8 & 0.7 + magic(limit ) & 05 - 07 - 13 - 04:29:42 & 05 - 07 - 13 - 05:06:45 & 2223 & 40 + swift - xrt & 05 - 07 - 13 - 04:30:14 & 05 - 08 - 01 - 04:37:02 & 167740 & 72 + xmm - newton & 05 - 07 - 13 - 10:17:00 & 05 - 07 - 13 - 18:22:00 & 20900 & 21000 + cccccc 79 & 101 & 22 & na & 5.6 @xmath102 1.8 & @xmath135 ( from bat ) + 101 & 161 & 60 & 23.3 @xmath102 5 & 8.4 @xmath102 1.8 & @xmath136 + 161 & 304 & 143 & 8.9 @xmath102 3 & 6.1 @xmath102 1.2 & @xmath137 + 9751 & 11840 & 2089 & 5.76 @xmath102 1.8 & 11.0 @xmath102 2.4 & @xmath53 + cccc mcdonald obs , tex & t@xmath0 + 22.4 s & unfilt & 17.7 ( lim ) + raptor - s , lanl & t@xmath0 + 99.3 s & r & 18.4 @xmath102 0.18 + liverpool robotic telescope , canary islands & t@xmath0 + 180 s & r & 19.2 + _ swift _ & t@xmath0 + 252 s & v & 17.98 + _ swift _ & t@xmath0 + 309 s & u & 17.81 + _ swift _ & t@xmath0 + 311 s & uvm2 & 17.13 + _ swift _ & t@xmath0 + 325 s & uvw1 & 16.85 + _ swift _ & t@xmath0 + 326
s & uvw2 & 17.08 + _ swift _ & t@xmath0 + 351 s & b & 18.08 + red buttes obs , wy & t@xmath0 + 27 m & r & 19.4 ( lim ) + red buttes obs , wy & t@xmath0 + 31 m & i & 18.2 ( lim ) + nordic optical tel & t@xmath0 + 47 m & r & @xmath5 dss limit + galileo national telescope , canary islands & t@xmath0 + 48 m & i & @xmath5 dss limit + arc telescope , apache point obs & t@xmath0 + 53 m & j , h , k & detected + red buttes obs , wy & t@xmath0 + 93 m & r & 19.4 ( lim ) + red buttes obs , wy & t@xmath0 + 98 m & i & 18.7 ( lim ) + lulin telescope , taiwan & t@xmath0 + 10.3h & r & 22.4 ( lim ) + vla , nrao & t@xmath0 + 4.3d & 8.5 ghz & 96 microjan + cccc _ swift _ orbit 1 & 2.28 @xmath102 0.04 & @xmath149 & pre- energy injection phase + _ swift _ orbit 2 & 1.90 @xmath102 0.13 & @xmath150 & energy injection phase + _ swift _ orbit 3 & 2.25 @xmath102 0.23 & @xmath151 & flare during energy injection + _ swift _ after orbit 3 & 2.8 @xmath102 0.6 & @xmath152 & post - break + _ xmm - newton _ & 2.1 @xmath102 0.05 & @xmath153 & post - break + + instr & bat & and &
k - w & & bat & & xrt & and & bat & + pl : & na & na & na & na & na & 0.52 & 0.59 & 0.60 & 0.54 & 0.38 + pl : phind & 1.26 & 1.44 & 1.61 & 1.47 & 1.83 & 2.47 & 1.72 & 2.68 & 2.52 & 2.55 + pl:@xmath154 & 1.54 & 1.96 & 1.39 & 1.95 & 0.98 & 1.20 & 1.21 & 1.07 & 1.04 & 1.00 + pl : dof & 12 & 101 & 91 & 98 & 26 & 74 & 98 & 125 & 22 & 37 + cutoffpl : phind & 0.913 & 1.26 & 1.31 & 1.25 & na & na & na & na & na & na + cutoffpl : e@xmath114 & 270 & 421 & 312 & 400 & na & na & na & na & na & na + cutoffpl:@xmath35 & 1.20 & 1.60 & 1.02 & 1.29 & na & na & na & na & na & na + cutoffpl : dof & 11 & 110 & 90 & 100 & na & na & na & na & na & na + bknpl : & na & na & na & na & na & 0.58 & 0.43 & 0.45 & na & na + bknpl : phind1 & na & na & na & na & 0.95 & 2.69 & 1.26 & 1.93 & na & na + bknpl : e@xmath155 & na & na & na & na & 5.8 & 4.1 & 3.4 & 1.8 & na & na + bknpl : phind2 & na & na & na & na & 2.1 & 1.5 & 2.0 & 2.8 & na & na + bknpl:@xmath154 & na & na & na & na & 1.06 & 1.31 & 1.03 & 1.04 & na & na + bknpl : dof & na & na & na & na & 23 & 73 & 97 & 124 & 21 & 36 + band : & na & na & na & na & na & 0.33 & 0.49 & 0.54 & 0.41 & 0.33 + band:@xmath32 & -1.13 & -1.27 & -1.27 & -1.29 & -2.01 & -1.69 & -1.46 & -2.32 & -1.98 & -2.17 + band:@xmath156 & -1.26 & -9.36 & -2.39 & -9.29 & -9.07 & -1.88 & -9.36 & -9.32 & -9.04 & -8.78 + band : e@xmath114 & 101 & 761 & 244 & 636 & 994 & 10.6 & 30.7 & 10.6 & 10.6 & 10.6 + band:@xmath154 & 1.85 & 1.61 & 1.03 & 1.28 & 1.06 & 2.11 & 1.11 & 1.13 & 1.26 & 1.12 + band : dof & 10 & 109 & 89 & 101 & 23 & 73 & 97 & 124 & 21 & 36 + pl+bb : & na & na & na & na & na & 0.59 & 1.17 & 0.63 & 0.50 & 0.44 + pl+bb : kt & 3.59 & 59.4 & 26.4 & 44.97 & na & 7.1 & 0.1 & 4.4 & 200 & 5.2 + pl+bb : bb@xmath157 & 0.068 & 2.74 & 0.46 & 1.47 & na & 0.07 & 2.48 & 0.01 & 2.90 & 0.01 + pl+bb : phind & 1.00 & 1.64 & 1.76 & 1.63 & na & 2.75 & 2.00 & 2.77 & 2.43 & 2.83 + pl+bb : pl@xmath157 & 0.053 & 13.97 & 7.69 & 11.65 & na & 1.53 & 2.32 & 2.03 & 0.51 & 0.44 + pl+bb:@xmath154 & 1.15 & 1.79 & 1.03 & 1.60 & na & 1.33 & 1.04 & 1.07 & 0.90 & 0.98 + pl+bb : dof & 10 & 109 & 89 & 101 & na & 73 & 97 & 124 & 21 & 36 + lc pl : & @xmath158 @xmath12 +
pl : phind & 2.14 + pl:@xmath154 & 10 ( 65 dof ) + bknpl : & @xmath159 @xmath12 + bknpl : phind1 & 1.1 @xmath102 0.1 + bknpl : e@xmath155 & 1.3 @xmath102 0.2 kev + bknpl : phind2 & 2.2 @xmath102 0.1 + bknpl:@xmath154 & 1.20 ( 63 dof ) + band : & @xmath160 @xmath12 + band:@xmath32 & 1.3 + band:@xmath156 & 2.2 + band : e@xmath114 & 10.6 kev + band:@xmath154 & 2.97 ( 63 dof ) + magic observations . filled circles : number of excess events for 20 s intervals , in the 37 min window after the burst onset .
open circles : number of background events in the signal region .
no significant source signal is detected above the background . ] | _ swift _ discovered grb 050713a and slewed promptly to begin observing with its narrow field instruments 72.6 seconds after the burst onset , while the prompt gamma - ray emission was still detectable in the bat .
simultaneous emission from two flares is detected in the bat and xrt .
this burst marks just the second time that the bat and xrt have simultaneously detected emission from a burst and the first time that both instruments have produced a well sampled , simultaneous dataset covering multiple x - ray flares . the temporal rise and decay parameters of the flares are consistent with the internal shock mechanism .
in addition to the _ swift _ coverage of grb 050713a , we report on the konus - wind ( k - w ) detection of the prompt emission , an upper limiting gev measurement of the prompt emission made by the magic imaging atmospheric cherenkov telescope and _ xmm - newton _ observations of the afterglow . simultaneous observations between _ swift _ xrt and _ xmm - newton _ produce consistent results , showing a break in the lightcurve at t@xmath0+@xmath115 ks .
together , these four observatories provide unusually broad spectral coverage of the prompt emission and detailed x - ray follow - up of the afterglow for two weeks after the burst trigger .
simultaneous spectral fits of k - w with bat and bat with xrt data indicate that an absorbed broken powerlaw is often a better fit to grb flares than a simple absorbed powerlaw .
these spectral results together with the rapid temporal rise and decay of the flares suggest that flares are produced in internal shocks due to late time central engine activity . |
for oxygen and calcium we used the pseudopotentials ( pps ) optimized by shulenburger and mattsson @xcite , who also demonstrated their good quality by performing numerous tests .
of a much greater concern in the present study was the proper performance of the cu pp since the magnetic properties of ca@xmath2cuo@xmath3 are largely determined by the behavior of the cu @xmath13 electrons .
therefore , we subjected our candidate cu pps to a comprehensive selection process , as presented below .
the candidate cu pps were generated using the opium code @xcite .
our final choice is the hard ne - core cu pp [ fig .
[ f.ne_hardsoft ] ( a ) , ( c ) ] that satisfies the most stringent accuracy criteria and thus ensures the validity of the bulk calculations .
[ [ bulk - dft - calculations - and - rejection - of - ar - core - cu - pseudopotentials ] ] bulk dft calculations and rejection of ar - core cu pseudopotentials ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ using our hard ne - core cu pp , we are able to accurately reproduce with quantum espresso the lda results of the all - electron ( ae ) code wien2k @xcite for bulk ca@xmath2cuo@xmath3 in an antiferromagnetic ( afm ) and a ferromagnetic ( fm ) states .
thus , in fig . [ f.dos ] the ca@xmath2cuo@xmath3 densities of states ( dos ) obtained from pp and ae calculations are compared , for an afm [ fig .
[ f.dos ] ( a ) ] and an fm [ fig
. [ f.dos ] ( b ) ] cu spin configurations . for both configurations ,
the agreement with the ae code is very good .
interestingly , we were able to equally well reproduce the ae dos when also using properly optimized mg - core cu pps .
this , however , did not hold for any of the ar - core cu pps we tried : in this case , the bandwidth of the pp states as well as the conduction gap ( afm configuration ) are systematically larger than in the ae calculations .
this allowed us to discard ar - core cu pps already at this stage of pp validation .
as for the spin superexchange coupling constant @xmath1 , quantum espresso with the hard ne - core cu pp gives 0.64 ev in lda , while wien2k gives 0.72 ev . in lda+u , the wien2k @xmath1 is rapidly decreasing as @xmath35 . in pp dft
, we find only a weak dependence on @xmath8 .
this may be a peculiarity of the implementation of the lda+u scheme within the plane - wave basis method of quantum espresso .
[ [ fp - dmc - atom - ionization - energies - and - rejection - of - mg - core - cu - pseudopotentials ] ] fp - dmc atom ionization energies and rejection of mg - core cu pseudopotentials ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ as we have pointed out , in dft calculations for bulk ca@xmath2cuo@xmath3 , the ne - core and the mg - core cu pps appear to be equally good , provided proper optimization has been carried out .
however , this does not necessarily mean that they will perform well in diffusion monte carlo calculations . in order to test the latter , we calculated the cu atom ionization energy with fp - dmc , using lda for generating trial wave functions .
the fp - dmc computational parameters are the following : @xmath36 ha@xmath19 for the imaginary time step and 4000 for the number of dmc walkers . with the hard ne - core and with the mg - core cu pps , we obtain , respectively , 7.724(37 ) ev and 8.302(36 ) ev for the cu atom ionization energy .
the former number is in a much better agreement with the experimental result of 7.72638(1 ) ev .
one of the reasons of the poor performance of the mg - core pp in this test , is that the @xmath37 orbital , which is treated as core here , has a significant overlap in space with the @xmath38 orbital , treated as valence [ see fig .
[ f.ne_hardsoft ] ( c ) ] .
this causes less trouble in dft which is formulated in terms of kohn - sham orbitals so that such a division based on orbital character is natural .
dmc , on the other hand , operates with a full many - electron wave function where a removal of the @xmath37 electrons negatively affects the representation of the motion of the nearby @xmath38 electrons .
this issue has also been discussed in the context of gw @xcite .
we would like to point out that using a mg - core cu pp instead of a ne - core one in bulk ca@xmath2cuo@xmath3 fp - dmc calculations could provide a speedup of more than 30% , owing to the fact that the deeply lying @xmath37 electrons are a significant source of the energy variance .
[ [ equation - of - state - of - cuo - dimer - and - hard - vs .- soft - ne - core - cu - pseudopotentials ] ] equation of state of cuo dimer and hard vs. soft ne - core cu pseudopotentials ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ although the hard ne - core pp has been proven to be of a good quality for both dft and dmc , it has a disadvantage in terms of computational load and memory demands as it requires a minimum of 500 ry energy cut - off for the quantum espresso plane - wave basis .
this results from the quite small cut - off radii of the @xmath39- , @xmath40- , and @xmath0-channels , as shown in fig .
[ f.ne_hardsoft ] ( a ) . in view of this
, we constructed and tested an alternative ne - core cu pp with a soft core and low energy cut - off requirements of less that 200 ry [ fig .
[ f.ne_hardsoft ] ( b ) ] .
it demonstrated excellent characteristics in dft tests but , unfortunately , gives worse results in dmc than the hard - core cu pp .
in particular , with the soft - core pp the equilibrium interatomic separation distance in a cuo molecule is overestimated by more than 3% in fp - dmc [ holds for lda , lda+u ( @xmath8=3.5 , 6 ev ) , and b3lyp wave functions ] , whereas with the hard - core pp it is overestimated by only 0.6% ( fig .
[ f.cuo_eos ] ) . also the cu atom ionization energy is slightly underestimated : 7.548(42 ) ev .
thus all calculations in dmc reported in the paper were obtained with the hard ne - core pseudopotential .
the surprising sensitivities to pseudopotential formulation that we find even for small core pseudopotentials indicates that careful testing is essential and results from large core pseudopotentials must be treated with caution .
a. c. walters , t. g. perring , j. caux , a. t. savici , g. d. gu , c. lee , w. ku , and i. a. zaliznyak , _ effect of covalent bonding on magnetism and the missing neutron intensity in copper oxide compounds _ , http://dx.doi.org/10.1038/nphys1405[nat .
phys . * 5 * , 867 ( 2009 ) ] .
t. ami , m. k. crawford , r. l. harlow , z. r. wang , d. c. johnston , q. huang , and r. w. erwin , _ magnetic susceptibility and low - temperature structure of the linear chain cuprate @xmath42@xmath43 _ , http://dx.doi.org/10.1103/physrevb.51.5994[phys .
b * 51 * , 5994 ( 1995 ) ] .
n. motoyama , h. eisaki , and s. uchida , _ magnetic susceptibility of ideal spin 1@xmath442 heisenberg antiferromagnetic chain systems , @xmath45 and @xmath46 _ , http://dx.doi.org/10.1103/physrevlett.76.3212[phys .
lett . * 76 * , 3212 ( 1996 ) ] . associating the qmc afm state with the bethe anzatz ground - state of the 1d heisenberg model ( instead of the neel / ising state
as is usually done @xcite ) gives a different representation of @xmath1 : @xmath47 ev . c. de graaf and f. illas , _ electronic structure and magnetic interactions of the spin - chain compounds @xmath48 and @xmath45 _ , http://dx.doi.org/10.1103/physrevb.63.014404[phys .
b * 63 * , 014404 ( 2000 ) ] .
j. schlappa , k. wohlfeld , k. j. zhou , m. mourigal , m. w. haverkort , v. n. strocov , l. hozoi , c. monney , s. nishimoto , s. singh , a. revcolevschi , j. caux , l. patthey , h. m. rnnow , j. van den brink , and t. schmitt , _ spin orbital separation in the quasi - one - dimensional mott insulator sr@xmath2cuo@xmath3 _ , http://dx.doi.org/10.1038/nature10974[nature * 485 * , 82 ( 2012 ) ] . from the fit
to the temperature dependence of magnetic susceptibility , the superexchange constant @xmath1 between the nearest - neighbor cu spins in sr@xmath2cuo@xmath3 was estimated as 0.146(13 ) @xcite and 0.189(17 ) @xcite ev .
for the same compound , analysis of the angle - resolved photoemission spectroscopy data leads to a @xmath1 value between 0.13 and 0.16 ev @xcite while the midinfrared spectrum interpretation in terms of phonon - assisted magnetic excitations to a value between 0.246 and 0.26 ev @xcite .
resonant inelastic x - ray scattering @xcite and inelastic neutron scattering @xcite data point to 0.249 and 0.241(11 ) ev , respectively .
finally , from the @xmath49cu neutron magnetic resonance data one infers the @xmath1 value of 0.24 ev @xcite .
r. q. hood , p. r. c. kent , and f. a. reboredo , _ diffusion quantum monte carlo study of the equation of state and point defects in aluminum _ , http://dx.doi.org/10.1103/physrevb.85.134109[phys .
b * 85 * , 134109 ( 2012 ) ] .
k. p. esler , r. e. cohen , b. militzer , j. kim , r. j. needs , and m. d. towler , _ fundamental high - pressure calibration from all - electron quantum monte carlo calculations _ , http://dx.doi.org/10.1103/physrevlett.104.185702[phys .
lett . * 104 * , 185702 ( 2010 ) ] .
p. giannozzi , s. baroni , n. bonini , m. calandra , r. car , c. cavazzoni , d. ceresoli , g. l. chiarotti , m. cococcioni , i. dabo , a. dal corso , s. de gironcoli , s. fabris , g. fratesi , r. gebauer , u. gerstmann , c. gougoussis , a. kokalj , m. lazzeri , l. martin - samos , n. marzari , f. mauri , r. mazzarello , s. paolini , a. pasquarello , l. paulatto , c. sbraccia , s. scandolo , g. sclauzero , a. p. seitsonen , a. smogunov , p. umari , and r. m. wentzcovitch , _ quantum espresso : a modular and open - source software project for quantum simulations of materials _ , http://dx.doi.org/10.1088/0953-8984/21/39/395502[j .
matter * 21 * , 395502 ( 2009 ) ] .
i. i. mazin , m. d. johannes , l. boeri , k. koepernik , and d. j. singh , _ problems with reconciling density functional theory calculations with experiment in ferropnictides _ , http://dx.doi.org/10.1103/physrevb.78.085104[phys .
b * 78 * , 085104 ( 2008 ) ] .
h. akamatsu , y. kumagai , f. oba , k. fujita , h. murakami , k. tanaka , and i. tanaka , _
antiferromagnetic superexchange via 3@xmath0 states of titanium in eutio@xmath3 as seen from hybrid hartree - fock density functional calculations _ , http://dx.doi.org/10.1103/physrevb.83.214421[phys .
b * 83 * , 214421 ( 2011 ) ] .
h. seo , a. posadas , and a. a. demkov , _ strain - driven spin - state transition and superexchange interaction in lacoo@xmath3 : _ ab initio _ study _ , http://dx.doi.org/10.1103/physrevb.86.014430[phys .
b * 86 * , 014430 ( 2012 ) ] . c. duan , r. f. sabiryanov , w. n. mei , p. a. dowben , s. s. jaswal , and e. y. tsymbal , _ magnetic ordering in gd monopnictides : indirect exchange versus superexchange interaction _ , http://dx.doi.org/10.1063/1.2200767[appl . phys . lett . *
88 * , ( 2006 ) ] .
p. reinhardt , m. p. habas , r. dovesi , i. de p. r. moreira , and f. illas , _ magnetic coupling in the weak ferromagnet cuf@xmath2 _ , http://dx.doi.org/10.1103/physrevb.59.1016[phys .
b * 59 * , 1016 ( 1999 ) ] .
j. kim , k. p. esler , j. mcminis , m. a. morales , b. k. clark , l. shulenburger , and d. m. ceperley , _ hybrid algorithms in quantum monte carlo _ , http://dx.doi.org/10.1088/1742-6596/402/1/012008[j .
. ser . * 402 * , 012008 ( 2012 ) ] . l. k. wagner and p. abbamonte , _ the effect of electron correlation on the electronic structure and spin - lattice coupling of the high - t@xmath50 cuprates : quantum monte carlo calculations _ , http://www.arxiv.org/abs/1402.4680[arxiv:1402.4680 ( 2014 ) ] .
p. blaha , k. schwarz , g. k. h. madsen , d. kvasnicka , and j. luitz , _
wien2k , an augmented plane wave + local orbitals program for calculating crystal properties _ , karlheinz schwarz , techn .
universitt wien , austria , ( 2001 ) h. dixit , r. saniz , d. lamoen , and b. partoens , _ the quasiparticle band structure of zincblende and rocksalt zno _ , http://dx.doi.org/10.1088/0953-8984/22/12/125505[j .
phys . : condens .
matter * 22 * , 125505 ( 2010 ) ] .
h. dixit , r. saniz , d. lamoen , and b. partoens , _ accurate pseudopotential description of the gw bandstructure of zno _ , http://dx.doi.org/10.1016/j.cpc.2011.02.001[comput .
. commun . * 182 * , 2029 ( 2011 ) ] .
h. fujisawa , t. yokoya , t. takahashi , s. miyasaka , m. kibune , and h. takagi , _ spin - charge separation in single - chain compound sr@xmath2cuo@xmath3 studied by angle - resolved photoemission _ , http://dx.doi.org/10.1016/s0038-1098(98)00099-4[solid state commun . * 106 * , 543 ( 1998 ) ] .
j. lorenzana and r. eder , _ dynamics of the one - dimensional heisenberg model and optical absorption of spinons in cuprate antiferromagnetic chains _ , http://dx.doi.org/10.1103/physrevb.55.r3358[phys .
b * 55 * , r3358 ( 1997 ) ] .
h. suzuura , h. yasuhara , a. furusaki , n. nagaosa , and y. tokura , _ singularities in optical spectra of quantum spin chains _ , http://dx.doi.org/10.1103/physrevlett.76.2579[phys .
* 76 * , 2579 ( 1996 ) ] .
m. takigawa , o. a. starykh , a. w. sandvik , and r. r. p. singh , _ nuclear relaxation in the spin-@xmath51 antiferromagnetic chain compound sr@xmath2cuo@xmath3 : comparison between theories and experiments _ , http://dx.doi.org/10.1103/physrevb.56.13681[phys .
b * 56 * , 13681 ( 1997 ) ] .
m. takigawa , n. motoyama , h. eisaki , and s. uchida , _ dynamics in the @xmath52 one - dimensional antiferromagnet sr@xmath2cuo@xmath3 via @xmath49cu nmr _ , http://dx.doi.org/10.1103/physrevlett.76.4612[phys .
lett . * 76 * , 4612 ( 1996 ) ] . | in view of the continuous theoretical efforts aimed at an accurate microscopic description of the strongly correlated transition metal oxides and related materials , we show that with continuum quantum monte carlo ( qmc ) calculations it is possible to obtain the value of the spin superexchange coupling constant of a copper oxide in a quantitatively excellent agreement with experiment .
the variational nature of the qmc total energy allows us to identify the best trial wave function out of the available pool of wave functions , which makes the approach essentially free from adjustable parameters and thus truly _
ab initio_. the present results on magnetic interactions suggest that qmc is capable of accurately describing ground state properties of strongly correlated materials . for decades ,
transition metal oxides have been amongst the most intriguing materials due to the complex correlated behavior of the 3@xmath0 or 4@xmath0 electrons of a transition metal ion . in particular ,
strong electronic correlations often give rise to non - trivial magnetism , such as quantum spin - liquid states in low - dimensional mott insulating oxides .
high - temperature superconductivity in copper oxides ( cuprates ) is also believed to originate from magnetic spin excitations that bind cooper pairs @xcite .
electronic correlations , however , also make this class of materials among the most difficult to describe theoretically , both from model as well as _
ab initio _ perspectives .
one of the practical challenges of _ ab initio _ electronic structure theory is to accurately predict the strength of magnetic coupling between localized spins of transition metal ions @xcite . for solids , a natural method of choice
is periodic density functional theory ( dft ) .
dft gives access to the system s ground state energy corresponding to different configurations of localized spins , which can be mapped onto the eigenstates of the a spin model to extract the magnetic couplings @xmath1 .
this approach relies on the accuracy of the description of the ground state .
unfortunately , the presently available approximations to the exchange - correlation functional in dft either poorly account for the exchange and correlation effects [ local density approximation ( lda ) ] or depend on empirical input parameters ( lda+u , hybrid functionals ) .
often , the only way to find an appropriate approximation for the system of interest is by comparing theoretical calculations with experiment , compromising the predictive nature of such calculations . . the nearest - neighbor spin superexchange coupling constant @xmath1 of ca@xmath2cuo@xmath3 and sr@xmath2cuo@xmath3 obtained with different theoretical ( ca@xmath2cuo@xmath3 ) and experimental ( sr@xmath2cuo@xmath3 ) methods .
the abbreviations used stand for : ins = inelastic neutron scattering , fp - dmc = fixed - phase diffusion monte carlo , ddci3=difference dedicated configuration interaction with three degrees of freedom , uhf = unrestricted hartree - fock , lda = local density approximation .
@xmath4 denotes a temperature dependent magnetic susceptibility .
[ cols= " < , > , < , < " , ] this problem is generic to a broad class of transition metal oxides . let us exemplify the aforementioned limitations of dft by considering the case of the mott insulators ca@xmath2cuo@xmath3 and sr@xmath2cuo@xmath3 .
these systems are one of the best realizations of the one - dimensional ( 1d ) spin-@xmath5 antiferromagnetic heisenberg chain model , demonstrating spin - liquid behavior and separation of spin and orbital degrees of freedom @xcite .
the crystal structure of ca@xmath2cuo@xmath3 and sr@xmath2cuo@xmath3 is similar to that of the superconducting two - dimensional cuprates , with the difference that in the cuo@xmath2 plane the oxygen atoms along the crystallographic @xmath6 direction are missing , so that the cu chains run along the @xmath7 direction [ fig .
[ f.js ] ( a ) ] .
the cu - o - cu bridge provides a favorable path for superexchange coupling between cu spins , resulting in a particularly strong coupling constant @xmath1 .
the experimental estimate of @xmath1 has been extracted from various probes , performed mostly on sr@xmath2cuo@xmath3 , and ranges between 0.13 and 0.26 ev ( see table [ t.js ] and ref . for details ) . temperature dependent magnetic susceptibility measurements , supported by rigorous theoretical modeling @xcite , narrow this window to 0.15 - 0.19 ev .
theoretical predictions of @xmath1 , in turn , vary drastically by as much as an order of magnitude depending on the method used . for example , periodic dft with lda gives 0.64 ev , whereas the periodic unrestricted hartree - fock ( uhf ) method gives 0.04 ev ( table [ t.js ] ) .
in addition , lda+u calculations give results that depend strongly on @xmath8 .
cluster calculations with the configuration interaction method may give a more reasonable result , as in this case ( table [ t.js ] , ref . ) , but , generally speaking , their applicability to condensed phase systems is intrinsically limited .
in view of this , developing a new , more universal and accurate , _ ab initio _ approach to computing magnetic interactions is critically important .
needless to say , a method capable of accurately describing magnetic interactions in transition metal oxides will also provide an improved _ ab initio _ description of many other ground state properties of these complex systems , and hence may yield new physical insights . here
, we apply the diffusion monte carlo method @xcite within the fixed - phase approximation ( fp - dmc ) to compute the value of the spin superexchange interaction constant in a transition metal oxide . to the best of our knowledge ,
this is among the first calculations of magnetic couplings in complex oxides that has been attempted with a method capable of chemical accuracy in full periodic boundary conditions .
the fixed - phase error is controlled by scanning over a set of trial wave functions and using the variational nature of dmc energy , as explained below .
this way , the variational principle determines the choice of the initial dft functional to generate a trial wave function and thus eliminates empiricism from the calculations .
we choose the 1d cuprate antiferromagnet ca@xmath2cuo@xmath3 as our test system because of the simplicity of its underlying spin model and relatively light constituent atoms , as compared to other cuprates , including sr@xmath2cuo@xmath3 , to minimize the potential role of relativistic effects .
our result for the nearest - neighbor cu spin coupling , @xmath9 ev @xcite , is in excellent agreement with the value extracted from the temperature dependence of magnetic susceptibility ( refs . ,
table [ t.js ] ) .
historically , quantum monte carlo has played a fundamental role in advancing electronic structure theory .
released - node dmc calculations on the homogeneous electron gas by ceperley and alder @xcite were used to construct the local density approximation to the exchange - correlation functional which is at the core of modern dft methods . in fp - dmc ,
the schrdinger equation is rewritten in a form of an integral diffusion equation , which is stochastically solved via quantum monte carlo sampling by iteratively propagating the wave function in imaginary time . as a result ,
the ground state ( gs ) wave function is projected out . in order to handle the fermion sign problem ,
the complex phase of the target gs wave function is fixed to those of an input trial wave function .
the fixed - phase approximation introduces a variational systematic error in the dmc energy which can be assessed and controlled by comparing results obtained with different trial wave functions .
exchange and correlation effects are fully accounted for within the fixed phase approximation . even though the computational costs of fp - dmc grow as @xmath10 , where @xmath11 is the number of particles , the costs of calculating quantities _ per particle _ grow only as @xmath12 , which is a great advantage in the present case since @xmath1 is a characteristic of a _ cu - cu bond_. due to improvements in algorithms and available computer power
, qmc has achieved a growing success in accurately predicting the properties of complex materials @xcite .
we used the dft plane - wave code quantum espresso @xcite in order to self - consistently generate trial wave functions in a single slater determinant form .
the lda+u method was applied , with the values of the on - site coulomb repulsion between cu @xmath13 electrons , @xmath8 , varying as 1 , 2 , 3.5 , 5.5 , and 7 ev .
the energy cut - off was set to 500 ry due to the use of very hard pseudopotentials and inclusion of semi - core electrons in the valence ( described below ) .
the corresponding dmc energies were subsequently compared to determine the best trial wave function .
we present results that follow procedures in the literature @xcite to obtain the spin superexchange coupling constant @xmath1 .
the coupling constant can be computed from a single total energy difference : @xmath14 .
here @xmath15 is the z - component of an electron s spin and @xmath16 is the number of nearest neighbor cu - cu bonds in a given supercell .
this results from a mapping to the total energy differences of an ising hamiltonian .
the estimate for @xmath1 changes somewhat using a different approach @xcite , though the result remains close to the experimental range .
although the total energies are variational , @xmath1 is not and care must be taken to fully optimize the trial wavefunctions .
wavefunctions of the standard spin - assigned slater - jastrow type @xcite used here are eigenstates of total @xmath17 , but not @xmath18 .
although it is possible to construct eigenstates of @xmath18 , this is seldom done in practice with dmc because the slater - jastrow form gives accurate total energies within the fixed - node / phase approximation @xcite for spin - independent hamiltonians .
for all considered @xmath8 values , both fm and afm solutions are insulating in dft .
fixed phase dmc calculations were performed with qmcpack @xcite . the dmc imaginary time step and the number of walkers
have been converged to , respectively , 0.005 ha@xmath19 and 2000 per boundary twist ( in the @xmath20 supercell ) . to assess finite size errors we considered @xmath20 ( @xmath21 ) and @xmath22 ( @xmath23 ) supercells , defined with respect to the conventional @xmath24 crystallographic unit cell of ca@xmath2cuo@xmath3 [ fig .
[ f.js ] ( a ) ] .
the @xmath20 supercell contains four formula units , _
i.e. _ , 28 atoms with 228 electrons .
we also performed averaging over twisted boundary conditions on a @xmath25 @xmath26-point grid for the @xmath20 supercell ( @xmath22 for the @xmath22 supercell ) .
the necessity of twist averaging indicates that cluster model calculations , such as those in ref .
, are under converged with respect to finite size effects .
the ionic potentials were approximated by employing pseudopotentials ( pps ) . through an extensive investigation , we have found the inclusion of semi - core electrons to be essential for high quality results .
core sizes used for the pseudopotentials are as follows : he - core for oxygen atoms ( 6 electrons in valence ) , ne - core for calcium atoms ( 10 electrons in valence ) , and ne - core for copper atoms ( 19 electrons in valence ) . the quality of the pps has been carefully tested within both dft and dmc , as reported in detail in appendix [ app ] .
we first present the fp - dmc results obtained for the @xmath20 supercell , which contains two cu atoms along the chain direction @xmath7 .
figure [ f.js ]
( b ) displays the fp - dmc total energies of the fm and afm states as a function of the trial wave function , characterized by the lda+u parameter @xmath8 . in these calculations we stress that the u is simply a convenient optimization parameter for generating fp - dmc wave functions .
both the fm and afm curves follow a non - linear @xmath8 dependence , reaching minima and leveling off in the region between @xmath8=1 and 3.5 ev , within the available statistical resolution . from the difference between the afm and fm fp - dmc total energies we compute the spin superexchange constant @xmath1 , shown in fig .
[ f.js ] ( c ) together with the respective lda+u @xmath1 values for comparison . also indicated are the ranges of experimentally determined @xmath1 values of sr@xmath2cuo@xmath3 : the light gold band represents all reported experimental estimates @xcite , while the dark gold band represents susceptibility measurements @xcite , which is one of the most reliable probes .
since , unfortunately , no equivalent experimental data on ca@xmath2cuo@xmath3 are available , we can only compare our theoretical calculations with the experiments performed on sr@xmath2cuo@xmath3 .
this is a valid approach as the spin exchange couplings of the two cuprates should differ by no more than a few percent @xcite . from fig .
[ f.js ] ( c ) , one readily sees that in the @xmath8 region between 1 and 3.5 ev , corresponding to the minimal fp - dmc energies in fig .
[ f.js ] ( b ) , the fp - dmc results for @xmath1 are in good agreement with the susceptibility data , within statistical resolution . in contrast , all electron lda+u ( lapw ) results strongly depend on @xmath8 , requiring a large value of @xmath27 to obtain reasonable @xmath1-values .
we now assess the finite size error associated with these results by performing fp - dmc calculations on a @xmath22 supercell , obtained by doubling the original @xmath20 supercell in the direction perpendicular to the cu chains .
the @xmath8=3.5 ev lda+u trial wave function is used here as the one to provide a good complex phase for fp - dmc , as has been established above .
the resulting @xmath28 ev is to be compared with the @xmath20 result of @xmath29 ev . from this , we conclude that the finite size error must be within @xmath30 ev , which is the statistical accuracy of the @xmath20 calculations of fig .
[ f.js ] ( c ) .
we would also like to give a brief comment here on the computational costs involved .
the @xmath1 values for 2 ev @xmath31 7 ev in fig . 3 ( c ) cost @xmath32100 - 130k cpu hours each ( error bar @xmath320.045 ev ) . calculating @xmath1 for
the @xmath22 supercell took 1.8 m cpu hours .
we note that such costs are not insignificant , but are affordable on modern supercomputers such as titan at ornl . in conclusion ,
we have presented a theoretical determination of the value of the spin superexchange constant in a transition metal oxide with the fp - dmc method .
our results for the 1d antiferromagnetic cuprate ca@xmath2cuo@xmath3 are in excellent agreement with experiment . moreover ,
this is a purely _ ab initio _ approach , where the fixed phase error is controlled via the variational principle , with no empirical adjustable parameters . in this sense ,
fp - dmc is superior to dft where in order to improve the description of exchange and correlations one often resorts to lda+u or hybrid functionals and chooses the `` best '' functional empirically .
the success of fp - dmc in the present case implies that this method is capable of accurately describing the complicated spin superexchange processes between the correlated cu @xmath13 orbitals and oxygen @xmath33 orbitals , involving on - site coulomb correlations and @xmath34 orbital hybridization .
we hope that our present successful application of fp - dmc will stimulate future studies of magnetic and other properties of strongly correlated transition metal oxides with this highly competitive _
ab initio _ method .
note : at the time of submitting this manuscript , we learned of similar dmc calculations , performed independently , published on ` arxiv.org ` @xcite .
the authors would like to thank l. shulenburger for sharing expertise in pseudopotential construction and for providing access to pseudopotential datasets prior to publication .
the work was supported by the materials sciences & engineering division of the office of basic energy sciences , u.s .
department of energy .
prck was supported by the scientific user facilities division , office of basic energy sciences , u.s .
department of energy . computational time used resources of the oak ridge leadership computing facility at the 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 . |
from the phase diagram of the cuprate superconductors it is evident that as a function of doping , antiferromagnetic ( af ) order gives way to @xmath0-wave superconductivity ( dsc ) . this has led to the natural view that these two ordered phases generally compete in these materials . at an overall global scale
this certainly seems to be the case .
however , when including disorder the opposite can also occur : dsc can function as a catalyst for induced spin - density wave ( sdw ) order@xcite .
this takes place because of the generation of low - energy impurity resonance states which are spin split by local magnetic ordering@xcite .
a qualitatively similar effect happens in the vortex phase generated by an external applied magnetic field@xcite .
the @xmath0-wave symmetry of the superconducting state is crucial for this cooperative effect to occur between sdw and dsc phases .
experimentally , it is well known that disorder may cause slowing down and eventual freezing of spin fluctuations@xcite .
for example , substitution of nonmagnetic zn ions for cu in near - optimally doped la@xmath3sr@xmath4cuo@xmath5 ( lsco ) has been shown to shift spectral weight into the spin gap , and eventually , for enough zn ( @xmath6 ) , generate elastic magnetic peaks in the neutron response@xcite . in yba@xmath7(cu@xmath8zn@xmath9)@xmath10o@xmath11 with @xmath12 a similar in - gap zn - induced spin mode was observed@xcite .
upon increased temperature @xmath13 the elastic signal decreases and eventually vanishes near @xmath14 which is similar to an equivalent disorder signal in zn - free lsco@xcite , these results of disorder - induced freezing of spin fluctuations are consistent with @xmath1sr data on underdoped cuprates@xcite .
recent neutron scattering off detwinned yba@xmath7cu@xmath10o@xmath15 ( ybco ) with 2% zn found induced short - range magnetic order and a redistribution of spectral weight from the resonance peak to uniaxial incommensurate ( ic ) spin fluctuations at lower energies@xcite . applied magnetic fields
introduce vortices into the system and cause much of the same phenomenology as described above for the disorder .
theoretically this can be ascribed to the fact that vortices in dsc also generate low - energy resonance states which may favorably split due to electronic correlations and generate local sdw order@xcite . for samples without static sdw order ,
experiments show that at low magnetic fields the vortices slow down the spin fluctuations and generate an in - gap mode@xcite . for larger fields the ground state exhibits static local sdw order which disappears above @xmath14 .
enhancement of static order was reported first in neutron diffraction experiments@xcite on lsco , with a correlation length of several hundred , and has been confirmed in other underdoped cuprates@xcite .
the freezing of spin fluctuations is also relevant even in the absence of substitutional disorder or external applied magnetic fields , as evidenced by the presence of a spin - glass phase in the underdoped regime .
this seems to apply to both `` clean '' cuprates like ybco where quasi - static sdw order is found@xcite , and to intrinsically disordered materials like lsco where the static spin correlations are long - range , and persist for a large doping range well into the dsc dome@xcite .
the size of the spin - glass phase in temperature and doping is clearly enhanced by disorder . here , we review a picture that has emerged for the description of the induced magnetism in underdoped cuprate materials@xcite .
we focus on the disorder- and field - induced sdw phase coexisting with dsc , and calculate the associated dynamical spin susceptibility .
the basis for our model analysis is the bcs hamiltonian for a dsc with orbital coupling to an applied magnetic field * b * , to which we add site - centered disorder and a local hubbard repulsion treated in an unrestricted hartree - fock approximation : @xmath16 here , @xmath17 creates an electron on site @xmath18 with spin @xmath19 on a @xmath20 square lattice .
the hopping matrix elements between nearest and next - nearest neighbor sites are denoted by @xmath21 and @xmath22 , respectively . the magnetic field is included through standard peierls phase factors @xmath23 , where @xmath24 and @xmath25 is the vector potential in landau gauge .
the chemical potential @xmath1 is adjusted to fix the electron density @xmath26 , where @xmath27 is the hole concentration .
the @xmath0-wave pairing amplitude @xmath28 is determined by the strength of an attractive nearest - neighbor interaction @xmath29 .
the impurity potential @xmath30 arises from a concentration @xmath31 of nonmagnetic pointlike scatterers at random positions .
all fields , @xmath28 , the local charge density @xmath32 , and the local magnetization @xmath33 are calculated self - consistently from the bogoliubov - de gennes ( bdg ) equations@xcite . in the following we fix parameters to @xmath34 , @xmath35 .
the model ( [ hamiltonian ] ) has been used extensively to study the competition between dsc and sdw phases@xcite , the origin of field - induced magnetization@xcite , and moment formation around nonmagnetic impurities@xcite . in the case of many impurities , eq.([hamiltonian ] ) was used to explain the origin of static disorder - induced magnetic droplet phases@xcite , and to study how these may form a quasi - long range ordered sdw phase .
the underlying idea is that when magnetic droplets begin to overlap , either when the magnetic correlation length grows at low doping , or when the concentration of scatterers is increased , the effective interaction between droplets allows them to align their staggered patterns coherently@xcite .
more recently eq.([hamiltonian ] ) was used to obtain semi - quantitative description of the @xmath13-dependence of the elastic neutron response in lsco in an applied magnetic field@xcite .
finally , eq.([hamiltonian ] ) has been used to explain various transport measurements@xcite such as , for example , the experimental observation of a non - universal low-@xmath13 limit of the thermal conductivity @xmath36 @xcite in terms of a disorder - induced sdw phase@xcite .
the sdw droplet phase has an enhanced scattering rate while maintaining the same low energy density of states ( dos ) as the homogeneous case , thus breaking the cancellation between the residual quasiparticle dos and relaxation rate which gave rise to the universal @xmath36 in the first place@xcite .
thus , for underdoped cuprates it is not @xmath36 but rather the ( spatially averaged ) low - energy dos which is universal , in agreement with stm experiments@xcite .
theoretically a universal low - energy dos can be traced to suppressed charge modulations caused by the electronic coulomb repulsion@xcite .
the hamiltonian ( [ hamiltonian ] ) supports both a correlation- and disorder - induced sdw phase .
specifically , in the clean case above a critical repulsion @xmath37 a global stripe phase is the favorable state , and disorder acts mainly to scramble the stripes .
below @xmath37 the ground state is a homogeneous dsc but nonmagnetic disorder or vortices may locally induce sdw order if @xmath38 where @xmath39 is another critical interaction strength@xcite . in this low-@xmath40 regime ,
both disorder and the @xmath0-wave symmetry of the pairing are crucial for generating static magnetism . in the absence of a magnetic field ,
[ fig1](a - c ) show the magnetization in real - space arising from a weak disorder potential from the dopant ions@xcite .
figure [ fig1](a ) shows that not all impurities in the correlated system need to `` magnetize '' for a given @xmath40 ; in the disordered system , the effective criterion to drive the impurity through the local magnetic phase transition is different for each impurity . increasing the repulsion
@xmath40 then increases the concentration of impurities which induce a local magnetization droplet , as shown in fig . [
fig1](b ) and [ fig1](c ) . with further increase of @xmath40 ,
the system evolves from a state with dilute non - overlapping af droplets to connected spin textures .
a similar process takes place in 1d where shender and kivelson@xcite pointed out that the interactions between impurities in a quantum spin chain are non - frustrating : if an impurity creates a local af droplet , a second one can always orient itself to avoid losing exchange energy . in 2d
this continues to apply for spin models with nearest neighbor exchange , but may break down in the presence of mobile charges . , @xmath41 , @xmath42 , @xmath43 and @xmath44 ( a ) , @xmath45 ( b ) , and @xmath46 ( c ) .
( d ) schematic illustration of how a single nel phase gets stabilized by disorder.,title="fig : " ] ( -126,87)*a * , @xmath41 , @xmath42 , @xmath43 and @xmath44 ( a ) , @xmath45 ( b ) , and @xmath46 ( c ) .
( d ) schematic illustration of how a single nel phase gets stabilized by disorder.,title="fig : " ] ( -126,87)*b * + , @xmath41 , @xmath42 , @xmath43 and @xmath44 ( a ) , @xmath45 ( b ) , and @xmath46 ( c ) .
( d ) schematic illustration of how a single nel phase gets stabilized by disorder.,title="fig : " ] ( -126,87)*c * , @xmath41 , @xmath42 , @xmath43 and @xmath44 ( a ) , @xmath45 ( b ) , and @xmath46 ( c ) .
( d ) schematic illustration of how a single nel phase gets stabilized by disorder.,title="fig : " ] ( -126,87)*d * the present mean field treatment of the hubbard interaction does not capture the band narrowing due to correlations near half - filling leading to the mott transition .
however , it is expected that underdoped systems are characterized by larger effective interactions . within the present scenario for dirty cuprates , the @xmath27 dependence of the spin order
is therefore described qualitatively by the sequence [ fig1](c ) @xmath47 [ fig1](b ) @xmath47 [ fig1](a ) , until it disappears completely at effective @xmath40 s below @xmath39 near optimal doping .
increasing @xmath27 should also be accompanied by a weakening of the dopant disorder potential due to enhanced screening . within our model , increasing @xmath40 or @xmath48 leads to qualitatively similar results , and we can not determine from this approach which effect is dominant in real systems . note from fig .
[ fig1 ] that af droplets are induced for @xmath49 , a substantially reduced critical value compared to the 1-impurity case where @xmath50 .
this is because the hubbard correlations induce charge redistributions which alter the effective local chemical potential , such that the criterion for magnetization of each impurity site depends on its local disorder environment .
some regions containing impurities have charge densities closer to the phase boundary for af order , thus enhancing local moment formation relative to the single impurity case . in the limit of large @xmath40 ,
the magnetic order becomes qualitatively similar to that arising in a stripe state with quenched disorder@xcite . in the presence of a magnetic field
, vortices may lead to an enhancement of the sdw order .
this is illustrated in fig .
[ fig2 ] which shows the magnetization for a random disorder potential with @xmath51 [ [ fig2](b ) ] and @xmath52 [ [ fig2](c ) ] . in fig .
[ fig2](c ) two flux quanta penetrate the system corresponding to a large magnetic field of @xmath53 .
focussing on the @xmath13-dependence , fig .
[ fig2](a ) shows the structure factor @xmath54 integrated near @xmath55 as a function of @xmath13@xcite .
the main result of fig .
[ fig2](a ) is seen by the dashed and dot - dashed lines displaying @xmath54 including disorder with and without an applied magnetic field , respectively . in both cases
the magnetic order sets in at @xmath14 below which the @xmath13-dependence is in remarkable agreement with neutron diffraction data on underdoped lsco@xcite .
the origin of the qualitatively different curvature of the @xmath13-dependence of @xmath54 in the presence of a magnetic field can be traced to a larger concentration of anti - phase domain walls when @xmath51@xcite .
-dependence of the peak intensity of the magnetic structure factor integrated around @xmath56 in zero field and finite field with the density of impurities @xmath57 ( dash - dotted and dashed curves , respectively ) .
for the data with @xmath58 and @xmath59 , the zero - field data were subtracted . for comparison , the structure factor in a clean system is also included for the same magnetic - field strength ( solid curve ) .
@xmath54 ( integrated ) translates directly to the ordered spin moment squared in units of @xmath60 per cu@xmath61 ion .
( b , c ) real - space magnetization at @xmath42 for a system with 10% weak impurities modeling a dopant potential ( b ) and with the addition of two vortices ( c ) . in ( c )
the vortices are located at sites @xmath62 and @xmath63 .
for all results in this figure @xmath64.,title="fig:",width=302 ] ( -230,139)*a * + + -dependence of the peak intensity of the magnetic structure factor integrated around @xmath56 in zero field and finite field with the density of impurities @xmath57 ( dash - dotted and dashed curves , respectively ) .
for the data with @xmath58 and @xmath59 , the zero - field data were subtracted . for comparison ,
the structure factor in a clean system is also included for the same magnetic - field strength ( solid curve ) .
@xmath54 ( integrated ) translates directly to the ordered spin moment squared in units of @xmath60 per cu@xmath61 ion .
( b , c ) real - space magnetization at @xmath42 for a system with 10% weak impurities modeling a dopant potential ( b ) and with the addition of two vortices ( c ) . in ( c ) the vortices are located at sites @xmath62 and @xmath63 .
for all results in this figure @xmath64.,title="fig : " ] ( -126,87)*b * -dependence of the peak intensity of the magnetic structure factor integrated around @xmath56 in zero field and finite field with the density of impurities @xmath57 ( dash - dotted and dashed curves , respectively ) . for the data with @xmath58 and @xmath59 , the zero - field data were subtracted . for comparison , the structure factor in a clean system is also included for the same magnetic - field strength ( solid curve ) .
@xmath54 ( integrated ) translates directly to the ordered spin moment squared in units of @xmath60 per cu@xmath61 ion .
( b , c ) real - space magnetization at @xmath42 for a system with 10% weak impurities modeling a dopant potential ( b ) and with the addition of two vortices ( c ) . in ( c )
the vortices are located at sites @xmath62 and @xmath63 .
for all results in this figure @xmath64.,title="fig : " ] ( -126,87)*c * we now turn to a discussion of the transverse dynamical susceptibility @xmath65\right\rangle$ ] which can be expressed in terms of the bdg eigenvalues @xmath66 and eigenvectors @xmath67 @xmath68 where @xmath69 , and @xmath70 . including the electronic interactions within rpa we find for the full susceptibility @xmath71_{\vec{r}_i,\vec{r}_l}^{-1}\chi^{xx}_{0}(\vec{r}_l,\vec{r}_j,\omega).\ ] ] fourier transforming with respect to the relative coordinate @xmath72 defines the spatially resolved momentum - dependent susceptibility @xmath73 . averaging over the center of mass coordinate @xmath74 , this expression gives the susceptibility @xmath75 relevant for comparison with neutron scattering measurements .
we have checked that at half - filling in the clean af state without pairing , @xmath75 is characterized by spin - wave excitations on a cone centered at @xmath56 as expected . although we are restricted to a mean - field / rpa approach
, a strength of the present calculation is that we can include both spin and charge degrees of freedom within an unrestricted method which is capable of describing realistic inhomogeneous situations , and thus naturally includes both the response from regions dominated by the dsc condensate or local sdw order . for a homogeneous dsc
our model reduces to a system of bogoliubov quasiparticles whose magnetic response displayed in fig . [
fig3](a ) depends crucially on the presence of a dsc gap . in the remainder
, we investigate the role of spatially inhomogeneous local moments induced in this quasiparticle system .
figure [ fig3](b ) shows how the magnetic droplet phase freezes the low - energy spin fluctuations by removing the spin - gap of the hour glass spectrum@xcite .
this is also seen from the @xmath76-summed local susceptibility @xmath77 plotted in fig . [
fig3](c ) for different concentrations of the nonmagnetic disorder . in fig .
[ fig3](b ) and [ fig3](c ) we have not included the weaker disorder potential arising from the dopants which will lead to additional freezing , but only included disorder from a few percent strong scatterers relevant for the modeling of e.g. zn substitutional disorder . in figs .
[ fig3](a , b , c ) the spectrum is @xmath78 symmetric both in the static and dynamical response . in ref .
@xcite is was shown how the stripe phase breaks this @xmath78 symmetry in the inelastic response even in the presence of substantial disorder concentrations . for a clean ( dirty )
dsc phase without ( with ) disorder using @xmath79 , @xmath51 , @xmath80 , and in ( b ) @xmath81 strong scatterers with @xmath82 .
( c ) local susceptibility @xmath83 for the same parameters as in ( b ) except for different disorder concentrations of strong scatterers . in ( b ,
c ) we have averaged over 5 different impurity configurations .
( d ) comparison of @xmath83 in the clean dsc phase with ( without ) an applied magnetic field shown by dashed ( solid ) line.,title="fig : " ] ( -126,87)*a * for a clean ( dirty ) dsc phase without ( with ) disorder using @xmath79 , @xmath51 , @xmath80 , and in ( b ) @xmath81 strong scatterers with @xmath82 .
( c ) local susceptibility @xmath83 for the same parameters as in ( b ) except for different disorder concentrations of strong scatterers . in ( b ,
c ) we have averaged over 5 different impurity configurations .
( d ) comparison of @xmath83 in the clean dsc phase with ( without ) an applied magnetic field shown by dashed ( solid ) line.,title="fig : " ] ( -126,87)*b * + for a clean ( dirty ) dsc phase without ( with ) disorder using @xmath79 , @xmath51 , @xmath80 , and in ( b ) @xmath81 strong scatterers with @xmath82 . ( c )
local susceptibility @xmath83 for the same parameters as in ( b ) except for different disorder concentrations of strong scatterers . in ( b ,
c ) we have averaged over 5 different impurity configurations .
( d ) comparison of @xmath83 in the clean dsc phase with ( without ) an applied magnetic field shown by dashed ( solid ) line.,title="fig : " ] ( -126,87)*c * for a clean ( dirty ) dsc phase without ( with ) disorder using @xmath79 , @xmath51 , @xmath80 , and in ( b ) @xmath81 strong scatterers with @xmath82 .
( c ) local susceptibility @xmath83 for the same parameters as in ( b ) except for different disorder concentrations of strong scatterers . in ( b ,
c ) we have averaged over 5 different impurity configurations .
( d ) comparison of @xmath83 in the clean dsc phase with ( without ) an applied magnetic field shown by dashed ( solid ) line.,title="fig : " ] ( -126,87)*d * we end by discussing the dynamical spin susceptibility when @xmath84 .
since the vortices can also nucleate local sdw order , we expect a similar slowing down of the fluctuations as for the point - like disorder . that this is the case can be seen from fig .
[ fig3](d ) which compares @xmath83 for the clean dsc with and without an applied magnetic field@xcite .
clearly , the same filling - in of the low - energy weight takes place because of the vortices .
we have reviewed a theoretical scenario for the understanding of disorder- and magnetic field - induced sdw phases in underdoped cuprate superconductors . in the weak coupling limit ,
nonmagnetic impurities nucleate static sdw order .
the @xmath0-wave nature of the pairing state is crucial for generation of local magnetism and works as a catalyst for the sdw phase . in the strong coupling limit disorder
mainly acts to scramble correlation - generated stripe order .
we demonstrated explicitly how both disorder and magnetic fields slow down the spin excitations as observed in experiments .
b.m.a . acknowledges support from the danish council for independent research @xmath85 natural sciences .
m.s . , s.g . , and
acknowledge support from the dfg through trr 80 .
p.j.h . received partial support from the u.s .
dept . of energy under grant de - fg02 - 05er46236 .
m. schmid _
et al . _ , new j. phys .
* 12 * , ( 2010 ) 053043 .
h. tsuchiura _
et al . _ ,
b * 64 * , ( 2001 ) 140501(r ) .
z. wang and p. a. lee , phys .
lett . * 89 * , ( 2002 ) 217002 .
j. w. harter _
et al . _ ,
b * 75 * , ( 2007 ) 054520 .
k. hirota _
et al . _ , physica b ( amsterdam ) * 241 - 243 * , ( 1998 ) 817
. b. lake _
et al . _ , nature ( london ) * 415 * , ( 2002 ) 299 .
h. kimura _
et al . _ ,
lett . * 91 * , ( 2003 ) 067002 .
a. t. savici _
et al . _ ,
lett . * 95 * , ( 2005 ) 157001 .
et al . _ ,
b * 53 * , ( 1996 ) 6811 .
h. kimura _
_ , phys . rev .
b * 59 * , ( 1999 ) 6517 .
mendels _ et al .
b * 49 * , ( 1994 ) 10035 . c. bernhard _
b * 58 * , ( 1998 ) r8937 . ch .
et al . _ ,
lett . * 80 * , ( 1998 ) 3843 .
i. watanabe _
b * 65 * , ( 2002 ) 180516(r ) . c. panagopoulos
_ et al . _ ,
b , * 66 * , ( 2002 ) 064501 .
a. suchaneck _
et al . _ ,
b. m. andersen , h. bruus , and p. hedegrd , phys .
b * 61 * , ( 2000 ) 6298 .
a. himeda _
et al . _ , j. phys .
japan * 66 * , ( 1997 ) 3367 .
j. x. zhu and c. s. ting , phys .
lett . * 87 * , ( 2001 ) 147002 . j. x. zhu , i. martin , and a. r. bishop , phys .
lett . * 89 * , ( 2002 ) 067003 .
a. ghosal , c. kallin , and a. j. berlinsky , phys .
b * 66 * , ( 2002 ) 214502 .
b. m. andersen , p. hedegrd , and h. bruus , phys .
rev b * 67 * , ( 2003 ) 134528 .
_ , science * 291 * , ( 2001 ) 1759 .
et al . _ , phys .
lett . * 102 * , ( 2009 ) 177006
. b. m. andersen , o. f. syljusen , and p. hedegrd , phys .
b * 80 * , ( 2009 ) 052509
. s. katano _
et al . _ ,
b * 62 * , ( 2002 ) r14677 .
b. khaykovich _
et al . _ ,
b * 66 * , ( 2002 ) 014528
. b. khaykovich _
et al . _ ,
b * 71 * , ( 2005 ) 220508 .
et al . _ , phys .
b * 78 * , ( 2008 ) 104525 .
d. haug _ et al .
lett . * 103 * , ( 2009 ) 017001 . c. stock _ et al .
b * 73 * , ( 2006 ) 100504(r ) .
j. sonier _
b * 76 * , ( 2007 ) 064522 . c. stock _ et al . _ ,
b * 77 * , ( 2008 ) 104513 .
v. hinkov _
et al . _ , science * 319 * , ( 2008 ) 597 .
s. wakimoto _
b * 63 * , ( 2001 ) 172501 . for a review , see m .- h .
julien , physica b * 329 - 333 * , ( 2003 ) 693 .
b. m. andersen _
et al . _ ,
* 99 * , ( 2007 ) 147002
. b. m. andersen and p. j. hirschfeld , physica ( amsterdam ) * 460c * , ( 2007 ) 744 .
b. m. andersen and p. j. hirschfeld , phys .
* 100 * , ( 2008 ) 257003
. b. m. andersen , s. graser , and p. j. hirschfeld , arxiv:1005.0520v1 .
i. martin _
et al . _ ,
* 14 * , ( 2000 ) 3567
. b. m. andersen and p. hedegrd , phys .
lett . * 95 * , ( 2005 ) 037002 .
m. ichioka , m. takigawa , and k. machida , j. phys .
. jpn . * 70 * , ( 2001 ) 33 .
g. alvarez _
et al . _ , phys . rev .
b * 71 * , ( 2005 ) 014514 .
w. a. atkinson , phys .
b * 75 * , ( 2007 ) 024510 .
w. chen , b. m. andersen , and p. j. hirschfeld , phys .
b * 80 * , ( 2009 ) 134518 .
m. sutherland _
et al . _ ,
94 * , ( 2005 ) 147004 .
x. f. sun _
* 96 * , ( 2006 ) 017008 .
m. j. graf _
et al . _ ,
b * 53 * , ( 1996 ) 15147 .
a. c. durst and p. a. lee , phys .
b * 62 * , ( 2000 ) 1270
. k. m. lang _
_ , nature ( london ) * 415 * , ( 2002 ) 412 .
a. garg , m. randeria , and n. trivedi , nat . phys .
* 4 * , ( 2008 ) 762 .
e. f. shender and s. a. kivelson , phys .
lett . * 66 * , ( 1991 ) 2384 .
m. vojta , t. vojta , and r. k. kaul , phys .
lett . * 97 * , ( 2006 ) 097001 . j. a. robertson _
et al . _ ,
b * 74 * , ( 2006 ) 134507 .
a. del maestro , b. rosenow , and s. sachdev , phys .
b * 74 * , ( 2006 ) 024520 . note that in fig .
[ fig3](d ) we have a slightly larger @xmath86 , and u=2.15 for the results shown there . | we review a theoretical scenario for the origin of the spin - glass phase of underdoped cuprate materials . in particular it
is shown how disorder in a correlated @xmath0-wave superconductor generates a magnetic phase by inducing local droplets of antiferromagnetic order which eventually merge and form a quasi - long range ordered state . when correlations are sufficiently strong , disorder is unimportant for the generation of static magnetism but plays an additional role of pinning disordered stripe configurations .
we calculate the spin excitations in a disordered spin - density wave phase , and show how disorder and/or applied magnetic fields lead to a slowing down of the dynamical spin fluctuations in agreement with neutron scattering and muon spin rotation ( @xmath1sr ) experiments .
high - t@xmath2 superconductors , disorder - induced magnetism , field - induced magnetism , neutron response , stripes 74.25.ha , 74.72.-h , 75.40.gb |
mergers of galaxies is a driving factor in their evolution over cosmic times .
several schemes to identify merging galaxy pairs have been developed in the past decade , many based on the number of physically close pairs ( in both sky coordinates as well as redshift ) or on the characterization of the disturbed appearance of galaxies due to gravitational interaction ( often through visual inspection , e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
both these techniques have been used to determine the interaction fraction in the local universe as well as out to high redshift in hubble images . using n - body simulations ,
one can determine how long a merger will be identified as such by both techniques ; the galaxy pair is close enough , the galaxies look sufficiently disturbed . because the volume probed increases with redshift
, there were until recently paradoxically better measures of the interaction fraction for higher redshift than for the local universe .
the sdss search for close ( and disturbed looking ) pairs of galaxies @xcite added the valuable local universe interaction fraction , improving on the estimate by @xcite , from close pairs in the uppsala general catalog @xcite .
a compilation of merger fractions determined as a function of redshift is shown in figure [ f : frac ] and table [ t : cites ] .
the scatter in the merger fractions , even for those determined over the same data is striking and it is tied to the definition of what constitutes a pair or disturbed morphology ( see for instance the discussion in * ? ? ?
the definition of morphologically disturbed became quantified in several schemes of morphological parameterisation schemes ( e.g. , * ? ? ?
* ; * ? ? ?
observational uncertainties are the time scale a merger is identifiable as one , the completeness of the various samples for each technique and the volumes considered .
similarly , a current substantial theoretical effort is to map the dark matter halo merger rates onto actual observable galaxy mergers ( * ? ? ? * and reference therein ) .
.the reference , data - set , and method for merger fractions in the local and distant universe . [ cols="<,<,<",options="header " , ] [ t : mrg ]
from the number of galaxies with a companion or the number of disturbed looking galaxies in a given volume ( @xmath6 and @xmath7 respectively ) , one can calculate the volume merger rate ( @xmath8 ) , provided one has an estimate of the merger time scale ( @xmath9 ) , the merger rate from pairs ; @xmath10 or the merger rate from morphology @xmath11 . from @xcite , we have an estimate of the _ mean _ merger time scale with some variance due to differences in merger conditions ( type of feedback physics in interstellar matter , type of encounter , and gas masses of the disks ) and perspective ( face - on versus edge - on ) .
@xcite note similarly that time scales depend on mass ratio and gas fraction for optical morphological selection .
mergers were on average visible for 40% of the 2.5 gyr of the merger simulation , making our typical time scale @xmath12 gyr.(see * ? ? ?
* this paper s companion paper . ) , very similar to those used in the literature for morphological selection .
@xcite use a merger time scale for pairs of @xmath13 gyr . the volume represented by the whisp sample
was computed above ( [ s : vol ] ) as 6835 mpc@xmath14 .
following our simple merger fraction of 7% from the number of whisp galaxies with companions in the datacube , we obtain a volume merger rate of @xmath8(pairs ) = @xmath15 mergers mpc@xmath5 gyr@xmath4 .
merger rates based on morphology can use a variety of selection criteria ( table [ t : mrg ] ) and the merger rate from the concentration-@xmath0 criterion is @xmath8(morph ) = @xmath3 mergers mpc@xmath5 gyr@xmath4 . in comparison to
the fraction found from galaxy pairs ( 7% ) with the number found from morphology ( 13% ) , both seem to be in reasonable agreement with the morphological selection on the high side .
this is in line with the discrepancy found between pair selection and morphological selection as can be seen in figure [ f : frac ] .
after factoring in the relative timescales a merger is visible as a close pair ( 0.5 gyr * ? ? ?
* ; * ? ? ?
* ) or above the concentraton-@xmath0 criterion ( @xmath161 gyr ; * ? ? ?
* ) , the volume merger rates agree very well : 7 and @xmath3 mergers gyr@xmath4 mpc@xmath5 for paired and morphologically disturbed disks respectively .
the whisp sample represents only a very small volume of the universe and the resulting merger fraction and rates are uncertain as a result of that .
however , the reasonable consistency with much larger samples such as the sloan digital sky survey ( e.g. , * ? ? ?
* see figure [ f : frac ] ) are cause for optimism as morphology as a tracer of the merger fraction and rate of galaxies .
volume merger rates in the literature for the local universe vary somewhat with sample and survey .
@xcite finds for luminous red galaxies in sdss a volume merger rate of@xmath17 = @xmath18 gpc@xmath5 gyr@xmath4 = @xmath19 h@xmath14 mpc@xmath5 gyr@xmath4 .
@xcite found for galaxies of all types in the millennium galaxy catalogue ( mgc ) a volume merger rate of @xmath17 = @xmath20 h@xmath14 mpc@xmath5 gyr@xmath4 , and @xcite find a volume merger rate for all galaxy types based on sdss and mgc of @xmath17 = @xmath21 h@xmath14 mpc@xmath5 gyr@xmath4 for _ major _ mergers .
in contrast , we find a volume merger rate of @xmath8(morph ) = @xmath22 mergers h@xmath14 mpc@xmath5 gyr@xmath4 ( h = 0.73 or @xmath23=73 km / s / mpc ) , an order of magnitude more than those above . since our merger fractions are similar within a factor two to those in the literature for the local universe , the issue for the volume merger rate would either have to be the inferred whisp volume , the timescale or a bias in the selection of galaxies .
the merger time scale is unlikely to be the issue .
the visibility time of the merger starts earlier in than in the stellar perspective but it is not substantially different from what other authors have found .
substituting any other visibility time - scale from the literature for morphological or pair selection would not reduce the merger rate ( selection times are typically less or equal to @xmath2 1 gyr ) .
we are more sensitive to minor mergers and the implied shorter visibility time - scales but this is unlikely to be an order of magnitude effect .
alternatively , we may have to consider the possibility that the morphologically disturbed galaxies are not all gravitationally disturbed but may suffer from effects unique to the gas perspective , for example ram - pressure stripping affecting the appearance of the disk .
compared to the observed fraction of mergers from other sources ( e.g. , * ? ? ?
* for the sdss ) , this is of order a factor two discrepancy .
the agreement between volume merger rates from whisp from the pairs and morphology contradict this however .
the whisp volume estimate is a likely source of the discrepancy as it is the most uncertain of our numbers .
however , even with the naive , larger estimate of the whisp volume ( @xmath24 mpc@xmath14 ) , this is only a factor three , not an order of magnitude .
the whisp selection process favours late - type galaxies ( figure [ f : type ] ) and local small irregular galaxies ( to complement the spirals ) and it was never intended as a volume - limited sample . hence an intentional or unintentional selection bias may well have been introduced .
mergers identifiable by their morphology are more likely to happen to the gas - rich late - types and the irregulars are confined to a local smaller volume , and many of them will be tidally affected .
the perspective is likely to be more sensitive to unequal mass mergers as these can be identified much more readily ( the contrast in gas surface densities is not as great as it is in stellar surface brightness ) .
@xcite points out how one expects a much higher merger fraction in lower - mass systems and @xcite points out that a factor two discrepancy can easily be expected if lower mass systems are included in even a pair statistical analysis .
in addition , @xcite and @xcite identify gas - rich mergers are the most easily identified by their morphology .
therefore , we suspect our result points to a higher merger fraction and volume merger rate for spirals and irregulars in the local universe , and less to a gross error in our whisp volume , merger fraction or the merger time - scale .
if this is the case , merger fractions and rates for at least these types may not evolve with redshift as dramatically as previously thought .
however , to confirm this , one would need a volume - limited large survey with sufficient resolution and sensitivity for both morphological selection as well as accurate pair identification .
the perspective can be reliably used in the local universe(z@xmath20 ) , where a spatial resolution can be achieved in large , all - sky surveys ( e.g. , wallaby on askap , koribalski et al . _ in prep . _ and wnshis on wsrt , jzsa et al . in prep . ) but any morphological identification of interacting gas disks at higher redshift will have to wait for ska .
the resolution of the pathfinder instruments ( meerkat and askap ) may well be enough to identify close pairs in the proposed deep surveys ( dingo on askap , @xcite , meyer et al . _ in prep _ , and laduma with meerkat , @xcite , holwerda et al .
_ in prep .
in this paper , we explored the merger fraction and rate based on the observations of the whisp sample of galaxies .
the sample is still small compared to other local references based on for instance the sdss or the millennium galaxy catalogue @xcite but provides us with an indication how well the surveys of the near future will perform in this respect . from the quantified morphologies of the whisp column density maps , we conclude : * the merger fraction in the whisp sample is 7% based on pairs , and 13 % based on disturbed morphology .
these percentages are consistent if one takes into account how long a merger is visible as a close pair of galaxies and how long as a morphologically disturbed disk .
* assuming the representative volume of the whisp sample is 6835 mpc@xmath14 , and a merger visibility time scale of 1 gyr , the merger rate for our selection criterion is @xmath8(morph ) = @xmath3 mergers gyr@xmath4 mpc@xmath5 in the local universe , very close to the value of @xmath8(pairs ) = @xmath25 mergers gyr@xmath4 mpc@xmath5 for galaxy pairs in whisp .
* while the whisp merger fractions and especially rates mutually agree , the merger rates are much higher than those reported in the literature .
selection effects in the whisp survey , preferring dwarf and irregulars , rather than a gross error in the whisp volume , could well account for the difference as well as variation in the quality of the whisp maps across the sample . upcoming , volume - limited surveys should provide an accurate measurement of the local merger rate from both morphology as well as close pairs .
the 21 cm window on the universe is set to revolutionise our understanding of the merger rate of spiral and irregular galaxies as three independent measures of merging or gravitational interaction are available in the data : morphology , kinematic signatures in of interaction ( e.g. , lopsidedness of the profile , non - circular motions and an irregular velocity field ) and the easy detection of physically close companions in the data - cube . planned surveys with the ska pathfinder instruments , include an all - sky survey ( wallaby with askap ) and ( wnshis with wsrt / apertif ) , and a medium deep survey to @xmath26 ( dingo ) with askap , and an extremely deep survey ( laduma ) with meerkat . combined ,
these will revolutionise the volume probed with and help shed light on the merger fraction using all three tracers ; morphological , dynamic and pair identification in the local universe surveys , and dynamical and close pairs identification our to higher redshift .
the authors would like to thank the anonymous referee for his or her diligence , effort and inquiry , which resulted in a great improvement of the manuscript .
we acknowledge support from the national research foundation of south africa . the work of b.w .
holwerda and w.j.g .
de blok is based upon research supported by the south african research chairs initiative of the department of science and technology and the national research foundation .
a. bouchard acknowledges the financial support from the south african square kilometre array project .
we acknowledge the usage of the hyperleda database http://leda.univ-lyon1.fr . we have made use of the wsrt on the web archive .
the westerbork synthesis radio telescope is operated by the netherlands institute for radio astronomy astron , with support of nwo .
d. w. , kaviraj s. , lintott c. j. , schawinski k. , sarzi m. , bamford s. , silk j. , proctor r. , andreescu d. , murray p. , nichol r. c. , raddick m. j. , slosar a. , szalay a. s. , thomas d. , vandenberg j. , 2009 , arxiv e - prints w. j. g. , jonas j. , fanaroff b. , holwerda b. w. , bouchard a. , blyth s. , van der heyden k. , pirzkal n. , 2009 , in conference proceedings of the `` panoramic radio astronomy : wide - field 1 - 2 ghz research on galaxy evolution '' , june 02 - 05 , 2009 groningen , the netherlands l. , le fvre o. , tresse l. , bottini d. , garilli b. , le brun v. , maccagni d. , scaramella r. , scodeggio m. , vettolani g. , zanichelli a. , adami c. , arnouts s. , bardelli s. , bolzonella m. , cappi a. , charlot s. , ciliegi p. , contini t. , foucaud s. , franzetti p. , gavignaud i. , guzzo l. , ilbert o. , iovino a. , lamareille f. , mccracken h. j. , marano b. , marinoni c. , mazure a. , meneux b. , merighi r. , paltani s. , pell r. , pollo a. , pozzetti l. , radovich m. , vergani d. , zamorani g. , zucca e. , bondi m. , bongiorno a. , brinchmann j. , cucciati o. , de la torre s. , gregorini l. , memeo p. , perez - montero e. , mellier y. , merluzzi p. , temporin s. , 2009 , , 498 , 379 j. e. , van gorkom j. h. , rupen m. p. , schiminovich d. , 2001 , in astronomical society of the pacific conference series , vol . 240 , gas and galaxy evolution , hibbard j. e. , rupen m. , van gorkom j. h. , eds . , pp .
657+ b. w. , de blok w. j. g. , bouchard a. , blyth s. , van der heyden k. , pirzkal n. , 2009 , in conference proceedings of the `` panoramic radio astronomy : wide - field 1 - 2 ghz research on galaxy evolution '' , june 02 - 05 , 2009 groningen , the netherlands s. , miller s. h. , penner k. , skelton r. e. , conselice c. j. , somerville r. s. , bell e. f. , zheng x. z. , rix h. , robaina a. r. , barazza f. d. , barden m. , borch a. , beckwith s. v. w. , caldwell j. a. r. , peng c. y. , heymans c. , mcintosh d. h. , huler b. , jahnke k. , meisenheimer k. , sanchez s. f. , wisotzki l. , wolf c. , papovich c. , 2009 , , 697 , 1971 s. , bailes m. , bartel n. , baugh c. , bietenholz m. , blake c. , braun r. , brown j. , chatterjee s. , darling j. , deller a. , dodson r. , edwards p. g. , ekers r. , ellingsen s. , feain i. , gaensler b. m. , haverkorn m. , hobbs g. , hopkins a. , jackson c. , james c. , joncas g. , kaspi v. , kilborn v. , koribalski b. , kothes r. , landecker t. l. , lenc e. , lovell j. , macquart j .-
, manchester r. , matthews d. , mcclure - griffiths n. m. , norris r. , pen u .-
phillips c. , power c. , protheroe r. , sadler e. , schmidt b. , stairs i. , staveley - smith l. , stil j. , taylor r. , tingay s. , tzioumis a. , walker m. , wall j. , wolleben m. , 2007 , publications of the astronomical society of australia , 24 , 174 s. , feain i. j. , gupta n. , 2009 , in astronomical society of the pacific conference series , vol .
407 , astronomical society of the pacific conference series , d. j. saikia , d. a. green , y. gupta , & t. venturi , ed . , pp .
446+ s. , taylor r. , bailes m. , bartel n. , baugh c. , bietenholz m. , blake c. , braun r. , brown j. , chatterjee s. , darling j. , deller a. , dodson r. , edwards p. , ekers r. , ellingsen s. , feain i. , gaensler b. , haverkorn m. , hobbs g. , hopkins a. , jackson c. , james c. , joncas g. , kaspi v. , kilborn v. , koribalski b. , kothes r. , landecker t. , lenc a. , lovell j. , macquart j .- p .
, manchester r. , matthews d. , mcclure - griffiths n. , norris r. , pen u .-
l . , phillips c. , power c. , protheroe r. , sadler e. , schmidt b. , stairs i. , staveley - smith l. , stil j. , tingay s. , tzioumis a. , walker m. , wall j. , wolleben m. , 2008 , experimental astronomy , 22 , 151 j. m. , davis m. , faber s. m. , guhathakurta p. , gwyn s. , huang j. , koo d. c. , le floch e. , lin l. , newman j. , noeske k. , papovich c. , willmer c. n. a. , coil a. , conselice c. j. , cooper m. , hopkins a. m. , metevier a. , primack j. , rieke g. , weiner b. j. , 2008 , , 672 , 177 p. , 1973
, uppsala general catalogue of galaxies .
acta universitatis upsaliensis .
nova acta regiae societatis scientiarum upsaliensis - uppsala astronomiska observatoriums annaler , uppsala : astronomiska observatorium , 1973 a. r. , bell e. f. , skelton r. e. , mcintosh d. h. , somerville r. s. , zheng x. , rix h .- w
. , bacon d. , balogh m. , barazza f. d. , barden m. , boehm a. , caldwell j. a. r. , gallazzi a. , gray m. e. , haussler b. , heymans c. , jahnke k. , jogee s. , van kampen e. , lane k. , meisenheimer k. , papovich c. , peng c. y. , sanchez s. f. , skibba r. , taylor a. , wisotzki l. , wolf c. , 2009 , arxiv e - prints c. , carollo c. m. , lilly s. , sargent m. t. , feldmann r. , kampczyk p. , porciani c. , koekemoer a. , scoville n. , kneib j .-
, leauthaud a. , massey r. , rhodes j. , tasca l. , capak p. , maier c. , mccracken h. j. , mobasher b. , renzini a. , taniguchi y. , thompson d. , sheth k. , ajiki m. , aussel h. , murayama t. , sanders d. b. , sasaki s. , shioya y. , takahashi m. , 2007 , , 172 , 406 j. m. , 2002 , in astronomical society of the pacific conference series , vol .
276 , seeing through the dust : the detection of hi and the exploration of the ism in galaxies , taylor a. r. , landecker t. l. , willis a. g. , eds . ,
84+ m. a. w. , oosterloo t. a. , van cappellen w. a. , bakker l. , ivashina m. v. , van der hulst j. m. , 2008 , in american institute of physics conference series , vol . 1035 , the evolution of galaxies through the neutral hydrogen window , r. minchin & e. momjian , ed .
, pp . 265271 | the morphology of the atomic hydrogen ( ) disk of a spiral galaxy is the first component to be disturbed by a gravitational interaction such as a merger between two galaxies .
we use a simple parametrisation of the morphology of column density maps of westerbork spiral project ( whisp ) to select those galaxies that are likely undergoing a significant interaction .
merging galaxies occupy a particular part of parameter space defined by asymmetry ( a ) , the relative contribution of the 20% brightest pixels to the second order moment of the column density map ( @xmath0 ) and the distribution of the second order moment over all the pixels ( @xmath1 ) . based on their morphology
, we find that 13 % of the whisp galaxies are in an interaction ( concentration-@xmath0 ) and only 7% based on close companions in the data - cube .
this apparent discrepancy can be attributed to the difference in visibility time scales : mergers are identifiable as close pairs for 0.5 gyr but @xmath2 1 gyr by their disturbed morphology .
expressed as volume merger rates , the two estimates agree very well : 7 and @xmath3 mergers gyr@xmath4 mpc@xmath5 for paired and morphologically disturbed disks respectively .
the consistency of our merger fractions to those published for bigger surveys such as the sloan digital sky survey , shows that morphology can be a very viable way to identify mergers in large survey . the relatively high value for the volume merger rate may be a bias in the selection or whisp volume .
the expected boon in high - resolution data by the planned meerkat , askap and wsrt / apertif radio observatories will reveal the importance of mergers in the local universe and , with the advent of ska , over cosmic times .
[ firstpage ] |
the authors are very grateful to v.n .
pomerantsev for many fruitful discussions on the topics discussed here .
one of the authors , ( v.i.k . ) is grateful to willi gruebler for supplying him the full tables of experimental data of the zrich group .
we are also most grateful to the uk epsrc for grant gr / h00895 supporting s.g .
cooper , the russian foundation for basic research ( grant 97 - 02 - 17265 ) for financial assistance and to the royal society ( uk ) for supporting a visit by v.i . kukulin to england .
kukulin , v.n .
pomerantsev and s.b .
zuev , yad .
* 59 * , 428 ( 1996 ) ; english translation : physics of atomic nuclei , * 59 * , 403 ( 1996 ) ; + v.i.kukulin and v.n.pomerantsev , yad
. fiz . * 51 * , 376(1990 ) ; * 60 * , 1228(1997 ) ; english translation : physics of atomic nuclei , * 51 * , 240 ( 1990 ) ; * 60 * , 1103 ( 1997 )
. r.s . mackintosh and a.m. kobos ,
phys . lett . * 116b * , 95 ( 1982 ) ; a.a .
ioannides and r.s .
mackintosh , nucl .
a438 * , 354 ( 1985 ) .
kukulin , v.n .
pomerantsev and j. horaek , phys . rev . * a 42 * , 2719 ( 1990 ) s.g .
cooper and r.s .
mackintosh , phys .
* c 43 * , 1001 ( 1991 ) ; nucl .
a517 * , 285 ( 1990 ) ; nucl . phys . *
a576 * , 308 ( 1994 ) ; nucl . phys . * a582 * , 283 ( 1995 ) ; phys . rev .
* c 54 * , 3133 ( 1996 ) s.g .
cooper , nucl .
phys . * a 618 * , 82 ( 1997 ) e.v .
kuznetsova and v.i .
kukulin , yad . fiz . * 60 * , 608 ( 1997 ) ; english translation : physics of atomic nuclei , * 60 * , 528 ( 1997 ) v.m .
krasnopolsky , v.i .
kukulin , e.v .
kuznetsova , j. horaek and n.m .
queen phys .
* c43 * , 822 ( 1991 ) b. jenny , w. gruebler , v. knig , p.a .
schmelzbach , and c. schwizer , nucl . phys . *
a397 * , 61 ( 1983 ) g. breit , m.h .
hull , k.e .lassila , and k.d .
pyatt , phys.rev .
* 120*,2227 ( 1960 ) .
green , k.w .
kemper , p.l .
kerr , k. mohajeri , e.g. myers , d. robson , k. rusek , and i.j .
thompson , phys . rev . *
c53 * , 2862 ( 1996 ) see e.g. , g. r. satchler , direct nuclear reactions , ( clarendon press , oxford , 1983 ) .
w. gruebler , p.a .
schmelzbach , v. knig , and p. marmier , nucl . phys . * a 134 * , 686 ( 1969 ) l.s . senhouse and t.a .
tombrello , nucl . phys . * 57 * , 624 ( 1964 ) r. frick , h. clement , g. graw , p. schiemenz and n. seichert , phys .
44 * , 14 ( 1980 ) ; z. phys
. * a314 * , 49 ( 1983 ) ; * a319 * , 133 ( 1984 ) y. wang , c.c .
foster , e.j .
stephenson , li yuan and j. rapaport , phys .
* c45 * , 2891 ( 1992 ) r.s . mackintosh and s.g .
cooper , nucl .
a625 * 651 h. kanada , t. kaneko , s.saito and y.c .
tang , nucl .
* a444 * ( 1985 ) 209 r.s . mackintosh and s.g .
cooper , _ `` using inverse scattering methods to study inter - nucleus potentials '' _ , invited contribution at the international workshop on physics with radioactive nuclear beams , puri , india , january 12 - 17 , ( 1998 )
blokhintsev , v.i .
kukulin , a.a .
sakharuk , a.a .
savin , and e.v .
kuznetsova , phys . rev . *
c48 * , 2390 ( 1993 ) v.i .
kukulin , v.n .
pomerantsev , s.g .
cooper , and s.b .
dubovichenko , phys.rev.c .
( 1998 ) , to be published .
cooper and r.s .
mackintosh , phys .
* c 54 * , 3133 ( 1996 ) v.i .
kukulin and v.n .
pomerantsev , progress in theoretical physics , * 88 * , 159 ( 1992 ) .
kukulin , v.n.pomerantsev , s.b .
dubovichenko , s.g .
cooper and r.s .
mackintosh , work in progress . | we demonstrate a new approach to the analysis of extensive multi - energy data . for the case of d + @xmath0he
, we produce a phase shift analysis covering for the energy range 3 to 11 mev .
the key idea is the use of a new technique for data - to - potential inversion which yields potentials which reproduce the data simultaneously over a range of energies .
it thus effectively regularizes the extraction of phase shifts from diverse , incomplete and possibly somewhat contradictory data sets . in doing so
, it will provide guidance to experimentalists as to what further measurements should be made .
this study is limited to vector spin observables and spin - orbit interactions .
we discuss alternative ways in which the theory can be implemented and which provide insight into the ambiguity problems .
we compare the extrapolation of these solutions to other energies .
majorana terms are presented for each potential component .
a well known problem confronting any phase shift analysis ( psa ) , both for a single energy and for multiple energies , is the absence of complete sets of experimental data .
a complementary problem is the occurrence of apparent inconsistencies between data from different experiments .
these problems are particularly acute for projectiles of spin @xmath1 . for example , spin one projectiles require , at each energy , eight or nine independent measurements ( cross sections @xmath2 , vector @xmath3 and tensor @xmath4 analysing powers , etc . ) since the psa solutions based on incomplete data will be far from unique , we must find a way to apply constraints .
apart from certain smoothness requirements , it is highly nontrivial to find _ general _ restrictions which are convenient to apply within the framework of existing psa methods .
in this paper we present a new approach to phase shift analysis , psa .
in essence , the idea is to find a single multi - component potential to describe the experimental data over the energy range in question .
this is made possible by a recently developed @xcite direct data - to - potential inversion technique , the generalized iterative perturbative method , hereafter gip , which we describe below .
gip is a generalization of the established ip @xmath5-matrix - to - potential inversion method @xcite .
it enables data for many energies to be fitted with great computational efficiency by a single energy dependent potential that is as flexible as required .
a psa based on a potential , unlike conventional psas , leads in a natural way to sets of phase shifts bearing physically reasonable relationships between the different partial waves .
moreover , the model itself will now reveal any inconsistent data and indicate where new data are required and thereby be a useful source of experimental guidance .
the gip potential will have a small energy dependence which , unlike the rapid energy dependence of the phase shifts , can be compared with the energy dependence predicted by theory . we demonstrate our approach by applying the method to d + @xmath0he and will show that an integrated picture of d + @xmath0he scattering can be obtained from a diverse range of data covering a substantial energy range .
the need for such an approach may be seen from ref .
@xcite where a huge amount of experimental data for this reaction , including cross section and vector and tensor analysing powers , are analysed at length by elaborate forms of psa .
the methods used are described by krasnopolsky _
et al _ @xcite .
although the authors of @xcite derived much important information from their psa ( e.g. exact resonance widths , vertex constant etc . ) many results are still on a preliminary and qualitative level ( e.g. the complex tensor mixing parameters , odd - parity phase shifts ) . therefore we believe that the very considerable experimental effort devoted to this system , see @xcite and many other papers cited in reference @xcite , motivates a new approach .
since d + @xmath0he is a perfect theoretical test case , the rewards will be physical insight of general relevance to nuclear physics . in principle
the potential searched for should include _ all _ necessary components of nuclear interactions , including central ( wigner ( w ) and majorana ( m ) ) terms , spin - orbit terms ( again , both w and m ) and the various possible tensor terms , once more both w and m ; all terms may be complex as required . in determining a suitable potential , one can impose constraints such as : conformity to known behaviour of higher partial waves ( as in , e.g. @xcite ) ; smooth energy dependence of underlying potentials ; consistency with established theories ; reproduction of bound- and resonant - state energies . in principle
it is possible to find a potential fitting data at many energies by applying standard searching procedures to the parameters of a sufficiently flexible potential model , whether of standard multi - parameter or model independent form ( e.g. the so - called fourier - bessel analysis ) . to do
this generally entails computationally expensive and highly non - linear multi - parameter fitting , often leading to many local minima@xcite .
the gip procedure for direct data - to - potential inversion solves many of these problems .
the advantages of ip over other methods for @xmath5-matrix - to - potential inversion apply here too and are particularly relevant .
the first advantage is the power to control the exactness of the inversion so that noisy , incomplete or even partly erroneous data can be fitted with ( one - channel or multi - channel ) potentials which do not have spurious oscillatory features .
the second advantage is its virtually unlimited generalizability .
here we illustrate this feature by including in our analysis the four majorana components , normally omitted in optical model fits .
a further feature of psa using the gip method is its speed and simplicity of application enabling a thorough exploration of ambiguities .
these ambiguities are _ not _ a matter of shallow valley floors in parameter hyper - space , but appear in the form of apparently disconnected minima . in the present case
we apply the procedure to @xmath6 projectiles , although for clarity we suppress spin - related subscripts .
the method involves the following three key elements : + * ( i * ) expansion of components of the potential ( central ( c ) , spin - orbit ( s - o ) , tensor ( t ) , etc . ) in a suitable basis . for potential component @xmath7c , s - o , t
@xmath8 where @xmath9 are coefficients to be determined , @xmath10 are the basis functions and @xmath11 is the starting potential .
note that this expansion applies to both real and imaginary components and that the notation @xmath10 embodies the possibility that it might be appropriate for different components of the potential to be expanded in different bases .
in particular , real and imaginary terms , or central and spin - orbit terms , or the majorana terms might well require different bases .
+ [ 2 mm ] * ( ii ) * the linear response of the complex @xmath5-matrix @xmath12 to small changes @xmath13 in the potential:@xmath14 with @xmath12 defined in terms of the asymptotic form of the regular radial wave function as @xmath15 where @xmath16 and @xmath17 are incoming and outgoing coulomb wave functions of ref .
@xcite .
the formulation @xcite in terms of @xmath18 , where @xmath19 , is exactly equivalent .
note that the energy @xmath20 is implicit in these equations and , for simplicity , we have labelled the channels only by the orbital angular momentum @xmath21 although we do include spin in our calculations .
equation [ second ] can be recast as @xcite : @xmath22 where any required superscript @xmath23 , labelling the potential component , is implicit .
+ * ( iii ) * the @xmath24 function is defined from : @xmath25 where @xmath26 and @xmath27 are the input experimental values of cross sections and analyzing powers of type @xmath28 respectively .
since we are fitting data for many energies at once , the index @xmath29 indicates the energy as well as angle .
the data normalising factors can be introduced as an additional contribution to equation [ fourth ] .
we must now expand @xmath24 in terms of the @xmath9 . to do this
we first linearize the theoretical cross sections and analyzing powers , by expanding @xmath30 ( and @xmath31 ) about some current point @xmath32 : @xmath33 which applies at each iterative step @xmath34 and the correction ( to be determined ) for the @xmath35-th amplitude is @xmath36 .
equivalent relations are applied for the @xmath31 , linear equations result from demanding that @xmath24 be locally stationary with respect to variations in the potential coefficients @xmath37 , i.e. the derivatives of @xmath24 with respect to the potential components @xmath37 must vanish . solving these linear equations
is straightforward for any reasonable number of them and yields corrected values @xmath38 [ 8,10 ] .
we then iterate the whole procedure , with wave - functions @xmath39 in equation [ third ] calculated using the corrected potentials from equation [ first ] , until convergence is reached .
this algorithm almost always converges very rapidly [ 8,10 ] , in general diverging only when highly inconsistent or erroneous data have been used or when the iterative process involves a very unsuitable starting point .
+ multi - energy inversion is thus reduced to the solution of simultaneous equations at a series of iterative steps . to show how effective this is , we present the results of a multi - energy psa for the d + @xmath0he system . for this initial study
, we have selected a small subset of the experimental data tabulated in @xcite , in particular the data of jenny _ et al _
@xcite and that of @xcite . at this stage , we have fitted only the cross sections and vector analysing powers and correspondingly limited ourselves to the following potential components : wigner central ; majorana central ; wigner spin - orbit ; majorana spin - orbit .
all terms are complex so that there are eight components to be determined .
the neglect of the various complex tensor components is justified because their primary effect is on the tensor analysing powers .
it is well known @xcite that tensor interactions in the d + @xmath0he system play a moderate role , mainly influencing the @xmath40s@xmath41 @xmath40d@xmath41 and @xmath40p@xmath42 @xmath40f@xmath42 mixing parameters which are not significant here .
the generalisation of gip to yield tensor interactions is under development and we expect a full psa , including all off - diagonal terms , to be presented in due course .
data renormalization was not considered here since its effect is small compared to neglect of the tensor force , particularly for the data sets fitted here @xcite . in order to get some understanding of the ambiguity problems , we consider here two extreme approaches to the fitting process which we label a and b. the question of the meaningfulness of the potentials that are found we leave to later publications . _
approach a _ begins the iterative procedure with a starting potential reflecting very little _ a priori _ information concerning the potential and consists of two components only : simple real and imaginary central wigner terms of gaussian form .
the data is fitted in stages , adding a further potential component at each step with basis dimensions restricted to two or three gaussian functions .
generally convergence results from two or three inversion iterations at each stage . by applying a criterion of visual smoothness ,
an optimum solution was found , ` potential a ' , corresponding to @xmath43 . fits giving a lower @xmath44 are possible with a larger basis , but the corresponding @xmath45 also show a significant unitarity breaking for certain @xmath46 .
this case involves about 20 independent parameters .
_ approach b _ starts the iterative procedure with a potential derived by inversion @xcite of @xmath47 from the multi - configuration rgm calculations of kanada _
et al _
@xcite which include s - wave deuteron breakup .
this approach gave ` potential b ' with @xmath48 but is accompanied by a significant breaking of unitarity in the s wave .
( the results are described in detail in ref .
@xcite . ) in both approaches energy dependence is included only in the imaginary components .
the procedure used follows ref .
@xcite , which applies for shape invariant energy dependent potentials .
since the inelastic threshold is at @xmath49 mev , we expect the imaginary components to increase rapidly as the energy rises above @xmath50 and so we assume that all parts of the imaginary potential increase linearly with @xmath51 .
in fact , the results are insensitive to this energy dependence
. both the detailed form of the imaginary potentials and the imaginary phase shifts are less well determined than the corresponding real quantities and qualitative features of the data can be reproduced with a real potential alone . in figure 1
we display , for representative energies over the complete energy range of 3 11.5 mev , typical fits to cross sections @xcite and in figure 2 , analyzing powers @xcite . both @xmath2 and @xmath3
are very well fitted over the entire energy range .
closely compatible fits to the data of ref .
@xcite were found , both visually and in the values of @xmath24 .
all the quoted @xmath44 values apply to the fit over the full energy range , but are only relative since the tabulated data did not include all the sources of error discussed in the original papers .
we have found that although the contribution of the mixing parameters to the cross - section is almost negligible , there is a more noticeable effect on the fit to @xmath3 .
the bound state energy of the @xmath0he d system , which can be identified as the ground state energy of @xmath52li in the @xmath0he d channel , is not included in these inversions .
potential a gives @xmath53 mev ( @xmath54 mev ) .
note that this energy is extremely sensitive to the form of the potentials and to the energy dependence of the d
@xmath0he @xmath40s@xmath41 phase shifts @xcite . in figure 3 we present the real parts of potentials a. known ambiguity problems suggest this potential is almost certainly not unique . within either approach , a or b
, certain potential components are more reliably determined than others , the real central wigner term being the best determined .
its volume integral is consistent with global potentials and also with volume integrals of the corresponding potential derived by @xmath5-matrix to potential inversion for the theoretical @xmath12 of resonating group model ( rgm ) calculations @xcite .
the phase shifts corresponding to the solution a for @xmath55 are displayed in figure 4 for an energy range of 0 to 15 mev laboratory energy , i.e. extrapolating outside the range of the data .
this figure also includes the results of a previous analysis @xcite .
the really difficult problem for all previous ( standard ) psas was to achieve a low energy description of odd partial waves ( i.e. @xmath40p@xmath56 with @xmath57 and @xmath40f@xmath56 with @xmath58 ) , due to the weak sensitivity of cross sections and analysing powers to the odd partial waves @xcite .
thus , by fitting all significant partial waves independently in the course of a standard psa @xcite , a range of solutions are possible which are consistent with the data .
the resulting odd - parity phase shifts have very large error bars . in the present method for phase shift analysis a further restriction
is applied by demanding a smooth underlying potential and therefore the approach should lead , in principle , to much more reliable and accurate values for all phase shifts than found in previous psas @xcite . the comparison in fig . 4 of our new psa solution with previous results
, shows that the agreement for even partial waves is quite close while there is less agreement for odd partial waves .
this is probably due our neglect of tensor forces .
reliable knowledge of the odd partial wave phase shifts is crucially important @xcite , since the nature of the deuteron
nucleus interaction , particularly for d + @xmath0he , is different for even and odd partial waves .
the even parity d + @xmath0he interaction is determined by an intermediate state in which two nucleons in the incident deuteron occupy two ( 1p)-orbitals beyond the @xmath0he core . however , for odd parity , the two outer nucleons occupy non - overlapping 1p2s or 1p2d orbits ( designating orbits @xmath59 , with @xmath60 the number of oscillator quanta ) . thus , since the n
n interaction is short ranged compared to the range of d
@xmath0he interaction , the contribution of virtual breakup should be higher for odd than for even partial waves and the sensitivity to the n + @xmath61 interaction should also be higher . due to this feature of the d + @xmath0he interaction , the p- and f - wave phase shifts
have been shown @xcite to give a strong test of supersymmetrical aspects of composite particle interactions and the structure of tensor interactions of deuterons .
a further step now is to include in our potential terms which have never previously been considered for nucleus - nucleus interactions : complex majorana tensor forces .
preliminary results @xcite show that the majorana tensor force is approximately as strong as the wigner tensor force . in summary
: we have demonstrated a new approach to psa based on a linearized iterative approach to direct inversion from multi - energy data to potentials .
the example presented , approach a , involved far fewer parameters than a conventional psa ( about a hundred for this case ) .
the new method is computationally efficient and avoids many drawbacks and instabilities of conventional psas , especially in cases of projectile of spin 1 or greater when one generally has an incomplete data set with data at many relevant energies absent or having large error bars . as well as correct phase shifts ,
the potential itself is of great interest since it can be used as input for other calculations and can also be compared with potentials found by double folding procedures or by inversion from @xmath12 obtained from rgm and other theoretical models . |
a typical star spends 90@xmath4 of its lifetime burning hydrogen on the main sequence . when hydrogen is exhausted in the center , the remaining helium core contracts transforming gravitational potential energy into thermal energy . as the temperature increases helium burning starts .
the first step of this process is the triple-@xmath3 capture to form @xmath2c , followed by the @xmath2c radiative capture of @xmath3 particles to form @xmath1o @xcite . in the helium - burning phase of stellar evolution the triple-@xmath3 process and
@xmath2c(@xmath3,@xmath0)@xmath1o are the most important reactions .
their relative reaction rates determine the @xmath2c@xmath7@xmath1o ratio at the end of helium burning and beyond .
the triple-@xmath3 reaction rate is well known for stellar evolution calculations .
however the rate of @xmath2c(@xmath3,@xmath0)@xmath1o is poorly determined as the cross section was measured experimentally down to energies around 1 mev and needs to be extrapolated to helium - burning energies around 300 kev .
the @xmath2c(@xmath3,@xmath0)@xmath1o cross section is composed of 3 major parts : e1 , e2 ground state transitions and cascade transitions .
ground state transitions provide most of the total cross section while cascade transitions have contributions of the order of 10 - 20@xmath4 . determining the @xmath8-mev branching ratio will result in a better extrapolation of the cascade and @xmath6 ground state cross sections to lower energies @xcite .
the 7.12-mev excited state in @xmath1o is formed via the @xmath9f(p,@xmath3)@xmath1o@xmath10 reaction ; a level diagram is shown in fig . [
fig : levels ] .
a 100-@xmath11g/@xmath12-thick caf@xmath13 layer evaporated onto a 1-mm - thick carbon backing was bombarded by protons produced at the 4.5-mv tandem van de graaf accelerator at ohio university .
an angular distribution study for the 7.12-mev @xmath0-ray at 4 different proton energies was performed .
these measurements offered a better understanding of the @xmath0-ray angular distribution for the strong transitions and information on the dependence of the relative intensity of the three @xmath0-ray components with proton energy . in fig .
[ fig : peakratio ] the relative intensities of the strong transitions in the direction of the beam are shown .
for the branching - ratio measurements the proton beam energy was chosen to be e@xmath14=2.0025 mev in order to maximize the relative population of the 7.12-mev state @xcite .
the coincidence setup uses a two - detector configuration . for the detection of the @xmath15-mev transition
, a 22.9-cm @xmath16 22.9-cm nai(tl ) split annulus consisting of two optically - isolated sections is used .
the large volume of the crystal offers a 30@xmath4 efficiency in detecting the 6 - 7 mev @xmath0-rays .
the inside cylindrical hole in the crystal is able to accommodate an extended target holder as sketched in fig .
[ fig : upview ] . for the @xmath5-mev transition a hpge detector is used .
this detector has a planar geometry to lower the noise and achieve better time resolution for low - energy gamma rays .
the efficiency of different detector geometries and thicknesses was simulated using the geant3 simulation package .
the simulation looked for the optimum ge thickness to maximize the photopeak efficiency at 200 kev while minimizing the sensitivity to 7-mev @xmath0-rays .
a planar crystal of 1.5-cm thickness and 2000-mm@xmath17 area was chosen . to further lower the background in the 0.2-mev region of interest a tungsten shield
was utilized to minimize the scattering of @xmath0-rays between detectors .
the design of the shield was also guided by the geant3 simulation package .
data on @xmath0-@xmath0 coincidences was taken for 4 days , in 24 hour shifts , with beam currents ranging from 10 to 20 na . a standard coincidence setup fed the signal from the detectors and the time - to - amplitude converter ( tac ) into a 4096 channels data acquisition system .
the time resolution of the system ( for energy signals of interest ) was 5 ns .
the spectrum shown in fig .
[ fig : spectrum ] was obtained by the hpge detector using gate settings on the tac coincidence peak and the 6 - 7 mev region in the nai(tl ) detector .
-mev transition is indicated .
right figure : expanded 200-kev region with the 197-kev peak from the @xmath9f(p , p@xmath18 ) reaction and the position of a possible @xmath5-mev transition.,title="fig : " ] -mev transition is indicated .
right figure : expanded 200-kev region with the 197-kev peak from the @xmath9f(p , p@xmath18 ) reaction and the position of a possible @xmath5-mev transition.,title="fig : " ] for the @xmath19-mev transition the peak is clearly visible between channels 3000 and 3250 in fig .
[ fig : spectrum ] .
the region surrounding the peak was fitted to subtract the background and obtain the number of counts in the @xmath19-mev transition peak .
calibrated sources and geant3 simulations were used to find the efficiencies of the detectors .
the result obtained for this branching ratio is @xmath20 , in good agreement with a previous measurement by wilkinson et al .
@xcite .
the coincident hpge spectrum near 200 kev is dominated by the 197.1-kev line which results from random - coincidence @xmath9f(p , p@xmath18 ) events detected in the hpge .
no peak is evident at 199.8 kev corresponding to the @xmath5-mev transition .
it should be noted that the hpge detector resolution is 1.3 kev at 200 kev , so energy range of interest is well resolved from the 197.1-kev line . from the spectrum shown in fig .
[ fig : spectrum ] we have obtained the following upper limit with a 2-@xmath21 confidence level : @xmath22 note that if the @xmath5-mev transition were to have the same weisskopf strength as the @xmath23-mev transition this branching ratio would be @xmath24 .
the results are preliminary and further analysis of the data may reveal more information .
histogram manipulation and improved geant3 simulations will be performed to further understand the limitations of the present setup and perhaps discover improvements to the experimental approach .
it is a pleasure to thank j. bevington , d. carter , c. dodson , d. jacobs , t. massey , j. odonnell , y. parpottas , a. salas , and r. wheeler for assistance with the experiments .
this work was supported in part by the u.s .
department of energy , under grant no .
de - fg02 - 88er40387 . | knowledge of the @xmath0-ray branching ratios of the 7.12-mev state of @xmath1o is important for the extrapolation of the @xmath2c(@xmath3,@xmath0)@xmath1o cross section to astrophysical energies .
ground state transitions provide most of the @xmath2c(@xmath3,@xmath0)@xmath1o total cross section while cascade transitions have contributions of the order of 10 - 20@xmath4 . determining the @xmath5-mev branching ratio will also result in a better extrapolation of the cascade and @xmath6 ground state cross section to low energies .
we report here on measurements on the branching ratio of the @xmath5-mev transition in @xmath1o . |
gabaix et al .
@xcite present statistical evidence that appears to show that the nyse and paris data are consistent with the hypothesis that the average market impact follows a square root law . in this section
we show that their test may have problems in circumstances ( such as those of the real data ) in which orders have long - memory properties .
this weakens their test , so that it lacks the power to reject reasonable alternative hypotheses and may give misleading results .
their method to test the hypothesis of square root price impact is to investigate @xmath10 $ ] over a given time interval , e.g. @xmath11 minutes , where @xmath6 is the price shift and @xmath12 is the sum of the volumes of the @xmath13 transactions occurring in that time interval . they have chosen to analyze @xmath14 rather than @xmath6 because of its properties under time aggregation . to see why this might be useful , assume the return due to each transaction
@xmath15 is of the form @xmath16 , where @xmath17 is an iid noise process that is uncorrelated with @xmath18 , and @xmath19 is the sign of the transaction .
the squared return for the interval is then of the form @xmath20 under the assumption that @xmath18 , @xmath21 , @xmath19 , and @xmath22 are all uncorrelated , when @xmath8 it is easy to show that @xmath10=a + b~v$ ] , where @xmath23 and @xmath24 are constants .
the problem is that for the real data @xmath18 , @xmath21 , @xmath19 , and @xmath22 are strongly correlated , and indeed , the sequence of signs @xmath19 is a long - memory process @xcite . to demonstrate the gravity of this problem
, we use real transactions @xmath18 , but introduce an artificial and deterministic market impact function of the form @xmath25 with @xmath26 .
we first fix the number of transactions , and then repeat the same procedure using a fixed time period .
we examine blocks of trades with @xmath13 transactions , @xmath27 , @xmath28 , where @xmath29 ( @xmath30 ) for buyer ( seller ) initiated trades and @xmath18 is the volume of the trade in number of shares . for each trade
we create an artificial price return @xmath31 , where @xmath7 is a constant .
then for each block of @xmath13 trades we compute @xmath32 and @xmath33 .
since we are using the real order flow we are incorporating the correct autocorrelation of the signs @xmath19 and transaction sizes @xmath18 .
figure [ exprsq](a ) shows @xmath10 $ ] for different values of @xmath13 and @xmath34 for the british stock vodafone in the period from may 2000 to december 2002 , a series which contains approximately @xmath35 trades .
$ ] , is plotted as a function of total transaction size @xmath36 , where @xmath18 is the size of transaction @xmath15 .
each transaction causes a simulated market impact of the form @xmath37 , to generate total return @xmath38 .
the transaction series @xmath18 and @xmath19 are from the real data from the electronic market for the british stock vodafone , and contain roughly @xmath35 events .
the error bars are the @xmath39 confidence intervals computed following the procedure specified by gabaix et al .
( a ) shows the results for a fixed number of transactions , with @xmath13 varying from @xmath40 to @xmath41 ; the curves are in ascending order of @xmath13 ; ( b ) is the same using a fixed time interval of @xmath11 minutes , with variable @xmath13 ; and ( c ) is the same as ( a ) with the order of the transactions randomly shuffled . for ( a ) and ( b )
we see straight lines for large @xmath13 , indicating that the test is passed , even though by construction the market impact does not follow the @xmath42 hypothesis , whereas for the shuffled data the test quite clearly shows us that the hypothesis is false.,title="fig : " ] $ ] , is plotted as a function of total transaction size @xmath36 , where @xmath18 is the size of transaction @xmath15 .
each transaction causes a simulated market impact of the form @xmath37 , to generate total return @xmath38 .
the transaction series @xmath18 and @xmath19 are from the real data from the electronic market for the british stock vodafone , and contain roughly @xmath35 events .
the error bars are the @xmath39 confidence intervals computed following the procedure specified by gabaix et al .
( a ) shows the results for a fixed number of transactions , with @xmath13 varying from @xmath40 to @xmath41 ; the curves are in ascending order of @xmath13 ; ( b ) is the same using a fixed time interval of @xmath11 minutes , with variable @xmath13 ; and ( c ) is the same as ( a ) with the order of the transactions randomly shuffled . for ( a ) and ( b )
we see straight lines for large @xmath13 , indicating that the test is passed , even though by construction the market impact does not follow the @xmath42 hypothesis , whereas for the shuffled data the test quite clearly shows us that the hypothesis is false.,title="fig : " ] $ ] , is plotted as a function of total transaction size @xmath36 , where @xmath18 is the size of transaction @xmath15 .
each transaction causes a simulated market impact of the form @xmath37 , to generate total return @xmath38 .
the transaction series @xmath18 and @xmath19 are from the real data from the electronic market for the british stock vodafone , and contain roughly @xmath35 events .
the error bars are the @xmath39 confidence intervals computed following the procedure specified by gabaix et al .
( a ) shows the results for a fixed number of transactions , with @xmath13 varying from @xmath40 to @xmath41 ; the curves are in ascending order of @xmath13 ; ( b ) is the same using a fixed time interval of @xmath11 minutes , with variable @xmath13 ; and ( c ) is the same as ( a ) with the order of the transactions randomly shuffled . for ( a ) and ( b )
we see straight lines for large @xmath13 , indicating that the test is passed , even though by construction the market impact does not follow the @xmath42 hypothesis , whereas for the shuffled data the test quite clearly shows us that the hypothesis is false.,title="fig : " ] we see that for small values of @xmath13 the quantity @xmath10 $ ] follows the artificial market impact functional form @xmath10\sim v^{2\beta}= v^{0.6}$ ] , but when @xmath13 is large the relation between @xmath10 $ ] and @xmath3 becomes linear .
the value @xmath43 is roughly the average number of trades in a @xmath11 minute interval .
we also show error bars computed as specified by gabaix et al .
we can not reject the null hypothesis of a linear relation between @xmath10 $ ] and @xmath3 with @xmath39 confidence , even though we have a large amount of data , and we know by construction that @xmath44 is quite different from @xmath45 .
we have also performed tests on other stocks , which give similar results .
one can ask whether it makes a difference that we used a fixed number of transactions rather than a fixed time interval . to test this we repeat the procedure using a fixed time interval of @xmath11 minutes .
figure [ exprsq](b ) shows the result .
we see an even clearer linear relation between @xmath10 $ ] and @xmath3 than before , so that the test once again fails . why does nt this test work ? to gain some understanding of this , we repeat the same test but shuffle the order of the data , which breaks the correlation structure . as shown in figure [ exprsq](c ) , the result in this case is far from linear even when @xmath43 , and the test easily shows that the market impact does not follow a square root law .
thus , we see that the problem lies in the autocorrelation structure of the real data . in conclusion
our numerical simulations show that the linearity test of @xmath10 $ ] lacks power to test for a square root market impact with data containing the correlation structure of real data .
in fact , even a deterministic market impact like @xmath46 is consistent with the relation @xmath10=a+b~v$ ] for a sufficiently large number of trades . doing this for a fixed time interval rather than a fixed number of trades time
makes this even more evident .
thus the test of gabaix et al .
provides no evidence that the average market impact follows a square root law .
while there have been many previous studies of average market impact , they have not included the statistical analysis needed to assign good error bars . in this section
we present results about average market impact at the level of individual ticks .
we show that it does not generally follow a square root law , and that it varies from market to market and in some cases from stock to stock in a substantial and statistically significant way .
realistic error bars for the average market impact are difficult to assess due to the fact that volatility is a long - memory process @xcite .
that is , its time series has a slowly decaying power law autocorrelation function that is asymptotically of the form @xmath47 , with @xmath48 so that the integral is unbounded .
this makes error analysis complicated , since data from the distant past have a strong effect on data in the present . because volatility is long - memory , the price returns that fall in a given volume bin @xmath49 , which are by definition all of the same sign , are also long - memory .
this means that the errors in measuring market impact are much larger than one would expect from intuition based on an iid hypothesis .
we analyze the market impact only for orders ( or portions of orders ) that result in immediate transactions .
we call the portion of an order that results in an immediate transaction an effective market order , and for the remainder of the paper @xmath18 represents effective market order size rather than transaction size .
each order of size @xmath18 generates a price return @xmath50 , where @xmath51 is the midpoint price quote just before the order is placed and @xmath52 is the midpoint price quote just after .
we analyze buy and sell orders separately .
the electronic ( sets ) data for the lse has the advantage that the data set contains a record of orders , and so we can distinguish buy and sell orders unambiguously , but has the disadvantage that it omits trades made in the upstairs market . for the nyse data we use the trades and quotes ( taq ) data to infer orders and their signs using the lee and ready algorithm @xcite ; to identify orders we lump together all trades with the same timestamp and order code . to estimate the average market impact we sort the events @xmath53 , @xmath54 ) with the same sign @xmath19 into bins based on @xmath18 and plot the average value of @xmath18 for each bin against the average value of @xmath54 , as shown in figure [ mktimpactfig ] .
we choose the bins so that each bin has roughly the same number of points in it .
are binned together ; on the horizontal axis we show the average volume of the orders in each bin , and on the vertical axis the average size of the logarithmic price change for the orders in that bin . in both cases comparison to the dashed black line in the corner , which has slope @xmath45 ,
makes it clear that the behavior for large volume does not follow a law of the form @xmath55 .
error bars are computed using the variance plot method @xcite as described in the text . ] to assign error bars for each bin we use the variance plot method @xcite . for each bin we split the events into @xmath56 subsamples with @xmath57 points , where @xmath58 is the number of records in the bin .
the subsamples are chosen to be blocks of values adjacent in time .
for each subsample @xmath15 we compute the mean @xmath59 , @xmath60 .
then we compute the standard deviation of the @xmath59 which we indicate as @xmath61 . by plotting @xmath61 versus @xmath62 in a log - log plot
we compute the hurst exponent @xmath63 by fitting the data with a power - law function @xmath64 .
we compute the error in the mean of the entire sample of @xmath58 points by extrapolating the fitted function to the value @xmath65 , i.e. @xmath66 where @xmath67 and @xmath68 are the ordinary least square estimate of the parameters @xmath69 and @xmath63 .
interestingly , for smaller values of @xmath18 we find hurst exponents substantially larger than @xmath45 , whereas for large values of @xmath18 the hurst exponents are much closer to @xmath45 .
when @xmath70 the error bars are typically much larger than standard errors can not be due to a difference in the mean interval between samples . ] . in figure [ mktimpactfig ]
we show empirical measurements of the average market impact for the new york stock exchange and for the london stock exchange .
we consider three highly capitalized stocks for each exchange , lloyds ( lloy ) , shell ( shel ) and vodafone ( vod ) for the lse , and general electric ( ge ) , procter & gamble ( pg ) and at&t ( t ) for the nyse . for lse stocks we consider the period may 2000- december 2002 , while for nyse stocks we consider the time period 1995 - 1996 .
the data for the nyse are consistent with results reported earlier without error bars @xcite , while the lse market impact data is new .
the nyse data clearly do not follow a power law across the whole range , consistent with earlier results in references @xcite . while @xmath71 for small @xmath18 , for larger @xmath18 it appears that @xmath72 . as shown in reference @xcite ,
this transition occurs for smaller values of @xmath18 for stocks with lower capitalization .
thus , the assumption that @xmath73 breaks down for high volumes , precisely where it is necessary in order for the theory of gabaix et al . to hold . for the london data the power law assumption seems more justified across the whole range , but
the exponent is too low ; a least squares fit gives @xmath74 .
while we have not attempted to compute error bars for the regression , a visual comparison with the error bars of the individual bins makes it quite clear that @xmath8 is inconsistent with either the london or the nyse data .
a separate study of eleven lse stocks gives @xmath75 for buy orders and @xmath76 for sell orders @xcite ; in as yet unpublished work this has been extended to @xmath77 stocks , with similar results .
our earlier study for the nyse was based on 1000 stocks @xcite .
it is clear that the average market impact functions are qualitatively different for lse and nyse stocks , and that for nyse stocks the functional form varies with market capitalization @xcite .
even if we abandon the prediction that the average market impact is a square root law , one might imagine that we could explain fluctuations in prices in terms of fluctuations in volume modulated by average market impact of the form @xmath78 . however , if this were true , for the nyse the predicted exponent for price fluctuations would be @xmath79 , which is much too large to agree with the data .
( a typical value @xcite is @xmath80 ) . to make matters even worse , the power law hypothesis for volume or market impact appears to fail in some other markets . in the paris stock exchange bouchaud et al .
@xcite have suggested that the average market impact function , but breaks down for higher @xmath18 ] is of the form @xmath82 , yielding @xmath83 in the limit as @xmath84 .
for the london stock exchange the power law hypothesis for average market impact seems reasonable , but with an exponent significantly smaller than @xmath45 .
moreover , the volume for the electronic market is not power law distributed , as discussed in the next section .
note that we are making all the above statements for individual orders , whereas many studies have been done based on aggregated data over a fixed time interval .
aggregating the data in time complicates the discussion , since the functional form of the market impact generally depends on the length of the time interval .
hence it is more meaningful to do the analysis based on individual transactions .
the theory of gabaix et al . explains the power law of returns in terms of the power law of volume , so if volume does nt have a power law , then returns should nt either .
the existence of a power law tail for volume seems to vary from market to market . for the nyse
we confirm the observation of power law tails for volume reported earlier @xcite .
however in figure [ voldist ] we show the distribution of volumes for three stocks in the electronic market of the lse . in order to compare different stocks we normalize the data by dividing by the sample
mean for each stock .
data points for each stock .
( b ) @xmath85 , where @xmath86 is the cumulative density function of returns induced by the same transactions in ( a ) . for the normalized volume there is no clear evidence for power law tails ; in contrast for returns this is quite plausible .
furthermore , the volume distributions are essentially identical , whereas the return distribution for vod decays more steeply than the others .
, title="fig : " ] data points for each stock .
( b ) @xmath85 , where @xmath86 is the cumulative density function of returns induced by the same transactions in ( a ) . for the normalized volume
there is no clear evidence for power law tails ; in contrast for returns this is quite plausible .
furthermore , the volume distributions are essentially identical , whereas the return distribution for vod decays more steeply than the others .
, title="fig : " ] all three stocks have strikingly similar volume distributions ; this is true for the roughly twenty stocks that we have studied .
there is no clear evidence for power law scaling , even though the power law scaling of the corresponding return distributions shown in figure [ voldist](b ) is rather clear .
if one attempts to fit lines to the larger volume range of the curve ( roughly @xmath87 ) , the exponent of the cumulative distribution corresponding to figure [ voldist](a ) is highly uncertain but it is at least @xmath88 , which together with the measured values of @xmath44 would imply @xmath89 . in contrast , the measured exponents for figure [ voldist](b ) are roughly @xmath90 , @xmath91 , and @xmath92 for shel , lloy , and vod respectively . it is noteworthy that vod has a much larger @xmath93 than the other stocks , even though it has essentially the same volume distribution and a similar volume distribution ; if anything from figure [ mktimpactfig ] it s @xmath44 is larger than that of the other stocks , which according to @xmath94 would imply a smaller @xmath93 .
this provides yet more evidence that the power law tails of returns are not driven by those of volume .
note that one of the differences between the nyse and the lse data examined here , which may be the underlying cause of the difference in their distributions , is that the data from the nyse includes upstairs market trades , whereas the lse data does not .
we have shown that the conclusions of gabaix et al .
@xcite are suspect for three different reasons : first , their statistical analysis in claiming the existence of a square root law for average market impact lacks power to reject alternative hypotheses in the presence of the strong autocorrelations that are present in real data ; second , new measurements of the average market impact with proper error bars show that it does not follow a square root law ; third , for electronic data the london stock exchange the distribution of volumes does not have a power law tail , and there are substantial variations between the return distributions that are not reflected in variations in volume or average market impact .
thus , it seems that the distribution of large price fluctuations can not be explained as a simple transformation of volume fluctuations .
this leaves open the question of what really causes the power law tails of prices .
we believe that the correct explanation lies in the extension of theories based on the stochastic properties of order placement and price formation @xcite , which naturally give rise to fluctuations in the response of prices to orders .
further work is clearly needed .
note added in press : in a recent study it has been shown that large price fluctuations in the nyse and the electronic portion of the lse are driven by fluctuations in liquidity @xcite . that is , if one matches up returns with the orders that generate them , the conditional distribution of large returns is essentially independent of order size .
this has been confirmed for the nyse and island by weber and rosenow @xcite .
the idea that the tail of prices is driven by fluctuations in liquidity rather than fluctuations in the number of trades was implicitly suggested earlier by results of plerou et al .
@xcite .
we would like to thank the mckinsey corporation , credit suisse first boston , the mcdonnel foundation , bob maxfield , and bill miller for supporting this research , and janos kertesz , moshe levy , rosario mantegna , and ilija zovko for valuable conversations .
poon and c.w.j .
granger , forecasting volatility in financial markets : a review , _
j. of economic literature _ 41 , 478 - 539 ( 2003 ) . c. m. c. lee and m. j. ready , _ journal of finance _ , * 46 * 733 - 746 ( 1991 ) daniels , m , farmer , j.d . ,
gillemot , l. , iori , g. , and smith , d.e . quantitative model of price diffusion and market friction based on trading as a mechanistic random process , _ physical review letters _ * 90 * , 108102 ( 2003 ) . | in a recent nature paper , gabaix et al .
@xcite presented a testable theory to explain the power law tail of price fluctuations .
the main points of their theory are that volume fluctuations , which have a power law tail with exponent roughly @xmath0 , are modulated by the average market impact function , which describes the response of prices to transactions .
they argue that the average market impact function follows a square root law , which gives power law tails for prices with exponent roughly @xmath1 .
we demonstrate that the long - memory nature of order flow invalidates their statistical analysis of market impact , and present a more careful analysis that properly takes this into account .
this makes it clear that the functional form of the average market impact function varies from market to market , and in some cases from stock to stock .
in fact , for both the london stock exchange and the new york stock exchange the average market impact function grows much slower than a square root law ; this implies that the exponent for price fluctuations predicted by modulations of volume fluctuations is much too big .
we find that for lse stocks traded in the electronic market the distribution of transaction volumes does not even have a power law tail .
this makes it clear that volume fluctuations do not determine the power law tail of price returns .
gabaix _ et al .
_ @xcite have recently proposed a testable theory for the origin of power law tails in price fluctuations .
in essence , their proposal is that they are driven by fluctuations in the volume of transactions , modulated by a deterministic market impact function .
more specifically , they argue that the distribution of large trade sizes scales as @xmath2 , where @xmath3 is the volume of the trade and @xmath4 . based on the assumption that agents are profit optimizers , they argue that the average market impact function is a deterministic function of the form @xmath5 , where @xmath6 is the the change in the logarithm of price resulting from a transaction of volume @xmath3 , @xmath7 is a constant , and @xmath8 .
this implies that large price returns @xmath6 have a power law distribution with exponent @xmath9 .
they argue that their theory is consistent with the data , even though their hypothesis about market impact appears to contradict several other previous studies @xcite in the same markets they study ( the new york and paris stock exchanges ) . |
genome - wide association studies have identified hundreds of dna variants associated to complex traits including disease in human alone @xcite . to understand how these variants affect disease risk , genotype and organismal phenotype data are integrated with intermediate molecular phenotypes to reconstruct disease networks @xcite .
a first step in this procedure is to identify dna variants that underpin variations in expression levels ( eqtls ) of transcripts @xcite , proteins @xcite or metabolites @xcite . as modern technologies routinely produce genotype and expression data for a million or more single - nucleotide polymorphisms ( snps ) and ten - thousands of molecular abundance traits in a single experiment , often repeated across multiple cell or tissue types ,
the number of statistical tests to be performed when testing each snp for association to each trait is huge .
furthermore , multiple testing correction requires all tests to be repeated several times on permuted data to generate an empirical null distribution . despite being trivially parallelisable ,
the computational burden of testing snp - trait associations one - by - one quickly becomes prohibitive .
recently a new approach ( `` matrix - eqtl '' ) was developed which uses the fact that the test statistics for the additive linear regression and anova models can be expressed as multiplications between rescaled genotype and expression data matrices , thereby realising a dramatic speed - up compared to traditional qtl - mapping algorithms @xcite .
a limitation of these models is their assumption that the expression data is always normally distributed within each genotype group .
for this reason , qtl and eqtl studies have frequently used non - parametric methods which are more robust against variations in the underlying genetic model and trait distribution @xcite . in particular , the non - parametric kruskal - wallis one - way analysis of variance @xcite does not assume normal distributions and reports small @xmath0-values if the median of at least one genotype group is significantly different from the others @xcite .
here we report a matrix - based algorithm ( `` krux '' ) , implemented in matlab , python and r , to simultaneously calculate the kruskal - wallis test statistics for several millions of snp - trait pairs at once that is more than ten thousand times faster than calculating them one - by - one on a human test dataset with more than 500,000 snps and 20,000 expression traits .
additional benefits of krux include the explicit handling of missing values in both genotype and expression data and the support of genetic markers with any number of alleles , including variable allele numbers within a single dataset .
krux takes as input genotype values of @xmath1 genetic markers and expression levels of @xmath2 transcripts , proteins or metabolites in @xmath3 individuals , organised in an @xmath4 genotype matrix @xmath5 and @xmath6 expression data matrix @xmath7 .
genetic markers take values @xmath8 , where @xmath9 is the maximum number of alleles ( @xmath10 for biallelic markers ) , while molecular traits take continuous values .
we use built - in functions of matlab , python and r to convert the expression data matrix @xmath7 to a matrix @xmath11 of data ranks , ranked independently over each row ( i.e. molecular trait ) .
krux assumes that the input expression data has been adjusted for covariates if it is necessary to do so @xcite and all data quality control has been performed .
the genotype matrix @xmath5 is first converted to sparse logical index matrices @xmath12 of the same size , where @xmath13 if @xmath14 and @xmath15 otherwise ( @xmath16 ) .
next observe that the @xmath17 vector @xmath18 with entries @xmath19 and @xmath20 matrices @xmath21 with entries @xmath22 are respectively the number of individuals and the sum of ranks for the @xmath23th trait in the @xmath24th genotype group of the @xmath25th marker .
we can then calculate an @xmath20 matrix @xmath26 with entries @xmath27 using efficient vectorised operations . if none of the rows in @xmath7 contain ties , then each entry @xmath28 equals the kruskal - wallis test statistic for testing trait @xmath23 against marker @xmath25 @xcite . for markers with less than the maximum of @xmath9 genotype values
, @xmath29 division will result in nan columns in the intermediate matrices with entries @xmath30 for the empty genotype groups . by replacing all nan s by zeros before making the sum in eq .
( [ eq:2 ] ) , the corresponding entries in @xmath26 will be the correct statistics for a test with fewer than @xmath9 degrees of freedom .
thus we need @xmath31 matrix multiplications and the associated element - wise operations to calculate the test statistic values for all marker - trait combinations .
krux takes as input a @xmath0-value threshold @xmath32 , calculates the corresponding test statistic thresholds for @xmath33 degrees of freedom ( @xmath34 ) , and identifies the entries in @xmath35 which exceed the appropriate threshold value . for these entries
only a @xmath0-value is calculated using the @xmath36 distribution .
empirical false - discovery rate ( fdr ) values are computed by repeating the @xmath0-value calculation ( with the same @xmath32 ) multiple times on data where the columns of the expression data ranks are randomly permuted .
the fdr value for any value @xmath37 is defined as the ratio of the average number of associations with @xmath38 in the randomised data to the number of associations with @xmath38 in the real data .
when data values are missing for some marker or trait , all test statistics for that marker or trait need to be adjusted for a smaller number of observations . for the expression data , missing values
are easily handled since the ranking algorithms will give nan s the highest rank . by setting the entries corresponding to missing values in @xmath7 to zero in @xmath11 , eq .
( [ eq:4 ] ) still produces the correct sums of ranks , while the matrix multiplication @xmath39 where @xmath40 is the @xmath6 matrix with @xmath41 whenever @xmath42 and @xmath43 otherwise , produces the corrected number of individuals in the @xmath24th group of the @xmath25th marker for the @xmath23th trait . replacing the constant @xmath3 in eq .
( [ eq:2 ] ) by a @xmath20 matrix @xmath44 where @xmath45 is the number of non - missing samples for trait @xmath23 and performing element - wise division and substraction operations then gives the correct test statistic for all pairs .
handling missing genotype data is less easy because the expression ranks that need to be adjusted are specific to each marker - trait combination ( e.g if a marker has a missing value where a trait has rank @xmath46 , then all samples with ranks @xmath47 need to be lowered by @xmath43 ) .
krux uses the fact that missing genotype values are generally due to sample quality and therefore patterns of missing values are often repeated among markers . for each unique missing value pattern , a new genotype matrix for all markers with that pattern and a new expression data matrix with the corresponding samples removed are constructed to calculate the test statistics for all affected marker gene combinations .
missing genotype data increases the computational cost of the algorithm considerably and it is recommended to limit the number of missing values by only considering markers with a sufficiently high call rate . in the presence of tied observations , the statistic in eq .
needs to be divided by a factor @xmath48 , where the summation is over all groups of ties and @xmath49 for each group of ties , with @xmath50 the number of tied data in the group @xcite .
the factor @xmath51 is automatically computed for each trait during the ranking step and the matrix @xmath35 is therefore easily corrected using element - wise matrix operations ( matlab version only ) .
whereas ties are usually rare in standard gene expression datasets , the ability to handle tied data expands the scope of krux to count - based , discretised or qualitative data types . since krux needs to create intermediate matrices of size @xmath52 , where @xmath2 is the number of traits and @xmath1 the number of markers , which do not usually fit into memory for large datasets , krux supports the use of data ` slices ' to divide the complete data into manageable chunks .
in typical applications , the number of markers is one or two orders of magnitude larger than the number of traits .
therefore the default behaviour of krux is to keep the expression data as a single matrix and simultaneously test all traits against subsets of markers
. the user can provide either a slice size and krux will process marker blocks of this size serially , or a slice size and initial marker and krux will process a single slice starting from that marker .
the latter option allows trivial parallelisation across multiple processors .
to test krux we provide example analysis scripts and a small anonymised dataset of 2,000 randomly selected genes and markers from 100 randomly selected yeast segregants @xcite . here
we describe an application of krux on a human dataset of 19,610 genes and 530,222 snp markers measured in 102 whole blood samples from the stockholm atherosclerosis gene expression ( stage ) study @xcite .
all snps in the dataset had minor allele frequency greater than @xmath53 , no missing values and probability to be in hardy - weinberg equilibrium greater than @xmath54 .
we first confirmed that krux produces the same results as testing marker - trait combinations one - by - one using the built - in kruskal - wallis functions to verify the correctness of our implementations . to test the performance of krux we divided the genotype data into slices of variable size and extrapolated the total run time from running a single genotype data slice against all expression traits and multiplying by the number of slices needed to cover the entire set of 530,222 snps .
the total run time rapidly decreases until a genotype slice contains about 1,000 snps and stays almost constant thereafter . on a laptop with 8 gb ram
, the limit is reached at around 3,000 snps per slice after which run time sharply increases again due to memory limitations ( fig .
[ fig : cpu ] ) .
we therefore recommend a genotype slice size of around 2,000 markers , resulting for this dataset in around 250 separate jobs , which will take around 2,500 seconds ( 42 minutes ) when run serially on a single processor . by comparison , the total extrapolated run time when computing all 19,610 @xmath55 530,222 associations one - by - one using the built - in kruskal - wallis function on the same hardware as in fig .
[ fig : cpu ] are respectively @xmath56 ( 256 gb , 2.20 ghz server ) and @xmath57 ( 8 gb , 2.70 ghz laptop ) seconds such that krux is respectively 17,000 and 11,000 times faster on this particular dataset . on the same dataset and hardware , the comparatively simpler matrix operations for the parametric tests in matrix - eqtl took respectively 5 minutes ( linear model ) and 7.4 minutes ( anova model ) .
next we compared the output of krux and matrix - eqtl s parametric anova and linear model ( henceforth called `` anova '' and `` linear '' ) methods .
the kruskal - wallis test is more conservative than the anova and linear methods , i.e. it has a higher nominal @xmath0-value for almost all marker - trait combinations ( fig .
[ fig : pval ] ) . since
random data will be subjected to the same biases , nominal @xmath0-values can not be directly compared to assess significance .
we therefore performed empirical fdr correction for multiple testing using three randomised datasets ( cf .
implementation ) .
surprisingly , after fdr correction only a limited number of associations remained for anova even at an fdr threshold of @xmath58 , whereas the number of associations detected by krux and the linear method was comparable ( fig .
[ fig : fdr](a ) ) .
detailed analysis showed that this is due to pairing of snps with rare homozygous minor alleles ( one or two samples ) to genes with outlier expression levels , resulting in extremely low @xmath0-values for the anova method in real as well as randomised data ( see also below ) . to reduce the incidence of chance associations between singleton genotype groups and outlying expression values in the anova method we repeated the empirical fdr correction , this time keeping only marker - trait combinations within 1mbp of each other ( `` cis - eqtls '' ) . at an fdr threshold of @xmath59
the number of significant _ cis_-eqtl - gene pairs is indeed comparable between the three methods , with a large proportion of pairs detected by all three of them ( fig .
[ fig : fdr](b ) ) .
-values calculated by krux vs. parametric anova * ( a ) * and linear models * ( b ) * , showing all _ cis_-acting eqtl - gene pairs with @xmath60 detected by both methods ( blue dots ) and by only one of the methods ( red crosses ) .
the black line indicates the line with slope @xmath61 . ]
we classified eqtl - gene pairs as `` skewed group sizes '' ( smallest genotype group less than 5 elements ) , non - skewed `` non - linear '' [ median of heterozygous and homozygous samples significantly different ( wilcoxon rank sum @xmath62 ) ] and non - skewed `` other '' ( all others ) . _ cis_-associations identified exclusively by the kruskal - wallis test are more often non - linear and the overall distribution of eqtl - types is more similar to associations identified by all three methods , compared to the anova and linear methods ( fig . [ fig : groups ] and fig .
[ fig : box](a - b ) ) . of the 701 associations
exclusively identified using the parametric anova method , 657 ( @xmath63 ) had skewed group sizes , including 426 ( @xmath64 ) with a singleton genotype group ( the aforementioned ` outliers ' , cf .
[ fig : box](c ) ) .
the associations exclusively identified by the linear method also contained a much higher proportion of snps with skewed group sizes than the corresponding krux associations ( @xmath65 vs. @xmath66 ) and , as expected , a reduced number of non - linear associations ( fig .
[ fig : groups ] and fig .
[ fig : box](d ) ) .
we have developed krux , a software tool that uses matrix multiplications to simultaneously calculate the kruskal - wallis test statistics for millions of marker - trait combinations in a single operation , thereby realising a dramatic speed - up compared to calculating the test statistics one - by - one .
the availability of a fast method to identify eqtl associations using a non - parametric test allowed us to assess in more detail how differences in model assumptions compared to parametric methods lead to differences in identified eqtls .
our results on a typical human dataset indicate that the the parametric anova method is highly sensitive to the presence of outlying gene expression values and snps with singleton genotype groups .
we caution against its use without prior filtering of such outliers .
linear models reported the highest number of eqtl associations after empirical fdr correction .
these are understandably biased towards additive linear associations and were also sensitive to the presence of skewed genotype group sizes , albeit to a much lesser extent than the parametric anova method .
the kruskal - wallis test on the other hand is robust against data outliers and heterogeneous genotype group sizes and detects a higher proportion of non - linear associations , but it is more conservative for calling additive linear associations than linear models , even after fdr correction .
13ifxundefined [ 1 ] ifx#1 ifnum [ 1 ] # 1firstoftwo secondoftwo ifx [ 1 ] # 1firstoftwo secondoftwo `` `` # 1''''@noop [ 0]secondoftwosanitize@url [ 0 ]
+ 12$12 & 12#1212_12%12@startlink[1]@endlink[0]@bib@innerbibempty @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) | the kruskal - wallis test is a popular non - parametric statistical test for identifying expression quantitative trait loci ( eqtls ) from genome - wide data due to its robustness against variations in the underlying genetic model and expression trait distribution , but testing billions of marker - trait combinations one - by - one can become computationally prohibitive .
we developed krux , an algorithm implemented in matlab , python and r that uses matrix multiplications to simultaneously calculate the kruskal - wallis test statistic for several millions of marker - trait combinations at once .
krux is more than ten thousand times faster than computing associations one - by - one on a typical human dataset .
we used krux and a dataset of more than 500k snps and 20k expression traits measured in 102 human blood samples to compare eqtls detected by the kruskal - wallis test to eqtls detected by the parametric anova and linear model methods .
we found that the kruskal - wallis test is more robust against data outliers and heterogeneous genotype group sizes and detects a higher proportion of non - linear associations , but is more conservative for calling additive linear associations . in summary , krux enables the use of robust non - parametric methods for massive eqtl mapping without the need for a high - performance computing infrastructure and is freely available from http://krux.googlecode.com . |
as a general rule the dynamic range of precision detectors is limited to roughly 2 to 3 orders of magnitude in energy . in case of the charged cosmic rays ( cr ) with an energy spectrum extending over more than 13 orders of magnitude ,
this neccesitates a large number of different experimental setups in order to cover the full spectrum .
space - borne , i.e. direct , experiments , cover the spectrum from @xmath0 10@xmath1 ev to @xmath0 10@xmath2 ev / nucleon , and ground - based experiments operate above total energies of a few 10@xmath3 ev up to more than 10@xmath4 ev . in the following
we will concentrate on the ground - based measurements . here
various experiments which are sensitive in the energy region around 10@xmath2 ev consistently show a significant steepening of the all - particle spectrum around this energy .
when studying the data more closely , however , the agreement between the experiments turns out to be not so good , i.e. , differences well above the fluctuations given by the individual errors .
this is shown in fig . 1 where the data on the knee in the all - particle spectrum are collected @xcite .
from these data one must conclude that the absolute position , the sharpness of the knee , and also the absolute flux in this energy region are more uncertain than expected from the individual errors .
ground - based experiments use a detector , i.e. the atmosphere as absorber with some added readout elements , like scintillators , cherenkov detectors , etc . , which can only be calibrated in the laboratory to a very limited degree .
the calibration therefore has to rely very heavily on mc simulations of the development of the extensive air showers and of the performance of the detectors .
one possible reason for the deviations in the measured spectra might thus be the use of different monte carlo ( mc ) generators in the data analyses .
this will be briefly discussed below for the most recent data .
currently the two experiments hegra @xcite , located within the observatorio del roque de los muchachos on the canary island la palma , and kascade @xcite , located near karlsruhe , germany , are analysing their airshower data with respect to the all - particle spectrum and the chemical composition of the cr near the knee. in the following we shall discuss some of the preliminary hegra results and compare to kascade where appropriate .
the hegra setup for the detection of cr with energies above @xmath5 15 tev has 3 components : ( i ) 243 scintillator stations of 1 m@xmath6 each distributed over 40,000 m@xmath6 , ( ii ) 77 airobicc cherenkov detectors spread out over the same area , and ( iii ) 17 geiger towers within the central 15,000 m@xmath6 .
the measured distribution of particle densities in the scintillator array is fitted to yield the shower size n@xmath7at detector level and the correlated shower size at the maximum ( x@xmath8 ) of the shower development . the lateral cherenkov light density , @xmath9 , is analysed in the intervall 7.5 @xmath10 100 m and can be described by an exponential @xmath11 where the shape parameter , @xmath12 , is inversely proportional to the _ absolute _ penetration depth , x@xmath13 , of the analysed shower . as mc investigation show
, this proportionality is essentially independent of the mass , a , of the incident primary particle .
the achieved resolution for x@xmath13 varies between 1.0 and @xmath0 0.5 radiation lenghts ( x@xmath14 = 36 g/@xmath15 ) for energies between 100 tev and a few pev .
the reconstruction of the energy in two independent analyses relies either on a combination of n@xmath7 and @xmath12 , or on a combination of n@xmath7 , n@xmath16(the corresponding muon shower size ) , and @xmath12 . in both cases
an energy resolution of @xmath030% is achieved . using the corsika simulation program @xcite different hadronic interaction generators ( i.e. , venus , qgsjet , hdpm , dpmjet , sibyll ) can be used .
mc investigation have shown that the qualitative sensitivity for physics parameters and e.g. the energy resolution is independent of the hadronic interaction generator .
the absolute energy scale , however , turns out to be dependent on the mc generator and the chemical composition . using the venus model within corsika
, hegra performed a regularized unfolding of the measured n@xmath17 distributions on a run - by - run basis ( see @xcite for details ) .
the result of the unfolding is shown in fig . 2 . the spectral index of the all - particle spectrum _ below _ the knee is found to be @xmath18 where the last error is an estimation of the dependence on the hadronic interaction generator in the mc . in order to get a first indication of the behaviour of the chemical composition around the knee and in order to check the mc generators , hegra made in one analysis the simple ansatz of using parametrized mc distributions for the energy determination @xcite . a more involved unfolding procedure based on the full
mc and detector information is used in a second analysis decribed in detail in @xcite . in both analyses the composition around the knee is observed to be consistent with an unchanged or slightly heavier composition . for illustrating the results and the observed mc dependence the first approach is discussed in the following . in the energy range from 300 tev to 3 pev
the height of the individual shower maxima can be reconstructed with an accuracy comparable to the rms spread for iron showers ( rms(x@xmath19 ) = 33 @xmath20 2 g/@xmath15 ) and thus much better than the rms spread for proton showers ( rms(x@xmath21 ) = 84 @xmath20 6 g/@xmath15 ) .
since the mean x@xmath13 values of protons and iron showers differ by about 140 g/@xmath15 , the fractions of light ( hydrogen and helium ) and heavy elements ( oxygen and iron ) contributing to the measured distribution can thus be obtained from fitting model x@xmath13distributions to the data for different energy bins .
the details of this analysis can be found in @xcite . under the assumption that the chemical composition is the same for the different energy bins , the mean fraction of light elements is fitted to be 0.52 @xmath20 0.10 ( total error ) , which within errors
, is compatible with direct measurements below 100 tev . besides the shape of the x@xmath13 distribution the mean value may serve to determine the elemental composition .
the measured mean values , however , turn out to be systematically smaller by 45 @xmath20 24 g/@xmath15 than expected from mc for the composition compatible with the fitted fractions which are based on the _ detailed shape _ of the x@xmath13 distribution .
this apparant discrepancy between data and mc can be due to detector , atmospheric , or mc generator effects .
after careful investigations hegra now suspects the longitudinal shower development in the mc to yield too large x@xmath13 values due to a non - perfect simulation of the longitudinal shower development .
the investigations , however , are not yet completed .
note that the airobicc - type detectors as employed e.g. by hegra are the only airshower detectors which are sensitive to the _ absolute _ position of the shower maximum due to the translation of the longitudinal development into the radial cherenkov photon density .
this is a consequence of the varying refractive index and height of emission along the shower . like hegra
, the kascade collaboration also finds little change in the chemical composition around the knee @xcite . the absolute determination which relies on the mc is , however , still biased by large discrepancies between some of the hadronic interaction generators . here
still considerable effort is needed in order to achive reliable results in the future .
since the detection of the crab nebula as a source of gamma - rays ( @xmath22-rays ) with energies above 500 gev in 1989 @xcite , about 10 galactic and extragalactic sources of very high energetic photons ( e @xmath23 300 gev ) have been detected by about as many experiments .
the status of the field is indicated in fig .
3 which shows the skymap of the very high energy ( vhe ) , i.e. , @xmath24 300 gev , @xmath22-ray sources as of august 1997 .
note that contrary to former beliefs also extragalactic sources of vhe @xmath22-rays were discovered during the last 5 years .
the fast progress of the field will be illustrated by the most recent hegra results concerning the brightest source on the @xmath22-ray sky in 1997 , the active galactic nucleus ( agn ) mkn 501 at a distance of @xmath25 = 0.034 ( @xmath0 600 million light years , for h@xmath14= 50 km sec@xmath26mpc@xmath26 ) . in 1996
the hegra experiment completed the installation of six imaging air cherenkov telescopes ( iacts ) .
four identical iacts are operated in coincident mode , i.e. for each airshower four different views are recorded .
hereby the energy threshold can be lowered and the @xmath22/hadron separation efficiency can be raised .
this is an advantage for the study of weak and especially for extended sources but is less than optimal for strong point - like sources due to the restricted effective collection area .
for the current energy range of iacts , i.e. above 300 gev , and the current stage of @xmath22-ray astronomy , i.e. the discovery age in which we learn how to optimize the detectors , this setup , however , has its merits due to the redundancy of the information . for energies below 100 gev where air showers only produce very little cherenkov light ,
however , the redundancy will be lost to a large degree due to the very limited number of particles above the cherenkov threshold , i.e. , different telescopes will view different tracks in the shower . in this energy regime therefore very sensitive telescopes like the magic telescope discussed below will be needed and which will not compromise on the effective photo collection area .
mkn 501 belongs to the blazar class of agn , i.e. it is an agn with a large plasma jet which is pointing along the line - of - sight , and it has been observed at tev energies since 1995 @xcite when its activity level corresponded to about 8% of the flux of the strongest galactic source , the crab nebula . due to its steady emission level the crab nebula has become the standard candle for @xmath22-ray astronomy .
the activity level of mkn 501 in 1996 rose to about 30% of the crab flux @xcite and in 1997 hegra recorded on average a flux corresponding to about 200% of the crab flux , and , in addition to this much higher average activity level , huge flares on time - scales as short as a day or less were recorded .
as summarized in @xcite a great number of cherenkov telescopes ( whipple , cat , tactic , telescope array , hegra ) were able to detect this signal and record a light curve .
the most complete light curve was obtained by the stand - alone telescopes ct1 and ct2 of the hegra collaboration , because ct1 was also operational during moonshine . the light curve extending from march until september 1997 is shown in fig . 4 . the short - term flares point towards very compact sources , i.e. regions in the vicinity of the supermassive ( @xmath27(10@xmath28)m@xmath29 ) black hole thought to be in the centre of the galaxy and powering the very strong non - thermal emission of the agn .
these large fluxes , which for the first time were recorded from an extragalactic source , allowed the determination of the energy spectrum extending beyond 5 tev .
the energy spectrum as measured by the hegra iact system telescopes in 1997 is shown in fig . 5 @xcite .
as the analysis of systematic effects at higher energies is not yet completed , the spectrum is only shown up to 10 tev .
but already the unabsorbed spectrum extending up to 10 tev allows to extract a fundamentally important upper limit on the density of infrared ( ir ) relic photons in intergalactic space .
this density is dependent on the era of galaxy formation and thus on the nature of the dark matter .
one extraction of an upper limit on the infrared photon density was performed in @xcite where it is assumed that the ir background is uniformly distributed in the universe .
the result , together with existing upper limits , lower limits , and an average model for galaxy evolution based on a mixed dark matter ansatz is shown in fig .
the resulting low density of infrared photons limits the parameter space for dark matter models and at the same time opens up the deeper regions of the universe ( i.e. @xmath25 = @xmath27(0.1 ) ) for tev astronomy .
currently space - borne @xmath22-astronomy is limited to energies _ below _ 10 gev and ground - based @xmath22-astronomy to energies _ above _ 300 gev . in both cases the reasons are limitations of sensitivity and/or effective collection areas .
the result is an observation gap between 10 gev and 300 gev where the universe has not been observed in and where we expect to find hints or answers to important physics questions in astrophysics , cosmology , and particle physics . in order to bridge this gap the ground - based 17 m diameter magic telescope ( see fig .
7 ) has been designed during the last 2 years @xcite . using innovative elements it will be possible to close the last observation gap for about 1% ( ! ) of the cost of a satellite experiment , which until now was believed to be necessary in order to do measurements in this energy domain . at the same time
the sensitivity in the energy region of current cherenkov telescopes will be improved by up to an order of magnitude . | selected results from the hegra experiment on charged cosmic rays and on very high energy gamma - rays are presented .
the magic telescope is presented as an outlook to the future of gamma - ray astronomy .
* selected results from * + * ground - based cosmic ray and * + * gamma - ray experiments * + n. magnussen + _ bergische universitt wuppertal + gaustr . 20 , d-42097 wuppertal , germany _ |
the scattering of electrons by ionized impurities in solids has been studied for more than half a century .
the problem is related to the nature of long - range action of the electrostatic potential , which complicates the reduction of the many - body problem to the problem of two bodies .
the basic idea for solving the problem is to perform statistical averaging of the field source over the all charged particles .
thus averaging leaves only effective near - field , while the far - field averages to zero .
it was suggested many different theoretical models during the last 70 years .
the most famous of them are the conwel and weisskopf model , models that reduce the problem to the third body exclusion method@xcite , brooks - herring model @xcite and model of the partial wave phase shifts @xcite .
the conwel and weisskopf model was a simple attempt to avoid divergence in collusion integral by using cutoff parameter therefore it is not quite consistent .
model of the partial wave phase shifts has been developed for the case of extremely high carrier concentrations realized in metals . in this case
the born approximation for transition probability is not justified and angular momentum became essential quantum number for the scattering process .
the idea of third body exclusion methods is that scattered by the given center particle must not be scattered by another one . on our opinion
, the cross section must not be subjected by this additional condition because it follows from the boltzmann kinetic equation .
given the above , the brooks - herring model is the most consistent for low and moderate carriers concentrations ( criterion can be found in 9 , ref . ) .
the validity of the brooks - herring approach was discussed several times , but to the best of our knowledge this investigation mathematically not rigorous or just comparable . moreover , previous authors consider only limiting cases of non - degenerate or degenerate electron gases .
it is well known that brooks - herring screened potential @xmath3 can be obtained as the solution of the poisson equation with linearized right - hand side . the parameter @xmath4 is named screening length . the main issues that arise in this model is the validity of the expansion of the right - hand side of poisson s equation in power series and accuracy which implements a linear term of the expansion . in the present investigations , we obtain the approximate solution of poisson s equation that contains the following non - linear terms of the expansion . as an example
, the calculation of the ionized impurity scattering mobility is performed for n - gaas parameters .
the results of the calculation give an opportunity to determine which areas and with what precision brooks - herring approach is valid .
in addition , the results also provide an opportunity to assess the accuracy of the formulas for mobility with the following non - linear terms .
the poisson equation for the screened potential has the form @xmath5-n[\zeta , t]\right\},\label{eq1 } % \int\rho(e)\left\{\frac1{1+\exp[(e+e\phi({\bf r})-\zeta)/t]}-\frac1{1+\exp[(e-\zeta)/t]}\right\}de \end{aligned}\ ] ] where @xmath6^{3/2}{\cal f}_{1/2}(\zeta / t)$ ] is the electron concentration as the function of temperature @xmath7 and chemical potential @xmath8 , @xmath9 is the fermi - dirac integral @xcite , @xmath10 is the effective mass , @xmath11 is the static dielectric constant and @xmath12 is the electron charge .
we will restrict the investigation to the centrally symmetric case for the bulk semiconductor electron gas ( @xmath13 ) . leaving only the first non - zero term of the right - hand side power - series expansion
, one can obtain the brooks - herring approach for the screened potential @xmath14 with the screening length @xmath4 : @xmath15 below we use dimensionless coordinate @xmath16 and potential @xmath17 , where @xmath18 is the charge of the scattering center in @xmath19 units .
preceding the expansion of the right - hand side of eq .
( [ eq1 ] ) one obtains : @xmath20^{l-1},\label{eq3}\nonumber\\ r_l(x)=\frac1{(l+1)!}\frac{\frac{\partial^{l+1 } n[\zeta+e\xi\phi(x),t]}{\partial \zeta^l}}{\frac{\partial n(\zeta , t)}{\partial \zeta}}\left[\frac{ze^2}{\epsilon\lambda}\right]^{l},\qquad 0<\xi<1 ; \end{aligned}\ ] ] where we use the remainder term in the lagrange form @xcite . note that since the functions @xmath21 have a maximum in the region @xmath22 , in the degenerate case strict inequality holds @xmath23 ( @xmath24 ) , whereas in the case of boltzmann electron gas we can speak only about the order of magnitude @xmath25 .
we proceed the consideration , omitting the remainder term @xmath26 and supposing that tolerance is estimated by the value @xmath27 . supposing that @xmath28 are low parameters one can replace @xmath29 with @xmath30 in the right hand side .
the latter designation contains dimensionless parameters @xmath31 and @xmath32 ( here @xmath33 is the number of iterations and @xmath34 , @xmath35 mean small values in order of argument ) .
we define the parameters @xmath36 and @xmath37 from the iteration procedure for the special case considered the sec .
the obtained non - homogeneous equation is simply resolved @xmath38-\exp[x - x'(\alpha_0l+1)]\}}{x'^{l-1 } } dx',\label{eq4 } \end{aligned}\ ] ] where upper index @xmath39 identify the number of taken into account nonlinear terms and lower index identify the number of iterations was done .
it is easy to see that for @xmath40 iteration procedure is divergent in the region @xmath41 , because @xmath42 . on the other hand ,
relaxation processes are introduced through the fourier components @xmath43 of the potential in born approximation , and the region of divergence @xmath41 may became are not essential for them for the certain range of parameters .
the fourier transformation of the eq.([eq4 ] ) is possible for @xmath44 and has the next form @xmath45 where @xmath46 and @xmath47 is the wave - vector transfer value . to obtain usual dimension one has to multiply the fourier component by the factor @xmath48 .
the coefficients @xmath49 are defined as the next @xmath50,&l=2 ; \\ \frac1{1+q^2}\left\{6\alpha_0\arctan\left(\frac{q}{3\alpha_0}\right)+q\ln|9\alpha_0
^ 2+q^2|-q\ln\left[\frac{(3\alpha_0 + 1)^{3\alpha_0 + 1}}{(3\alpha_0 - 1)^{3\alpha_0 - 1}}\right]\right\},&l=3 ; \\
\arctan\left(\frac{q}{4\alpha_0}\right)\left(\frac{q^2 - 16\alpha_0 ^ 2}{1+q^2}\right)+\frac{q}{1+q^2}\left\{\frac12\ln\left[\frac{(4\alpha_0 + 1)^{(4\alpha_0 + 1)^2}}{(4\alpha_0 - 1)^{(4\alpha_0 - 1)^2}}\right]-4\alpha_0\ln|16\alpha_0 ^ 2+q^2|\right\},&l=4 ; \end{array } \right.\label{eq6 } \end{aligned}\ ] ] one can see from eq .
( [ eq5 ] ) that for any @xmath51 coefficients @xmath52 . therefore ,
if @xmath53 are small enough , then eqs .
( [ eq4 ] ) and ( [ eq5 ] ) with @xmath54 and @xmath55 give valid additives to the fourier components of the brooks - herring screened potential .
next , we consider bulk gaas as an example . to calculate @xmath28 we need the equation of electroneutrality @xmath56 + 1\},&n_d<10^{16}\mbox{cm}^{-3 } ; \\ n_d,&n_d\geq10^{16}\mbox{cm}^{-3 } ; \\ \end{array } \right.\label{eq7 } \end{aligned}\ ] ] where @xmath57 is the donor concentration and @xmath58 is the donor energy level . in the latter formula we suppose that donor concentration higher than @xmath59 @xmath2 leads to the mott transition in bulk gaas ( see for example pg .
41 , fig . 21 in ref .
@xcite ) . figures [ fig1 ] ( a - c ) demonstrate coefficients @xmath60 for n - gaas parameters @xcite versus temperature @xmath7 and donor concentration @xmath57 as the contours of equivalent values .
the calculations performed for @xmath61 . obviously , in the case @xmath62 ( acceptors ) @xmath63 change their signs to opposite [ see eq.([eq3 ] ) ] .
therefore , screened potential loose its symmetry relative to the sign of the centers charge , when the nonlinear terms are taken into account . for n - gaas parameters versus temperature @xmath7 and donor concentration @xmath57 ( logarithmic scale ) as the contour plots .
frames a - c correspond to @xmath60 , @xmath61 .
[ fig1],width=288 ] and donor concentration @xmath57 ( logarithmic scale ) as the contour plots .
frames a - c correspond to @xmath64 .
[ fig2],width=288 ]
in the present section we calculate the nonlinear additives to the ionized impurity scattering mobility in n - gaas@xcite .
the transport time for the electron - impurity scattering can be calculated according to the formula @xmath65 where @xmath66 is the electron energy and @xmath67 is defined by the formulas ( [ eq5 ] ) , and ( [ eq6 ] ) with @xmath55 and @xmath54 . the low - field mobility of the electron gas can be introduced in the form @xcite : @xmath68 ^ 2}.\label{eq9 } \end{aligned}\ ] ] to demonstrate the influence of the non - linear additives on mobility we introduce the next dimensionless parameter @xmath69 where @xmath70 is the mobility calculated using the brooks - herring approach .
the parameters @xmath71 may be treated as the relative mobility changes due to the @xmath39-th non - linear term .
the results of calculation for gaas parameters one can see in fig .
[ fig2 ] ( a - c ) .
it is easy to see that outside the region @xmath72 , @xmath73 non - linear additives change the mobility no more then 20% . whereas , inside the region @xmath72 , @xmath73 the changes in mobility with the addition of the quadratic term @xmath74 is very significant see [ fig .
[ fig2 ] ( a ) ] . on the other hand , with increasing amounts of non - linear terms from @xmath75 to @xmath76 ,
the mobility does not change substantially [ see figs .
[ fig2 ] ( b ) and ( c ) ] .
correspondingly , one may expect that for @xmath75 the mean inaccuracy of the formulas ( [ eq5 ] ) and ( [ eq6 ] ) caused by the power - series cutoff at most 10% in low the temperature region and below 5% for @xmath77 [ see fig .
1(b ) ] . in the case
@xmath78 inaccuracy decreases further more [ see fig . [ fig2](c ) ] .
considering eqs . ( [ eq5 ] ) and ( [ eq6 ] ) , one can see that for @xmath79 fourier components in eq .
( [ eq5 ] ) can be rewritten as @xmath80q^2+o_1(q^4),\label{eq11 } \end{aligned}\ ] ] where we suppose that @xmath81 and the functions @xmath82 describe small values in order of the argument .
it is easy examining that functions @xmath83 conserve this form for arbitrary number of iterations @xmath33 . neglecting the small values
@xmath84 and supposing that the number of convergent iterations is infinite , one could obtain the algebraic system of equations for the parameters @xmath85 and @xmath86 : @xmath87 where the coefficients @xmath88 are defined by the equality @xmath89,&l=3 ; \\ -4\alpha-8\alpha\ln|4\alpha|+\frac12\ln\left[\frac{(4\alpha+1)^{(4\alpha+1)^2}}{(4\alpha-1)^{(4\alpha-1)^2}}\right],&l=4 ; \end{array } \right.\label{eq13 } \end{aligned}\ ] ] and the coefficients @xmath90 are defined by the next equality @xmath91 - 2 - 2\ln|3\alpha|,&l=3 ; \\ \frac1{12\alpha}-\frac12\ln\left[\frac{(4\alpha+1)^{(4\alpha+1)^2}}{(4\alpha-1)^{(4\alpha-1)^2}}\right]+4\alpha+4\alpha\ln|16\alpha^2|,&l=4 . \end{array } \right.\label{eq14 } \end{aligned}\ ] ] if the algebraic system ( [ eq12 ] ) with given @xmath39 and coefficients ( [ eq13 ] ) , ( [ eq14 ] ) has real positive solutions then the iteration procedure is supposed to be convergent .
choosing the solution that is closest to the pair @xmath92 and @xmath93 one can use the fourier components @xmath94 the latter approach valid for processes with a small wave - vector transfer @xmath95 , but for arbitrary @xmath96 that provide eqs .
( [ eq12])-([eq14 ] ) with real positive roots . to analyze the system we consider limiting case @xmath97 and put @xmath75 .
under such conditions we obtain @xmath92 and only one equation @xmath98 .
the real root @xmath99/\gamma_2 $ ] that satisfies the mentioned above conditions exists only if @xmath100 .
note , that in the case @xmath101 the latter inequality is always satisfied . on the other hand , in the case
@xmath102 it fails in the low temperature region [ see fig . [ fig1](a ) ] .
the linearization of the system ( [ eq12 ] ) leads to the approximate solution @xmath103 and @xmath104 .
therefore , one can expect that if @xmath24 then @xmath105 and @xmath106 , whereas if @xmath101 then @xmath107 and @xmath108 .
one also can estimate that inaccuracy of the first iteration , considered in sec .
iii is about @xmath109% for slow particles [ see fig . [ fig1](a ) ] .
the physical interpretation of the approach ( [ eq12])([eq15 ] ) is simple : the brooks - herring screened potential remains valid for slow particles , but the screening length became in @xmath110 times larger and the charge of the ionized impurity became @xmath111 times larger .
the validity of the obtained results restricted by the conditions for the born and the effective mass approximations .
the latter supposes @xmath112 ( @xmath113 is the lattice constant ) that is well satisfied for the considered doping levels , whereas born approximation is questionable for some regions of parameters .
we also omit consideration of the nonparabolicity effect that is not essential for the investigation performed .
the main theoretical results are next : ( i ) the screened potential loose its symmetry relative to the sign of the scattering centers charge , when the nonlinear terms are taken into account ; ( ii ) in the region @xmath72
, @xmath73 the nonlinear terms changes the ionized impurity scattering mobility up to 50% in n - gaas ; ( iii ) the brooks - herring screened potential is valid in low temperature region under the condition @xmath114 , but with the different screening length @xmath115 and the different charge of the scattering center @xmath116 , where @xmath103 , @xmath104 and @xmath117[{\partial n(\zeta , t)}/{\partial \zeta}]^{-1}$ ] .
100 e. m. conwell and v. f. weisskopf , theory of impurity scattering in semiconductors . _ phys rev_. vol .
388 - 390 ( * 1950 * ) b. k. ridley , reconciliation of the conwell - weisskopf and brooks - herring formulae for charged - impurity scattering in semiconductors : third - body interference .
_ j. phys .
c : solid state phys_. vol .
1589 - 1593 , ( * 1977 * ) b. k. ridley , _ quantum processes in semiconductors _ , oxford university press , new york ( * 1999 * ) d. poznyakov , scattering of electrons by ionized impurities in semiconductors : quantum - mechanical approach to third body exclusion .
_ j comput .
electron_. vol .
338 - 351 ( * 2014 * ) h. brooks , scattering by ionized impurities in semiconductors .
rev_. vol .
83 , pp . 879 ( * 1951 * ) h. brooks , theory of the electrical properties of germanium and silicon .
electron phys_. vol . 7 , pp .
85 - 182 ( * 1955 * ) f. j. blatt , residual resistivity of copper and silver alloys : dependence on periodic table .
rev_. vol .
285 - 290 ( * 1957 * ) j. fridel , metalic alloys .
_ niovo cimento , suppl_. vol . 7 ( 2 ) , pp .
287 - 311 ( * 1958 * ) v. f. gantmakher and y. b. levinson , _ carrier scaterring in metals and semiconductors _ , nauka , moskow , 1084 ; north - holland , the netherlands ( * 1987 * ) l. m. falicov and m. cuevas , mobility of electrons in compensated semiconductors .
ii . theory .
rev_. vol .
1025 - 1032 ( * 1967 * ) d. chattopadhyay , h. j. queisser , electron scattering by ionized impurities in semiconductors .
phys_. vol .
745 - 768 ( * 1981 * ) k. seeger , _ semiconductor physics _ ,
springer - verlag , wien ( * 1973 * ) m. kline , _ calculus : an intuitive and physical approach _ , dover ( * 1998 * ) gaas parameters : effective mass is @xmath118 , static dielectric constants is @xmath119
. l. j. brillson _ contacts to semiconductors .
fundamental technology _ , noyes publications , new jersey ( * 1993 * ) | in the present work we calculate the non - linear additives to the brooks - herring electrostatic screened potential .
we also calculate the corresponding additives to the ionized impurity scattering mobility in n - gaas over a wide range of doping levels and temperatures .
it is demonstrated that the additives to the mobility changes its value up to 50% .
the most dramatically mobility is changed near the mott transition , when the doping level is about @xmath0@xmath1 @xmath2 and temperature is below 50 k. also , the results allow us to conclude that the non - linear additives break the cross section symmetry with respect to the sign of the charge of the scattering center . |
the objectives of quantum control are to find ways to manipulate the time evolution of a quantum system such as to drive an initial given state to a pre - determined final state , the target state ; or optimize the expectation value of a target observable . among a wealth of applications
are those to quantum computing , where it is clearly essential to be able to start off a quantum procedure with a given initial state , and to problems involving the population levels in atomic systems , such as the laser cooling of atomic or molecular systems .
the mathematical tools necessary for the theoretical investigation of these control problems are diverse , involving algebraic , group theoretic and topological methods . the questions that one may ask include : when is a given quantum system completely controllable ?
if a system is not completely controllable , how does this affect optimization of a given operator ?
how near can you get to a target state for a not completely controllable system ?
the answer to the first question depends on a knowledge of the lie algebra generated by the system s quantum hamiltonian , that to the second arises from properties of the lie group structure , while the last clearly involves ideas of topology . especially in the area of lie group theory
, there is a large corpus of classical mathematics which can supply answers to questions arising in quantum control .
in particular , for the type of controllability known as _ pure state controllability _ classical lie group theory has already given the basic results .
the quantum control system we shall consider is typically of the form @xmath0 where @xmath1 is the internal hamiltonian of the unperturbed system and @xmath2 are interaction terms governing the interaction of the system with an external field . the dynamical evolution of the system is governed by the unitary evolution operator @xmath3 , which satisfies the schrodinger equation @xmath4 with initial condition @xmath5 , where @xmath6 is the identity operator . by use of the magnus expansion , it can be shown that the solution @xmath7 involves all the commutators of the @xmath8 .
the operators @xmath8 , @xmath9 , in ( [ eq : h ] ) are hermitian .
their skew - hermitian counterparts @xmath10 generate a lie algebra @xmath11 known as the dynamical lie algebra of the control system which is always a subalgebra of @xmath12 , or for trace - zero hamiltonians , @xmath13 .
the degree of controllability is determined by the dynamical lie algebra generated by the control system hamiltonian @xmath14 . if @xmath15 then _ all _
the unitary operators are generated and we call such a system _ completely controllable_. a large variety of common quantum systems can be shown to be completely controllable @xcite .
the interesting cases arise when @xmath11 is a proper subalgebra of @xmath12 .
such systems may still exhibit _ pure state controllability _ , in that starting with any initial pure state any target pure state may be obtained , as distinct from the completely controllable case , when all ( kinematically admissible ) states pure or mixed may be achieved .
we shall restrict our attention here to finite - level quantum systems with @xmath16 discrete energy levels .
the pure quantum states of the system are represented by normalized wavefunctions @xmath17 , which form a hilbert space @xmath18 .
however , the state of a quantum system need not be represented by a pure state @xmath19 .
for instance , we may consider a system consisting of a large number of identical , non - interacting particles , which can be in different internal quantum states , i.e. , a certain fraction @xmath20 of the particles may be in quantum state @xmath21 , another fraction @xmath22 may be in another state @xmath23 and so forth .
hence , the state of the system as a whole is described by a discrete ensemble of quantum states @xmath24 with non - negative weights @xmath25 that sum up to one .
such an ensemble of quantum states is called a _ mixed - state _ , and it can be represented by a density operator @xmath26 on @xmath18 with the spectral representation @xmath27 where @xmath28 is an orthonormal set of vectors in @xmath18 that forms an ensemble of independent pure quantum states .
the evolution of @xmath29 is governed by @xmath30 with @xmath31 as above .
clearly if _ all _ the unitary operators can be generated we have the optimal situation , complete controllability .
however , classical lie group theory tells us that even when we only obtain a subalgebra of @xmath12 we can obtain pure state controllability .
the results arise from consideration@xcite of the transitive action of lie groups on the sphere @xmath32 .
the classical `` orthogonal '' groups @xmath33 where the field @xmath34 is either the reals @xmath35 , the complexes @xmath36 or the quaternions @xmath37 , are defined to be those that keep invariant the length of the vector @xmath38 ; the squared length is given by @xmath39 , where @xmath40 refers to the appropriate conjugation .
these compact groups are , essentially , the only ones which give transitive actions on the appropriate spheres , as follows : @xmath41 transitive on @xmath42 @xmath43 transitive on @xmath44 @xmath45 transitive on @xmath46 .
since we may regard our pure state as a normalized vector in @xmath47 and thus as a point on @xmath48 , we obtain pure state controllability only for @xmath49 ( or @xmath50 if we are not too fussy about phases ) and @xmath51 , the latter for even @xmath16 only .
( note that we can not get @xmath52 as a subalgebra of @xmath49 . )
complete controllability is clearly a stronger condition than pure state controllability . to illustrate our theme of the limitations on quantum control , we now give two examples based on a truncated oscillator with nearest - level interactions for which the algebras generated are @xmath53 and @xmath54
both these examples are generic .
consider a three - level system with energy levels @xmath55 , @xmath56 , @xmath57 and assume the interaction with an external field @xmath58 is of dipole form with nearest neighbor interactions only .
then we have @xmath59 , where the matrix representations of @xmath1 and @xmath60 are @xmath61 , \quad { h}_1 = \left[\begin{array}{ccc } 0 & d_1 & 0 \\ d_1 & 0 & d_2 \\ 0 & d_2 & 0 \end{array}\right].\ ] ] if the energy levels are equally spaced , i.e. , @xmath62 and the transition dipole moments are equal , i.e. , @xmath63 then we have @xmath64 , \quad { h}_1 = d \left[\begin{array}{ccc } 0 & 1 & 0 \\ 1 & 0 & 1 \\ 0 & 1 & 0 \end{array}\right]\ ] ] where @xmath65 is the traceless part of @xmath1 .
both @xmath66 and @xmath67 satisfy @xmath68 where @xmath69\ ] ] which is a defining relation for @xmath70 .
the dynamical lie algebra in this case is in fact @xmath70 .
it is easy to show that the matrix @xmath71 is a real anti - symmetric representation of @xmath70 if @xmath72 .
explicitly , a suitable unitary matrix is given by @xmath73.\ ] ] since the dynamical algebra and group in the basis determined by @xmath7 consists of _ real _ matrices , real states can only be transformed to real states ; this means that for any initial state there is a large class of unreachable states .
this example is generic as it applies to @xmath16-level systems , although for even @xmath16 we need other than dipole interactions to generate @xmath53 .
the analogous dipole interaction generates @xmath54 in the even @xmath16 case , as we now illustrate .
consider a four - level system with hamiltonian @xmath59 , @xmath74 and @xmath75 note that @xmath76 and @xmath77 satisfy @xmath78 for @xmath79 where @xmath80 is unitarily equivalent to @xmath81 which is a defining relation for @xmath54 .
consider an initial state of the form @xmath82 where @xmath83 satisfies ( [ eq : sp ] ) , it can only evolve into states @xmath84 where @xmath85 satisfies ( [ eq : sp ] ) , under the action of a unitary evolution operator in exponential image of @xmath11 .
hence , any target state that is not of the form ( [ eq : rho_1 ] ) is not accessible from the initial state ( [ eq : rho_0 ] ) .
note that the initial state @xmath86 is of the form @xmath87 and that @xmath88 satisfies ( [ eq : sp ] ) . consider the target state @xmath89 which is clearly kinematically admissible since @xmath90 but note that @xmath91 and @xmath92 does not satisfy ( [ eq : sp ] ) .
hence , @xmath93 is not dynamically accessible from @xmath94 for this system . given a target state @xmath93 that is not dynamically accessible from an initial state @xmath94 , we can easily construct observables whose kinematical upper bound for its expectation value can not be reached dynamically .
simply consider @xmath95 .
the expectation value of @xmath96 assumes its kinematical maximum only when the system is in state @xmath93 .
since @xmath93 is not reachable , the kinematical upper bound is not dynamically realizable .
in this short introduction to quantum control theory , we have described the goals of the subject briefly , and then illustrated the limitations by generic examples where complete control is not possible .
the tools we have used are , in the main , those of classical lie group theory .
theoretical problems that remain to be tackled include to what extent these non - controllable systems can , in fact , be controlled ; and , of course , the paramount problem of implementing these controls in practice .
it gives us great pleasure to acknowledge many conversations with colleagues , including drs .
hongchen fu and gabriel turinici .
a.i.s would also like to thank the laboratoire de physique thorique des liquides , of the university of paris vi , for hospitality during the writing of this note . | in this note we give an introduction to the topic of quantum control , explaining what its objectives are , and describing some of its limitations . |
at high temperatures qcd is expected to be found in a new phase , the quark - gluon plasma . while the thermal excitations are hadrons and glueballs at low @xmath0 , the degrees of freedom in the plasma phase are the quarks and gluons .
this new state of matter is believed to have existed during the first microseconds after the big bang , and much of the recent interest stems from the fact such conditions may be produced in heavy - ion collisions .
already the results from cern - sps seem to hint in that direction , and the experiments at higher energies at bnl - rhic have provided a wealth of new interesting results after its first year of running @xcite . to understand and interpret the experimental signatures in terms of the evolution of the initial stage after
a heavy - ion collision is clearly a very challenging theoretical task .
the most convincing theoretical results that a drastic change in the degrees of freedom takes place at a certain @xmath0 come from lattice studies . in the pure glue theory ,
there is a phase transition between the confined and deconfined phases at a critical temperature @xmath1 mev @xcite .
when massless quarks are added , there is similarly a phase transition to a chirally symmetric phase at @xmath2 mev @xcite , where the precise value depends on the number of flavors .
recent lattice simulations also suggest that the chiral transition is simultaneous with the deconfining one @xcite . at physical quark masses
the situation is not completely clear , in the sense that there may not be a true phase transition but only a rapid cross - over @xcite .
nevertheless , lattice simulations have shown that the pressure , divided by the ideal gas result and plotted against @xmath3 , is almost independent of the number of flavors @xcite . due to asymptotic freedom
, the quark - gluon plasma behaves as an ideal gas at asymptotically high temperatures . up to corrections of the order of 20 percent
, this behavior holds down to temperatures @xmath4 , even though each higher order term in a straightforward perturbative expansion gives widely different contributions in this temperature regime @xcite .
instead , at @xmath5 one needs a resummed effective theory in terms of quasi particles , the htl effective action @xcite .
such an effective description correctly reproduces thermodynamic quantities like the pressure , as measured by the lattice , down to approximately @xmath6 @xcite . despite this progress , it is of course highly desirable to actually have an analytical description at @xmath7 , close to the critical temperature . since the qcd coupling constant @xmath8 at @xmath9 ( using a renormalization scale @xmath10 ) ,
one is presumably forced to consider effective models that go beyond the fundamental qcd lagrangian . in a recent paper @xcite ,
such an effective theory was constructed in terms of the thermal wilson line @xmath11 , @xmath12 , \label{wilsonline}\end{aligned}\ ] ] where @xmath13 denotes path ordering , @xmath14 is the inverse temperature and @xmath15 the time component of the gluon field , with @xmath16 the generators of the fundamental su(3 ) representation , @xmath17 .
the trace of the wilson line is proportional to the polyakov loop @xmath18 , @xmath19 . in the pure yang - mills theory ,
the polyakov loop is an order parameter for a global z(3 ) symmetry separating the confined and deconfined phases , with @xmath20 ( @xmath21 ) above ( below ) the phase transition @xcite .
when dynamical quarks are introduced , @xmath18 ceases to be an order parameter in the strict sense , but the susceptibility of @xmath18 still peaks strongly at @xmath9 @xcite . in the effective theory @xcite , the pressure of the quark - gluon plasma at @xmath22 is completely due to the condensate of @xmath18 , and below @xmath9 , where @xmath21 , the pressure vanishes .
moreover , the effective potential @xmath23 changes extremely rapidly around @xmath9 . hence ,
as the system cools it may find itself trapped at the wrong value of @xmath24 . by coupling the effective field @xmath18 to hadronic degrees of freedom , e.g. the pions ,
hadrons can be produced as @xmath18 evolves from @xmath20 and subsequently oscillates around @xmath21 .
this scenario is somewhat reminiscent of reheating after inflation @xcite , and much attention has lately been paid to that aspect of the model @xcite
. remarkably , many qualitative features observed at rhic are in accordance with the model predictions
. however , when it comes to questions related to the change of the expectation value of @xmath18 and particle production around @xmath9 , one has to take into account the variation of @xmath18 in space - time . in this paper
we address the question of radiative corrections to the spatial variation , by considering the leading one - loop qcd contribution to the spatial derivatives of @xmath18 .
previous work @xcite took into account only the classical kinetic term in the ( euclidean ) effective action , @xmath25 .
while such an approach certainly is justified at these preliminary stages , it is important to estimate how much the radiative qcd - effects can affect @xmath26 around @xmath9 , where the qcd coupling constant becomes large . as for the parameters in the potential @xmath23
, they can be fitted by comparing to qcd lattice results and so are well defined at all @xmath0 .
the kinetic term , on the other hand , has to be matched to perturbatively calculated terms in qcd , and could therefore receive large radiative corrections .
the magnitude of the first one - loop qcd correction can then hopefully serve as a guideline to the importance of loop effects , and indicate how reliable the above form of @xmath26 is at @xmath9 .
we want to stress that the radiative corrections to be discussed come from qcd , and not from fluctuations in @xmath26 . to study the correction to the kinetic term @xmath27
, we first consider the one - loop induced quartic terms in qcd , that contain four powers of the external field @xmath28 and two powers of the external momenta .
these terms contribute to the high @xmath0 , dimensionally reduced qcd effective action @xmath29 @xcite , and apart from providing a correction to @xmath29 they can also be related to the kinetic term in @xmath26 . as a byproduct we obtain some additional derivative interactions in @xmath26 .
the paper is organized as follows . in the next section ,
we give the perturbative qcd calculation that corresponds to the leading derivative interactions in @xmath29 . in sec .
iii we make the actual matching from an effective theory in terms of @xmath28 to the one in @xmath18 , and discuss the validity of the results .
we end with our conclusions and an outlook .
our conventions and some technical details can be found in the appendix .
in the high temperature regime , long distance phenomena ( i.e. @xmath30 ) are dominated by the static sector of qcd . at high
@xmath0 it therefore makes sense to use dimensional reduction and integrate out all the nonstatic modes in the theory @xcite . with only the static modes left ,
the full qcd lagrangian is reduced to a three - dimensional theory . in principle
the integrating - out procedure gives rise to an infinite number of interaction terms , but higher dimensional operators become more suppressed by powers of the qcd coupling constant @xmath31 and/or @xmath0 . in full qcd , the following terms in the resulting effective action @xmath29 have been calculated @xcite , @xmath32[d_i , a_0 ] + g^2t^2\left ( 1+\frac{n_{\ !
f}}{6}\right ) { \rm tr}\ , a_0 ^ 2 + \frac{g^4(9-n_{\ !
f})}{24\pi^2 } \left ( { \rm tr}\ , a_0 ^ 2\right ) ^2 \right ] \ , \label{knownaction}\end{aligned}\ ] ] where @xmath33 , @xmath34 , @xmath35 and @xmath36 .
the next term in @xmath29 contains two derivatives and four powers of @xmath28 , and corresponds to the following part in the euclidean effective action for @xmath37 , @xmath38 a_0^a(\vec{k_1})a_0^b(\vec{k_2})a_0^c(\vec{k_3})a_0^d(\vec{k_4 } ) \ , \label{effactiona0}\end{aligned}\ ] ] where @xmath39 is the four - point function of order @xmath40 , obtained by integrating out all the non - static modes .
such higher dimensional terms have been calculated in the pure yang - mills theory using the background field method @xcite , as well as in qed @xcite , but not in qcd with quarks . in this paper we will use a diagrammatic approach to the four - point function .
perturbatively , the four - point function receives contributions from the diagrams shown in fig . 1 , together with the additional permutations of the external legs
. there are five permutations adding to the graphs ( a ) , ( b ) and ( d ) , and two to the diagrams ( c ) and ( e ) , where ( e ) has a symmetry factor @xmath41 .
even though all diagrams are superficially logarithmically divergent , it is well known that both the fermion diagram and the sum of the pure yang - mills diagrams are ultra - violet finite .
we will therefore only give explicit results for the finite @xmath0 part of these diagrams , where we use the imaginary time formalism @xcite combined with the particular technique described in the appendix .
= 8 cm consider first the contribution from the @xmath42 massless quarks , with the particular ordering of momenta shown in fig .
1(a ) . using the conservation of momentum , @xmath43 , with @xmath44
, we find @xmath45\beta^{-z } \times \nonumber \\ & & \left [ f_1^{-z/2}\gamma ( z/2 ) \left \ { \frac{5}{2}\gamma ( ( 1\!-\!z)/2 ) -6\gamma ( ( 3\!-\!z)/2 ) + \frac{2}{3 } \gamma ( ( 5\!-\!z)/2 ) \right \ } + f_1^{-(2+z)/2}\gamma ( ( 2\!+\!z)/2 ) \times \right .
\nonumber \\ & & \left . \left
\ { f_3\gamma ( ( 1\!-\!z)/2)-\frac{2}{3}f_2\gamma ( ( 3\!-\!z)/2)\right \ } + \frac{2}{3}f_4f_1^{-(4+z)/2 } \gamma ( ( 1\!-\!z)/2)\gamma ( ( 4\!+\!z)/2 ) \right ] \ , \label{fermionloop}\end{aligned}\ ] ] where @xmath46 is the riemann zeta - function and @xmath47 , given explicitly in the appendix , are functions of the feynman parameters @xmath48 and the external momenta @xmath49 . in the high temperature limit ,
where @xmath50 , we can use the residue theorem to evaluate the @xmath51-integral by closing the contour on the left side in the complex @xmath51-plane .
although there is seemingly a logarithmic dependence on @xmath0 from a double - pole at @xmath52 , coming from the product @xmath53 , the coefficient is actually proportional to @xmath54 = 0 $ ] .
this is in accordance with the fact that the fermion loop does not have any logarithmic enhancements @xcite . for the term of order @xmath55
we get , from the poles at @xmath56 , @xmath57 when the additional five permutations are added and the resulting four - point function inserted into eq .
( [ effactiona0 ] ) , the contribution to the effective action from the quark loop becomes , @xmath58 \ , \label{fermionpart1}\end{aligned}\ ] ] after a partial integration . in the qed case
we have @xmath59 , and then our result for the two - derivative part of the qed effective action agrees with the earlier calculation in @xcite .
we now turn to the pure yang - mills contribution , i.e. the diagrams ( b)-(e ) in fig . 1 .
to evaluate the finite @xmath0 part of these diagrams we will use the gauge condition @xmath60 , and work in feynman gauge .
as in the fermion case , the functions @xmath61 below are all functions of the relevant feynman parameters and the external momenta .
their explicit forms can also be found in the appendix .
proceeding in a way similar to the previous section , we have for the ghost loop depicted in fig .
1(b ) , @xmath62(\beta^2g_1)^{-z/2 } \gamma ( ( 5\!-\!z)/2)\gamma ( z/2 ) , \label{ym1}\end{aligned}\ ] ] where the color structure is @xmath63 , with @xmath64 and @xmath65 the completely symmetric structure constant . for the graph in ( c )
we find , @xmath66\beta^{-z } \times \nonumber \\ & & \left [ g_2^{-z/2}\gamma ( z/2 ) \left \ { 5\gamma ( ( 1\!-\!z)/2 ) + 16\gamma ( ( 3\!-\!z)/2 ) + 32 \gamma ( ( 5\!-\!z)/2 ) \right \ } -g_2^{-(2+z)/2}\gamma ( ( 2\!+\!z)/2 ) \times \right .
\nonumber \\ & & \left . \left
\ { g_4\gamma ( ( 1\!-\!z)/2)-\frac{16}{3}g_3\gamma ( ( 3\!-\!z)/2)\right \ } -\frac{2}{3}g_5g_2^{-(4+z)/2 } \gamma ( ( 1\!-\!z)/2)\gamma ( ( 4\!+\!z)/2 ) \right ] \ , \label{ym2}\end{aligned}\ ] ] where the color structure is as for the ghost loop . for the triangle diagram ( d )
we get , @xmath67(\beta^2g_6)^{-z/2 } \times \nonumber \\ & & \left [ g_6^{-1}(g_6 - 3g_7 ) \gamma ( ( 1\!-\!z)/2)\gamma ( ( 2\!+\!z)/2 ) -15\gamma ( ( 3\!-\!z)/2)\gamma ( z/2 ) \right ] \ , \label{ym3}\end{aligned}\ ] ] where @xmath68 .
finally , we find for graph ( e ) , @xmath69(\beta^2g_8)^{-z/2}\gamma ( ( 1\!-\!z)/2)\gamma ( z/2 ) \ , \label{ym4}\end{aligned}\ ] ] with @xmath70 .
contrary to the fermion case , at order @xmath71 each diagram contains a logarithmic dependence on @xmath0 and the external momenta @xmath49 .
in addition , the @xmath72 part depends logarithmically on @xmath49 and an ultraviolet cut - off @xmath73 , that has to be introduced to regularize the loop - momentum integral .
the logarithmic dependence on @xmath49 cancels out between the @xmath72 and @xmath74 parts for each permutation of each individual diagram , whereas the terms containing @xmath75 ( @xmath76 ) only cancel out in the total @xmath74 ( @xmath72 ) result , i.e. when all the different diagrams are added together .
this was basically noted already in @xcite , and we have checked that it holds true in our calculations as well .
there is also a linear divergence when @xmath77 , at order @xmath78 , in all of the eqs .
( [ ym1])-([ym4 ] ) .
this divergence originates from static propagators running in the loop , i.e. the propagators with a vanishing matsubara frequency , @xmath79 .
if only the non - static modes are integrated out , the static terms should be subtracted and our remaining result is then finite in the limit @xmath77 , as it should @xcite .
we emphasize that the @xmath79 modes do not influence the two - derivative term .
taking into account all the permutations of eqs .
( [ ym1])-([ym4 ] ) and adding the different contributions , we find for the @xmath80 term , @xmath81 \ , \label{gluepart}\end{aligned}\ ] ] where we have performed an integration by parts .
this result for the pure yang - mills contribution disagrees slightly with the previous finding from the background field method in @xcite , in that the first two term in eq .
( [ gluepart ] ) are a factor @xmath82 smaller , and the last a factor @xmath83 . by combining the results in eqs .
( [ fermionpart1 ] ) and ( [ gluepart ] ) , the complete contribution to the effective action becomes , @xmath84 - \nonumber \\ & & \frac{11g^4\xi ( 3)\beta^2}{1536\pi^4}\left ( \frac{28n_{\ ! f}}{33}+1\right ) \int_0^{\beta}\!d\tau\!\int \ ! d^3x \ !
\left [ f^{acm}f^{bdm } ( { \partial}_ia_0^a)\!\cdot \!({\partial}_ia_0^b ) a_0^ca_0^d \right ] \ . \label{effaction}\end{aligned}\ ] ] in the dimensionally reduced theory of qcd @xcite , eq .
( [ effaction ] ) provides the leading derivative interactions between the @xmath28-fields .
how large the calculated derivative term is , compared to both the ones already present in eq .
( [ knownaction ] ) as well as the omitted higher dimension operators , depends on the scales of interest .
for instance , at the soft scale where @xmath85 and @xmath86 , we have @xmath87 , a factor @xmath88 higher than the term @xmath89 . in general , this effective theory is interesting on length scales @xmath90 in the high temperature regime , especially when combined with nonperturbative lattice methods @xcite .
it is for example possible to study the non - perturbative debye mass @xcite , and the 3d effective theory is also useful for calculations of the pressure in the quark - gluon phase , both perturbatively @xcite and nonperturbatively @xcite .
although the derived correction in eq .
( [ effaction ] ) presumably gives a minor effect only , it is somewhat interesting to note that the terms are rather sensitive to the number of quark flavors . in eq .
( [ effaction ] ) , the coefficient of the first term is proportional to @xmath91 , which goes from @xmath92 to @xmath93 between @xmath94 and @xmath95 .
similarly , the second term in eq .
( [ effaction ] ) increases by more than 250% in the same range of @xmath42 .
this should be contrasted with the constant part of the @xmath96-contribution to eq .
( [ knownaction ] ) , that depends on @xmath42 as @xmath97 and therefore only changes by 33% when going from @xmath94 to @xmath95 . due to this strong behavior of the number of quark flavors ,
it is not inconceivable that the derivative interactions will make a small but noticeable difference between e.g. the pure glue theory and the three - flavor case .
the qcd dimensionally reduced theory describes accurately static phenomena at very high @xmath0 , but the approximations break down around a few times @xmath9 @xcite .
in addition , one is by construction omitting all dynamical information . to understand the features around @xmath9 a ginzburg - landau type of effective theory was proposed in @xcite . in this model , the potential is written in terms of @xmath18 , @xmath98 \ .
\label{robpotential}\end{aligned}\ ] ] the constants @xmath99 are then used to fit the pressure above @xmath9 , with @xmath100 a function of temperature so that the global minimum of the potential is at @xmath101 ( @xmath102 ) above ( below ) @xmath9 @xcite .
one of the important aspects of the potential in eq .
( [ robpotential ] ) is the extremely rapid change around @xmath9 , due to a very sensitive dependence of @xmath100 on @xmath3 @xcite . in a dynamical scenario
one can therefore assume an instantaneous quench , where the value of @xmath18 suddenly no longer corresponds to the correct minimum .
the @xmath18-field then rolls down the potential , and by coupling the @xmath18-field to a linear sigma model the potential energy is converted into pions @xcite .
even though the model is of phenomenological origin , it thus makes predictions that can be compared to experimental results .
after the quench , the evolution of the euler - lagrange equations from the initial conditions requires , apart from the potential and the coupling to the chiral field , also a kinetic term for @xmath18 @xcite .
although the time dependence is beyond the calculation presented in this paper , we can provide the first perturbative qcd - correction to the spatial derivatives .
the leading coefficient is the classical contribution to the derivative term , and comes from the kinetic term of @xmath103 in eq .
( [ knownaction ] ) , as can be seen from the following argument @xcite : decomposing the wilson line in eq .
( [ wilsonline ] ) into an octet @xmath104 and the singlet @xmath18 , @xmath105 where @xmath104 is traceless and @xmath64 , we have @xmath106 on the other hand , by a direct calculation in the static limit , @xmath107 when the commutator terms in eq .
( [ commutators ] ) are rewritten in terms of @xmath11 they can only involve the adjoint field , or products of @xmath18 and @xmath104 , since @xmath18 ( times the identity matrix ) by itself commutes with all su(3 ) matrices .
thus , by combining eqs .
( [ decomposel ] ) and ( [ commutators ] ) , we have , @xmath108 + f(\tilde{{\bf
l } } ) \ , \label{leadingderivative}\end{aligned}\ ] ] where @xmath109 corresponds to the commutator terms , rewritten as a function of @xmath104 . at @xmath9 , @xmath110 so that the leading coefficient for the kinetic term of @xmath18 is @xmath111 , which is reasonably close to the canonical value @xmath41 .
given the unknown function @xmath109 it is not clear whether the kinetic term for the adjoint field actually is unique .
however , @xmath104 does not play any important role around @xmath9 , and can therefore be neglected on physical grounds @xcite .
the procedure to obtain the kinetic term for @xmath18 is thus to match terms in the effective theory @xmath29 to a corresponding @xmath27 piece in @xmath26 .
this means that the classical coefficient for @xmath27 will change when radiative corrections are taken into account in @xmath29 .
to lowest order , the kinetic term for @xmath28 can receive corrections from the polarization tensor @xcite , that would affect the @xmath27 term via eq .
( [ leadingderivative ] ) .
however , with the optimal choice for the counterterms @xcite there is in fact no @xmath112-contribution , so the renormalized kinetic term in @xmath29 remains @xmath113 . even though the kinetic term for @xmath28 is unchanged at one - loop
, this does not mean that @xmath27 is so .
since @xmath114 , the @xmath27-term contains at least four powers of @xmath28 when @xmath11 is expanded in powers of @xmath28 . the two - derivative term in @xmath26 can therefore receive a perturbative correction from the four - point function in eq .
( [ effactiona0 ] ) . indeed , by using the relations @xmath115 \left ( { \rm tr } |{\partial}_i \tilde{{\bf l}}|^2 + 3|{\partial}_i l|^2 \right ) + o(a_0 ^ 6 ) \ , \nonumber \\
( { \partial}_i(a_0^aa_0^a))^2 = & & \frac{144}{g^4\beta^4}|{\partial}_i l|^2 + o(a_0 ^ 6)\nonumber \\
f^{acm}f^{bdm } ( { \partial}_ia_0^a)\!\cdot \!({\partial}_ia_0^b ) a_0^ca_0^d = & & 2{\rm tr}| [ { \partial}_i \tilde{{\bf l } } , \
\tilde{{\bf l } } ] |^2 + o(a_0 ^ 6 ) \ , \end{aligned}\ ] ] we can rewrite eq .
( [ effaction ] ) as @xmath116 \right .
- \nonumber \\ & & \left .
\frac{11\xi ( 3)t^2}{768\pi^4 } \left ( \frac{28n_{\ !
f}}{33}+1\right ) { \rm tr}|[{\partial}_i \tilde{{\bf l } } , \
\tilde{{\bf l } } ] |^2 \right \ } \ . \label{gammainls}\end{aligned}\ ] ] from the calculation of the four - point function we thus find the leading perturbative correction to the kinetic term of @xmath18 . including the contribution from eq .
( [ gammainls ] ) , the kinetic term now becomes @xmath117 |{\partial}_i l|^2 = \frac{3t^2}{2g^2 } \left [ 1+c_{{\rm corr}}\right ] |{\partial}_i l|^2 \ .\end{aligned}\ ] ] since @xmath18 is dimensionless , the correct dimension of the operator has to be supplied by some other scales . in the perturbative calculation
the only scale is @xmath0 , and hence the one - loop induced coefficient has the same @xmath0-dependence as the leading term in eq .
( [ leadingderivative ] ) .
however , the perturbative correction does not depend on the qcd coupling constant and is therefore just a fixed number at this order .
for example , for three flavors the coefficient is @xmath118 .
compared to the classical contribution @xmath119 , the fraction of the one - loop correction is only @xmath120 . even at @xmath9 , this is merely of the order of 5% , and at higher @xmath0 even less due to the logarithmic decrease of @xmath31 .
having derived the first correction to the kinetic term for @xmath18 from perturbation theory , let us now discuss to what extent , and in what temperature range , the terms in eq .
( [ gammainls ] ) can be trusted .
first of all , it should be noted that the form of the effective action in eq .
( [ gammainls ] ) is not completely unique .
the reason is that by a partial integration , and discarding any surface terms , we can always trade factors of @xmath121 $ ] and @xmath122 for a term like @xmath123 $ ] , but this equality does not hold at the level of @xmath18 .
in fact , if a term @xmath123 $ ] is kept in the action , not only do the coefficients in eq .
( [ gammainls ] ) change , but there are also additional nonequivalent terms of the form @xmath124 and @xmath125 $ ] . nevertheless , the action in eq .
( [ gammainls ] ) is of course unique to order @xmath126 . since higher order operators in the dimensionally reduced theory are further suppressed at high @xmath0 , the predictions in eq .
( [ gammainls ] ) should at the very least be reliable down to @xmath127 , i.e. when the effective 3d theory itself is applicable .
when @xmath128 , the question is admittedly more subtle , as higher loop effects , higher dimensional operators and possibly nonperturbative effects become important .
however , eq . ( [ gammainls ] )
does not have to break down completely when the 3d theory does so .
the 3d theory becomes invalid because the procedure of integrating out the non - static modes is unreliable when @xmath129 @xcite .
in contrast , @xmath26 is by construction valid near @xmath9 , so the question is rather how much the coefficient for the spatial derivative term changes .
considering first the operators of higher dimensionality , it is certainly possible to imagine that their bulk part follows from an expansion of eq .
( [ gammainls ] ) .
the additional contributions that do not originate from these sources , e.g. terms like @xmath130 , that are at least of order @xmath131 ( with @xmath132 ) , would then be suppressed .
not because they are unimportant a priori , but because their numerical coefficients are small .
there are also higher derivative terms not accounted for in eq .
( [ gammainls ] ) , like @xmath133 , but they do not affect the kinetic term . to @xmath126 , they are in fact straightforward to obtain from eq .
( [ fermionloop ] ) and eqs .
( [ ym1])-([ym4 ] ) .
when it comes to higher loop and nonperturbative effects , they will naturally induce a @xmath0-dependence in the radiative corrections to the kinetic term .
thus , @xmath134 , which follows partly from the running of the qcd coupling constant in the higher loop effects .
how much this will affect the kinetic term is difficult to estimate , but the small correction from the four - point function may indicate that the perturbative qcd - contributions are not too important for constructing @xmath26 .
in this paper we calculated the leading momentum dependence of the four - point function in qcd with @xmath42 massless flavors , and related this contribution to terms in both the effective action @xmath29 and @xmath26 . as for the derivative terms in @xmath29 , they will only have a minor influence when @xmath135 .
as the temperature decreases and approaches @xmath9 the 3d theory becomes less reliable , but there could very well be a temperature region where the effective theory is still valid and the derivative interactions nonnegligible . since the contribution has a rather strong dependence on @xmath42 , a difference between the pure glue theory and e.g. @xmath95 qcd could perhaps be noticed .
from the four - point function , we also found the lowest one - loop qcd correction to the spatial derivative term in the effective theory @xmath26 .
the coefficient is independent of @xmath31 and much smaller than the classical term , the ratio between the two being of the order of @xmath136 at @xmath9 .
this derivation assumes that the strange quark mass @xmath137 can be neglected even at @xmath9 , which of course is an oversimplification , given that @xmath138 .
nevertheless , the influence of @xmath137 is not likely to change the fact that the correction is small even at @xmath9 . at one - loop
, there is an infinite number of terms that contribute to the coefficient of @xmath27 , to the same order in @xmath31 and @xmath0 , as the four - point function .
this follows from the fact the induced two - derivative interactions , with @xmath139 external fields @xmath28 , is of the functional form @xmath140 , which corresponds to an expansion of @xmath18 to at most order @xmath141 in the term @xmath142 .
some of these higher - dimensional contributions , maybe even the major parts , are already accounted for by rewriting the four - point function in @xmath29 in terms of @xmath18 , as in eq .
( [ gammainls ] ) . in any case , since the four - point contribution is very small , it is reasonable to assume that the higher @xmath139-point functions give even smaller corrections . in that case , eq .
( [ gammainls ] ) should give the correct order of magnitude for the total one - loop correction . as mentioned earlier , there are also higher loop effects that contribute to the kinetic term in @xmath26 . for example , taking into account the two - loop correction to the diagrams in fig
. 1 gives @xmath143 $ ] .
to understand the reliability of the canonical term it is then crucial to know the magnitude of @xmath144 .
surprisingly , studies of higher loop effects in the 3d effective theory @xmath29 indicate that they only give corrections of the order of 30% at @xmath9 @xcite .
if these conjectures can be taken over to the wilson line model , one could in fact expect the derivative term @xmath145 to change by perhaps at most a factor two , with all qcd - corrections taken into account .
of course , this has to be regarded as a highly speculative suggestion at the present stage . to complete the dynamical scenario
one also needs the time dependence of @xmath18 .
unfortunately , it is yet unclear how @xmath18 generalizes to a real time formulation @xcite . assuming that the form of the spatial derivatives can be extended to a lorentz invariant form , the predictions for pion production and the evolution of @xmath146 will remain almost unchanged @xcite . in particular ,
if the lorentz invariant kinetic term does not change by more than a factor of two , it can easily be compensated by a difference in e.g. the expansion rate of the plasma . finally , to obtain a decisive estimate of the qcd - effects in @xmath26
, one has to establish either a unique mapping from @xmath29 to @xmath26 , or find a way to determine @xmath26 directly , perhaps numerically .
hopefully , the calculations presented in this paper can serve as a first step in that direction .
* acknowledgments * the author thanks r. d. pisarski for discussions that initialized this project , and for reading the manuscript .
d. bdeker , a. dumitru and r. d. pisarski are greatfully acknowledged for useful and interesting discussions during the investigations , and k. kajantie and m. laine for useful comments on an early version of the manuscript .
this work was supported by the swedish foundation for international cooperation in research and higher education ( stint ) under contract 99/665 , and in part by doe grant de - ac02 - 98ch10886 .
in this appendix we outline our method for calculating the feynman diagrams shown in fig . 1 .
we first follow the feynman rules given in @xcite for minkowski space - time .
after all contractions and traces over spinor indices have been performed , we continue to a euclidean space compact in the imaginary time direction : @xmath147 where @xmath148 is the matsubara frequency , @xmath149 ( @xmath150 ) for fermions ( bosons ) .
next , we use a feynman parametrization to combine the denominators in the loop integral , and extract the @xmath74 part from the following relations @xcite : @xmath151 \ \ { \rm ( bosons ) } \cr & \mbox { } \cr & \int_{\delta - i\infty}^{\delta+i\infty}\frac{dp_0}{2\pi i } \ , n_f \ , [ f(p_0)+f(-p_0 ) ] \ \ { \rm ( fermions ) } } \right .
\ .\end{aligned}\ ] ] where @xmath152^{-1}$ ] , @xmath153^{-1}$ ] and @xmath154 .
the ghosts follow bose statistics , despite their anticommuting properties . by shifting the vector momentum in the loop , @xmath155 , to @xmath156 , with @xmath157 a linear function of the external momenta @xmath158
, we then integrate over @xmath159 .
finally , by performing a mellin transform , @xmath160 where @xmath161 , @xmath162 , and with the contour specified by @xmath163 , @xmath164 ( where @xmath165 ) , we can integrate over @xmath166 after a change of variables . to illustrate the above procedure , consider the diagram in fig .
1(e ) . omitting the color factors for simplicity , and using the notation @xmath167 , @xmath168
, we have to calculate the following integral : @xmath169 ^ 2 } \nonumber \\ & & \ \ \ \ = \frac{3ig^4}{2\pi^2}\int_0 ^ 1dy_1 \int_{\delta - i\infty}^{\delta+i\infty}\frac{dp_0}{2\pi i } n_b \int_0^{\infty}dq \frac{q^2}{[q^2+g_8-p_0 ^ 2]^2 } \ , \label{illustration}\end{aligned}\ ] ] where we made the shift @xmath170 and defined @xmath171 in the last integral . after performing the @xmath172-integral in eq .
( [ illustration ] ) , we are left with , @xmath173\beta^{-z } \ ! \int_0^{\infty}\ ! du \frac{u^{-z}}{\sqrt{g_8+u^2 } } \nonumber \\ & & \ \ \ = \frac{3ig^4}{16\pi^{5/2 } } \int_0 ^ 1dy_1 \int_{1+\epsilon -i\infty}^{1+\epsilon + i\infty } \frac{dz}{2\pi i}\gamma ( z ) \xi ( z ) \beta^{-z } \cos [ \pi z/2 ] g_8^{-z/2 } \gamma ( ( 1-z)/2 ) \gamma ( z/2 ) \ , \end{aligned}\ ] ] which is eq .
( [ ym4 ] ) , except for the color factors . to check our method we also calculated the @xmath174 contribution to the above diagram in a different way : we first performed the @xmath166-integral , by picking up the poles in the complex @xmath166-plane , and then did the @xmath175-integral , without any feynman parametrization , by expanding the integrand differently in different integration regions .
the two results of course agree with each other . for completeness , we also give the functions @xmath176 in eq .
( [ fermionloop ] ) and @xmath177 in eqs .
( [ ym1])-([ym4 ] ) . using the conservation of momentum , and the shorthand notation @xmath178 , their explicit forms are as follows :
@xmath179 ^ 2 \\ f_2 = & & 3k_1 ^ 2 + 4k_1k_2+k_2 ^ 2 + 2k_1k_3+k_2k_3 - 3(3k_1 + 2k_2+k_3)[y_2k_1+y_3k_{12}+(1-y_1-y_2-y_3)k_{123}]+\nonumber \\ & & + 6[y_2k_1+y_3k_{12}+(1-y_1-y_2-y_3)k_{123}]^2 \\ f_3 = & & ( 1/3)\left \ { 5k_1 ^ 2y_1(2y_1 - 1)+k_2 ^
2[3 + 10y_1 ^ 2 - 10y_2 + 10y_2 ^ 2 + 10y_1(2y_2 - 1)]+ + 5k_3 ^ 2[1 + 2y_1 ^ 2 + 2y_2 ^ 2 - 3y_3 + 2y_3 ^ 2+\right .
\nonumber \\ & & + y_2(4y_3 - 3)+y_1(4y_2 + 4y_3 - 3)]+k_2k_3[8 + 20y_1 ^ 2 + 20y_2 ^ 2 - 10y_3 + 5y_2(4y_3 - 5)+5y_1(8y_2 + 4y_3 - 5 ) ] + \nonumber \\ & & \left . k_1[k_2(3 + 20y_1 ^ 2 - 5y_2 + 5y_1(4y_2 - 3))+k_3(2 + 20y_1 ^ 2 - 5y_2 - 5y_3 + 20y_1(y_2+y_3 - 1 ) ) ] \right
f_4 = & & \left \ { [ ( 1-y_2)k_1-y_3k_{12}-(1-y_1-y_2-y_3)k_{123}]\cdot [ -y_2k_1+(1-z)k_{12}-(1-y_1-y_2-y_3)k_{123}]\right \ } \times \nonumber \\ & & \left \ { [ -y_2k_1-y_3k_{12}-(1-y_1-y_2-y_3)k_{123}]\cdot [ -y_2k_1-y_3k_{12}+(y_1+y_2+y_3)k_{123}]\right \ } - \nonumber \\ & & -\left \ { [ -y_2k_1-y_3k_{12}-(1-y_1-y_2-y_3)k_{123}]\cdot [ -y_2k_1+(1-z)k_{12}-(1-y_1-y_2-y_3)k_{123 } ] \right \ } \times \nonumber \\ & & \left \ { [ ( 1-y_2)k_1-y_3k_{12}-(1-y_1-y_2-y_3)k_{123}]\cdot [ -y_2k_1-y_3k_{12}+(y_1+y_2+y_3)k_{123}]\right \}+ \nonumber \\ & & + \left \ { [ -y_2k_1-y_3k_{12}-(1-y_1-y_2-y_3)k_{123}]\cdot [ ( 1-y_2)k_1-y_3k_{12}-(1-y_1-y_2-y_3)k_{123}]\right \ } \times \nonumber \\ & & \left \ { [ -y_2k_1+(1-z)k_{12}-(1-y_1-y_2-y_3)k_{123}]\cdot [ -y_2k_1-y_3k_{12}+(y_1+y_2+y_3)k_{123}]\right \ } \\ g_1 = & & f_1 \\ g_2 = & & f_1 \\
g_3 = & & 3k_1 ^ 2 + 2k_2 ^ 2 + 2k_3 ^ 2 + 3k_1k_2 + 3k_1k_3 + 2k_2k_3 + 2[(1-y_1)k_1+(1-y_1-y_2)k_2+(1-y_1-y_2-y_3)k_3]^2 - \nonumber \\ & & - ( 3k_1 + 2k_2+k_3)[(1-y_1)k_1+(1-y_1-y_2)k_2+(1-y_1-y_2-y_3)k_3 ] \\
g_4 = & & ( 1/3)\left \ { k_1 ^ 2(10y_1 - 20y_1 ^ 2 - 7)-2k_2 ^ 2[10y_1 ^ 2 - 10y_2 + 10y_2 ^ 2 + 10y_1(2y_2 - 1)-3 ] - \right .
\nonumber \\ & & - k_3 ^ 2[17 + 20y_1 ^ 2 + 20y_2 ^ 2 - 30y_3 + 20y_3 ^ 2 + 10y_2(4y_3 - 3)+10y_1(4y_2 + 4y_3 - 3 ) ] - \nonumber \\ & & -2k_2k_3[2 + 20y_1 ^ 2 + 20y_2 ^ 2 - 10y_3 + 5y_2(4y_3 - 5)+5y_1(8y_2 + 4y_3 - 5 ) ] - \nonumber \\ & & \left .
-2k_1[k_2(20y_1 ^ 2 - 5y_2 + 5y_1(4y_2 - 3)-3)+5k_3(3 + 4y_1 ^ 2-y_2-y_3 + 4y_1(y_2+y_3 - 1 ) ) ] \right \ } \\
g_5 = & & \left \ { [ y_1k_1+(1+y_1+y_2)k_2-(1-y_1-y_2-y_3)k_3]\cdot [ ( 2-y_1)k_1+(1-y_1-y_2)k_2+(1-y_1-y_2-y_3)k_3 ] \right \ }
\times \nonumber \\ & & \left \ { [ -y_1k_1-(y_1+y_2)k_2+(2-y_1-y_2-y_3)k_3]\cdot [ ( 2-y_1)k_1+(2-y_1-y_2)k_2 + ( 2-y_1-y_2-y_3)k_3 ] \right \ } + \nonumber \\ & & \left \ { [ ( 1+y_1)k_1-(1-y_1-y_2)k_2-(1-y_1-y_2-y_3)k_3]\cdot [ ( 1+y_1)k_1+(1+y_1+y_2)k_2+(1+y_1+y_2+y_3)k_3 ] \right \ } \times \nonumber \\ & & \left \ { [ -y_1k_1+(2-y_1-y_2)k_2+(1-y_1-y_2-y_3)k_3]\cdot [ y_1k_1+(y_1+y_2)k_2+(1+y_1+y_2+y_3)k_3 ] \right \ } \\ g_6 = & & y_1k_{12}^2+y_2k_{123}^2+(1-y_1-y_2)k_{1234}^2 -[k_1+k_2+(1-y_1)k_3+(1-y_1-y_2)k_4]^2 \label{eqg6 } \\ g_7 = & & [ ( 1+y_1)k_3-(1-y_1-y_2)k_4][(2-y_1-y_2)k_4-y_1k_3 ] \ , \label{eqg7}\end{aligned}\ ] ] where , in order to simplify the permutations of the triangle graph , we did not use @xmath180 in eqs .
( [ eqg6 ] ) and ( [ eqg7 ] ) . finally , @xmath181 | we calculate the time independent four - point function in high temperature ( @xmath0 ) qcd and obtain the leading momentum dependent terms . furthermore , we relate these derivative interactions to derivative terms in a recently proposed finite @xmath0 effective action based on the su(3 ) wilson line and its trace , the polyakov loop . by this procedure
we thus obtain a perturbative matching at finite @xmath0 between qcd and the effective model .
in particular , we calculate the leading perturbative qcd - correction to the kinetic term for the polyakov loop . |
the copious production of @xmath4 quarks in @xmath5 decays at the large electron - positron collider ( lep ) and in 1.8-tev proton - antiproton collisions at the fermilab tevatron opens for study the rich spectroscopy of mesons and baryons beyond @xmath6 and @xmath7 .
in addition to @xmath8 and @xmath9 , which have already been widely discussed , a particularly interesting case is the spectrum of @xmath2 states and its ground state , the @xmath10 meson @xcite .
even more than their counterparts in the @xmath11 and @xmath12 families , the @xmath2 states that lie below the ( @xmath13 ) threshold for decay into a pair of heavy - flavored mesons are stable against strong decay , for they can not annihilate into gluons .
their allowed decays , by e1 or m1 transitions or by hadronic cascades , lead to total widths that are less than a few hundred kev .
all decay chains ultimately reach the @xmath14s@xmath15 ground state @xmath16 , which decays weakly .
it may be possible , in time , to map out the excitation spectrum by observing photons or light hadrons in coincidence with a prominent decay of the @xmath16 @xcite .
this would test our understanding of the force between heavy quarks .
the weak decays of the @xmath2 ground state will be of particular interest because the influence of the strong interaction can be estimated reliably @xcite .
the deep binding of the heavy quarks within the @xmath16 means that the spectator picture is misleading . taking proper account of binding energy
, we expect a rather long lifetime that implies easily observable secondary vertices .
the deep binding also affects the @xmath16 branching fractions and leads us to expect that final states involving @xmath17 will be prominent .
the modes @xmath18 , @xmath19 , @xmath20 , @xmath21 , and @xmath22 will serve to identify @xmath16 mesons and determine the @xmath16 mass and lifetime . in this article , we present a comprehensive portrait of the spectroscopy of the @xmath16 meson and its long - lived excited states . in section
ii , we estimate the mass of the @xmath16 in the framework of nonrelativistic quarkonium quantum mechanics and calculate the spectrum of @xmath2 states in detail . in section iii
, we compute rates for the prominent radiative decays of the excited states and estimate rates and spectra of the hadronic cascades @xmath23 and @xmath24 . using this information
, we outline a strategy for partially reconstructing the @xmath2 spectrum .
a brief summary appears in section iv .
both in mass and in size , the mesons with beauty and charm are intermediate between the @xmath0 and @xmath1 states .
estimates of the @xmath16 mass can , consequently , be tied to what is known about the charmonium and @xmath12 families .
to predict the full spectrum and properties of @xmath2 states , we rely on the nonrelativistic potential - model description of quarkonium levels . the interquark potential is known rather accurately in the region of space important for the @xmath11 and @xmath12 families @xcite , which spans the distances important for @xmath2 levels .
this region lies between the short - distance coulombic and long - distance linear behavior expected in qcd .
we consider four functional forms for the potential that give reasonable accounts of the @xmath0 and @xmath1 spectra : the qcd - motivated potential @xcite given by buchmller and tye @xcite , with @xmath26 a power - law potential @xcite , @xmath27 with @xmath28 a logarithmic potential @xcite , @xmath29 with @xmath30 and a coulomb - plus - linear potential ( the `` cornell potential '' ) @xcite , @xmath31 with @xmath32 we solve the schrdinger equation for each of the potentials to determine the position of the 1s center of gravity for @xmath0 , @xmath2 , and @xmath1 . the @xmath33s@xmath34
@xmath14s@xmath15 splitting of the @xmath35 ground state is given by @xmath36 the hyperfine splitting observed in the charmonium family @xcite , @xmath37 fixes the strong coupling constant for each potential .
we neglect the variation of @xmath38 with momentum and scale the splitting of @xmath2 and @xmath1 from the charmonium value ( [ psisplit ] ) .
the resulting values of vector and pseudoscalar masses are presented in table [ mass1s ] .
predictions for the @xmath2 ground - state masses depend little on the potential .
the @xmath16 and @xmath39 masses and splitting lie within the ranges quoted by kwong and rosner @xcite in their survey of techniques for estimating the masses of the @xmath2 ground state .
they find @xmath40 and @xmath41 with @xmath42 we take @xmath43 as our best guess for the interval in which @xmath16 will be found @xcite .
we shall adopt the buchmller - tye potential @xcite for the detailed calculations that follow , because it has the correct two - loop short - distance behavior in perturbative qcd .
the interaction energies of a heavy quark - antiquark system probe the basic dynamics of the strong interaction . the gross structure of the quarkonium spectrum reflects the shape of the interquark potential . in the absence of light quarks , the static energy explicitly exhibits linear confinement at large distance .
further insight can be obtained by studying the spin - dependent forces , which distinguish the electric and magnetic parts of the interactions . within the framework of quantum chromodynamics , the nature of the spin - dependent forces
was first studied nonperturbatively by eichten and feinberg @xcite .
gromes @xcite subsequently added an important constraint that arises from boost - invariance of the qcd forms @xcite .
one - loop perturbative qcd calculations for the spin - dependent interactions in a meson composed of two different heavy quarks have also been carried out @xcite .
the spin - dependent contributions to the @xmath2 masses may be written as @xmath44 where the individual terms are @xmath45 and the tensor operator is @xmath46\;\;. \label{sij}\ ] ] in eq .
( [ t1to4 ] ) and ( [ sij ] ) , @xmath47 and @xmath48 are the spins of the heavy quarks , @xmath49 is the orbital angular momentum of quark and antiquark in the bound state , and @xmath50 is an arbitrary unit vector .
the total spin is @xmath51 .
the leading contributions to the @xmath52 have no explicit dependence on the quark masses . assuming that the magnetic interactions are short - range ( @xmath53 ) and thus can be calculated in perturbation theory
, we have @xmath54 the connection between @xmath55 and @xmath56 is gromes s general relation ; the other equations reflect the stated approximations . for quarkonium systems composed of equal - mass heavy quarks ,
the total spin @xmath57 is a good quantum number and @xmath58 coupling leads to the familiar classification of states as @xmath59 , where @xmath60 @xcite .
the calculated spectra are compared with experiment in table [ charmonium ] ( for the @xmath17 family ) and table [ upsilons ] ( for the @xmath12 family ) .
overall , the agreement is satisfactory .
typical deviations in the charmonium system are less than about @xmath61 ; deviations in the upsilon system are somewhat smaller .
the differences between calculated and observed spectra suggest that the excitation energies in the @xmath62 system can be predicted within a few tens of mev .
the leptonic decay rate of a neutral ( @xmath63 ) vector meson @xmath64 is related to the schrdinger wave function through @xcite @xmath65 where @xmath66 is the number of quark colors , @xmath67 is the heavy - quark charge , and @xmath68 is the mass of the vector meson .
the resulting leptonic widths , evaluated without qcd corrections , are tabulated in tables [ charmonium ] and [ upsilons ] . within each family ,
the leptonic widths are predicted in proper proportions , but are larger than the observed values .
the qcd correction reduces the magnitudes significantly ; the amount of this reduction is somewhat uncertain , because the first term in the perturbation expansion is large @xcite . for unequal - mass quarks ,
it is more convenient to construct the mass eigenstates by @xmath69 coupling , first coupling @xmath70 and then adding the spin of the heavier quark , @xmath71 . the level shifts @xmath72 for the @xmath73 states with @xmath74 and @xmath75 are @xmath76 for a given principal quantum number , the two @xmath77 @xmath2 states with @xmath78 and @xmath79 are mixed in general .
the elements of the mixing matrix are @xmath80 @xmath81 @xmath82 two limiting cases are familiar .
\(i ) with equal quark masses @xmath83 , the level shifts become @xmath84 while the mixing matrix becomes @xmath85 the mass eigenstates are the familiar @xmath14p@xmath34 and @xmath33p@xmath34 states of the @xmath58 coupling scheme . in this basis , they may be written as @xmath86 with eigenvalues @xmath87 the position of the @xmath88 level coincides with the centroid @xmath89/9 $ ] of the @xmath90 levels .
\(ii ) in the heavy - quark limit , @xmath91 , the level shifts of the @xmath92 levels become @xmath93 while the mixing matrix becomes @xmath94 the @xmath95 and @xmath78 states separate into degenerate pairs , as expected on the basis of heavy - quark symmetry @xcite . in the @xmath2 system , we label the mass eigenstates obtained by diagonalizing the matrix ( [ unequal ] ) as @xmath96 and @xmath97 . for the 2p@xmath34 levels ,
the mixing matrix is @xmath98 with eigenvectors @xmath99 for the 3p@xmath34 levels , the mixing matrix is @xmath100 with eigenvectors @xmath101 for the 4p@xmath34 levels , the mixing matrix is @xmath102 with eigenvectors @xmath103 the calculated spectrum of @xmath2 states is presented in table [ bcbar ] and figure [ spectre ] .
our spectrum is similar to others calculated by eichten and feinberg @xcite in the cornell potential @xcite , by gershten et al .
@xcite in the power - law potential ( [ martin ] ) , and by chen and kuang @xcite in their own version of a qcd - inspired potential .
levels that lie below the @xmath13 flavor threshold , i.e. , with @xmath104 , will be stable against fission into heavy - light mesons . for quarks bound in a central potential , it is convenient to separate the schrdinger wave function into radial and angular pieces , as @xmath105 where @xmath106 is the principal quantum number , @xmath107 and @xmath108 are the orbital angular momentum and its projection , @xmath109 is the radial wave function , and @xmath110 is a spherical harmonic @xcite .
the schrdinger wave function is normalized , @xmath111 so that @xmath112 the value of the radial wave function , or its first nonvanishing derivative at the origin , @xmath113 is required to evaluate pseudoscalar decay constants and production rates through heavy - quark fragmentation @xcite .
the quantity @xmath114 is presented for four potentials in table [ psizero ] . the stronger singularity of the cornell potential is reflected in spatially smaller states . the pseudoscalar decay constant @xmath115 , which will be required for the discussion of annihilation decays @xmath116 , is defined by @xmath117 where @xmath118 is the axial - vector part of the charged weak current , @xmath119 is an element of the cabibbo - kobayashi - maskawa quark - mixing matrix , and @xmath120 is the four - momentum of the @xmath16 .
the pseudoscalar decay constant is related to the ground - state @xmath2 wave function at the origin by the van royen - weisskopf formula @xcite modified for color , @xmath121 in the nonrelativistic potential models we have considered to estimate @xmath122 and @xmath123 , we find @xmath124 even after qcd radiative corrections of the size suggested by the comparison of computed and observed leptonic widths for @xmath11 and @xmath12 , @xmath115 will be significantly larger than the pion decay constant , @xmath125 @xcite .
the compact size of the @xmath62 system enhances the importance of annihilation decays .
as in atomic physics , it is the spectral lines produced in cascades from excited states to the readily observable @xmath16 ground state that will reveal the @xmath62 level scheme . as in the @xmath11 and @xmath12 quarkonium families , the transitions are
mostly radiative decays .
a few hadronic cascades , analogs of the @xmath126 transition first observed in charmonium , will also be observable . except for the magnetic - dipole ( spin - flip ) transition between the ground - state @xmath39 and @xmath16 ,
only the electric dipole transitions are important for mapping the @xmath62 spectrum .
the strength of the electric - dipole transitions is governed by the size of the radiator and the charges of the constituent quarks .
the e1 transition rate is given by @xmath127 where the mean charge is @xmath128 @xmath129 is the photon energy , and the statistical factor @xmath130 is as defined by eichten and gottfried @xcite .
@xmath131 for @xmath33s@xmath132p@xmath133 transitions and @xmath134 for allowed e1 transitions between spin - singlet states .
the statistical factors for @xmath135-wave to @xmath136-wave transitions are reproduced in table [ stat ] for convenience .
the e1 transition rates and photon energies in the @xmath62 system are presented in table [ e1rates ] .
the only decay mode for the 1@xmath33s@xmath34 @xmath137 state is the magnetic dipole transition to the ground state , @xmath16 .
the m1 rate for transitions between @xmath138-wave levels is given by @xmath139 where the magnetic dipole moment is @xmath140 and @xmath129 is the photon energy .
rates for the allowed and hindered m1 transitions between spin - triplet and spin - singlet @xmath138-wave @xmath2 states are given in table [ m1rates ] .
the m1 transitions contribute little to the total widths of the 2s levels .
because it can not decay by annihilation , the 1@xmath33s@xmath34 @xmath2 level , with a total width of 135 ev , is far more stable than its counterparts in the @xmath0 and @xmath1 systems , whose total widths are @xmath141 kev and @xmath142 kev , respectively @xcite .
a hadronic transition between quarkonium levels can be understood as a two - step process in which gluons first are emitted from the heavy quarks and then recombine into light hadrons .
perturbative qcd is not directly applicable , because the energy available to the light hadrons is small and the emitted gluons are soft . nevertheless , the final quarkonium state is small compared to the system of light hadrons and moves nonrelativistically in the rest frame of the decaying quarkonium state .
a multipole expansion of the color gauge field converges rapidly and leads to selection rules , a wigner - eckart theorem , and rate estimates for hadronic transitions @xcite .
the recombination of gluons into light hadrons involves the full strong dynamics and can only be modeled .
the general structure of hadronic - cascade transitions and models for the recombination of gluons into light hadrons can be found in a series of papers by yan and collaborators @xcite .
the hadronic transition rates for an unequal - mass @xmath143 system differ in some details from the rates for an equal - mass @xmath63 system with the same reduced mass .
the relative strengths of various terms that contribute to magnetic - multipole transitions are modified because of the unequal quark and antiquark masses .
the electric - multipole transitions are only sensitive to the relative position of the quark and antiquark and will be unchanged in form . as in the @xmath144 and @xmath145 systems , the principal hadronic transitions in the @xmath62 system involve the emission of two pions .
electric - dipole contributions dominate in these transitions , and so the equal - mass results apply directly .
the initial quarkonium state is characterized by its total angular momentum @xmath146 with @xmath147-component @xmath148 , orbital angular momentum @xmath149 , spin @xmath150 , and other quantum numbers collectively labelled by @xmath151 .
the corresponding quantum numbers of the final quarkonium state are denoted by the unprimed symbols . since the transition operator is spin - independent , the initial and final spins are the same : @xmath152 . because the gauge - field operators in the transition amplitude do not depend on the heavy - quark variables ,
the transition operator is a reducible second - rank tensor , which may be decomposed into a sum of irreducible tensors with rank @xmath153 .
the differential rate @xcite for the e1e1 transition from the initial quarkonium state @xmath154 to the final quarkonium state @xmath155 and a system of @xmath106 light hadrons , denoted @xmath156 , is given by @xmath157 where @xmath158 is the invariant mass squared of the light hadron system , @xmath159 is a 6-@xmath160 symbol , and @xmath161 is the contribution of the irreducible tensor with rank @xmath129 .
the wigner - eckart theorem ( [ wet ] ) yields the relations among two - pion transition rates given in table [ hadi ] .
the magnitudes of the @xmath161 are model - dependent .
since the @xmath162 contributions are suppressed in the soft - pion limit @xcite , we will set @xmath163 .
for some of the remaining rates we can use simple scaling arguments from the measured rates in @xmath164 systems @xcite .
the amplitude for an e1e1 transition depends quadratically on the interquark separation , so the scaling law between a @xmath143 and the corresponding @xmath164 system states is given by @xcite : @xmath165 up to possible differences in phase space .
the measured values for the @xmath166 , @xmath167 , and @xmath168 transition rates allow good scaling estimates for the @xmath169 and @xmath170 transitions in the @xmath62 system .
we have estimated the remaining transition rates by scaling the @xmath171 rates calculated by kuang and yan @xcite in their model c , which is based on the buchmller - tye potential @xcite .
the results are shown in table [ hadii ] .
chiral symmetry leads to a universal form for the normalized dipion spectrum @xcite , @xmath172 where @xmath173 and @xmath174 is the three - momentum carried by the pion pair . the normalized invariant - mass distribution for the transition @xmath175 is shown in figure [ hadfig ] for the @xmath144 , @xmath62 , and @xmath145 families .
the soft - pion expression ( [ bandc ] ) describes the depletion of the dipion spectrum at low invariant masses observed in the transitions @xmath176 @xcite and @xmath177 @xcite , but fails to account for the @xmath178 and @xmath179 spectra @xcite .
we expect the 3s levels to lie above flavor threshold in the @xmath62 system . by the wigner - eckart theorem embodied in eq .
( [ wet ] ) , the invariant mass spectrum in the decay @xmath180 should have the same form ( [ bandc ] ) as the @xmath181 transition .
braaten , cheung , and yuan @xcite have calculated the probability for a high - energy @xmath182 antiquark to fragment into the @xmath62 @xmath138-waves as @xmath183 , @xmath184 , @xmath185 , and @xmath186 .
given the excellent experimental signatures for @xmath187 decay and the favorable prospects for @xmath188 production in high - energy proton - antiproton collisions , it may be possible to observe the @xmath189 transition for the first time in the @xmath16 family .
the 2@xmath33s@xmath190 1@xmath33s@xmath191 transition has been observed in charmonium .
this transition proceeds via an m1m1 or e1m2 multipole . in the @xmath62 system the e1m2 multipole dominates and
the scaling from the @xmath144 system should be given by @xmath192^{1/2 } [ m_{\phi^\prime}^2-(m_\phi - m_\eta)^2]^{1/2 } } { [ m_{\psi^\prime}^2-(m_\psi+m_\eta)^2]^{1/2 } [ m_{\psi^\prime}^2-(m_\psi - m_\eta)^2]^{1/2}}\;\;\ ; , \ ] ] where @xmath193 and @xmath194 are the masses of the 2@xmath33s@xmath34 and 1@xmath33s@xmath34 @xmath62 levels , respectively .
because of the small energy release in this transition , the slightly smaller level spacing in the @xmath16 family compared to the @xmath11 family ( 562 mev _ vs. _ 589 mev ) strongly suppresses @xmath195-emission in the @xmath62 system . the observed rate of @xmath196 @xcite scales to @xmath197 .
the total widths and branching fractions are given in table [ totalwidth ] .
the most striking feature of the @xmath62 spectrum is the extreme narrowness of the states . a crucial element in unraveling the spectrum will be the efficient detection of the 72-mev m1-photon that , in coincidence with an observed @xmath16 decay , tags the @xmath39 .
this will be essential for distinguishing the @xmath198 transition from @xmath181 , which will have a nearly identical spectrum and a comparable rate . combining the branching fractions in table [ totalwidth ] with the @xmath4-quark fragmentation probabilities of ref .
@xcite , we expect the cross section times branching fractions to be in the proportions @xmath199 a reasonable but challenging experimental goal would be to map the eight lowest - lying @xmath62 states : the 1s , 2s , and 2p levels . a first step , in addition to reconstructing the hadronic cascades we have just discussed , would be the detection of the 455-mev photons in coincidence with @xmath16 , and of 353- , 382- , and 397-mev photons in coincidence with @xmath200 .
this would be a most impressive triumph of experimental art .
a meson with beauty and charm is an exotic particle , but prospects are good that it will be discovered in the near future .
as soon as @xmath16 has been identified , the investigation of competing weak - decay mechanisms , @xmath201 ( represented by @xmath202 , @xmath203 , etc . ) , @xmath204 ( represented by @xmath205 , @xmath206 , etc . ) , and @xmath207 ( represented by @xmath21 , @xmath208 , etc . ) , can begin .
the issues to be studied , and predictions for a wide variety of inclusive and exclusive decays , are presented in a companion paper @xcite . before the end of the decade
, it should prove possible to map out part of the @xmath62 spectrum by observing @xmath209- and @xmath210-coincidences with the ground - state @xmath16 or its hyperfine partner @xmath39 .
fermilab is operated by universities research association , inc .
, under contract de - ac02 - 76cho3000 with the united states department of energy .
thanks the cultural section of the vienna municipal government and members of the institute for theoretical physics of the university of vienna for their warm hospitality while part of this work was carried out .
the cdf collaboration has reconstructed the @xmath212 states by observing photons in coincidence with leptonic decays of @xmath11 .
see f. abe _
( cdf collaboration ) , phys .
* 71 , * 2537 ( 1993 ) .
m. baker , j. s. ball , and f. zachariasen , phys .
d * 45 , * 910 ( 1992 ) , use a dual - qcd potential to calculate @xmath213 and @xmath214 .
we note that their prediction for the @xmath215 1s level lies at @xmath216 , @xmath217 above the observed value .
r. roncaglia , a. r. dzierba , d. b. lichtenberg , and e. predazzi , indiana university preprint iuhet 270 ( january 1994 , unpublished ) , use the feynman - hellman theorem to predict @xmath218 .
s. godfrey and n. isgur , phys .
d * 32 , * 189 ( 1985 ) , estimate @xmath219 and @xmath220 .
general reviews of the eichten - feinberg - gromes ( efg ) formulation have been given by m. peskin , in _ dynamics and spectroscopy at high energy , _ proceedings of the eleventh slac summer institute on particle physics , slac report no .
267 , edited by p. m. mcdonough ( stanford linear accelerator center , stanford , ca , 1983 ) , p. 151 ; e. j. eichten , in _ the sixth quark , _ proceedings of the twelfth slac summer institute on particle physics , slac report no .
281 , edited by p. m. mcdonough ( stanford linear accelerator center , stanford , ca , 1985 ) , p. 1 . in eq .
( 2.13 ) , @xmath221 should read @xmath222 ; d. gromes , in _ spectroscopy of light and heavy quarks , _ edited by ugo gastaldi , robert klapisch , and frank close ( plenum press , new york and london , 1987 ) , p. 67 ; d. gromes , in _ the quark structure of matter , _ proceedings of the yukon advanced study institute , edited by n. isgur , g. karl , and p. j. odonnell ( world scientific , singapore , 1985 ) , p. 1 . for the unequal - mass case relevant to the @xmath62 system , y. j. ng , j. pantaleone , and s .- h .
henry tye , phys .
* 55 , * 916 ( 1985 ) [ see also j. pantaleone , s .- h .
h. tye , and y. j. ng , phys .
d * 33 , * 777 ( 1986 ) ] , introduced a new , spin - dependent form factor .
however , these changes can be absorbed into the efg formalism without introducing any new form factors by allowing the existing form factor to depend ( logarithmically ) on the heavy - quark masses .
recently , yu - qi chen and yu - ping kuang , `` general relations of heavy quark - antiquark potentials induced by reparametrization invariance , '' china center of advanced science and technology ( world laboratory ) preprint ccast-93 - 37 ( unpublished ) , extended the gromes analysis @xcite to show in general that no new spin - dependent structures appear to order @xmath223 . the expectation values of spin and orbital angular momentum operators are conveniently evaluated using @xmath224 , @xmath225 , @xmath226 $ ] .
the tensor operator can be written as @xmath227}$ ] , for which [ w. kwong and j. l. rosner , phys .
d * 38 , * 279 ( 1988 ) ] @xmath228/[(2l-1)(2l+3)]$ ] .
if , for example , we interpret the factor @xmath229 as the beginning of an expansion for @xmath230 with @xmath231 , then the predictions for the @xmath17 family agree with experiment , within errors , while those for the @xmath12 family are about @xmath232 low .
e. eichten , k. gottfried , t. kinoshita , k. d. lane , t .-
yan , phys .
d * 21 , * 203 ( 1980 ) , observed in the context of potential models that , in the heavy - quark limit , the state with @xmath233 is degenerate with the @xmath234 level , while the state with @xmath235 is degenerate with the @xmath236 level .
( see their appendix on charmed mesons . )
see also j. l. rosner , comments nucl .
* 16 , * 109 ( 1986 ) .
the observation in the heavy - quark limit of qcd is due to n. isgur and m. b. wise , phys .
* 66 , * 1130 ( 1991 ) .
yu - qi chen and yu - ping kuang , phys .
d * 46 , * 1165 ( 1992 ) .
see also yu - qi chen , `` the study of @xmath237 meson and its excited states , '' institute for theoretical physics , academia sinica , ph.d .
thesis , october 24 , 1992 ( unpublished ) .
we adopt the standard normalization , @xmath238 .
see , for example , the appendix of hans a. bethe and edwin e. salpeter , _ quantum mechanics of one- and two - electron atoms _ ( springer - verlag , berlin , 1957 ) . eric braaten , kingman cheung , and tzu chiang yuan , phys .
d * 48 , * r5049 ( 1993 ) ; kingman cheung , `` @xmath16 meson production at hadron colliders by heavy quark fragmentation , '' nuhep - th-93 - 19 ( unpublished ) ; yu - qi chen , `` perturbative qcd predictions for the fragmentation functions of the @xmath239-wave mesons with two heavy quarks , '' china center for advanced science and technology ( world laboratory ) preprint ccast-93 - 4 ( unpublished ) .
k. gottfried , in _ proceedings of the international symposium on lepton and photon interactions as high energies , _ edited by f. gutbrod , desy , hamburg ( 1978 ) ; phys .
* 40 , * 598 ( 1978 ) ; m. b. voloshin , nucl .
b * 154 , * 365 ( 1979 ) .
yan , phys .
d * 22 , * 1652 ( 1980 ) . y .-
kuang and t .- m .
yan , phys .
d * 24 , * 2874 ( 1981 ) . y .-
kuang , s. f. tuan and t .- m .
yan , phys .
d * 37 , * 1210 ( 1988 ) . y .-
kuang and t .- m .
yan , phys .
d * 41 , * 155 ( 1990 ) .
it must be remembered in applying these relations to the @xmath62 system that the physical eigenstates with @xmath240 are linear combinations of the equal - mass spin - singlet and spin - triplet states .
r. a. partridge , ph .
d. thesis , caltech report no .
calt-68 - 1150 ( 1984 , unpublished ) .
r. h. schindler , ph .
d. thesis , stanford linear accelerator laboratory report no .
slac-219 ( 1979 , unpublished ) .
j. adler , et al .
( mark iii collaboration ) , phys .
* 60 , * 89 ( 1988 ) .
l. s. brown and r. n. cahn , phys .
lett . * 35 , * 1 ( 1975 ) .
g. s. abrams _ et al . _
( mark i collaboration ) , phys .
* 34 , * 1181 ( 1975 ) ; g. s. abrams , in _ proceedings of the 1975 international symposium on lepton and photon interactions at high energies , _ edited by w. t. kirk ( slac , stanford , 1975 ) , p. 25 .
d. besson _ et al . _ ( cleo collaboration ) , phys .
d * 30 , * 1433 ( 1984 ) .
t. bowcock _
( cleo collaboration ) , phys .
* 58 , * 307 ( 1987 ) ; i. c. brock _
( cleo collaboration ) , phys .
d * 43 , * 1448 ( 1991 ) . decay mode & branching fraction ( percent ) + + 1@xmath14s@xmath242 & 100 .
+ + 1@xmath14s@xmath243 & 91 .
+ 2@xmath244 & 9 .
+ + 1@xmath33s@xmath245 & 55 .
+ 2@xmath33p@xmath246 & 20 .
+ 2@xmath247 & 16 .
+ 2@xmath33p@xmath242 & 9 .
+ + 1@xmath33s@xmath248 & 100 .
+ + 1@xmath33s@xmath248 & 100 .
+ + 1@xmath14s@xmath242 & 100 .
+ + 1@xmath33s@xmath248 & 100 .
+ + 1@xmath33s@xmath245 & 18 .
+ 2@xmath33p@xmath246 & 2 .
+ 2@xmath247 & 29 .
+ 2@xmath33p@xmath242 & 51 .
+ + 1@xmath33s@xmath245 & 22 .
+ 2@xmath33p@xmath246 & 17 .
+ 2@xmath247 & 61 .
+ + 1@xmath33s@xmath245 & 24 .
+ 2@xmath33p@xmath246 & 76 .
+ + 1@xmath14s@xmath243 & 26 .
+ 2@xmath244 & 74 .
+ + 2@xmath33p@xmath243 & 2 .
+ 1@xmath33s@xmath248 & 31 .
+ 2@xmath33s@xmath248 & 57 .
+ 3@xmath33d@xmath248 & 10 .
+ + 1@xmath33s@xmath248 & 26 .
+ 2@xmath33s@xmath248 & 63 .
+ 3@xmath33d@xmath246 & 11 .
+ + 2@xmath249 & 7 .
+ 1@xmath33s@xmath248 & 10 .
+ 2@xmath33s@xmath248 & 26 .
+ 3@xmath33d@xmath246 & 55 .
+ + 2@xmath33p@xmath250 & 1 .
+ 1@xmath33s@xmath248 & 21 .
+ 2@xmath33s@xmath248 & 60 .
+ 3@xmath33d@xmath251 & 15 .
+ 3@xmath33d@xmath246 & 3 . | applying knowledge of the interaction between heavy quarks derived from the study of @xmath0 and @xmath1 bound states , we calculate the spectrum of @xmath2 mesons .
we compute transition rates for the electromagnetic and hadronic cascades that lead from excited states to the @xmath3 ground state , and briefly consider the prospects for experimental observation of the spectrum .
= 10000 |
in this paper we consider the cauchy problem of the spatially inhomogeneous boltzmann equation without angular cutoff .
it reads , with a @xmath4 , as the following equation , @xmath5 , \\
f(0,x , v)=f_0(x , v ) , \end{array } \right.\end{aligned}\ ] ] for the density distribution function of particles @xmath6 , which are located around position @xmath7 with velocity @xmath8 at time @xmath9 .
the right - hand side of the above equation is the so - called boltzmann bilinear collision operator acting only on the velocity variable @xmath1 : @xmath10 above , we use the standard shorthand @xmath6 , @xmath11 , @xmath12 , @xmath13 .
the relations between the post- and pre - collisional velocities are described by the @xmath14-representation , that is , for @xmath15 , @xmath16 note that the collision process satisfies the conservation of momentum and kinetic energy , i.e. @xmath17 the collision cross section @xmath18 is a given non - negative function depending only on the interaction law between particles . from a mathematical viewpoint , that is to say , @xmath18 depends only on the relative velocity @xmath19 and the deviation angle @xmath20 defined through the scalar product @xmath21 without loss of generality , the cross section @xmath22 is assumed to be of the form : @xmath23 where the kinetic factor @xmath24 is given by @xmath25 and the angular part @xmath26 , with a singularity , satisfies , @xmath27 for some @xmath28 .
we remark that if the inter - molecule potential is given by the inverse - power law @xmath29 ( where @xmath30 ) , it holds @xmath31 , @xmath32 generally , the cases @xmath33 , @xmath34 , and @xmath35 correspond to so - called hard , maxwellian , and soft potential respectively . and the cases @xmath36 , @xmath37 correspond to so - called mild singularity and strong singularity respectively .
we begin with a brief review for the non - cutoff theory of the boltzmann equation .
we refer to villani s review book @xcite for the physical background and the mathematical theories of the boltzmann equation .
furthermore , in the non - cutoff setting , alexandre gave more details in @xcite .
our discussion is based on the following definition of gevrey spaces @xmath38 on an open subset @xmath39 ( see @xcite , for instance ) : [ def1 ] for @xmath40 , we say that @xmath41 , if @xmath42 , and there exist @xmath43 such that @xmath44 note that the gevrey scale measures regularity between analytic and @xmath45 .
more precisely , when @xmath46 , it is usual analytic function .
if @xmath47 , it is gevrey class function . and for @xmath48 , it is called ultra - analytic function . for the cauchy problem of the boltzmann equation in gevrey classes , ukai showed , in @xcite in 1984 , that there exists a unique local solution for both spatially homogeneous and inhomogeneous cases , with the assumption on the cross section : @xmath49 in particular , for the spatially inhomogeneous case , by introducing the norm of gevrey space @xmath50 ukai proved that , under some assumptions for @xmath51 and the initial datum @xmath52 , the cauchy problem ( [ be ] ) has a unique solution @xmath53 for @xmath54 $ ] .
on the other hand , desvillettes established in @xcite the @xmath45 smoothing effect for solutions of cauchy problem in spatially homogeneous case , and conjectured gevrey smoothing effect .
he also proved , without any assumptions on the decay at infinity in @xmath1 variables , the propagation of gevrey regularity for solutions ( see @xcite ) . in 2009 morimoto et al . considered in @xcite the gevrey regularity for the linearized boltzmann equation around the absolute maxwellian distribution , by virtue of the following mollifier : @xmath55 we remark that the same operator was used in many related models such as the fokker - planck equation , the kac s equation , the landau equation , and so on . in the mild singularity setting @xmath36 , huo et al .
proved in @xcite that any weak solution @xmath56 to the cauchy problem ( [ be ] ) satisfying the natural boundedness on mass , energy and entropy , namely , @xmath57dv < + \infty,\end{aligned}\ ] ] belongs to @xmath58 for any @xmath59 , and moreover , @xmath60;h^{+\infty}(\mathbb{r}^n)\big),\end{aligned}\ ] ] for any @xmath61 and @xmath62 .
in paper @xcite , the five authors proved the smoothing effect on the solution with weight .
more percisely , if the non - negative @xmath63 belongs to @xmath64 , solves the boltzmann equation ( [ be ] ) in the above domain in the classic sense , and meantime satisfies the non - vacuum condition @xmath65 , then it follows that , @xmath66 and hence it holds , @xmath67 therein the five authors also considered a kind of solution having the maxwellian decay , based on which we introduce the following definition : [ def2 ] we say that @xmath53 is a smooth maxwellian decay solution to the cauchy problem ( [ be ] ) if @xmath68;~h^{+\infty}(\mathbb{r}^3_x \times \mathbb{r}^3_v ) \right ) .
\end{array } \right.\end{aligned}\ ] ] note that the theorem 1.2 of @xcite shows the uniqueness of the smooth maxwellian decay solution to the cauchy problem ( [ be ] ) . in 2010
morimoto - ukai considered the gevrey regularity of @xmath45 solutions with the maxwellian decay to the cauchy problem of spatially homogeneous boltzmann equation ( see @xcite ) .
motivated by their results , we studied this problem in @xcite in a more general framework .
more precisely , we considered the general kinetic factor @xmath69 instead of the moderate form @xmath70 in @xcite , and a wider range of the parameter of @xmath71 such that @xmath72 which applies for both hard potential and soft potential
. in the present paper , we study still in the general framework @xmath73 with @xmath72 in the mild singularity assumption @xmath36 . beyond that
, we focus here on the spatially inhomogeneous case , which is much more complicate than homogeneous case ( because , in the spatially inhomogeneous case , the interaction between the kinetic part and nonlinear collision part is very complicate ) .
the full , spatially inhomogeneous model is more closely related to the real physical setting , thus , is more meaningful and interested , and in particular is a cornerstone of statistical physics .
since the estimates obtained in @xcite can carry over to the inhomogeneous framework , it seems easy to begin our justification
. we can handle with the space variables @xmath3 by virtue of one more integrations .
however , it is difficult to clarify distinctly the process when taking double supremum on space variables @xmath3 and velocity variables @xmath1 , which will be related to a more intricate technique , as we will see later .
we aim in this work at expressing the whole process explicitly . combining with the previous result in @xcite
, we can get a characterization of the gevrey regularity of smooth maxwellian decay solutions to the non - cutoff boltzmann .
we wish these results will be useful for the forthcoming research . when considering the gevrey regularity we may assume @xmath74 in the above statement by translation .
now we are in a position to state our main result of propagation of gevrey regularity , as follows : [ propagation of gevrey ] let @xmath75(which are independent of s ) and assume that @xmath76 , @xmath77 . let @xmath53 be a smooth maxwellian decay solution to the cauchy problem ( [ be ] ) . if there exist @xmath78 , @xmath79 such that @xmath80 then there exist @xmath81 and @xmath82 with @xmath83 such that @xmath84}\sup_{\alpha,\beta } \frac{\rho^{|\alpha|+|\beta| } \|e^{(\delta-\kappa t ) \langle v \rangle ^2}\partial_x^\alpha \partial^{\beta}_{v } f(t)\|_{l^2_{x , v}}}{\{\alpha!\}^{\nu_1 } \{\beta!\}^{\nu_2 } } < + \infty.\end{aligned}\ ] ] it should be noted that the above theorem is similar as theorem 1.3 in @xcite , but here we consider the spatially inhomogeneous case . by using the arguments in section 4 below , we obtain the gevrey smoothing effect of order @xmath0 in variable @xmath1 and order @xmath2 in variable @xmath3 as follows : [ gevrey regularity ] assume that @xmath85 and @xmath86 .
let @xmath53 be a smooth maxwellian decay solution to the cauchy problem ( [ be ] ) , and further , @xmath87 be analytic with respect to space variable @xmath3 , then for any @xmath62 , there exist @xmath88 and @xmath89 with @xmath83 such that @xmath90}\sup_{\alpha,\beta } \frac{\rho^{|\alpha|+|\beta| } \|e^{(\delta-\kappa t ) \langle v \rangle ^2}\partial^{\alpha}_{x } \partial_v^\beta f(t)\|_{l^2_{x , v}}}{\{\alpha!\}^{\nu_1 } \{\beta!\}^{\nu_2 } } < + \infty.\end{aligned}\ ] ] the remainder of the paper proceeds as follows . in the next section
we give some preliminaries and a key lemma , by which we can complete immediately the proof of theorem [ propagation of gevrey ] .
section 3 is devoted to the proof of the key lemma .
section 4 is arranged for the justification of theorem [ gevrey regularity ] about the gevrey smoothing effect .
we give some notations and fundamental facts here .
( see @xcite for details ) . denote @xmath91 throughout this paper .
let @xmath92 whose values will be chosen later .
for @xmath93 we define : @xmath94 where @xmath95 are multi - index of @xmath96 , i.e. , @xmath97 .
we denote @xmath98 and @xmath99 we introduce the definition : @xmath100 with fixed @xmath101 satisfying @xmath102 . here
@xmath103 is a fixed large integer .
analogous to the argument of @xcite , we can obtain that , for @xmath104 , @xmath105 setting @xmath106 and taking a large enough @xmath107 , then it follows from the assumption ( [ initial assumption ] ) that , we may let @xmath108 be as small as possible , where @xmath109 can be chosen any positive less than @xmath110 in ( [ initial assumption ] ) .
therefore , in order to prove the result ( [ propag ] ) , it suffices to prove that , @xmath111 } \|f\|_{l,\rho , r , n}(t ) < \infty,\end{aligned}\ ] ] under the assumption that @xmath112 is sufficiently small . above ,
@xmath113 .
now we give the key lemma , which will play an important role in the following sections .
[ key lemma ] if @xmath114 and @xmath115 , then for any @xmath95 satisfying @xmath116 we have @xmath117 where @xmath118 .
following along the same lines as that of section 2 in @xcite , we can prove theorem [ propagation of gevrey ] , and so omit it .
the proof of this lemma will be given in the next section .
applying @xmath119 to eq.([be ] ) , we have , for @xmath120 , @xmath121 and for @xmath122 , @xmath123 where @xmath124 , and @xmath125 .
note that if @xmath126 , the @xmath127-th component of @xmath128 , equals to @xmath129 , then @xmath130 will be zero , the above equation will still hold true .
let @xmath131 with @xmath102 .
multiplying by @xmath132 both sides of the above equation , we obtain @xmath133 set @xmath134 and denote @xmath135 for @xmath136 . we can rewrite eq.([eq1 ] ) as follows : @xmath137 noticing that @xmath138 , we then get the following formula , @xmath139 thus we obtain , @xmath140 set @xmath141 . taking the @xmath142 inner product with @xmath143
, we have , @xmath144 we notice that the third term on the left - hand side equals to @xmath145 , i.e. @xmath146 dx \right ) dv = 0 .\end{aligned}\ ] ] multiplying by @xmath147 both sides , and integrating in time from @xmath145 to @xmath54 $ ] , we obtain finally , @xmath148 * step i : * firstly we have @xmath150 this yields @xmath151 where we have used the cauchy - schwartz inequality in the last inequality .
* step ii : * as for @xmath152 , we recall the result in our preceding paper @xcite : under the assumption @xmath153 , we have @xmath154 here we mention that , the corresponding term in @xcite is @xmath155 , and we can consider the space variable @xmath3 as parameter . then if @xmath156 $ ] , we have @xmath157 and @xmath158 where in the second inequality we have used the embedding @xmath159 .
on the other hand , if @xmath160 + 1 $ ] , which implies that @xmath161 , we have @xmath162 therefore , we obtain if @xmath156 $ ] , then @xmath163 and if @xmath160 + 1 $ ] , then @xmath164 specifically , we point out that if @xmath165 , i.e. , @xmath166 and @xmath167 , the above two estimates remain valid .
* step iii : * next , concerning @xmath168 , we use the coercivity estimate in @xcite : @xmath169 then it follows @xmath170 observing that @xmath171 , we have @xmath172 * step iv : * now we consider the term @xmath173 with the restriction @xmath174 . by virtue of the upper bound estimate for collision operator @xmath175 from proposition 2.9 of @xcite : [ upper bound-2 ] let @xmath176 and @xmath177 .
for any @xmath178 and @xmath179 $ ] , there exists a @xmath180 such that @xmath181 applying above lemma with @xmath182 , @xmath183 , @xmath184 , @xmath185 , and @xmath186 , by setting @xmath187 , we can infer that , @xmath188 in view of @xmath189 and @xmath190 whereas in the case @xmath191 $ ] , by setting @xmath192 $ ] , we have @xmath193 note that we have the same estimate on @xmath194 in both cases @xmath195 and @xmath196 , then if @xmath156 $ ] , by recalling @xmath197 , it holds @xmath198 , hence we can write that @xmath199 since the hlder inequality yields @xmath200 we obtain if @xmath156 $ ] , @xmath201 or alternatively , when @xmath160 + 1 $ ] , we can arrive at @xmath202 + 1 ) + 2 + 1 \le n$ ] , and further , @xmath203 above , @xmath187 , or @xmath204 if @xmath191 $ ] .
thus , if @xmath156 $ ] , we get that @xmath205 and @xmath206 similarly , when @xmath160 + 1 $ ] , we can have @xmath207 and @xmath208 * step v : * before continuing the proof , we state a useful lemma , as follows : if @xmath209 and @xmath210 , then there exists a constant @xmath211 depending only on @xmath212 such that for any @xmath213 , @xmath214 and @xmath215 furthermore , if @xmath216 and @xmath217 , then there exists a constant @xmath218 depending only on @xmath51 and @xmath212 such that for any @xmath219 , @xmath220 note that the first and third result were established in @xcite , and the proof of the second share a similar scheme as the first , so we omit it here .
additionally , we will consider @xmath22 , @xmath221 and @xmath222 as @xmath2 in later proof .
now we resume the proof of lemma [ key lemma ] . as for the integral including @xmath223 in ( [ eq2 ] ) , from ( [ psi_2 ] ) we can infer that , if @xmath156 $ ] , @xmath224 and if @xmath160 + 1 $ ] , we have , correspondingly , @xmath225 we then introduce the following notations : @xmath226 } \sup_{|\alpha|+|\beta| \leq 8r } \| f(t)\|_{\delta-\kappa t , l,\rho , r,\alpha,\beta}. % & [ f]_{l,\rho , r , n}(t,\alpha_0 ) \triangleq \sup_{\overset { |\alpha'| \leq 3r}{3r \leq |\beta'| \leq
n } } % \| f(t)\|_{\delta-\kappa t , l,\rho , r,\alpha',\beta'},\end{aligned}\ ] ] now , we shall give the different bounds on the factor @xmath227 with respect to the values of @xmath228 and @xmath229 : * if @xmath230 , we have @xmath231 * if @xmath232 , we have @xmath233 * if @xmath234 , we have @xmath235 * if @xmath236 , we have @xmath237 thus , together with the above inequalities , we get , @xmath238 which yields that , @xmath239 on the other hand , in the case @xmath160 + 1 $ ] , a similar scheme ensures us to get the same estimate as above .
considering the integral including @xmath240 in ( [ eq2 ] ) , we can write that , @xmath241 first , it is easy to check that , from ( [ j_2-original ] ) and ( [ j_2-original ] ) , the integral including @xmath242 has the same estimate as the inequality ( [ k(t ) ] ) .
furthermore , the last factor of ( [ j_1-original ] ) ( or ( [ j_1-original ] ) ) is bounded by @xmath243 or @xmath244 a similar procedure as that of handling with the estimate of ( [ k(t ) ] ) gives that , @xmath245 observing that @xmath246 implies @xmath247 and @xmath248 , so we obtain the estimate on @xmath240 as follows : @xmath249 * step i : * combining ( [ eq2 ] ) , ( [ r(t ) ] ) , ( [ psi(t ) ] ) , ( [ k(t ) ] ) and ( [ j(t ) ] ) , we finally deduce that , @xmath250 } \sup_{8r\le |\alpha| \le n } \| f \|^2_{\delta-\kappa \tau , l,\rho , r,\alpha+1,0}(\tau ) \\\nonumber & \hspace*{6.5em } + \frac{\kappa}{10}\sup_{8r \leq |\alpha|+|\beta| \leq n } \int_0^t \|f\|^2_{\delta-\kappa \tau , l+1,\rho , r,\alpha,\beta}(\tau)d\tau,\end{aligned}\ ] ] where @xmath246 .
* step ii : * now we need to control the term @xmath251 with @xmath252 . for sake of simplicity , we will treat with @xmath253 . indeed
, it is a matter repeating the argument that we did for the case @xmath254 , except that we need to replace the weight @xmath147 by @xmath255 , and we can get @xmath256 where we emphasize especially that , the integral including @xmath242 in the case of @xmath257 can be bounded by @xmath258 as a consequence , we have @xmath259 thanks to the facts @xmath260 and @xmath261 , we get @xmath262 * step iii : * plugging this into entails that @xmath263 this leads to the desired estimate ( [ basic inequa ] ) including the extra second term of the left - hand side and so completes the whole proof of lemma [ key lemma ] .
we hope to show the orders of gevrey regularity are @xmath0 for @xmath1 and @xmath2 for @xmath3 .
if @xmath265 , the above definition goes back to the definition ( [ double norm ] ) . obviously , the previous results are fit for the modified norm , which is just the reason why we only use @xmath266 in the previous process .
we now introduce a new norm @xmath267 where @xmath268 and @xmath269 .
it should be pointed out that , by the assumption of theorem [ gevrey regularity ] that @xmath87 is analytic in the variable @xmath3 and the definition of the smooth maxwellian decay solution ( see def.[def2 ] ) , we have , for sufficiently small @xmath266 , @xmath270 as for the above norm , we can get the following integral corresponding to ( [ eq2 ] ) : @xmath271 because of the formula @xmath272 note that , when considering a sufficiently small @xmath266 , @xmath273 is well - defined . considering the third term on the left - hand side of the above equation , with @xmath278 we have @xmath279 then the third term can be ignored .
thus , one can obtain immediately the inequality similar to ( [ final inequa ] ) , which yields the gevrey smoothing effect in a short interval .
that leads us to the conclusion together with theorem [ propagation of gevrey ] .
* acknowledgements*. this work was partially supported by nnsfc ( no . 11271382 and no .
10971235 ) , rfdp ( no .
20120171110014 ) , and the key project of sun yat - sen university ( no .
the authors thank the referee for valuable comments and suggestions .
r. alexandre , y. morimoto , s. ukai , c .- j .
xu and t.yang , _ boltzmann equation without angular cutoff in the whole space : qualitative properties of solutions _ , arch .
rational mech .
, * 202(2 ) ( 2011 ) * , 599 - 661 . | in this paper we consider the non - cutoff boltzmann equation in spatially inhomogeneous case .
we prove the propagation of gevrey regularity for the so - called smooth maxwellian decay solutions to the cauchy problem of spatially inhomogeneous boltzmann equation , and obtain gevrey regularity of order @xmath0 in the velocity variable @xmath1 and order @xmath2 in the space variable @xmath3 .
the strategy relies on our recent results for spatially homogeneous case ( j. differential equations 253(4 ) ( 2012 ) , 1172 - 1190 .
doi : 10.1016/j.jde.2012.04.023 ) .
rather , we need much more intricate analysis additionally in order to handle with the coupling of the double variables . combining with the previous result mentioned above , it gives a characterization of the gevrey regularity of the particular kind of solutions to the non - cutoff boltzmann .
_ 2000 mathematics subject classification _
: 35a05 , 35b65 , 35d10 , 35h20 , 76p05 , 82c40 .
_ keywords _ : boltzmann equation ; spatially inhomogeneous ; non - cutoff ; gevrey regularity . ' '' '' |
in the sir model , the fixed population @xmath0 is divided into three distinct groups : susceptible @xmath1 , infected @xmath2 and removed @xmath3 .
those at risk of the disease are susceptible , those that have it are infected and those that are either quarantined , dead , or have acquired immunity are removed .
the following flow chart shows the basic procession of the sir model daley1.@xmath4here @xmath5 and @xmath6 are the infection coefficient and removal rate , respectively .
a discrete model was adapted by gani from the original sir model through the coarse - graining process and was applied to successfully predict outbreak of influenza epidemics in england and walesgani1,spicer1 .
@xmath7 i_{i } \label{sir2 } \\
r_{i+1 } & = & r_{i}+ai_{i } \notag\end{aligned}\ ] ] during the epidemic process , @xmath8 is fixed .
initially , we examine the data for the first 15 days to estimate the parameters @xmath5 and @xmath6 of sars for hk .
the only data are new cases ( removal @xmath9 ) announced everyday by hk dept .
of health from march 12 , 2003 followed by a revised version later@xcite . to avoid inadvertently using future information we do not use the revised data at this stage .
population @xmath10 ; since @xmath11 it is reasonable to let @xmath12 in right hand side of ( [ sir2 ] ) .
@xmath13 and @xmath14 is set as the initial condition .
@xmath15 is replaced with @xmath16 whereas @xmath15 is uncertain .
this assumption implies the incubation period is only one day , in spite of the fact that the true incubation period of the coronavirus is 2 - 7 days@xcite .
the parameters @xmath5 and @xmath6 are scanned for the best fit for the stage .
for every @xmath17 , a sequence of @xmath18 is obtained by numerical simulations .
an euclidean norm of @xmath19 , which indicates a distance between the true and simulated data , is applied to measure the fit . in fig .
[ graph1 ] @xmath17 of the lowest point is the best fit parameters for this stage and this value is used for the following prediction .
we get @xmath20 and @xmath21 .
the method is applied to get parameters of @xmath6 and @xmath5 in figs .
[ graph2 ] and [ graph3 ] . vs. @xmath5 and @xmath6
is plotted as dots for the first 15 days sars data for hk .
the natural logarithm is applied . @xmath3 and
@xmath22 are original and simulated data , respectively .
the thick red curve shows the bottom of the sharp valley clearly .
the lowest point of the valley is according to the best fit : @xmath20 and @xmath21.,width=316 ] the prediction is available for the trend based on parameters @xmath5 and @xmath6 of this stage , the middle day ( march 20 , 2003 ) of the stage is applied as the first day and @xmath23 is assumed as @xmath24 .
a curve of squares is plotted in fig .
[ graph2 ] for the first stage prediction .
in the same way the best fit and prediction is applied for the next two stages and plotted in fig .
[ graph2 ] .
the best fit is also processed weekly for detail discussion later . and @xmath21 ) ,
circles ( @xmath25 and @xmath26 ) and triangles ( @xmath27 and @xmath28 ) are prediction for the 3 stages described in the text .
each stage has 15 days .
the curve of black dots is for revised data.,width=316 ] the curves for prediction clearly shows the 3 stages in fig .
[ graph2 ] . the first stage ( march 12 - 27 , 2003 ) exhibits dangerous exponential growth .
it shows that the extremely infectious sars coronavirus spread quickly in the public with few protections during this early stage .
more seriously , it leads to a higher infection peak although in this stage the averaged number of new cases @xmath29 is below 30 .
this prediction gets confirmation in the second stage ( march 28 - april 11 , 2003 ) .
the peak characterized by the amoy garden outbreak comes earlier and higher .
it indicates appearance of a new transmission mode that differs from intimate contact route observed in the first stage .
we name the new transmission mode @xmath30 growth in the contrast to transmission in the first stage which we refer to as @xmath31 growth . outbreaks at prince wales hospital ( pwh ) ( where sars patients received treatment ) and amoy gardens ( a high - density housing estate in hong kong ) represent infections of @xmath32 and @xmath30 modes in the full epidemic , respectively .
the revised data provided by hk dept . of health in fig .
[ graph2 ] present a clearer indication for these two transmission modes .
the outbreak of pwh sent the clear message that intimate contacts with sars patient led to infection@xcite .
detailed investigations of the propagation at amoy garden suggested that faulty of sewerage pipes allowed droplets containing the coronavirus to enter neighboring units vertically in the building@xcite .
furthermore , poor ventilation of lifts and rat infestation where also suggested as possible modes of contamination @xcite .
however , control measures bring the spread of the disease under control in the second stage .
contrary to the increasing trend in the first stage , the prediction curve of the 2nd stage ( circle ) declines .
and it anticipates that new cases drop below 10 before the 60th day ( may 10 , 2003 ) . also on april 12 , 2003
we predicted number of whole infection cases would reach 1700 . up to april 11 , 2003
there were 32 deaths and 169 recoveries@xcite .
we calculated the mortality of sars as the ratio of death to sum of both deathes and recoveries , and it was 15.9% .
therefore we predicted approximately 270 fatalities in total . in the third stage
( april 12 - 27 , 2003 ) , triangle curve in fig . [ graph2 ] refines results and gives more accurate prediction .
this stage predicts that new cases per day drops to 5 before the 62th day , may 12 , 2003 .
the travel warning for hk was cancelled by who because hk had kept new cases below 5 for 10 days since may 15 , 2003 . and
, finally , we predict that the total cases reaches 1730 and nearly 287 deaths ( up to april 27 , 2003 there were 668 recovered and 133 , the mortality increased to 16.6% ) .
these numbers are very close to the true data@xcite .
precisely the method drawn from the sir model has been verified for prediction for full epidemic .
however , the accuracy is only possible by the first dividing the epidemic into separate `` stages '' .
this problem of determining to what extent an epidemic is under control is of greater strategic significance .
information on the efficacy of epidemic control will help determine whether to apply more control policies or not , and balance cost and benefit from them . for each individual the same question will also inform the degree to which precautions are taken : i.e. wearing a surgical mask to prevent the spread ( acquisition ) of sars . obviously a way to estimate control level is required .
actually this is a difficult problem because of significant statistical fluctuations in the data . a quite simple method to evaluate control efficiency
is discussed below . in the sir model ( [ sir2 ] ) , if @xmath33 a disease is regards as being controlled as new cases will decrease .
this inequality leads to a control criterion for some diseases in epidemiology research . applying the approximation of @xmath12 we get a threshold @xmath34 from ( [ sir2 ] ) .
we rescale @xmath35 ( @xmath5 is called infection rate in place of infection coefficient now ) and then get the threshold that is free from population @xmath0.@xmath36this indicates that the removal rate @xmath6 exceeds the infection rate @xmath5 . in fig .
[ graph3 ] the circles show the 3 stages evaluations with dash line of @xmath37 . and the parameters @xmath5 and @xmath6 of sars data for hk are estimated weekly as squares also in fig .
[ graph3 ] .
the line of @xmath37 is regarded as the critical line since number of infected cases increases when @xmath38 passes through it from below .
it is possible to apply the diagram of @xmath5 and @xmath6 to compare control level for different countries and areas even for different diseases .
this provides organizations like who with a simple and standard method to supervise infection level of any disease .
the limitation of the method comes from the assumptions of the sir model .
more accurate models may provide better estimate of epidemic state and future behaviours . .
in this panel the line of dots @xmath37 is regarded as alert line or critical line .
the parameters of @xmath5 and @xmath6 below it indicate that the epidemic is controlled , above the line indicates uncontrolled growth .
the parameter @xmath5 is rescaled by @xmath35.,width=316 ] in summary , a discrete sir model gives good predictions for epidemic of sars for hk .
two distinct methods are described for the disease propagation dynamics .
particularly the @xmath30 mode is much more hazardous than the @xmath39 mode .
we have introduced a simple method to evaluate control levels .
the method is generic and can be widely applied to various epidemiological data .
contrary to the long established sir model , epidemiology research using small - world ( sw ) network models is young and growing area .
the concept of sw was imported from the study of social network into nature sciences in 1998@xcite .
however , it provides a novel insight for networks andarouses a lot of explorations in the brain , social networks and the internet.@xcite .
some sw networks also exhibit a scale - free ( power law ) distribution : a node in this network has probability @xmath40 to connect @xmath41 nodes , @xmath42 is one basic character for the system@xcite .
most researches of virus spreading with the sw model , concern the propagation of computer viruses on the internet , however , a few studies have been published relating to epidemiologyebel1,liljeros1,strogatz1,kuperman1,huerta1 .
the dynamics of sw models enriches our realization of epidemic and possibly provides better control policies@xcite . from the first sars patient to the last one , an epidemic chain is embedded on the scale free sw network of social contacts .
it is of great importance to discover the underlying structure of the epidemic network because a successful quarantine of all possible candidates for infection will lead to a rapid termination of the epidemic . in hk e - sars , an electronic database to capture on - line and in real time clinical and administrative details of all sars patients , provided invaluable quarantine information by tracing contacts@xcite .
unfortunately a full epidemic network of sars for hk is still unavailable . because data representing the underlying network structure is currently not available
, we have no choice other than numerical simulations .
therefore , our analysis of the sw model is largely theoretical .
the only confirmation of our model we can offer is that the data appear to be realistic and exhibit the same features as the true epidemic data . to simulate an epidemic chain , a simple model of social contacts
is proposed .
the model is established on a grid network weaved by @xmath43 parallel and @xmath43 vertical lines .
every node in the network represents a person .
we set @xmath44 with population @xmath45 .
all nodes are initiated with a value of 0 ( named @xmath46 nodes ) .
every node has 2 , 3 or 4 nearest neighbors as short range contacts for corner , edge and center , respectively . for every node
there are two long range contacts with 2 other nodes randomly selected in the whole system everyday .
these linkages model the social contact between individuals ( i.e. social contacts that are sufficiently intimate to bring individuals at risk of spreading the disease ) .
one random node of the system is set to 1 ( called the @xmath47 node ) , through its short and long range contacts , the value of the nodes linked with it turns into 1 according to probability of @xmath48 and @xmath49 , respectively .
an infection happens if a node changes its value from 0 to 1 .
this change is irreversible . during the full simulation process
, the bad node is not removed from the whole system .
we make this assumption because the number of deaths is small in comparision of the population , and there is no absolute quarantine even the sars patient in hospital can affect the medical workers . moreover , the treatment period for sars is relatively length , and during this time infected indviduals are highly infectious . to reflect the true variation in control strategy and individual behaviour , the control parameters @xmath50 ( @xmath51 ) vary with time .
nodes with fixed infection probability @xmath52 which eventually leads to a full infection .
a ) . the full epidemic curve of the simulated epidemic .
b ) . while time is 45 , the infection cases clusteringly scatter in the map .
the infection happens in full scale map because of random connections .
the clusters show effect from the nearest neighbour connections .
, width=316 ] in same model it would be interesting to compare with the complete infection epidemic . a small system with @xmath53 nodes
is chosen .
a fixed probability @xmath54 leads to eventual infection of the entire population since all nodes are linked and the infected are not removed .
the epidemic curve for this process is plotted in fig .
[ graph4 ] a ) and is typical of many plagues . in fig .
[ graph4 ] b ) various sizes of clusters with infected nodes ( black dots ) scatter over the geographical map .
it contains all infection facts in first 45 days . comparing to the true sars infection distribution in hk
only slight similarity is observed .
the short and long range linkages give good infection dynamics as we expected . the epidemic chain is easily drawn by recording infection fact from the seed to last patient .
however , during simulations there is a problem : for a @xmath46 node linked by more than one @xmath47 nodes , which @xmath47 node infects it ? a widely accepted preferential attachment of _ rich - get - richer _ is a good answer@xcite . a linear preference function is applied here . with presence of both growth and preferential attachment ,
it is general to ask whether the chain is a scale - invariant sw network .
we plot the distribution in fig .
[ graph5 ] a ) and b ) in log - log and log - linear coordinates , respectively .
the hollow circle is for the system with @xmath55 nodes .
the scaling behaviour of it looks more like a piecewise linear ( i.e. bilinear ) in b ) rather than a power law in a ) . to confirm this we tried a larger system with @xmath56 nodes and the same fixed @xmath50 in the rest curves in figs .
[ graph5 ] .
the solid square curve is distribution of the whole epidemic network .
the linear fit for the solid square gives a correlation coefficient @xmath3 of @xmath57 in fig .
[ graph5 ] a ) . in b ) piecewise linear fits have @xmath3 of @xmath58 and @xmath59 .
these provide positive support for the original model .
the ratio of the cases of first 30% and 1% of whole process for this bigger system are plotted as hollow triangle and solid circle curves in figs .
[ graph5 ] , respectively .
the case of 1% , early stage of full infection epidemic , suggests a better fit for scale - free than the other cases since it has correlation coefficients @xmath3 of @xmath60 and ratio @xmath61 for linear fit in a ) . for the solid circle curve piecewise linear
fits give @xmath3 of @xmath62 and @xmath63 in fig .
[ graph5 ] b ) .
so , what scaling behaviour is true during full infection process ?
with absence of rigorous proof this problem is hard to answer correctly in simulations .
we may only draw conclusions based on which simulations most closely matches the qualitative features of the observed data .
the curves are for one with @xmath64 nodes .
the curves of dots , triangles and squares are for the epidemic networks of 1% , 30% and 100% for full infection process .
the two dash lines are piecewise linear fits for the full infection process with @xmath56 nodes .
the ratio is @xmath65 and @xmath66 with @xmath67 and @xmath68 , respectively .
it clearly shows the curves of full epidemic has two piecewise linear parts in log - linear graph . in the early stage of a full infection , the scaling behaviour of 1% infected ( dots )
has a part which might be regards as a log - log linear .
the ratio of it is @xmath69 with @xmath70 . in other words ,
it has scale - free part with @xmath71.,width=316 ] let s return to the system with @xmath72 nodes for modelling sars for hk .
the probability @xmath50 is fitted to the true epidemic data .
however , it is fruitless to obtain an exact coincidence between the simulated results and the true data as the model evolution is highly random ( moreover this would result in overfitting ) .
the control parameters @xmath50 ( dots and dashes curves ) and a simulated epidemic curve ( black dots ) with column diagram of sars for hk is plotted in figs .
[ graph6 ] a ) and b ) , respectively . for the simulated data ,
the total number of cases is 1830 that has a 4.3% deviation from the true data of 1755 .
contrary to the above full infection with fixed parameters , @xmath50 are believed to drop exponentially and lead to small part infection epidemic without quarantine or removal . in any case , the high probability of infection in the early stage is indicative of the critical ability of the sars coronavirus to attack an individual without protection .
if sars returns , the same high initial infection level is likely to occur .
the only hope to avoid a repeat of the sars crisis of 2003 is to shorten the high infection stage by quick identification , wide protection and sufficient quarantining . in other words , the best time to eliminate possible epidemic
is the moment that the first patient surfaces .
any delay may lead to a worsening crisis .
the long range infections in the model and the world also imply an efficient mechanism to respond rapidly to any infectious disease is required to establish global control .
a ) the short and long range linkages infection probability @xmath50 generally declines exponentially .
b ) the simulated epidemic curve ( black dots ) in the model is plotted with the original sars data ( grey column ) for hk.,width=316 ] data on the geographical distribution of sars cases in hk is much easier to collect than the full epidemic chain .
numerical simulations provide both simply . the full geographical map marked with all infected nodes ( black dots ) and an amplified window of a cluster
is plotted in figs .
[ graph7 ] a ) and b ) , respectively .
similarity is expected and verified in cluster patterns of figs .
[ graph4 ] b ) .
the scaling behaviour of the epidemic chain is plotted in figs .
[ graph8 ] .
the curve in log - log diagram exhibits a power - law coefficient of @xmath73 and gives the linear fit correlation coefficient @xmath74 .
the piecewise linear fit for the log - linear case gives correlation coefficients @xmath3 of @xmath75 and @xmath76 , respectively .
for a sw network often few vertexes play more important roles than others@xcite .
sars super - spreaders found in hk , singapore and china are consistent with this@xcite .
data for the early spreading of sars in singapore@xcite show definite sw structure with a small number of nodes with a large number of links .
the average number of links per node also shows a scale - free structure @xcite , but the available data is extremely limited ( the linear scaling can only be estimated from three observations ) .
this character is also verified in our model .
the first few nodes have a high chance of infecting a large number of individuals . in fig . [ graph8 ]
, a single node has 40 links .
clearly the index node has many long range linkages .
it has been suggested that travelling in crowded public places ( train , hospital , even an elevator ) without suitable precautions can cause an ordinary sars patient to infect a significant number of others .
again , this is an indication that to increase an individual s ( especially a probable sars patient s ) personal protection is key to rapidly control an epidemic .
actually , in our model , if duration of the early stage with high probability @xmath50 is reduced to less than 10 , the infection scale decreases sharply .
an engrossing phenomenon is the points of inflection in curves in log - linear diagrams of figs .
[ graph8 ] and figs .
[ graph5 ] b ) .
all are located near linkages number of 6 - 7 . on average a nodes
has about 6 contacts ( 2 - 4 short plus 2 long range ) every day , although in whole process there is no limitation for linkages . for a growing random network
, a general problem always exists among its scaling behaviour , preferential attachment and dynamics , even embedding a geographical mapcohen1,warren1,krapivsky1,dorogovtsev1,moore1 .
more work is required to address this issue . in conclusion
a sw epidemic network is simulated to model sars spreading in hk .
a comparison of the simulations with full infection data is presented .
our discussion of the infection probability and occurrence of super spreaders lead to the obvious conclusion : rapid response of an individual and government is a key to eliminating an epidemic with limited impact and at minimal cost . | a simplified susceptible - infected - recovered ( sir ) epidemic model and a small - world model are applied to analyse the spread and control of severe acute respiratory syndrome ( sars ) for hong kong in early 2003 . from data available in mid april 2003
, we predict that sars would be controlled by june and nearly 1700 persons would be infected based on the sir model .
this is consistent with the known data . a simple way to evaluate the development and efficacy of control
is described and shown to provide a useful measure for the future evolution of an epidemic .
this may contribute to improve strategic response from the government .
the evaluation process here is universal and therefore applicable to many similar homogeneous epidemic diseases within a fixed population .
a novel model consisting of map systems involving the small - world network principle is also described .
we find that this model reproduces qualitative features of the random disease propagation observed in the true data .
unlike traditional deterministic models , scale - free phenomena are observed in the epidemic network .
the numerical simulations provide theoretical support for current strategies and achieve more efficient control of some epidemic diseases , including sars . during 2003
sars killed 916 and infected 8422 globally@xcite . in hong kong ( hk ) , one of the most severely affected regions , 1755 individuals were infected and 299 died@xcite .
sars is caused by a coronavirus , which is more dangerous and tenacious than the aids virus because of its strong ability to survive in moist air and considerable potential to infect through close personal contact@xcite .
unlike other well - known epidemic diseases , such as aids , sars spreads quickly . although significant , its mortality rate is , fortunately , relatively low ( approximately 11%)@xcite .
researchers have decoded the genome of sars coronavirus and developed prompt diagnostic tests and some medicines , a vaccine is still far from being developed and widely usedmarra1,rota1,stadler1 .
the danger of a recurrence of sars remains .
irrespective of pharmacological research , the epidemiology study of sars will help to prevent possible spreading of similar future contagions .
generally , current epidemiological models are of two types .
first , the well - known susceptible - infected - recovered ( sir ) model proposed in 1927kermack1,daley1 .
second , the concept of small - world ( sw ) networkswatts1 . arousing a new wave of epidemiological research
, the sw model has made some progress recently@xcite .
our work aims to model sars data for hk .
practical advice for a better control are drawn from both the sir and sw models .
in particular , a generalized method to evaluate control of an epidemic is promoted here based on the sir model with fixed population .
using this method , measuring the spread and control of various epidemics among different countries becomes simple .
quick action in the early stage is highlighted for both government and individuals to prevent rapid propagation . |
the successful creation of @xmath3k@xmath4rb polar molecules with high space density @xcite has sparked a lot of research interests because of its potential applications in various areas @xcite .
an ultracold gas of polar molecules can provide a platform to study strongly - correlated systems @xcite , due to the confinement flexibility and the control over effective interactions @xcite .
the dipole - dipole interaction can be turned into different effective interactions in the presence of a dc or ac electric field . when the dipole direction is fixed by a dc electric field ,
a p - wave superfluid state with the dominant @xmath5 symmetry may appear in a 3d dipolar fermi gas @xcite .
when an ac electric field is applied to couple different rotational levels of molecules , a @xmath6 superfluid state @xcite and a ferro - electric state @xcite may show up in quasi-2d dipolar fermi gases . creating the antiferromagnetic state and exotic superfluid states such as p - wave and d - wave states
has been a focused goal in the research with optical lattices . in this work
we predict that some of these phases can be created in polar molecules trapped in a 2d square optical lattice in the presence of a resonant ac electric field . for the @xmath3k@xmath4rb molecule ,
rotational excitation energies are much smaller than electronic and vibrational excitation energies . in the presence of the resonant ac electric field for the first two rotational states ,
electronic and vibrational excitations can be ignored , and the rotational state of the dipole along or against the electric - field direction can represented by pseudospin - up or pseudospin - down state .
the dipole - dipole interaction is thus effectively a dipolar ising interaction between pseudospins
. the strength of this ising interaction can be much larger than the super - exchange interaction in a unpolarized two - component fermi gas trapped in the same optical lattice . in this aspect it is easier to create antiferromagnetic and superfluid states with exotic symmetries in the dipolar fermi gas .
this work is organized as follows .
first we derive the effective hamiltonian for this system .
then the antiferromagnetic state and superfluid states with different symmetries are studied in the mean - field approximation .
the ground state is determined by comparing energies of these states .
we find that the antiferromagnetic state is the ground state at half filling .
when the filling factor is reduced to certain values , the superfluid state takes over . depending on the parameters , different superfluid states , such as d - wave ( @xmath0 ) ,
extend s - wave ( @xmath1 ) , and the mixed - symmetry ( @xmath2 ) states , can appear .
phase separation occurs between superfluid and antiferromagnetic phases .
based on these results , the phase diagram at zero temperature is obtained and its experimental implication is discussed .
the hamiltonian of a single @xmath3k@xmath4rb molecule trapped in a two - dimensional optical lattice in the presence of an ac electric field is given by @xmath7 where the first r .- h .- s .
term is the molecule kinetic energy for the center - of - mass motion , and the second r .-
term is the potential energy in the optical lattice .
the rotational energy of the molecule in the presence of the ac electric field is given by @xmath8 , where @xmath9 is the angular momentum operator , @xmath10 is the rotational constant , * d * is dipole operator , and @xmath11 is the ac electric field .
the eigenstate of the angular momentum is given by @xmath12 , where @xmath13 and @xmath14 are quantum numbers of the total angular momentum and its z - component . in this work
we consider a linearly polarized ac field in the @xmath15 direction with amplitude @xmath16 , @xmath17 .
the frequency @xmath18 is set to couple rotational states @xmath19 and @xmath20 , @xmath21 , where @xmath0 is the dipole moment . under the rotating wave approximation @xcite in the basis of @xmath19 and @xmath22 , the molecule rotational energy is given by @xmath23 where @xmath24 . in the following we concentrate on the resonance limit @xmath25 with small electric field , where the two rotational states are nearly degenerate .
the dipole - dipole interaction is given by @xmath26 where @xmath27 is a dipole at @xmath28 , and @xmath29 is the direction of @xmath30 . in the presence of an resonant ac electric field , only rotational states @xmath19 and
@xmath20 are important .
the matrix elements of the dipole operator in this subspace are given by @xmath31 , and @xmath32 .
for this two - level system , we can define the pseudospin states @xmath33 and @xmath34 , so that the dipole matrix elements are @xmath35 , @xmath36 , and @xmath37 . in this new basis
, the interaction between two dipoles can be conveniently written as a dipolar ising interaction between pseudospins , @xmath38 where @xmath39 is the @xmath15 component of the pseudospin @xmath40 , and @xmath41 is the angle between @xmath29 and the @xmath15 direction .
for this system , the optical lattice is set in the plane perpendicular to the electric field , @xmath42 .
for a dipolar fermi gas trapped in a two - dimensional optical lattice in the presence of a resonant ac electric field , the effective hamiltonian is therefore given by @xmath43 where @xmath44 is the fermion annihilation operator for site @xmath45 and spin @xmath46 , @xmath47 , and the spin coupling constant between fermions at positions @xmath28 and @xmath48 is given by @xmath49 .
for simplicity we have assumed that in the optical lattice the fermion can only hop between nearest neighbors with hopping amplitude @xmath50 .
the onsite - interaction @xmath51 in eq .
( [ hamiltonian 1 ] ) includes both the s - wave scattering between different spin components and the onsite dipole interaction , and it can be tuned either by using feshbach resonance or by changing the vertical confinement @xcite . here
we concentrate on the effect of the dipolar ising interaction and set @xmath52 in the rest of this paper .
since the dipolar ising interaction is a short - range interaction in two dimension and the spin coupling constant is the largest between nearest - neighbor spins , the system tends to lower its energy by placing fermions with opposite spins at nearest - neighbor sites , which can lead to two different types of ground states .
the first one is the antiferromagnetic state in which spins align oppositely at nearest neighbors .
the other type is the superfluid state with exotic symmetries where a fermion attracts another fermion with the opposite spin at nearest neighbors to form a cooper pair . under the current experimental condition @xcite ,
the strength of the dipolar ising interaction between nearest neighbors is approximately @xmath53 for the singlet rovibrational state , the hopping amplitude @xmath50 is close to @xmath54 @xmath55 @xcite for a typical optical lattice with the strength @xmath56 , where @xmath57 , @xmath58 , and the lattice constant is @xmath59 .
in contrast , the super - exchange interaction of an ordinary fermi gas is usually much smaller than the hopping amplitude @xmath50 , which makes it more difficult to produce ordered states .
the dipolar fermi gas has another advantage that the spin fluctuation in ising systems is much less than that in heisenberg systems , which is helpful to stabilize ordered states .
for these reasons , the superfluid and antiferromagnetic states may be easier to be created in the dipolar fermi gas than in an ordinary fermi gas . in the following we focus on the properties of
these ordered states at zero temperature .
to study the antiferromagnetic state , we divide the square lattice into two sublattices a and b so that nearest neighbors of one sublattice all belong to the other sublattice .
the antiferromagnetic order is described by the finite average of the pseudospin , @xmath60 for sublattice a and @xmath61 for sublattice b. in the @xmath62-space , the mean - field hamiltonian for the antiferromagnetic state is given by @xmath63 where @xmath64 is the total number of lattice sites , the @xmath62-summation is over the first brillouin zone of a sublattice , @xmath65 and @xmath66 are fermion annihilation operators for sublattice a and b , @xmath67 , and @xmath68 is the lattice constant .
the constant @xmath69 can be regarded as the strength of the effective antiferromagnetic interaction , @xmath70 .
the mean - field hamiltonian in eq .
( [ afh ] ) can be diagonalized by a standard canonical transformation , @xmath71 where the quasi - particle operators are given by @xmath72 , @xmath73 , @xmath74 , @xmath75 , and the transformation coefficients are given by @xmath76/2 $ ] .
the quasi - particles form two bands with energies given by @xmath77 the magnetic moment @xmath14 can be determined self - consistently from the equation @xmath78 which leads to the equation @xmath79 , \label{afsc}\ ] ] where @xmath80 is the chemical potential .
( [ afsc ] ) is equivalent to the energy - extreme condition @xmath81 , and can be solved together with the equation for the fermion number per site @xmath82 . \label{afne}\ ] ] the total energy per site is given by @xmath83+\frac{h m^2}{2}. \label{afen}\ ] ] at half filling @xmath84 , the r .-
h .- s . of eq .
( [ afsc ] ) is infraredly divergent at @xmath85 , indicating that the antiferromagnetic order always exists for any finite dipolar ising interaction .
the antiferromagnetic solution can also be obtained at other filling factors as long as the following condition is satisfied , @xmath86.\ ] ] however , as shown in fig .
[ figure1 ] , the compressibility of the antiferromagnetic state turns negative , indicating that the antiferromagnetic state is dynamically unstable beyond half filling . at @xmath87 .
the downward curvature shows that the compressibility is negative and antiferromagnetic state is unstable beyond half filling .
, width=302 ]
at the superfluid state , two fermions with opposite spins form a pair . in the hamiltonian in eq .
( [ hamiltonian 1 ] ) , this pairing interaction is given by @xmath88 .
in the @xmath62-space , the mean - field hamiltonian for the superfluid state is given by @xmath89 where the pairing gap is given by @xmath90 , @xmath91 , and @xmath92.\ ] ] the mean - field hamiltonian eq .
( [ mfsfhm 1 ] ) can be diagonalized by bogoliubov transformation as in the bcs theory , @xmath93 where @xmath94 is the quasiparticle energy , @xmath95 is the quasi - particle annihilation operator , and @xmath96 is the ground - state energy per site , @xmath97 the gap @xmath98 can be determined from the gap equation @xmath99 the fermion number per site is given by @xmath100 we numerically solve the gap equation ( [ gap ] ) together with eq .
( [ density 1 ] ) , and find that the gap @xmath98 can display different symmetries depending on the parameters .
the d - wave superfluid state appears in the region with @xmath101 and @xmath102 . as shown in fig .
[ figure2](a ) , the d - wave symmetry is characterized by the sign change in the gap at @xmath103 . the extended - s - wave ( @xmath1 ) superfluid state may appear for @xmath104 with the sign change of the gap at @xmath105 as shown in [ figure2](b ) . between the d - wave and extended s - wave phases
, a superfluid phase appears with a mixed @xmath2 symmetry in the order - parameter . in this phase , a global unitary transformation can be always applied so that the real part of the gap has the extended - s - wave symmetry , and the imaginary part of the gap has the d - wave symmetry , as shown in fig . [
figure2](c ) and [ figure2](d ) .
unlike the d - wave or extend - s - wave state , this mixed - symmetry state breaks the time - reversal symmetry , and supports the spontaneous current @xcite . versus the @xmath106 component of the wavevector @xmath107 .
the dashed and solid lines are for @xmath98 at @xmath108 and @xmath109 respectively .
( a ) at @xmath87 and @xmath110 , the gap @xmath98 shows the d - wave symmetry .
( b ) at @xmath87 and @xmath111 , the gap shows the extended - s - wave symmetry . at @xmath87 and @xmath112 , ( c )
the real part of the gap shows the extended - s - wave symmetry , and ( d ) the imaginary part of the gap shows d - wave symmetry.,width=302 ]
by comparing the energies of the antiferromagnetic and superfluid states , we obtain the zero - temperature phase diagram of this system , as shown in fig .
[ figure3 ] . at half filling , the antiferromagnetic state is the ground state , but it is unstable beyond half filling . when @xmath102 , the d - wave superfluid state appears at a smaller filling factor .
phase separation between the antiferromagnetic and d - wave superfluid states takes place in between . as the filling factor further decreases
, the d - wave superfluid state evolves into the mixed - symmetry ( @xmath2 ) state .
as the filling factor continues to decrease , this mixed - symmetry eventually evolves into the extended - s - wave superfluid state .
when @xmath113 , the above picture slightly changes .
the d - wave superfluid state no longer shows up .
as the filling factor decreases from half filling , a superfluid state appears with the mixed - symmetry .
phase separation between the antiferromagnetic and mixed - symmetry superfluid states occurs in the middle .
as the filling factor further decreases , this mixed - symmetry evolves into the extended - s - wave superfluid state . in the current experimental situation@xcite ,
the strength of the dipolar ising interaction is approximately @xmath114 for the singlet rovibrational state , and the hopping amplitude is approximately @xmath115 @xcite for a typical optical lattice with lattice constant @xmath59 . when the filling factor changes from zero to half filling , extended - s - wave , mixed - symmetry , and d - wave superfluid phases , the phase - separation region , and the antiferromagnetic state at half filling can all be reached as shown in fig .
[ figure3 ] .
the hopping amplitude @xmath50 can be tuned by increasing or decreasing the strength of the optical - lattice potential , and other regions in the phase diagram can also be reached .
so far we have considered only the case for vanishing onsite interaction @xmath52 . with finite @xmath51 , the phase diagram can be more complicated . if @xmath116 and @xmath117 , the antiferromagnetic state is more energetically favorable , and we expect a wider region of the phase separation between the antiferromagnetic and superfluid states in the phase diagram .
if @xmath118 , for low energy states there is effectively a constraint of no more than one fermion per site as discussed in the @xmath119 model @xcite , which can strongly affect the superfluid states .
if @xmath120 , a pure s - wave superfluid state may appear depending on the parameters .
all these cases are worth to explore in future works .
experimentally the onsite interaction leads to losses of the krb molecules through chemical reaction , which pose difficulties for creating the antiferromagnetic and superfluid phases proposed in this paper .
progress has been made to suppress the reaction rate in a quasi-2d geometry @xcite .
in other polar molecules systems , such as kcs , nak and narb , there are no reactive channels@xcite , and the loss problem can be avoided . with more experimental progresses , it is hopeful for observing the phase diagram shown in fig .
[ figure3 ] . in summary
, we have shown that in the presence of a resonant ac electric field , the dipole - dipole interaction in a dipolar fermi gas can be turned to a dipolar ising interaction between pseudospins .
when trapped in a square optical lattice , the dominant antiferromagnetic interaction between nearest neighbors drives the system into different ordered states depending on parameters .
the antiferromagnetic state is the ground state at half filling .
the superfluid state appears as the ground state at a smaller filling factor . in the middle
is the phase - separated region .
depending on the parameters , the superfluid order parameter displays different symmetries , such as the d - wave , extended - s - wave , and the mixed symmetry .
these states mentioned above may be detectable under experimental conditions .
we would like to thank t .-
l . ho for helpful discussions .
this work is supported by nsfc under grant no 10974004 .
99 k .- k .
et al _ , science * 322 * , 231 ( 2008 ) . l. d. carr and j. ye , new j. phys . * 11 * , 055009 ( 2009 ) .
m. baranov , phys .
rep . * 464 * , 71 ( 2008 ) .
a. micheli , g. pupillo , h. p. bchler and p. zoller , phys .
a * 76 * , 043604 ( 2007 ) .
l. you and m. marinescu , phys .
a * 60 * , 2324 ( 1999 ) .
n. r. cooper and g. v. shlyapnikov , phys .
rev . lett . *
103 * , 155302 ( 2009 ) .
lin , y .- t .
hsu , h. lee and d .- w .
wang , phys .
a * 81 * , 031601(r ) ( 2010 ) .
t lahaye , c menotti , l santos , m lewenstein and t pfau , rep .
. phys . * 72 * , 126401 ( 2009 ) . c. trefzger , c. menotti , b. capogrosso - sansone and m. lewenstein , cond - mat/1103.3145 . l. han and c. a. r. s de melo , cond - mat/1006.2072 . m. l. wall and l. d. carr , new j. phys . * 11 * , 055027 ( 2009 ) .
a. v. gorshkov _
et al _ , phys .
107 , 115301 ( 2011 ) . m. h. g. de miranda _ et al _ , nature phys . *
7 * , 502 ( 2011 ) .
p. s. zuchowski and j. m. hutson , phys .
a * 81 * , 060703 ( 2010 ) . | in a dipolar fermi gas , the dipole - dipole interaction between fermions can be turned into a dipolar ising interaction between pseudospins in the presence of an ac electric field . when trapped in a 2d optical lattice
, such a dipolar fermi gas has a very rich phase diagram at zero temperature , due to the competition between antiferromagnetism and superfluidity . at half filling
, the antiferromagnetic state is the favored ground state .
the superfluid state appears as the ground state at a smaller filling factor . in between
there is a phase - separated region .
the order parameter of the superfluid state can display different symmetries depending on the filling factor and interaction strength , including d - wave ( @xmath0 ) , extended s - wave ( @xmath1 ) , or their linear combination ( @xmath2 ) .
the implication for the current experiment is discussed . |
very recently , we have developed an effective field theory to describe the nonperturbative qcd dynamics of heavy mesons @xcite .
this theory has incorporated in it heavy quark effective theory ( hqet ) , and is thus consistent with heavy quark symmetry ( hqs ) in the infinite quark mass ( @xmath5 ) limit . in this effective field theory ,
a heavy meson is considered as a composite particle consisting of a reduced heavy quark ( heavy quark in the @xmath5 limit ) coupled with the light degrees of freedom . a structure function @xmath6 , corresponding to the wave function of the heavy meson bound state ,
is introduced to describe the quark - meson vertex in the effective lagrangian . by combining such an effective lagrangian for the composite heavy mesons with the @xmath7 expansion of the heavy quark qcd lagrangian
, we can systematically evaluate various @xmath7 corrections to heavy meson properties in a covariant field - theoretic framework @xcite . in our previous works
@xcite , the heavy meson wave function @xmath6 is taken to be a phenomenological function .
specifically it is assumed to be a lorentzian function , @xmath8 , in which the parameters @xmath9 and @xmath10 are determined phenomenologically by fitting to the @xmath11 meson decay constant @xmath12 ; hence its relation to nonperturbative low - energy qcd dynamics is not clear . in this paper
, we attempt to determine the heavy meson wave function from a qcd - based bound state equation .
the bound state equation of heavy mesons is obtain from the heavy quark equation of motion in hqet which is incorporated in our effective field theory .
this heavy meson bound state equation is a covariant extension of the light - front bound state equation we obtained recently from light - front qcd and hqet@xcite . with the bound state equation for @xmath6
, we can use a variational procedure to determine its soluton .
the paper is organized as follows : in sec .
ii , we derive from the heavy quark equation of motion in the heavy quark limit and our effective heavy meson field theory the heavy meson bound state equation . then a variational method is used to determine the heavy meson wave function @xmath6 by minimizing the reduced mass ( binding energy ) of the heavy meson @xmath0 . in the literature ,
@xmath0 has only been estimated either from heavy meson mass spectrum or from inclusive heavy meson decay data @xcite .
a direct calculation of @xmath0 has not been carried out to date . in sec .
iii , we consistently calculate the other two basic hqet parameters @xmath1 and @xmath2 directly from the operators @xmath13 and @xmath14 in hqet .
having self - consistently calculated @xmath1 and @xmath2 , we can then extract from experimental heavy meson masses the heavy quark masses ( @xmath3 , @xmath4 ) at the low energy scale where our effective field theory applies .
such a procedure establishes a link between our heavy meson effective field theory and nonperturbative qcd dynamics .
finally a summary is given in sec .
before deriving the heavy meson bound state equation , we shall briefly review the heavy meson effecive field theory we developed recently @xcite . in heavy quark effective theory ( hqet ) @xcite , the qcd lagrangian is expanded in powers of @xmath7 : @xmath15 where the reduced heavy quark field @xmath16 is defined from the full heavy quark field @xmath17 by @xmath18 @xmath19 is the heavy quark four velocity which is defined to be equal to that of the parent heavy meson , and @xmath20 .
it is easily seen that the leading term @xmath21 in eq.([1me ] ) possesses a @xmath22 flavor - spin symmetry ( or simply heavy quark symmetry , hqs ) , whereas the symmetry breaking part @xmath23 contains @xmath7 corrections to @xmath24 .
the effective lagrangian we have constructed to describe the low energy dynamics of heavy mesons reads @xcite : @xmath25 where @xmath26 and @xmath27 represent respectively the composite pseudoscalar and vector heavy meson fields which appear only as external states , @xmath28 is their common residual mass in the heavy quark limit , @xmath29 represents collectively the light degrees of freedom in a heavy meson , where @xmath30 is heavy meson wave function mentioned earlier , @xmath31 is the light antiquark momentum , and @xmath32 is the normalization constant given by @xmath33 in eq.([hqel ] ) , @xmath34 obeys hqs and defines the nonperturbative effective coupling of heavy mesons with their constituent heavy and light quarks , while @xmath23 , which contains all the @xmath7 corrections , is considered as a perturbation to @xmath34 .
a practical calculational scheme has been developed in terms of the feynman diagrammatic rules summarized below @xcite : \(i ) a heavy meson bound state ( @xmath35 ) couples to its constituent heavy and light quarks via a three - point vertex @xmath36 @xmath37 where @xmath38 for pseudoscalar @xmath39 , @xmath40 for vector mesons @xmath41 , and @xmath31 is the momentum of the light degrees of freedom .
each vertex also carries a momentum conservation factor of @xmath42 .
\(ii ) the internal heavy quark and light quark propagators are , @xmath43 respectively , where @xmath44 is the residual momentum of the heavy quark , and @xmath45 the constituent mass of the light quark .
\(iii ) an integration factor @xmath46 is associated with each internal momentum @xmath47 , and a factor @xmath48 should be included for each fermion loop .
\(iv ) for all other lines and vertices not attached to heavy mesons ( mostly from @xmath23 ) , standard feynman rules apply .
hence @xmath7 corrections to various quantities can be calculated using standard field theoretic perturbative approach @xcite . in hqet
, the reduced heavy quark field @xmath16 obeys the following equation of motion @xmath49 where @xmath50 is the qcd covariant derivative . to ensure that our effective field theory , eq.([hqel ] ) , is consistent with low energy qcd dynamics of hqet , eq.([1me ] ) , we require that in the heavy quark limit , the heavy meson state @xmath51 satisfies @xmath52 strickly speaking , this matrix element must be calculated in a nonperturbative manner of qcd , which is however beyond our present day knowledge of qcd . to circumvent this predicament , we shall adopt an effective gluon propagator which contains a long - distance contribution as well as the familiar perturbative one @xcite : @xmath53 where @xmath54 is the qcd coupling constant , and the coefficient @xmath55 characterizes the confining strength .
note that the regularized infrared singular term @xmath56 can be effectively considered as a regularization of the linearly confining interaction @xmath57 @xcite . representing this effective gluon propagator by a double wavy line
, we can depict the left hand side of eq .
( [ hqem ] ) pictorially by feynman diagrams shown in fig . 1 . as discussed in @xcite , from the constraint that it forbids on - mass - shell decay of the heavy meson into @xmath58 , a possible form of the heavy meson wave function @xmath59
is given by @xmath60 where the function @xmath61 is regular at @xmath62 .
then , applying the diagrammatic rules stated in sec .
iia to fig . 1
, we can easily translate eq.([hqem ] ) into the following expression , @xmath63 where the color factor @xmath64 for @xmath65 .
the above equation can be rewritten as an equivalent bound state equation for heavy mesons : @xmath66 which is indeed a covariant generalization of the light - front bound state equation for heavy mesons we have previously derived @xcite , and it is also the heavy quark limit of the bethe - salpeter equation for heavy mesons . using the normalization condition of @xmath59
, we can simplified eq.([eqm2 ] ) to @xmath67 the physical meaning of the above result is clear .
the heavy meson binding energy @xmath0 defined by eq.([lmb ] ) consists of two parts : one is the light quark kinetic energy contribution @xmath68 , and the other is the interaction energy of the heavy - light system . note that , due to the effective gluon propagator ( [ egp ] ) used , the interaction energy contains a confining contribution , as well as the color coulomb contribution . in this section
we shall determine the heavy meson wave function @xmath69 through a variational procedure .
the first step is to choose an appropriate trial wave function . as discussed in @xcite , we demand
that the wave function should be relatively well behaved , by which we mean that , when continued into the complex @xmath31 plane , ( i ) @xmath61 is analytic except for isolated singularities , and ( ii ) it vanishes fast enough " as @xmath70 .
these two constraints also allow us to easily derive various relativisitic and/or light - front quark models results as special cases in our effective field theory @xcite . in various existing light - front or relativistic quark models , two types of distribution functions
have been widely adopted for phenomenological hadronic wave functions .
one is the exponential(gaussian)-type and the other the lorentzian - type .
the former can be excluded since it does not simultaneously satisfy the two constraints stated above @xcite . on the other hand , we find that lorentzian - type functions of the form , @xmath71 do meet our requirements @xcite , and will be taken as the working trial heavy meson wave function in this work .
it is interesting to note that such a choice of @xmath69 corresponds to a covariant light - front wave function @xcite of the form @xmath72 where @xmath73 is a normalization constant , @xmath74 , and @xmath75 is required to ensure convergence . for the sake of demonstration
, we first consider the naive case of ignoring of the confining interaction between the heavy quark and the light degrees of freedom , _
i.e. _ @xmath76 , so that color coulomb interaction is solely responsible for binding the quarks together .
we find that , for @xmath77 , a minimum with @xmath78 gev can be found at @xmath79 .
however a wave function with such a small @xmath9 is unrealistic because it contains too little high momentum components . in other words , the wave function
is too extended in configuration space , as would be expected since we have ignored the confining part of the inter - quark interaction . with an appropriate non - zero @xmath80 , physically sensible results can be obtained . in
what follows , we will take @xmath81 and @xmath82 gev , which are close to the values used in our previous work @xcite . with the confinement strength taken to be @xmath83 gev
, we determine the wave function parameter @xmath9 by minimizing @xmath0 .
the results are listed in the table i , where @xmath84 which is not sensitive to the choice of @xmath10 .
this value of @xmath0 is somewhat larger than our previous result of @xmath85 gev obtained in @xcite .
the difference is mainly due to the confinement effect which is not considered there .
0.5 cm it is worth emphasizing that , in our previous work @xcite , the parameter @xmath9 is fixed phenomenologically by fitting to @xmath86 mev from lattice gauge calculations ( it is experimentally unknown ) . in the present work , we have dynamically
determine the wave function @xmath87 by minimizing the blinding energy @xmath0 from the qcd - based bound state equation .
this is a significant improvement which helps to raise the predictive power of our heavy meson effective theory .
in this section , we shall use the bound state solution obtained in the previous section to calculate heavy meson masses up to order @xmath7 .
this will in turn yield more consistent results for @xmath1 and @xmath2 than those obtained in @xcite . in the heavy quark limit
, the heavy meson masses can be written as @xmath88 .
the correction to @xmath89 in hqet comes mainly from the leading hqs - breaking @xmath7 corrections [ see eq .
( [ hqel ] ) ] : @xmath90 where @xmath91 $ ] and @xmath92 $ ] . with these @xmath7 corrections
included , the heavy meson masses can be expressed as @xmath93 where @xmath1 comes from @xmath13 , @xmath94 from @xmath14 ( see fig .
3 ) , and @xmath95 for pseudoscalar ( vector ) mesons @xcite . in eq .
( [ mass ] ) , @xmath1 parametrizes the common mass shift , and @xmath2 accounts for the so called hyperfine mass splitting between the vector and pseudoscalar heavy meson .
they can be directly calculated once the heavy meson wave function @xmath96 is determined from the variational approach presented in sec .
. 0.5 cm table ii .
the wave function parameter @xmath9 is variationally determined by minimizing @xmath0 , and all the other hqet parameters are calculated self consistently in the heavy meson effective theory with @xmath82 gev and @xmath81 .
@xmath110 , and @xmath0 are in gev , and @xmath1 and @xmath2 in gev@xmath111 . [ cols="^,^,^,^,^,^,^,^,^",options="header " , ] @xmath112 in the literature , the so - called experimental " value @xmath113 gev@xmath111 is obtained by assuming @xmath114 .
one would get @xmath115 gev@xmath111 if the pole - mass @xmath116 gev is used in eq .
( [ msp ] ) for @xmath117 mev .
@xmath118 there are no experimental data for the constituent quark masses @xmath3 and @xmath4 , and the pole masses are displaced @xcite
. 0.5 cm
we know that hqs and confinement mechanism are crucial ingredients of nonperturbative qcd dynamics in heavy hadrons
. however they are rarely incorporated in quark models in a fully covariant manner . in this work ,
we attempt to establish a link between quark model and low - energy qcd dynamics in an effective field theoretic approach , in which the heavy meson wave function is determined variationally .
our effective theory preserves the simplicity of a conventional quark model ; in fact , for heavy quark transitions , we can reproduce the covariant light - front quark model @xcite as a special case .
the light - front quark model has been widely used to describe heavy meson structures . however , because its relation to qcd is not clear , it is incapable of describing the hadronic mass spectrum . moreover ,
the model is not fully covariant , so that it can only describe physical processes with space - like momentum transfers @xcite . in the present formalism
, we can study hadronic bound state physics in a covariant framework ; moreover , a connection to low - energy qcd dynamics has also been established .
thus our heavy meson effective theory , together with the nonperturbatively determinated heavy meson wave function , provides a useful qcd based framework in which to study heavy quark physics in the low energy regime .
we have adopted a self - consistent procedure to determine the heavy meson wave function , the binding energy @xmath0 , as well as its @xmath7 corrections @xmath1 and @xmath2 .
> from these hqet parameters , we can determine the heavy quark masses @xmath3 and @xmath4 in the low energy regime by fitting to the @xmath119 and @xmath120 hyperfine mass differences . for these static properties ,
our results are consistent with other studies on inclusive b - meson decays @xcite .
a great advantage of our formalism is that we can unambiguously study hadronic transition processes at arbitrary momentum transfers . using the heavy meson wave function determined in this paper
, we can further calculate dynamically the heavy meson decay constant ( which is usually considered as an input in various quark models for heavy mesons ) and other hadronic properties such as the isgur - wise function , @xmath121 and @xmath122 coupling constants , and also their @xmath7 corrections .
the results will be presented in forthcoming papers . c. y. cheung and w. m. zhang , mod .
. lett . * a13 * , 2163 ( 1998 ) .
h. y. cheng , c. y. cheung , and w. m. zhang , phys . rev . * d58 * , 74003 ( 1998 ) .
w. m. zhang , phys . rev . *
d56 * , 1528 ( 1997 ) .
m. gremm et al .
lett . * 77 * , 20 ( 1996 ) ; v. chernyak , phys . lett . * b387 * , 173 ( 1996 ) ; a. f. fack , m. like and m. j. savage , phys . rev . * d53 * , 6316 ( 1996 ) ; m. gremm and i. stewart , _ idib _ , * d55 * , 1226 ( 1997 ) ; h. li and h .- l .
yu , _ idib _ , * d55 * , 2833 ( 1997 )
. h. georgi , phys . lett . * b240 * , 447 ( 1990 ) ; e. eichten and b. hill , phys . lett . * b234 * , 511 ( 1990 ) ; * b243 * , 427 ( 1990 ) ; also see review article by m. neubert , phys .
* 245 * , 261 ( 1994 ) .
s. mandelstam , phys .
rev . * d20 * , 3223 ( 1979 ) ; u. bar - gadda , nucl . phys . * b163 * , 312 ( 1980 ) ; h. munczek and a. nemirocsky , phys . rev .
* d28 * , 181 ( 1983 ) h. y. cheng , c. y. cheung , c. w. hwang , and w. m. zhang , phys . rev . *
d57 * , 5598 ( 1998 ) . c. y. cheung , c. w. hwang and w. m. zhang , z. phys .
* c75 * , 657 ( 1997 ) . h. y. cheng , c. y. cheung , and c. w. hwang , phys . rev . *
d55 * , 1559 ( 1997 ) .
particle data group , phys . rev . *
d54 * , 1 ( 1996 ) .
m. gremm et al . , in ref.@xcite h. fusaoka and y. koide , phys . rev .
* d57 * , 3886 ( 1998 ) .
fig . 1 : : diagram for heavy meson bound states : ( a ) kinetic energy contribution and ( b ) effective one - gluon exchange contribution which contains the columbus interaction plus confining interaction in heavy quark symmetry . | in this paper we obtain a heavy meson bound state equation from the heavy quark equation of motion in heavy quark effective theory ( hqet ) and the heavy meson effective field theory we developed very recently .
the bound state equation is a covariant extention of the light - front bound state equation for heavy mesons derived from light - front qcd and hqet .
we determine the covariant heavy meson wave function variationally by minimizing the binding energy @xmath0 . subsequently the other basic hqet parameters @xmath1 and @xmath2 , and the heavy quark masses @xmath3 and @xmath4 can also be consistently determined . |
widespread attention has been paid in recent years to a class of optical field states that are called squeezed states @xcite , where the fluctuation in one quadrature is less than that of the vacuum state @xcite .
squeezed states have manifestly nonclassical properties , which may find application in low - noise optical communications as well as in the gravitational - wave detection @xcite .
the two - mode squeezed state @xcite , which is composed by idler mode and signal mode resulting from a parametric down conversion amplifier @xcite , is a typical entangled state of continuous variable@xcite .
recently , many groups have devoted to the research on nonclassical properties of two - mode squeezed state@xcite .
lee @xcite and cave group @xcite have studied the nonclassical photon statistics of two - mode squeezed states , respectively . theoretically , two - mode squeezed vacuum state is constructed by the two - mode squeezing operator @xmath2 $ ] acting on the vacuum state @xmath3 , @xmath4 where @xmath5 is a squeezing parameter . considering the coordinate and momentum operators , respectively , @xmath6 one can recast @xmath7 into the form @xmath8 .
\label{03}\ ] ] in the state @xmath9 , the variances of two - mode quadrature operators@xcite @xmath10 take the standard form , i.e. , @xmath11 in the present paper , we shall extend the two - mode squeezed states to the three - mode case and investigate their some nonclassical properties .
based on the fan s ideas @xcite , we construct the following unitary operator in the three - mode fock space@xmath12 \right \ } .
\label{06}\ ] ] then a question naturally arises : is the operator @xmath13 also a three - mode squeezing operator ?
if yes , what is its corresponding squeezed state ?
what is particular nonclassical properties of these squeezed states ? facing with these questions , this work is arranged as follows . in sec.2 , to answer the above questions , we first derive the normally ordered form of @xmath13 by virtue of the technique of integration within an ordered product ( iwop ) of operators and analyze if the squeezing exists
. then the corresponding squeezed states @xmath14 are obtained , where @xmath15 is a three - mode coherent state . in secs.3 and 4 , we concentrate our attention on studying the higher - order squeezing and higher - order sub - possonian photon statistics in the state @xmath16 . we find that @xmath17 can be squeezed to all even orders and the variances of the three - mode quadrature operators exhibit stronger squeezing comparing with the shown case in ref.@xcite .
so we call @xmath13 the three - mode enhancing squeezing operator
. we devote sec.5 to evaluate the violation of bell - inequality for the state @xmath14 using the formalism of the wigner representation in phase space based on the parity operator and the displacement operation .
finally , a brief summary is given in sec.6 .
in this section , we derive the normally ordered expansion @xmath13 and then obtain the corresponding squeezed state .
because these operators @xmath18 , @xmath19 , @xmath20 , @xmath21 , @xmath22 , @xmath23 in eq.([06 ] ) do not make up a closed lie algebra , we utilize the iwop technique @xcite to disentangle @xmath13 . for this purpose , we rewrite @xmath13 using the matrix form as @xmath24 = \exp \left [ i\lambda q_{j}a_{jk}p_{k}\right ] \label{b1}\ ] ] with @xmath25{ccc}0 & 1 & 1\\ 1 & 0 & 1\\ 1 & 1 & 0 \end{array } \right ) , \text { \ } j , k=1,2,3 , \label{b2}\ ] ] here and henceforth the repeated indices represent the einstein summation notation .
appealing to the baker - hausdorff formula@xmath26 + \frac{1}{2}\left [ c,\left [ c , b\right ] \right ] + \frac{1}{3}\left [ c,\left [ c,\left [ c , b\right ] \right ] \right ] + \cdots , \label{4}\ ] ] where @xmath27 and @xmath28 are operators , one can calculate out @xmath29 + \cdots \nonumber \\ & = q_{i}\left ( e^{-\mathcal{\lambda a}}\right ) _ { ik}=\left ( e^{-\lambda \mathcal{a}}\right ) _ { ki}q_{i } , \label{5}\ ] ] and @xmath30 + \cdots \nonumber \\ & = \left ( e^{\lambda \mathcal{a}}\right ) _ { ki}p_{i}. \label{6}\ ] ] it is implied from eqs.([5 ] ) and ( [ 6 ] ) that the action of @xmath13 on the three - mode coordinate eigenstate @xmath31,@xmath32whose expression in fock space is @xcite @xmath33\left \vert \vec{0}\right \rangle , \label{10}\ ] ] leads to @xmath34 where @xmath35 , @xmath36 and tilde represents transposition of matrix . according to properties of matrix @xmath37 in eq.([b2 ] ) , by expanding the exponential term @xmath38 , we get @xmath39{ccc}u_{1 } & v_{1 } & v_{1}\\ v_{1 } & u_{1 } & v_{1}\\ v_{1 } & v_{1 } & u_{1}\end{array } \right ) , \label{8 - 0}\ ] ] where@xmath40 similarly , @xmath41{ccc}u_{2 } & v_{2 } & v_{2}\\ v_{2 } & u_{2 } & v_{2}\\ v_{2 } & v_{2 } & u_{2}\end{array } \right ) , \label{8 - 1}\ ] ] where @xmath42 which will be used in the rest of this article . from eq.([7 ] ) , it is found that @xmath13 has natural representation in the coordinate basis @xmath31 @xmath43 it then follows that @xmath44 which is consistent with eq.([5 ] ) . substituting eq.([10 ] ) into eq.([9 ] ) and using the iwop technique as well as three - mode vacuum projector @xmath45 \colon,$ ] we easily get the normally ordered form of @xmath13 after integrating out of @xmath46 @xmath47 \nonumber \\ & \times \colon \exp \left [ \tilde{a}^{\dagger}\left ( \lambda n^{-1}-i\right ) a\right ] \colon \exp \left [ \frac{1}{2}\tilde{a}\left ( n^{-1}-i\right ) a\right ] , \label{11}\ ] ] where @xmath48 .
here we have used the mathematical formula @xmath49 = \pi^{n/2}(\det w)^{1/2}\exp \left ( \frac{1}{4}\tilde{v}w^{-1}v\right ) .\ ] ] finally , applying the operator @xmath13 in eq.([11 ] ) on the three - mode coherent state @xmath50 , we obtain the three - mode squeezed coherent state , denoted by @xmath51 , as follows @xmath52 \left \vert \vec{\alpha}\right \rangle .
\label{a1}\ ] ] when @xmath53 , eq.([a1 ] ) reduces to the three - mode squeezed vacuum state @xmath54 \left \vert \vec { 0}\right \rangle . \label{12}\ ] ]
it is well known that the concept of higher - order squeezing , introduced by hong and mandel @xcite , is another aspect for revealing nonclassical charactertics of a quantum state @xcite .
when @xmath55-th moment in a state is less than that in the coherent state , this state is said to be squeezed to order @xmath55 . here , we evaluate the higher - order squeezed behavior of the three - mode squeezed coherent state @xmath51 .
firstly , we derive the expression of the @xmath55-th moment of quadrature operators in the state @xmath14 .
the quadratures in the three - mode case are defined as @xmath56 and @xmath57 obeying @xmath58 = \frac{1}{2}i$ ] . letting @xmath59 and using @xmath60 = \delta_{jk}$ ]
, one can easily have @xmath61 , \label{b3}\ ] ] where @xmath62 , @xmath63 , @xmath64 and @xmath65 are @xmath27 number and @xmath66denotes normal ordering . by expanding both sides of eq.([b3 ] ) as power series in @xmath65 , and equating coefficients of @xmath67 , it is obtained that @xmath68 which is a basic formula in calculating @xmath55-th moment of two - mode fields . considering eqs.([5 ] ) and ( [ 6 ] ) , we calculate the @xmath55-th moment of the three - mode quadrature @xmath69 in the state @xmath17 , i.e. , @xmath70 \right \ } ^{2m}\left \vert \vec{\alpha}\right \rangle \nonumber \\ & = \left ( \frac{1}{12}\right ) ^{m}\left \langle \vec{\alpha}\right \vert ( \delta q)^{2m}\left \vert \vec{\alpha}\right \rangle , \label{15}\ ] ] where @xmath71 .
\label{15 - 1}\ ] ] as a result of eqs.([b3 - 1 ] ) and ( [ 8 - 0 ] ) , we obtain @xmath72 further , we have @xmath73 ^{2}-\left \langle q\right \rangle ^{2}\right \ } ^{m - k}. \label{18 - 1}\ ] ] on the other hand , @xmath74 \right \ } \colon \left \vert \vec{\alpha}\right \rangle \right ) ^{2}=\left [ \sum_{k=1}^{3}\left ( e^{-\lambda \mathcal{a}}\right ) _ { jk}\left ( \alpha_{j}+\alpha_{j}^{\ast } \right ) \right ] ^{2}. \label{18 - 2}\ ] ] by putting eq.([18 - 2 ] ) into eq.([18 - 1 ] ) , it is easily obtained that @xmath75 thus , considering eqs.([18 ] ) and ( [ 18 - 3 ] ) , the final expression for @xmath55-order squeezing of @xmath69 in eq.([15 ] ) is @xmath76 where @xmath77 and we have used the result of eq.([8 - 0 ] ) , i.e. , @xmath78 similarly , the @xmath55-order squeezing for @xmath79 is expressed as@xmath80 from eqs.([c1 ] ) and ( [ 24 ] ) , it is easily seen that when @xmath81 , @xmath82 which is just @xmath55-th moment in the coherent state . according to the definition of higher - order squeezing
, @xmath17 can be squeezed to all even orders .
in particular , when @xmath83 , eqs.([c1 ] ) and ( [ 24 ] ) reduce to the usual quadrature squeezing for the state @xmath51 , namely , @xmath84 and @xmath85 which shows that @xmath13 is also a three - mode squeezing operator .
in addition , eq.([b9 ] ) clearly indicates that the squeezed states @xmath86 exhibit stronger squeezing @xmath87 in one quadrature than that @xmath88 of the usual squeezed state in eq.([05 ] ) .
this is a way of enhanced squeezing .
that is the reason why we call @xmath13 three - mode enhanced squeezing operator .
in this section , we focus on investigating higher - order sub - poissonian statistics in the state @xmath14 .
the concept of higher - order sub - poissonian statistics was introduced in @xcite in terms of factorial moment . by using @xmath89 , with @xmath90
, a parameter @xmath91 can be defined as @xmath92 with @xmath93 is a positive integer . for all @xmath94
, @xmath95 corresponds to the poisson distribution .
the negative or positive values of parameter @xmath91 for @xmath96 indicate higher - order sub - poissonian or higher - order super - poissonian statistics , respectively . for the three - mode case of higher - order sub - poissonian statistics in @xmath14 , analogue to the single - mode case , we can define the joint operator @xmath97 to study the joint sub - poissonian distribution of @xmath16 . referring to eqs.([5 ] ) and ( [ 6 ] ) , the similar transformation of @xmath98 , @xmath99 under the operator @xmath13 is @xmath100 where @xmath101 according to coherent completeness @xmath102 with @xmath103 , we acquire the following equation @xmath104 ^{k}\left \vert \vec{\beta}\right \rangle \left \langle \vec{\beta}\right \vert
\left [ ua+va^{\dagger}\right ] ^{k}\left \vert \vec{\alpha}\right \rangle . \label{27}\ ] ] it is easy to check @xmath105 = 1 $ ] , which is fit for the following equation @xcite for @xmath98 in normal ordering form @xmath106 where @xmath107 is the @xmath108-order hermite polynomial . by putting eq([28 ] ) into [ 27 ] ) , we have @xmath109 \label{29}\ ] ] where @xmath110 and@xmath111 noticing that hermite polynomial can be gotten from differential of mother function , i.e. , @xmath112 eq.([29 ] ) can be simplified as @xmath113 \right \vert _ { t_{1},t_{2}=0}. \label{32}\ ] ] through integrating out variables @xmath114 it is obtained that @xmath115 where @xmath116 a careful observation of eq.([33 ] ) reveals that if we get rid of @xmath117 in the exponent , the rest just are the mother function of hermite polynomial .
fortunately we can express @xmath118 in the following way @xmath119 \right \vert _ { t_{1},t_{2}=0 } \label{35}\ ] ] then substituting eq.([35 ] ) into eq.([33 ] ) and considering the property of hermite polynomial eq.([31 ] ) , i.e. , @xmath120 we get the expression for the average of @xmath121 in the state @xmath14
@xmath122 ^{2}n!}h_{k - n}\left [ \frac{g}{2}\right ] h_{k - n}\left [ \frac{m}{2}\right ] .
\label{37}\ ] ] especially , when @xmath123 , noticing eqs.([26 - 1 ] ) and ( [ 37 ] ) , one has@xmath124 \sinh \left ( 4\lambda \right ) , \label{38}\ ] ] where @xmath125 for @xmath126 , @xmath127 \nonumber \\ & = \sinh^{2}\left ( -4\lambda \right ) \left [ 2^{-5}\tanh^{2}\left ( -2\lambda \right ) + 2gm\tanh \left ( 2\lambda \right ) + \left ( g^{2}-2\right ) \left ( m^{2}-2\right ) \right ] \label{40}\ ] ] where @xmath128 , \label{41}\ ] ] and @xmath129 .
\label{41 - 1}\ ] ] [ fig1 ] changing with @xmath130 by letting @xmath131 to reflect the sub - poissonian distribution of @xmath14,title="fig:",width=377 ] then introducing eqs.([38 ] ) and ( [ 40 ] ) into ( [ 25 ] ) for @xmath132 we easily obtain the result of @xmath133 . in order to clearly obverse the sub - poissonian statistics in the state @xmath134 we plot fig.1 to visualize the change of @xmath133 with @xmath130 by simply letting @xmath135 .
it clearly shows that the parameter @xmath133 @xmath136 when @xmath137 but unchanges along with @xmath138 .
the negative value of @xmath133 displays the sub - poissonian distribution of @xmath14 . according to ref.@xcite
, the higher - order sub - poissonian statistics may become useful for their use in detection of higher - order squeezing .
the quantum nonlocality for continuous variable states has attracted much attention .
wigner function representation of the bell - inequality has been developed using a parity operator as a quantum observable @xcite . in this section
, we turn our attention to evaluate the violation of the bell - inequality for squeezed coherent states @xmath17 using the formalism of the wigner representation in phase space based on the parity operator and the displacement operation .
for the three - mode system , the correlation function is the expectation of the operator @xmath139 which is a equivalent definition of the wigner function , namely @xmath140 where @xmath141 and @xmath142 $ ] is phase - space displacement operators acting on mode @xmath143 . thus @xmath144 is a product of displaced parity operators given as @xmath145 corresponding to the measurement of an even ( parity @xmath146 ) or an odd ( parity @xmath147 ) number of photons in mode @xmath143 . within the framework of local realistic theories ,
the wigner representation of bell inequality is of the form @xcite @xmath148 where @xmath149 is combination of @xmath150 .
for the three - mode case , @xmath151 @xmath152 , bell - inequality can be expressed as @xmath153 \label{45}\ ] ] next , in order to examine the violation of the bell - inequality for the state @xmath14 , we first derive the explicit expression of its wigner function . recalling that in refs.@xcite we have introduced the weyl ordering form of single - mode wigner operator @xmath154 @xmath155{c}\colon \\
\colon \end{array } \delta \left ( q_{1}-q_{1}\right ) \delta \left ( p_{1}-p_{1}\right ) \begin{array } [ c]{c}\colon \\
\colon \end{array } , \label{46}\ ] ] and its normal ordering form is @xmath156 \colon . \label{47}\ ] ] where symbols @xmath157{c}\colon:\\ \colon : \end{array } $ ] denotes weyl ordering .
note that the order of bose operators @xmath158 within a normally ordered product and a weyl ordered product can be permuted .
the weyl ordering has a remarkable property , i.e. , the order - invariance of weyl ordered operators under similar transformations which means @xmath159{c}\colon \\
\colon \end{array } f\left ( a , a^{\dagger}\right ) \begin{array } [ c]{c}\colon \\ \colon \end{array }
\mathcal{u=\begin{array } [ c]{c}\colon \\ \colon \end{array } } \left .
\mathcal{u}^{\mathcal{-}1}f\left ( a , a^{\dagger}\right ) \mathcal{u}\right . \begin{array } [ c]{c}\colon \\ \colon \end{array } , \label{48}\ ] ] as if the fence @xmath157{c}\colon:\\ \colon : \end{array } $ ] does not exist . for three - mode case , the weyl ordering form of the wigner operator
is @xmath160{c}\colon \\ \colon \end{array } } \delta \left ( \vec{q}-\vec{q}\right ) \delta \left ( \vec{p}-\vec{p}\right ) \mathcal{\begin{array } [ c]{c}\colon \\ \colon \end{array } } \right . , \label{49}\ ] ] where @xmath161 and @xmath162 .
thus , according to the principle of weyl ordering invariance under similar transformations as well as eqs.([5 ] ) , ( [ 6 ] ) , ( [ 49 ] ) , we have @xmath163{c}\colon \\
\colon \end{array } \delta \left ( \vec{q}-\vec{q}\right ) \delta \left ( \vec{p}-\vec{p}\right ) \begin{array } [ c]{c}\colon \\ \colon \end{array } s_{3}\left \vert \vec{\alpha}\right \rangle \nonumber \\ & = \left \langle \vec{\alpha}\right \vert \delta \left ( e^{-\lambda \mathcal{a}}\vec{q},e^{\lambda \mathcal{a}}\vec{p}\right ) \left \vert \vec{\alpha } \right \rangle \label{50}\ ] ] further , considering eqs.([8 - 0 ] ) , ( [ 8 - 1 ] ) and ( [ 47 ] ) , it is obtained that @xmath164 \colon \left \vert \vec{\alpha}\right \rangle \nonumber \\ & = \frac{1}{\pi^{3}}\exp \left [ -\left ( e^{-\lambda \mathcal{a}}\vec{q}-\frac{1}{\sqrt{2}}\left ( \vec{\alpha}+\vec{\alpha}^{\ast}\right ) \right ) ^{2}-\left ( e^{\lambda \mathcal{a}}\vec{q}-\frac{1}{i\sqrt{2}}\left ( \vec{\alpha}-\vec{\alpha}^{\ast}\right ) \right ) ^{2}\right ] \nonumber \\ & = \frac{1}{\pi^{3}}\exp \left [ -\lambda_{ji}\lambda_{ik}q_{j}q_{k}-\sum_{j=1}^{3}\sigma_{j}^{2}+2q_{k}\lambda_{kj}\sigma_{j}-\gamma_{ji}\gamma_{ik}p_{j}p_{k}-\sum_{j=1}^{3}\chi_{j}^{2}+2p_{i}\gamma_{kj}\chi _ { j}\right ] \nonumber \\ & = \frac{1}{\pi^{3}}\exp \left \ { -\sum_{j=1}^{3}\left [ \left ( u_{1}^{2}+2v_{1}^{2}\right ) q_{j}^{2}-\sigma_{j}^{2}+2u_{1}q_{j}\sigma_{j}\right ] -2\sum_{j > k=1}^{3}\left [ \left ( 2u_{1}v_{1}+v_{1}^{2}\right ) q_{j}q_{k}+2v_{1}q_{j}\sigma_{k}\right ] \right . \nonumber \\ & \left .
-\sum_{j=1}^{3}\left [ \left ( u_{2}^{2}+2v_{2}^{2}\right ) p_{j}^{2}-\chi_{j}^{2}+2u_{2}p_{j}\chi_{j}\right ] -2\sum_{j > k=1}^{3}\left [ \left ( 2u_{2}v_{2}+v_{2}^{2}\right ) p_{j}p_{k}+2v_{2}p_{j}\chi_{k}\right ] \right \ } , \label{52}\ ] ] where @xmath165 and @xmath166 .
especially , when @xmath167 , eq.([52 ] ) reduces to the expression of wigner function for the three - mode squeezed vacuum state @xmath168 in eq.([12 ] ) .
[ fig2 - 1 ] for the state @xmath16 by using parity measurements with @xmath169 , @xmath170 , @xmath171 , @xmath172 , @xmath173 , and @xmath174 . ,
title="fig:",width=302 ] now we put eq.([52 ] ) with corresponding variables into three - mode bell - inequality eq.([45 ] ) .
this bell - inequality has @xmath175 variables and it is highly nontrivial to find the global maximum values of @xmath176 for all @xmath175 variables . to deal with this problem ,
in fig.2 we plot the maximal bell violation of the state @xmath17 as a function of @xmath5 by defining @xmath169 , @xmath170 , @xmath171 , @xmath177 , @xmath173 , and @xmath178 ( @xmath179 is a positive constant associated with the displacement magnitude ) . from fig.2
, we see that when the @xmath180 , @xmath181 , the violation of the bell - inequality exists for positive squeezing parameter .
another interesting phenomenon reflected by fig.2 is that @xmath182 will convergent to a constant more than @xmath183 independent of change of @xmath5 .
in summary , we generalize the usual two - mode squeezed state to three - mode operator @xmath13 and present the normal ordered form by the iwop technique .
then , we apply the operator @xmath13 on three - mode coherent state @xmath184 to get a new squeezed state @xmath16 . it is found that the state @xmath16 displays enhanced higher - order squeezed property , which proves that @xmath13 is an enhanced squeezing operator .
we also formulate the explicit expression of @xmath91 in order to describe higher - order sub - poissonian statistics by taking @xmath126 as an example .
finally , we discover that the state @xmath14 violates the bell - inequality by using the formalism of the wigner representation .
this work was supported by the national natural science foundation of china under grant numbers 10775097 and 10874174 . | by extending the usual two - mode squeezing operator @xmath0 $ ] to the three - mode squeezing operator @xmath1 \right \ } $ ] , we obtain the corresponding three - mode squeezed coherent state . the state s higher - order properties , such as higher - order squeezing and higher - order sub - possonian photon statistics , are investigated .
it is found that the new squeezed state not only can be squeezed to all even orders but also exhibits squeezing enhancement comparing with the usual cases .
in addition , we examine the violation of bell - inequality for the three - mode squeezed states by using the formalism of wigner representation .
pacs : 03.65.-w ; 03.65.ud keywords : three - mode squeezed operator ; higher - order squeezing ; higher - order sub - poissonian statistics ; violation of bell - inequality |
in this note we will describe some of the properties of a @xmath0 symmetric generalization of dimensional reduction , applied to @xmath1 gauge theory in @xmath2 dimension .
the motivation is to construct an effective action , which is constrained by perturbative dimensional reduction at high temperature , but which can be applied all the way down to the confinement phase transition .
the partition function is defined as follows : @xmath3\text{d}[v]\exp\bigl(-s_{eff}\bigr)\\ s_{eff}&=s_u+s_{u , v}+s_v \end{split}\ ] ] @xmath4 the @xmath1 matrices @xmath5 and @xmath6 are defined on the sites and the links respectively of a two dimensional lattice .
the matrices @xmath7 represent the polyakov loops of the full @xmath8 theory .
similar models have been proposed in @xcite .
here we discuss the phase structure of the above model , and compare with results obtained within the standard reduction recipes @xcite .
for simplicity we set @xmath9 and discuss the phase structure in the @xmath10 plane .
the number @xmath11 of time slices in ( 2 + 1)d , which enters as a parameter in @xmath12 is fixed to 4 throughout this paper .
the model was studied using conventional lattice qcd monte - carlo techniques .
the multi hit metropolis algorithm was used both for updating the gauge fields @xmath13 and polyakov loops fields @xmath7 .
lattices of sizes ranging from @xmath14 to @xmath15 were studied .
the main quantities measured were the traces of polyakov loops @xmath16 and their two point correlators .
we started with simulations at @xmath17 for a range of @xmath18 values and got the expected signal of the phase transition between the low - temperature confined @xmath19 symmetric phase ( ia ) and a broken high temperature phase ( ii ) .
screening masses @xmath20 were obtained by fitting the two point polyakov loop unconnected correlators with the formula @xmath21 we plot the result of the fit in figure [ fig : massnp ] . on the same plot
we show the results obtained in @xcite with the naively reduced model ( no higgs potential ) . at sufficiently large temperature ( large @xmath18 ) the agreement is good , but the @xmath0 symmetric model also describes the vanishing of the screening mass associated with @xmath0 restoration at lower temperatures .
we then performed a series of scans in the @xmath18@xmath23 plane on @xmath14 lattices and looked for peaks in the susceptibility @xmath24 as indications of phase transitions , while the associated distributions of @xmath25 were used to characterize the phases .
after so identifying a tentative phase diagram ( see figure [ fig : pd ] ) , one sees a strong transition signal just above @xmath27 . there the size dependence of @xmath28 at its maximum , as well as the time history of @xmath29 and its doubly peaked histogram clearly indicate that the transition is first order .
the evidence for a second phase transition , slightly above @xmath30 is considerably weaker .
the peak is less pronounced and we had to go to @xmath15 lattices ( with much poorer statistics ) to make sure that it grows with the lattice size . in order to study the nature of the phases , we looked at the distributions of the polyakov loops @xmath25 .
the regions marked ia and ii on figure [ fig : pd ] are easily identified with standard @xmath19 symmetric and broken phases , with data concentrated respectively around @xmath31 and @xmath32 . at larger @xmath23 in region iii ,
a new pattern shows up , with again @xmath0 breaking peaks , but the argument of @xmath25 is now close to @xmath33 , as illustrated on figure [ fig : hist ] . finally ,
when @xmath23 is further increased ( region ib ) , the histogram is located around @xmath34 , no particular argument is selected , and we conjecture that the phase is the same as the low temperature confined one ( ia ) .
along @xmath26 , we have used the projected ( 0momentum ) correlators to extract the masses , by fitting them to the formula : @xmath37 our results are summarized in figure [ fig : mass ] .
the mass value is found to coincide with that measured at @xmath26 , @xmath38 in the original(2 + 1 ) model @xcite for @xmath23 close to 0.2 .
this point in parameter space is a stable point of the phase diagram , at variance with what happens in the conventionally reduced model , where the physical situation corresponds to a metastable point @xcite . in conclusion ,
the @xmath0 symmetric model opens a possibility to extend dimensional reduction down to the temperature where confinement is restored .
it exhibits an interesting phase diagram , with a new phase whose connection to qcd at finite temperature and/or chemical potential should be investigated . for this purpose ,
a study to constrain the model parameters is under way .
r. d. pisarski , phys .
d * 62 * , 111501 ( 2000 ) [ arxiv : hep - ph/0006205 ] .
p. bialas , a. morel , b. petersson , k. petrov and t. reisz , nucl .
b * 581 * ( 2000 ) 477 [ arxiv : hep - lat/0003004 ] .
p. bialas , a. morel , b. petersson , k. petrov and t. reisz , nucl .
b * 603 * , 369 ( 2001 ) [ arxiv : hep - lat/0012019 ] . | we present first results from a numerical investigation of a @xmath0 symmetric model based on dimensional reduction . |
in classical constituent quark models , baryons are ascribed as three - quark ( @xmath0 ) states and mesons are ascribed as quark - anti - quark ( @xmath1 ) states .
this picture is very successful in explaining properties of the spatial ground states of the flavor su(3 ) vector meson nonet , baryon octet and decuplet .
its predicted @xmath2 baryon with mass around 1670 mev was discovered by later experiments . however even for the lowest spatial excited states , this picture failed badly in both meson and baryon sectors . in the meson sector ,
the lowest spatial excited su(3 ) nonet is the scalar nonet composed of @xmath3 , @xmath4 , @xmath5 and @xmath6 . in the classical constituent quark models ,
these scalars should be @xmath7 states with @xmath3 as @xmath8 state , @xmath9 as @xmath10 state and @xmath6 as mainly @xmath11 state . then in this picture
, it can not explain why the mass of @xmath5 is degenerate with @xmath6 instead of close to @xmath3 as in the @xmath12-@xmath13 case in the vector nonet .
this made r.j.jaffe @xcite proposing these scalars are @xmath14 states instead of @xmath1 states . in the new picture , the @xmath3
is ascribed as @xmath15\overline{[ud]}$ ] state , @xmath9 as @xmath16\overline{[us]}-[ds]\overline{[ds]})/\sqrt{2}$ ] state and @xmath6 as @xmath16\overline{[us]}+[ds]\overline{[ds]})/\sqrt{2}$ ] state .
this gives a natural explanation of the degeneracy of @xmath5 and @xmath6 .
$ ] means a good diquark with configuration of flavor representation @xmath18 , spin 0 and color @xmath19 .
alternatively , these scalars are also proposed to be meson - meson dynamically generated states @xcite . in the baryon sector
, the similar thing seems also happening @xcite . in the classical quark models ,
the excited baryon states are described as excitation of individual constituent quarks , similar to the cases for atomic and nuclear excitations .
the lowest spatial excited baryon is expected to be a ( @xmath20 ) @xmath21 state with one quark in orbital angular momentum @xmath22 state , and spin - parity @xmath23 .
however , experimentally , the lowest negative parity @xmath21 resonance is found to be @xmath24 , which is heavier than two other spatial excited baryons : @xmath25 and @xmath26 .
this is the long - standing mass reverse problem for the lowest spatial excited baryons . in the simple 3q constituent quark models ,
it is also difficult to understand the strange decay properties of the @xmath24 , which seems to couple strongly to the final states with strangeness . besides a large coupling to @xmath27 , a large value of @xmath28
is deduced @xcite by a simultaneous fit to bes data on @xmath29 , @xmath30 , and cosy data on @xmath31 .
there is also evidence for large @xmath32 coupling from @xmath33 reaction at clas @xcite and @xmath34 reaction @xcite , and large @xmath35 coupling from @xmath36 , @xmath37 and @xmath38 reactions @xcite .
the third difficulty is the strange decay pattern of another member of the @xmath23-nonet , @xmath39 , which has its coupling to @xmath40 much larger than @xmath41 and @xmath42 according to its branching ratios listed in pdg @xcite .
all these difficulties can be easily understood by considering large 5-quark components in them @xcite .
the @xmath24 could be the lowest @xmath22 orbital excited @xmath43 state with a large admixture of @xmath44[us]\bar s>$ ] pentaquark component having @xmath15 $ ] , @xmath45 $ ] and @xmath46 in the ground state .
the @xmath26 could be the lowest radial excited @xmath43 state with a large admixture of @xmath44[ud]\bar d>$ ] pentaquark component having two @xmath15 $ ] diquarks in the relative p - wave . while the lowest @xmath22 orbital excited @xmath43 state should have a mass lower than the lowest radial excited @xmath43 state , the @xmath44[us]\bar s>$ ] pentaquark component
has a higher mass than @xmath44[ud]\bar d>$ ] pentaquark component .
the lighter @xmath47 is also understandable in this picture .
its main 5-quark configuration is @xmath44[qs]\bar q>$ ] which is lighter than the corresponding 5-quark configuration @xmath44[us]\bar s>$ ] in the @xmath48 .
the large mixture of the @xmath44[us]\bar s>$ ] pentaquark component in the @xmath24 naturally results in its large couplings to the @xmath27 , @xmath49 , @xmath50 and @xmath51 .
the main 5-quark configuration for the @xmath39 is @xmath52[ds]\bar s>$ ] which makes it heavier than other @xmath23 states and larger coupling to @xmath40 . besides the penta - quark configurations with the diquark correlation
, the penta - quark system may also be in the form of meson - baryon states .
the @xmath24 , @xmath25 and some other baryon resonances are proposed to be meson - baryon dynamically generated states @xcite .
however , a challenge for this meson - baryon dynamical picture is to explain the mass and decay pattern of the @xmath39 . from above facts and discussion for both meson and baryon sectors , one can see that unlike atomic and nuclear excitations where the number of constituent particles are fixed , the favorable hadronic excitation mechanism for the lowest spatial excited states in light quark sector seems to be dragging out a light @xmath1 pair from gluon field rather than to excite a constituent quark to be @xmath22 state .
a breathing mode of @xmath53 is proposed @xcite for the lowest @xmath23 baryon nonet .
each baryon is a mixture of the three - quark and five - quark components .
while the new picture gives a nice account of properties of scalar meson nonet and well - established members of the lowest @xmath23 baryon nonet , it is necessary to check its distinguishable predictions of other members in the @xmath23 baryon nonet .
while the classical quenched quark models @xcite predict the @xmath23 @xmath54 and @xmath55 to be around 1650 mev and 1760 mev , respectively , the unquenched quark models @xcite expect them to be around 1400 mev and 1550 mev , respectively , and meson - baryon dynamical models @xcite predict them to be around 1450 mev and 1620 mev , respectively . for @xmath56 resonance with @xmath57 ,
the pdg @xcite lists a two - star @xmath58 resonance , which seems to support the quenched quark models .
however , only four references listed in pdg show weak evidence for its existence . in ref .
@xcite , the total cross sections for @xmath59 and @xmath60 are analyzed , which indicates some @xmath56 resonance around 1600 mev without @xmath61 quantum number .
@xcite are based on multichannel analysis of the @xmath62 reactions .
both claim evidence for a @xmath63 resonance with mass around 1620 mev but give contradicted coupling properties to @xmath64 and to @xmath65 .
other later multichannel analyses of the @xmath62 reactions support the existence of an @xmath66 @xcite .
@xcite analyzes the reaction @xmath67 and gives two comparable solutions with and without @xmath68 . on the other hand , there are also some supports of the unquenched 5-quark models with @xmath69 of much lower masses
. the re - analysis of old data on @xmath70 finds hidden @xmath69 with mass around 1380 mev under the @xmath71 peak @xcite . from an analysis of the recent leps data on @xmath72
@xcite , there is also a possibility for the existence of such low mass @xmath69 @xcite .
an analysis of cebaf data on @xmath73 also suggests a possible @xmath69 around 1400 mev @xcite .
to clarify the situation for @xmath56 resonances , recently , a combined fit for the new cb data @xcite on @xmath74 together with the old data @xcite on @xmath75 for the energies from 1569 to 1676 mev was performed @xcite .
the @xmath76 reaction is the best channel available for the study of the @xmath56 resonances because the @xmath64 is a pure isospin 1 channel .
the high precision crystal ball @xmath77 polarization data @xcite are crucial for discriminating @xmath68 from @xmath78 .
it shows that the @xmath79 is definitely needed , while @xmath68 is not needed at all .
although @xmath80 is not demanded in this analysis , it can not be excluded .
therefore , no evidence to support the classical quenched quark models at all from the @xmath23 baryons .
additional @xmath81 , @xmath82 and @xmath83 may exist .
the @xmath81 is consistent with the resonance structure @xmath84 or @xmath85 in pdg @xcite and seems a good isospin partner of @xmath86 .
recently a very interesting narrow @xmath87 with a width about 1.5 mev was claimed from an analysis of @xmath88 data @xcite .
together with @xmath89 and either @xmath90 or @xmath91 , they fit in a nice @xmath92 baryon nonet with large penta - quark configuration , _
i.e. _ , @xmath93 as @xmath44\{uq\}\bar q>$ ] state , @xmath94 as @xmath44\{sq\}\bar q>$ ] state , @xmath95 as @xmath44\{ss\}\bar s>$ ] state , and @xmath96 as @xmath44\{ss\}\bar q>$ ] state
. here @xmath97 means a diquark with configuration of flavor representation @xmath98 , spin 1 and color @xmath19 .
the @xmath95 as @xmath44\{ss\}\bar s>$ ] state gives a natural explanation for its dominant @xmath99 decay mode with a very narrow width due to its very small phase space meanwhile a d - wave decay .
the available information on the hadron spectroscopy with strangeness strongly indicates that @xmath100 in s - state is more favorable than @xmath0 state with @xmath22 and @xmath101 in s - state is more favorable than @xmath1 state with @xmath22 .
the multi - quark components are very substantial and important for hadronic excited states .
even @xmath102 configuration may play dominant role for some baryon resonances @xcite . to further establish the multi - quark picture for hadronic excited states ,
it is very important to complete the low - lying hyperon spectrum , especially the @xmath23 and @xmath92 @xmath54 , @xmath55 and @xmath103 .
here the @xmath103 spectrum has a unique advantage that the favorable @xmath1 excitations from quark sea have different flavor from the valence strange quarks @xcite .
kaon beam experiment at jparc and hyperon production from charmonium decays at besiii may play very important role in this aspect .
it is also important to check the cases with @xmath104 quarks replaced by @xmath105 or @xmath106 quarks .
various pictures and dynamics for the spectroscopy with strangeness can be extended to and checked by its charm and beauty partners .
for example , if @xmath6 is a @xmath107 molecule mainly due to light vector meson exchange force @xcite , then with the same mechanism there should also exist @xmath108 , @xmath109 , @xmath110 and @xmath111 molecules @xcite .
the newly established @xmath112 is regarded as a @xmath108 molecule or tetra - quark state by many people @xcite .
the @xmath113 was proposed to be a @xmath114 molecule @xcite ; now the newly established @xmath115 is regarded as its @xmath116 partner @xcite .
the @xmath117 was proposed @xcite to be @xmath118 molecule as the charm partner of @xmath119 .
although many hadron resonances were proposed to be hadron - hadron dynamically generated states or multi - quark states , most of them can not be clearly distinguished from classical quark model states due to tunable ingredients and possible large mixing of various configurations in these models . even in 2010 ,
the pdg @xcite still claimed that
the clean @xmath120 spectrum has in fact been taken to settle the decades - long discussion about the nature of the @xmath119 true 3-quark state or mere @xmath121 threshold effect ?
unambiguously in favor of the first interpretation .
" a possible solution to this problem is to extend the penta - quark study to the hidden charm and hidden beauty sectors . if the @xmath24 is the @xmath122 quasi - bound state with hidden strangeness , then naturally by replacing @xmath123 by @xmath124 or @xmath125 one would expect super - heavy @xmath21 states with hidden charm and hidden beauty just below @xmath126 and @xmath127 thresholds , respectively .
following the valencia approach of ref.@xcite and extending it to the hidden charm sector , the interaction between various charmed mesons and charmed baryons were studied with the local hidden gauge formalism in refs.@xcite .
several meson - baryon dynamically generated narrow @xmath21 and @xmath128 resonances with hidden charm are predicted with mass around 4.3 gev and width smaller than 100 mev .
the s - wave @xmath129 and @xmath130 states with isospin i=1/2 and spin s=1/2 were also investigated by various other approaches @xcite .
they confirm that the interaction between @xmath131 and @xmath132 is attractive and results in a @xmath129 bound state not far below threshold .
the low - lying energy spectra of five quark systems @xmath133 ( @xmath134=@xmath135 , @xmath136=@xmath137 ) and @xmath138 ( @xmath134=@xmath137 , @xmath136=@xmath139 ) are also investigated with three kinds of schematic interactions : the chromomagnetic interaction , the flavor - spin dependent interaction and the instanton - induced interaction @xcite . in all the three models , the lowest five quark state ( @xmath133 or @xmath138 )
has an orbital angular momentum @xmath140 and the spin - parity @xmath141 ; the mass of the lowest @xmath138 state is heavier than the lowest @xmath133 state , which is different from the prediction of meson - baryon dynamical model @xcite .
the predicted new resonances definitely can not be accommodated by quark models with three constituent quarks . because these predicted states have masses above @xmath142 and @xmath143 thresholds , they can be looked for at the forthcoming panda / fair and jlab 12-gev upgrade experiments .
this is an advantage for their experimental searches , compared with those baryons with hidden charms below the @xmath142 threshold proposed by other earlier approaches @xcite .
the same meson - baryon coupled channel unitary approach with the local hidden gauge formalism was extended to the hidden beauty sector in ref.@xcite .
two @xmath144 states and four @xmath145 states were predicted to be dynamically generated .
because of the hidden @xmath146 components involved in these states , the masses of these states are all above 11 gev while their widths are of only a few mev , which should form part of the heaviest island for the quite stable @xmath21 and @xmath128 baryons . for the valencia approach
, the static limit is assumed for the t - channel exchange of light vector mesons by neglecting momentum dependent terms . in order to investigate the possible influence of the momentum dependent terms ,
the conventional schrodinger equation approach was also used to study possible bound states for the @xmath127 channel by keeping the momentum dependent terms in the t - channel meson exchange potential .
it was found that within the reasonable model parameter range the two approaches give consistent predictions about possible bound states .
this gives some justification of the simple valencia approach although there could be an uncertainty of 10 - 20 mev for the binding energies .
production cross sections of the predicted @xmath147 resonances in @xmath148 and @xmath149 collisions were estimated as a guide for the possible experimental search at relevant facilities in the future . for the @xmath150 reaction , the best center - of - mass energy for observing the predicted @xmath147 is @xmath151 gev , where the production cross section is about 0.01 nb . for the @xmath152 reaction ,
when the center - of - mass energy is larger than 14 gev , the production cross section should be larger than 0.1 nb .
nowadays , the luminosity for pp or ep collisions can reach @xmath153 , this will produce more than 1000 events per day for the @xmath147 production .
it is expected that future facilities , such as proposed electron - ion collider ( eic ) , may discover these very interesting super - heavy @xmath21 and @xmath128 with hidden beauty .
very recently , the observation of the iso - vector meson partners of the predicted @xmath147 , @xmath154 and @xmath155 , were reported by belle collaboration @xcite .
this gives us stronger confidence on the existence of the super - heavy island for the @xmath147 and @xmath156 resonances .
available information on hadron spectroscopy with strangeness and charm reveals unquenched quark picture . dragging out a @xmath1 from gluon field is a very important excitation mechanism for hadrons . to correctly describe the hadron spectrum ,
it is necessary to go beyond the classical quenched quark models which assuming a fixed number of constituent quarks .
distinguishable prediction for hyperon spectroscopy from the new picture is yelling for experimental confirmation .
kaon beam experiments at jparc and hyperon production data from charmonium decays at besiii can play very important role here .
super - heavy narrow @xmath21 and @xmath128 resonances are predicted by various models to exist around 4.3 gev and 11 gev for hidden charm and beauty , respectively .
their iso - vector meson partners @xmath154 and @xmath155 have recently been observed .
experimental confirmation of them will unambiguously establish multi - quark dynamics .
they can be looked for at cebaf-12gev - upgrade at jlab and panda at fair , maybe also at jparc , super - b , rhic , eic .
i thank c. s. an , p. z. gao , j. he , f. huang , t. s. h. lee , r. molina , e. oset , j. shi , w. l. wang , k. w. wei , j. j. wu , h. s. xu , s. g. yuan , l. zhao , z. y. zhang for collaboration works reviewed here .
this work is supported by the national natural science foundation of china under grant 11035006 , 11121092 , 11261130311 ( crc110 by dfg and nsfc ) , the chinese academy of sciences under project no.kjcx2-ew-n01 and the ministry of science and technology of china ( 2009cb825200 ) .
a. s. carroll _
et al . _ ,
* 37 * ( 1976 ) 806 .
j. k. kim , phys .
* 27 * ( 1971 ) 356 .
w. langbein , f. wagner , nucl .
phys . b47 ( 1972 ) 477 .
w. a. morris _
et al . _ ,
d17 ( 1978 ) 55 .
j. j. wu , s. dulat , b. s. zou , phys .
* d80 * ( 2009 ) 017503 ; phys .
* c81 * ( 2010 ) 045210 .
k. hicks _ et al _ ( leps collaboration ) , phys .
( 2009 ) 012501 .
p. gao , j. j. wu , b. s. zou , phys .
* c81 * ( 2010 ) 055203 .
a. gal , h. garcilazo , nucl .
* a864 * ( 2011 ) 153 .
a. m. torres , k.p .
khemchandani , e. oset , phys .
* c77 * ( 2008 ) 042203 .
s. g. yuan , c. s. an , k. w. wei , b. s. zou , h. s. xu , arxiv:1208.1742 [ hep - ph ] .
s. krewald , r. h. lemmer and f. p. sassen , phys .
d * 69 * ( 2004 ) 016003 .
y. j. zhang , h. c. chiang , p. n. shen , b. s. zou , phys .
d * 74 * ( 2006 ) 014013 .
f. k. guo , p. n. shen , h. c. chiang , r. g. ping and b. s. zou , phys .
b * 641 * ( 2006 ) 278 .
e. klempt and a. zaitsev , phys .
* 454 * ( 2007 ) 1 , and references therein .
n. a. tornqvist , phys .
* 67 * ( 1991 ) 556 .
x. liu , z. g. luo , y. r. liu and s. l. zhu , eur .
j. c * 61 * ( 2009 ) 411 ; phys .
d84 ( 2011 ) 054002 .
t. fernandez - carames , a. valcarce and j. vijande , phys .
* 103 * ( 2009 ) 222001 .
l. tolos et al . , chin
. phys .
c * 33 * ( 2009 ) 1323 ; j. haidenbauer et al .
j a47 ( 2011 ) 18 .
k. nakamura _
_ ( particle data group ) , j. phys .
g * 37 * ( 2010 ) 075021 .
e. oset , a. ramos , euro .
j. a * 44 * ( 2010 ) 445 .
j. j. wu , r. molina , e. oset and b. s. zou , phys . rev .
lett . *
105 * ( 2010 ) 232001 .
j. j. wu , r. molina , e. oset and b. s. zou , phys .
c * 84 * ( 2011 ) 015202 . | quarks of different flavors have different masses , which will cause breaking of flavor symmetries of qcd .
flavor symmetries and their breaking in hadron spectroscopy play important role for understanding the internal structures of hadrons .
hadron spectroscopy with strangeness reveals the importance of unquenched quark dynamics .
systematic study of hadron spectroscopy with strange , charm and beauty quarks would be very revealing and essential for understanding the internal structure of hadrons and its underlying quark dynamics . |
the study of manganese oxides is receiving considerable attention in recent years both in its theoretical and experimental aspects @xcite . from the technological viewpoint
, these materials could be used in the preparation of high - sensitive magnetic - field sensors due to their colossal magnetoresistance ( cmr ) phenomena .
in addition , researchers in the condensed matter field have been interested in the rich phase diagram of these materials originating from the competition and interplay among charge , spin , and orbital degrees of freedoms .
obtaining a unified picture for this rich phase diagram is a challenging open problem .
a prototype for the theoretical investigation of manganese oxides is the double - exchange ( de ) framework , describing the hopping motion of e@xmath2-electrons ferromagnetically coupled to localized t@xmath3-spins .
this idea has conceptually explained the appearance of ferromagnetism when holes are doped @xcite .
in addition , within the one orbital model , the existence of phase separation has been recently unveiled with the use of modern numerical techniques @xcite , leading to a potential explanation of the cmr effect @xcite .
however , in order to understand the fine details of the phase diagram of manganites , the one - orbital model is not sufficient since the highly nontrivial a - type spin antiferro ( af ) and c - type orbital structures observed experimentally in the undoped material lamno@xmath0@xcite can not be properly addressed in such a simple context .
certainly two - orbital models are needed to consider the nontrivial state of undoped manganites . in this framework the two - band model without phonons has been studied before , and the importance of the strong coulomb repulsion has been remarked for the appearance of the a - af state @xcite .
this is based upon the belief that the competition between kinetic and strong correlation effect determines the optimal orbital for e@xmath2-electrons and the lattice will be simply distorted to reproduce such optimal orbitals .
however , coulombic approaches have presented conflicting results regarding the orbital order that coexists with the a - type spin state , with several approaches predicting g - type orbital order , which is not observed in practice .
while it is certainly correct that the orbital degrees of freedom play an essential role for the stabilization of a - af , it should be noticed that our understanding is still incomplete . in particular , it is important to remark that the orbital structure is tightly related to the jahn - teller ( jt ) distortion of mno@xmath1 octahedron .
if each jt distortion would occur independently , optimal orbitals can be determined by minimizing the kinetic and interaction energy of e@xmath2-electrons .
however , oxygens are shared between adjacent mno@xmath1 octahedra , indicating that the jt distortions occurs cooperatively . especially in the undoped situation , all mno@xmath1 octahedra exhibit jt distortion , indicating that the cooperative effect is very important , as discussed by kanamori @xcite .
thus , in order to understand the magnetic and orbital structures in lamno@xmath0 , it is indispensable to optimize simultaneously the electron and lattice systems .
however , not much effort has been devoted to the microscopic treatment of the cooperative effect @xcite , although the jt effect in the manganese oxide has been studied by several groups @xcite .
then , in the present work , a careful investigation of this problem is performed with some numerical techniques , focusing on @xmath4 , where @xmath5 is the electron number per site . in this paper ,
the optimal oxygen positions are determined by a relaxation technique to obtain the lattice distortions corresponding to several t@xmath3-spin magnetic structures .
in addition , monte carlo ( mc ) simulations were also performed to investigate the spin and orbital structure without _ a priori _ assumptions for their order .
it is found that a - af , as well as the c - type orbital structure , occurs in realistic parameter regions for lamno@xmath0 , i.e. , large hund coupling between the e@xmath2-electron and t@xmath3-spin , small af interaction between t@xmath3-spins , and strong electron - lattice coupling .
it should be emphasized that our results are obtained without the coulomb interaction .
it is shown in a simple case that the optimized results are essentially unchanged even if the coulomb interaction is included explicitly in the model .
the organization of this paper is as follows .
section 2 is devoted to the formulation to include the cooperative effect in the two - orbital model tightly coupled to the jt distortion , and some technical points are briefly discussed . in sec . 3 ,
the results on the magnetic and orbital structures are provided and it is shown that the region of a - af in the magnetic phase diagram is reasonable for lamno@xmath0 , since the couplings needed agree with experiments . in sec . 4 ,
a prescription to obtain the c - type orbital order with the alternation of @xmath6 and @xmath7 orbitals is provided . finally , in sec . 5 ,
the effect of the coulomb interaction on the lattice distortion is discussed in the ferromagnetic state . throughout this paper , units such that @xmath8 are used .
let us consider the motion of e@xmath2-electrons tightly coupled to the localized t@xmath3-spins and the local distortions of the mno@xmath1 octahedra .
this situation is well described by @xmath9 here the first term indicates the two - orbital hamiltonian , given by , @xmath10 where @xmath11 ( @xmath12 ) is the annihilation operator for an e@xmath2-electron with spin @xmath13 in the @xmath14 ( @xmath15 ) orbital at site @xmath16 .
the vector connecting nearest - neighbor sites is @xmath17 , @xmath18 is the hopping amplitude between @xmath19- and @xmath20-orbitals connecting nearest - neighbor sites along the @xmath17-direction via the oxygen 2@xmath21-orbital , @xmath22 is the hund coupling , @xmath23 the localized classical t@xmath3-spin normalized to @xmath24 , and @xmath25 are the pauli matrices . the second term is needed to account for the afm character of the manganese oxide , given by @xmath26 where @xmath27 is the af - coupling between nearest - neighbor t@xmath3-spins . in the third term , the coupling of e@xmath2-electrons to the distortion of mno@xmath1 octahedron is considered as @xmath28 , \end{aligned}\ ] ] where @xmath29 is the electron - phonon coupling constant , @xmath30 denotes the distortion for the breathing mode of the mno@xmath1 octahedron , @xmath31 and @xmath32 are , respectively , jt distortions for the @xmath33- and @xmath34-type modes , and @xmath35 is the @xmath36 unit matrix .
spring constants for breathing- and jt - modes are denoted by @xmath37 and @xmath38 , respectively .
the final term indicates the coulomb interactions between e@xmath2-electrons . as mentioned in sec . 1
, since only the jt phonons are investigated , the coulomb interactions are neglected .
this point will be discussed later in the text ( sec .
5 ) where the explicit form of @xmath39 is provided and briefly analyzed .
1.5truein -0.8truein = 3truein -1.0truein as shown in fig .
1(a ) , oxygens are shared between adjacent octahedra , indicating that the local lattice distortions can not be treated independently and a cooperative analysis is needed for this problem .
for this purpose , the normal coordinates for distortions of the mno@xmath1 octahedron , shown in fig .
1(b ) , are written as @xcite @xmath40 for the breathing mode , @xmath41 and @xmath42 for the jt modes , where @xmath43 is given by @xmath44 the first term indicates the deviation from the cubic lattice , given by @xmath45 , where @xmath46 is the length between mn - ions along the @xmath17-axis and @xmath47 .
the second term is the contribution from the shift of oxygen position , expressed by @xmath48 , where @xmath49 is the deviation of oxygen from the equilibrium position along the mn - mn bond in the @xmath17-direction . by this consideration ,
the cooperative jt distortion as well as the macroscopic lattice deformation is reasonably taken into account .
note that the buckling and rotational modes of mno@xmath1 octahedron are not explicitly included in this work .
in general , @xmath46 can be different for each direction , depending on the bulk properties of the lattice .
since the present work focuses on the microscopic mechanism for a - af formation in lamno@xmath0 , the undistorted lattice with @xmath50 is treated first , and then corrections will be added . in the cubic undistorted lattice ,
the hopping amplitudes are given by @xcite @xmath51 for the @xmath52-direction , @xmath53 for the @xmath54-direction , and @xmath55 for the @xmath56-direction . throughout this paper ,
the energy unit is @xmath57 .
corresponding to this choice of the energy unit , the length in the lattice distortion is scaled by @xmath58 . as a result of this scaling ,
a non - dimensional electron - phonon coupling constant @xmath59 is defined as @xmath60 it is noted that this coupling constant can be related to the static jt energy , which is conventionally defined by @xmath61 , as @xmath62 note also that the present length scale is rewritten as @xmath63 where @xmath64 is the characteristic length for the jt distortion . from the experimental result , @xmath65 is estimated as @xmath66 @xcite , which is a typical length in this context . as for the spring constant for the breathing mode ,
it is expressed as @xmath67 and the ratio @xmath68 is treated as a parameter . if it is plausibly assumed that the reduced masses for those modes are equal , this ratio is given by @xmath69 , where @xmath70 and @xmath71 are the vibration energies for manganite breathing- and jt - modes , respectively . from experimental results and band - calculation data , @xmath70 and @xmath71 are , respectively , estimated as @xmath72@xmath73 and @xmath74-@xmath75@xmath73@xcite .
then , throughout this work , @xmath68 is taken as @xmath76 , although the results presented here are basically unchanged as long as @xmath68 is larger than unity . in this work ,
the change in @xmath57 due to the displacement of oxygen position is not taken into account , but such an effect is shown to be very small as follows . due to the pseudo - potential theory @xcite , the exact hopping amplitude , for example , along the @xmath52-direction between a - orbitals in @xmath16 and @xmath77 sites , is expressed as @xmath78 with @xmath79 .
it should be remarked that the change in @xmath80 due to the oxygen shift is of the order of @xmath81 , not of the order of @xmath82 .
since @xmath82 is estimated to be at most a few percent , the change is considered to be negligible .
thus , in the present work , such a change is not included to avoid unnecessary complication in the calculation .
however , when the distorted lattice is considered , namely , when the deviation from @xmath50 is taken into account , the change in the hopping matrix due to this distortion should be included , because the effect is of the order of @xmath83 in this case , not of the order of @xmath84 this point will be discussed again in sec .
4 . -1truein
-3.2truein = 2.8truein -6.0truein to study hamiltonian eq .
( [ hamiltonian ] ) without @xmath39 , two numerical techniques have been applied .
one is the relaxation technique , in which the optimal positions of the oxygens are determined by minimizing the total energy . in this calculation , only the stretching mode for the octahedron , is taken into account .
moreover , the relaxation has been performed for fixed structures of the t@xmath3-spins such as ferro ( f ) , a - type af ( a - af ) , c - type af ( c - af ) , and g - type af ( g - af ) , shown in fig . 2 .
the advantage of this method is that the optimal orbital structure can be rapidly obtained on small clusters .
however , the assumptions involved in the relaxation procedure should be checked with an independent method .
such a check is performed with the unbiased mc simulations used before for one- and two - dimensional clusters using non - cooperative jt - phonons @xcite .
the dominant magnetic and orbital structures are deduced from correlation functions . in the mc method , the clusters currently reachable are @xmath85 , @xmath86 , and @xmath87 . in spite of this size limitation , arising from the large number of degrees of freedom in the problem , the available clusters are sufficient for our mostly qualitative purposes .
in addition , the remarkable agreement between mc and relaxation methods lead us to believe that our results are representative of the bulk limit .
0.6truein -0.5truein = 5truein -2.5truein in fig . 3 ,
the mean - energy is presented as a function of @xmath27 for @xmath88 and @xmath89 , on a @xmath85 cluster with open boundary conditions .
the solid lines and symbols indicate the results obtained with the relaxation technique and mc simulations , respectively .
the agreement is excellent , showing that the relaxation method is accurate .
the small deviations between the results of the two techniques are caused by temperature effects .
as intuitively expected , with increasing @xmath27 the optimal magnetic structure changes from ferro- to antiferromagnetic , and this occurs in the order f@xmath90a - af@xmath90c - af@xmath90g - af . to check size effects ,
the t@xmath3-spin correlation function @xmath91 was calculated also in @xmath86 and @xmath87 clusters , where @xmath92 here @xmath93 is the number of sites and @xmath94 indicates the thermal average value . as shown in fig . 4 , with increasing @xmath27 the dominant correlation changes in the order of @xmath95 , @xmath96 , @xmath97 , and @xmath98 .
the values of @xmath27 at which the spin structures changes agree well with those in fig .
0.6truein -0.5truein = 5truein -2.5truein the shapes of the occupied orbital arrangement with the lowest energy for f , a - af , c - af , and g - af magnetic structures are shown in fig .
5 . for the f - case ,
the g - type orbital structure is naively expected , because it is believed that the ferromagnetic spin structure is favored by the af orbital configuration . however , a more complicated orbital structure is stabilized in the actual calculation , indicating the importance of the cooperative treatment for jt - phonons . for the a - af state ,
only the c - type structure is depicted in fig . 5 , but the g - type structure , obtained by a @xmath99/2-rotation of the upper @xmath100-@xmath101 plane of the c - type state , was found to have _ exactly _ the same energy .
small corrections will remove this degeneracy in favor of the c - type as described in the next section . for c- and g - af ,
the obtained orbital structures are g- and c - types , respectively .
note that there exists an additional triplet degeneracy in the a - type state due to the cubic symmetry for each magnetic structure : if axes are changed cyclically ( @xmath102 ) , the optimized orbital structure is also transformed by this cyclic change , but the energy is invariant .
then , the magnetic and orbital structure in lamno@xmath0 occurs through a _ spontaneous _ symmetry breaking process .
although the same change of the magnetic structure due to @xmath27 was already reported in the electronic model with purely coulomb interactions @xcite , the orbital structures in those previous calculations were g- , g-
, a- , and a - type for the f- , a - af , c - af , and g - af spin states , respectively . note that for the a - af state , of relevance for the undoped manganites , the g - type order was obtained @xcite , although in another treatment for the coulomb interaction , the c- and g - type structures were found to be degenerate @xcite , as in our calculation .
thus , the stabilization in experiments of the c - type orbital structure is still puzzling both in the jt and coulomb mechanisms . this point will be discussed later in the text . -1.7truein
-0.35truein = 6.5truein -2.8truein in figs .
6(a ) and ( b ) , the phase diagrams on the @xmath103-plane are shown for @xmath104 and @xmath105 , respectively .
the curves are drawn by the relaxation method . as expected , the f - region becomes wider with increasing @xmath22 .
when @xmath59 is increased at fixed @xmath22 , the magnetic structure changes from f@xmath90a - af@xmath90c - af@xmath90g - af . this tendency is qualitatively understood if the two - site problem is considered in the limit @xmath106 and @xmath107 .
the energy - gain due to the second - order hopping process of e@xmath2-electrons is roughly @xmath108 and @xmath109 for af- and f - spin pairs , respectively .
increasing @xmath110 , @xmath111 decreases , indicating the relative stabilization of the af - phase . in our phase diagram
, the a - af phase appears for @xmath112 and @xmath113 .
this region does not depend much on @xmath22 , as long as @xmath106 .
although @xmath59 seems to be large , it is realistic from an experimental viewpoint : @xmath110 is @xmath114ev from photoemission experiments @xcite and @xmath57 is estimated as @xmath115ev @xcite , leading to @xmath116 . as for @xmath27 ,
it is estimated as @xmath117 @xcite .
thus , the location in parameter - space of the a - af state found here is reasonable when compared with experimental results for lamno@xmath0 . -0.2truein
0.5truein = 3.6truein -1truein
let us now focus on the orbital structure in the a - af phase .
in the cubic lattice studied thus far , the c- and g - type orbital structures are degenerate , and it is unclear whether the orbital pattern in the @xmath100-@xmath101 plane corresponds to the alternation of @xmath6 and @xmath7 orbitals observed in experiments @xcite . to remedy the situation ,
some empirical facts observed in manganites become important : ( i ) the mno@xmath1 octahedra are slightly tilted from each other , leading to modifications in the hopping matrix . among these modifications ,
the generation of a non - zero value for @xmath118 is important .
( ii ) the lattice is not cubic , but the relation @xmath119 holds , where @xmath120 and @xmath121 . from experimental results
@xcite , these numbers are estimated as @xmath122 and @xmath123 , indicating that the distortion with @xmath124-symmetry occurs spontaneously .
0.5truein -0.2truein = 3.6truein -1.5truein 0.7truein -1.2truein = 3.0truein -1.7truein for the inclusion of the point ( ii ) , @xmath32 in eq .
( [ q3 ] ) is rewritten as @xmath125 where @xmath126 indicates the spontaneous distortion with @xmath124-symmetry , given by @xmath127 this length is rewritten in the non - dimensional form as @xmath128 , where @xmath129 is a numerical factor given by @xmath130 , estimated as @xmath131 by using the experimental data .
note that the hopping amplitude becomes different from those in the @xmath100-@xmath101 plane due to this distortion .
as discussed shortly in subsection 2.3 , it is obtained as @xmath132 . as for @xmath27 along the @xmath56-direction ,
it is given by @xmath133 , since the superexchange interaction is proportional to the square of the hopping amplitude .
motivated by these observations , the energies for c- and g - type orbital structures were recalculated including this time a nonzero value for @xmath118 in the magnetic a - af state ( see fig .
7a ) ) . in fig . 8 , the configuration of @xmath134 orbitals is depicted . from this figure , it is intuitively understood that if the @xmath134 orbitals in the upper and lower planes are tilted from the @xmath100-@xmath101 plane as showing by arrows in the figure , there appears finite a hopping integral between adjacent @xmath134 orbitals along the @xmath135-direction and the sign of this hopping amplitude is the same as that of @xmath80 or @xmath136 .
thus , in the real material , the tilting of the mno@xmath137 octahedra will always lead to a _ positive _ value for @xmath118 and the results of fig .
7(a ) suggest that the c - type orbital structure should be stabilized in the real materials .
the explicit shape of the occupied orbitals is shown in fig .
the experimentally relevant c - type structure with the approximate alternation of @xmath6 and @xmath7 orbitals is indeed successfully obtained by this procedure .
although the octahedron tilting actually leads to a change of all hopping amplitudes , effect not including in this work , the present analysis is sufficient to show that the c - type orbital structure is stabilized in the a - af magnetic phase when @xmath118 is a small positive number , as it occurs in the real materials .
our investigations show that this mechanism to stabilize the c - type structure works also for the purely electronic model in the hartree - fock approximation .
in this work , the coulomb interaction term @xmath39 has been neglected , but this detail needs further clarification . for this purpose , @xmath39 is written as @xmath138 where @xmath139 is the intra - orbital coulomb interaction , @xmath140 the inter - orbital coulomb interaction , and @xmath141 is the inter - orbital exchange interaction . for mn - oxides , they are estimated as @xmath142ev , @xmath143ev , and @xmath144ev @xcite , which are large compared to @xmath57 .
however , the result for the optimized distortion described in this paper , obtained without the coulomb interactions , is not expected to change , since the energy gain due to the jt - distortion is maximized when a single e@xmath2-electron is present per site .
this is essentially the same effect as produced by a short - range repulsion .
in fact , the mc simulations show that the probability of double occupancy of a single orbital is negligible in the window of couplings where the a - type spin and c - type orbital state is stable . in order to confirm the above statement ,
the jt- and breathing - distortions were calculated as a function of @xmath140 by using the exact diagonalization method on a @xmath36 cluster in the f - state in which @xmath139 and @xmath141 can be neglected .
the result is shown in fig .
9 , where @xmath145 and @xmath146 are defined as @xmath147 and @xmath148 respectively .
as expected , the mean value of the breathing - mode distortion is almost zero and only the jt - mode is active in the case of @xmath149 . it is noted that the dependence of @xmath145 on @xmath140 is very weak , indicating that the optimized distortion is not affected by the coulomb interaction .
the orbital arrangement in this @xmath36 lattice is exactly the same as that in the @xmath100-@xmath101 plane of the orbital structure for the a - af phase in fig . 5 and this arrangement is unchanged by the inclusion of @xmath140 .
note also that @xmath145 is gradually increased with the increase of @xmath140 , although the dependence is weak .
this suggests that the jt - distortion without @xmath140 is reproduced at smaller value of @xmath59 if @xmath140 is included explicitly @xcite .
thus , it is expected that the a - af state will be stabilized at smaller @xmath59 improving the comparison of our results with experiments .
based on all these observations , it is believed that the effect of the coulomb interaction is not crucial for the appearance of the a - af state with the proper orbital order .
another way to rationalize this result is that the integration of the jt - phonons at large @xmath59 will likely induce coulombic interactions dynamically .
-0.5truein = 5.5truein -2truein finally , let us briefly discuss transitions induced by the application of external magnetic fields on undoped manganites .
when the a - af state is stabilized , the energy difference ( per site ) obtained in our study between the a - af and f states is about @xmath150 . as a consequence , magnetic fields of @xmath151 t could drive the transition from a - af to f order accompanied by a change of orbital structure , interesting effect which may be observed in present magnetic field facilities . in summary , with the use of numerical techniques at @xmath4 , it has been shown that the a - af state is stable in models with jt - phonons , using coupling values physically reasonable for lamno@xmath0 .
our results indicate that it is not necessary to include large coulombic interactions in the calculations to reproduce experimental results for the manganites .
considering the small but important effect of the octahedra tilting of the real materials , the c - type orbital structure ( with the alternation pattern of @xmath6 and @xmath7 orbitals ) has been successfully reproduced for the a - af phase in this context .
t.h . is grateful to y. takada and h. koizumi for enlightening discussion .
t.h . is supported from the ministry of education , science , sports , and culture of japan .
is supported by grant nsf - dmr-9814350 .
jin _ et al .
_ , science * 264 * , 413 ( 1994 ) ; y. tokura _ et al .
_ , j. appl . phys . * 79 * , 5288 ( 1996 ) ; a. p. ramirez , j. phys .
: condens .
matter * 9 * , 8171 ( 1997 ) ; y. tokura , in _ colossal magnetoresistance oxides _
y. tokura , gordon & breach , monographs in cond .
science ( 1999 ) .
e. o. wallan and w. c. koeler , phys .
100 * , 545 ( 1955 ) ; g. matsumoto , j. phys .
* 29 * , 606 ( 1970 ) ; j. b. a. ellemaans _ et al .
_ , j. solid state chem . * 3 * , 238 ( 1971 ) .
the c - type orbital structure has been recently confirmed with the use of high resolution x - ray diffraction technique ( y. murakami _ et al .
81 * , 582 ( 1998 ) ) .
l. m. roth , phys . rev . * 141 * , 306 ( 1966 ) ; k. i. kugel and d. i. khomskii , usp . fiz . nauk . * 136 * , 621 ( 1981)[sov .
* 25 * , 231 ( 1982 ) ] ; s. inagaki , j. phys .
. jpn . * 39 * , 596 ( 1975 ) ; m. cyrot and c. lyon - caen , j. physique * 36 * , 253 ( 1975 ) ; w. koshibae _
et al . _ , j. phys .
jpn . * 66 * , 957 ( 1997 ) ; r. shiina _ et al .
_ , j. phys .
. jpn . * 68 * , 3159 ( 1997 ) ; l. f. feiner and a. m. ole , phys . rev . b*59 * , 3295 ( 1999 ) ; j. van den brink _
et al . _ ,
rev . b*59 * , 6795 ( 1999 ) . a. j. millis _
et al . _ ,
lett . * 74 * , 5144 ( 1995 ) ; h. rder _ et al .
. lett . * 76 * , 1356 ( 1996 ) ; a. j. millis _
et al . _ ,
lett . * 77 * , 175 ( 1996 ) ; a. j. millis , _ et al .
b * 54 * , 5389 ; _ ibid _ 5405 ( 1996 ) ; d. feinberg _ et al .
_ , phys . rev . b*57 * , r5583 ( 1998 ) .
a. e. bacquet _
et al . _ ,
phys . rev . b*46 * , 3771 ( 1992 ) ; a. chainani , _ et al .
rev . b*47 * , 15397 ( 1993 ) ; t. arima _
et al . _ ,
rev . b*48 * , 17006 ( 1993 ) ; t. saito _
et al . _ ,
rev . b*51 * , 13942 ( 1995 ) . | the magnetic structure of lamno@xmath0 is investigated on three - dimensional clusters of mno@xmath1 octahedra by using a combination of relaxation and monte carlo techniques .
it is found that the cooperative jahn - teller phonons lead to the stabilization of a - type antiferromagnetic and c - type orbital structures in the physically relevant region of parameter space for lamno@xmath0 with small corrections due to tilting effects .
the results suggest that strong coulomb interactions are not necessary for a qualitative description of undoped manganites .
in fact , it is shown that the present result is not essentially changed even if the coulomb interaction is explicitly included . |
the galactic circumnuclear disk ( cnd ) is a weakly - ionized ring of gas and dust which appears in various types of emission at galactocentric radii @xmath0@xmath1pc ( see morris & serabyn 1996 and davidson 1996 for recent reviews ) .
the origin and structure of the cnd magnetic field , , are particularly intriguing : mid- and far - infrared polarimetry ( @xcite ; @xcite ; hildebrand et al . 1990 , 1993 [ h90 , h93 ] ) show that lies primarily in the plane of the disk and
is roughly orthogonal to the poloidal field observed on larger scales .
werner et al . ( 1988 ) pointed out that a differentially - rotating disk would deform an initially poloidal field into a predominantly toroidal configuration .
subsequently , wardle and ( 1990 , [ wk ] ) described self - similar hydromagnetic accretion disk models of the cnd ; with the addition of a small nonaxisymmetric distortion , the magnetic field predicted by their preferred model , gc2 , is in reasonable agreement with the field inferred from far - infrared polarimetry ( h93 ) . in the wk models , the magnetic torques which remove angular momentum from the gas
are mediated by supersonic ambipolar diffusion . coincidentally ,
roberge , hanany , & messinger ( 1995 , [ rhm ] ) showed that ambipolar diffusion causes grain alignment : the partial coupling of charged dust grains to the drifting field lines produces gas - grain streaming , which causes alignment of the grains via gold s mechanism ( @xcite ; @xcite ; @xcite ; @xcite ; rhm ) .
the large magnitude of the `` intrinsic '' 100 polarization from grains aligned by supersonic ambipolar diffusion ( e.g. , @xmath2% for oblate silicate grains with @xmath3:@xmath4 axis ratios viewed with the most favorable magnetic field geometry ; see rhm ) leads us to conjecture that the ambipolar mechanism is the predominant cause of grain alignment in the cnd . in 23
, we test this conjecture by calculating the polarized thermal emission that would be caused by ambipolar alignment in model gc2 and comparing our results to the 100 polarimetry of h90 and h93 .
the largest optical depth through the cnd is @xmath5 at 100 ; consequently the stokes parameters , @xmath6 , @xmath7 , and @xmath8 , are well described by the transfer equations for optically thin emission , @xmath9 @xmath10 @xmath11 ( wk90 ) , where @xmath12 is distance along the line of sight . we integrated equations ( [ eq-2_1])([eq-2_3 ] ) to obtain the polarization magnitude , @xmath13 , and position angle if the electric vector is polarized along the east - west direction .
] , @xmath14 , using a model of the position - dependent dust density , , and temperature , , derived by davidson et al . ( 1992 ) from 50 and 90 photometry of the galactic center .
our results are only weakly sensitive to uncertainties in and because the cnd is optically and geometrically thin and the fir emission is in the rayleigh - jeans limit .
the other variables in equations ( [ eq-2_1])([eq-2_3 ] ) are determined by the alignment properties of the grains as follows . the quantities and are respectively the dust absorption cross sections for light polarized along the and directions , where is parallel to the projection of the local magnetic field onto the plane of the sky , points toward the observer , @xmath15 , and @xmath16 is the angle between and north , measured counterclockwise from north .
to compute and , we modeled the grains as oblate spheroids with semiaxes @xmath17 parallel to the symmetry axis and @xmath18 ( @xmath19 ) perpendicular to the symmetry axis .
we assumed that the angular momentum , , of a rotating spheroid is aligned perfectly with its symmetry axis by the barnett effect . in this approximation ,
@xmath20 and @xmath21 ( @xcite ) , where and are respectively the cross sections for light polarized parallel and perpendicular to the symmetry axis , @xmath22 , and @xmath23 is the angle between and the plane of the sky .
for a given @xmath24 ratio , we computed and in the electric dipole approximation using draine s ( 1987 ) dielectric function for astronomical silicates .
notice that @xmath25 and @xmath26 depend only on _ ratios _ of the grain cross sections , which are independent of grain size in the electric dipole limit .
thus , apart from its weak influence on the grain dynamics ( discussed below ) , the value of @xmath18 does not affect our calculations .
we assumed for simplicity that the grain shape is independent of location and let @xmath24 be a free parameter .
the angles @xmath23 and @xmath16 are functions of the position - dependent disk magnetic field .
we computed @xmath23 and @xmath16 from a numerical tabulation of model gc2 kindly supplied by mark wardle .
the model is self - similar and to determine the hydro variables uniquely it is also necessary to specify the magnetic field , @xmath27 , the density , @xmath28 , and the radial ( @xmath29 ) and azimuthal ( @xmath30 ) velocity components of the neutral gas at some fiducial radius , @xmath31 .
we set @xmath32pc and chose @xmath33 mg , consistent with h i and oh zeeman splitting measurements of the cnd ( reviewed by morris & serabyn 1996 ) .
we set @xmath34 , near the lower end of the range of estimates for the density ( @xcite ) , and assumed that @xmath35 and @xmath36 , consistent with kinematic models of the cnd ( @xcite ; @xcite ) .
the alignment of with respect to is characterized by the quantity @xmath37 , \label{eq-2_6}\ ] ] where @xmath38 is the angle between and and angle brackets denote the average for all grains .
we calculated @xmath39 from the theory of ambipolar alignment ( rhm ) . , and hence the polarizations , by an uncertain factor that depends on poorly - known magnetic properties of the grains .
the increase would be @xmath40% for grains composed of ordinary paramagnetic substances but could be larger for superparamagnetic grains . ] for supersonic drift speeds , @xmath39 depends only on @xmath24 and @xmath41 , where is the gas - grain drift speed and @xmath42 is the gas thermal speed ( rhm ) .
we computed at each point using model gc2 to predict the local hydro variables and a prescription for the grain dynamics ( draine 1980 ) which includes lorentz and gas drag forces on the grains .
for the purposes of calculating the grain dynamics only , we represented the grains as spheres with equivalent radii @xmath43 and arbitrarily set @xmath44 .
changing @xmath18 by a factor of 2 would have virtually no effect on the calculated polarizations .
the wk models do not predict the gas temperature ; we set @xmath45k , consistent with observations of various molecular lines ( @xcite ) .
the factor @xmath46 in equations ( [ eq-2_2])([eq-2_3 ] ) represents the possible effects of a random magnetic field component .
we set the total field to @xmath47 , where is the ordered field predicted by model gc2 and is a random component .
we assumed that is perpendicular to ( as would be the case if is due to alfvnic turbulence ) and that has an axisymmetric , gaussian distribution of amplitudes ( see @xcite ) .
for this model , it is straightforward to show that @xmath48 where the `` turbulence parameter , '' @xmath49 , is the ratio of the energies in the ordered and disordered field components .
we assumed for simplicity that this ratio is independent of position and let @xmath49 be a free parameter .
note that @xmath46 decreases monotonically from @xmath50 ( no disordered field ) to @xmath51 ( disordered field in equipartition with the ordered field ) .
we constructed synthetic maps of @xmath52 , the 100 polarization predicted by ambipolar alignment , for 3,136 parameter combinations sampled uniformly on the intervals @xmath53 and @xmath54 . for each @xmath55 combination , we calculated the stokes parameters on a grid of sightlines uniformly spaced by @xmath56 in right ascension and declination .
each map of @xmath6 , @xmath7 , and @xmath8 was then convolved with a @xmath57 ( fwhm ) gaussian beam to model the observed resolution ( h93 ) and maps of were computed from the convolved stokes parameters .
optimal values of @xmath24 and @xmath49 were obtained by performing a @xmath58 fit to , the 100polarization observed along @xmath59 lines of sight ( h90 , h93 ) .
our fits omitted 7 observed sightlines for which the fir emission is not associated with the cnd .
we did not attempt to fit @xmath26 , which is a good diagnostic of the magnetic field geometry but not of the alignment mechanism .
strictly speaking , the value of we calculate for each sightline is the _
polarization for a hypothetical ensemble of disks with identical distribution functions for .
that is , discrepancies between and are due not only to uncertainties in our theoretical model but also to random fluctuations in . to include the effects of random fluctuations on @xmath58
, we assumed that the polarization observed along the @xmath60th sightline has dispersion @xmath61 , where @xmath62 is the observational uncertainty quoted by h93 and represents the uncertainty due to fluctuations in .
we assumed for simplicity that is the same for every sightline and performed the @xmath58 fits independently for different choices of .
we found that @xmath58 per degree of freedom is @xmath63 in some subset of parameter space if @xmath64% .
that is , we are able to obtain a reasonable fit to the data if we assume that the random fluctuations in are about half a percent .
, since some of the errors that are ascribed to fluctuations in are undoubtedly due to idealizations in our model .
for example , h93 pointed out that the wk models predict polarizations that are symmetric in the northwest and southeast quadrants .
however , the observations clearly show departures from symmetry in the southeast quadrant ; not surprisingly , these points contribute about twice as much on average to @xmath58 as the rest of the data . ]
figure 1 is a contour plot of @xmath58 per degree of freedom ( @xmath65 ) for the case @xmath66% .
evidently the observations are consistent with a continuum of models with different combinations of @xmath24 and @xmath49 ; this is due to the fact that a small increase in @xmath24 ( which tends to reduce ) can always be compensated by a decrease in @xmath49 ( which tends to increase ) .
thus , the contours of constant @xmath65 slope downward with increasing @xmath24 .
it is interesting to note that the parameters allowed by our fits are consistent with independent constraints : for example , observations show nearby molecular clouds have @xmath67 ( jones 1989 ; myers & goodman 1991 ) .
also , zeeman splitting observations of the 21 cm hi ( @xcite ) and 18 cm oh ( @xcite ) lines toward the galactic center imply that the magnetic energy density of the cnd is comparable to the turbulent energy density inferred from atomic and molecular linewidths ( @xcite ; @xcite ) .
these observations are consistent with the presence of nonlinear alfvn waves in the cnd , for which @xmath67 .
nonlinear alfvn waves are predicted by some models of magnetic braking ( mouschovias & morton 1985 ) .
similarly , photopolarimetry of the 9.7 silicate resonance implies that the aligned grains in nearby molecular clouds are oblate with @xmath68 ( hildebrand & dragovan 1995 ) . a model with @xmath69 and @xmath70 ( indicated by the filled circle in fig .
1 ) is compared with the observations in figure 2 . for this model , @xmath71 and
the mean discrepancy between and is @xmath72% .
our results are insensitive to the precise values of the hydrodynamic variables ; for example , the polarizations we calculate would be virtually identical if we arbitrarily increased or decreased or @xmath73 by factors of @xmath74 .
this robustness is due to the fact that , for the highly supersonic drift speeds in the wk models ( @xmath75 for gc2 ) , the efficiency of ambipolar alignment saturates at a value which depends only on the grain shape ( see rhm , fig . 12 ) .
however , our results _ are _ sensitive to the magnetic field geometry .
for example , we attempted to fit the observations by arbitrarily replacing the magnetic field predicted by model gc2 with a purely toroidal field of the same magnitude at each point ; we were unable to find solutions with @xmath76 for these models .
it will be interesting to see whether one can model the cnd observations successfully with more detailed hydro models ( e.g. , wardle & knigl 1993 ) which calculate the vertical structure of the disk .
previous models of the cnd magnetic field geometry ( wk , h90 , h93 ) have accounted successfully for the observed polarization position angles by assuming that the efficiency of grain alignment is independent of position .
the saturation of ambipolar alignment at large drift speeds provides a natural justification for this assumption .
in contrast , it is difficult to reconcile the uniform efficiency required by the observations with other alignment mechanisms .
if the grains are superparamagnetic , then the efficiency of davis - greenstein alignment would be saturated but the saturation efficiency would be very sensitive to the local dust - to - gas temperature ratio : independent calculations on dg alignment ( lazarian 1995 ) show that factor of 2 variations in @xmath77 would cause factor of @xmath78 variations in . the efficiency of purcell s mechanism ( purcell 1979 ) is insensitive to but sensitive to the timescale for changes in the grain surface properties ; it is difficult to see why the latter should be uniform throughout the cnd and have just the right value to reproduce the observations .
we conclude that the degree and uniformity of the alignment required by the observations are natural consequences of ambipolar alignment but difficult to explain with other mechanisms .
1 . we have calculated the 100 linear polarization from the cnd using a model that attributes the grain alignment to ambipolar diffusion in the wardle- model accretion disks .
our polarization maps depend on just 2 adjustable parameters , the grain axis ratio , @xmath24 , and the mean ratio , @xmath49 , of the energies in the random and ordered magnetic field components
we estimated the values of @xmath24 and @xmath49 by performing a @xmath58 fit to the observed 100 polarizations ( h90 , h93 ) .
our models provide a reasonable fit to the observations if we assume that the @xmath79 fluctuations in the observed polarizations due to turblence in the cnd are @xmath80% .
3 . the observations are consistent with a continuum of models with different @xmath55 values .
a model with @xmath69 and @xmath70 , the parameter values implied by independent constraints , falls close to the minimum of @xmath58 per degree of freedom . for this model , @xmath71 and the mean error in the predicted polarizations is @xmath72% .
4 . our calculations imply that grains in the cnd must be aligned with an efficiency such that @xmath81 , nearly independent of position .
the degree and uniformity of alignment required by the observations are natural predictions of the ambipolar mechanism but require `` fine tuning '' of the parameters in other alignment theories .
aitken , d.k . ,
roche , p.f . ,
bailey , j.a .
, briggs , g.p . , hough , j.h .
, & thomas , j.a .
1986 , , 218 , 363 davidson , j.a .
1996 , in asp conf .
97 , polarimetry of the interstellar medium , ed . w. g. roberge & d. c. b. whittet ( san francisco : asp ) , 504 davidson , j.a . , et al .
1992 , , 387 , 189 draine , b.t .
1980 , , 241 , 1021 1987 , princeton observatory preprint 213 gold , t. 1952 , , 112 , 215 gsten , r. , genzel , r. , wright , m.c.h . ,
jaffe , d.t . ,
stutzki , j. , & harris , a.i .
1987 , , 318 , 124 harris , a.i . ,
jaffe , d.t .
, silber , m. , & genzel , r. 1985 , , 294 , l93 hildebrand , r.h . , gonatas , d.p . ,
platt , s.r .
, wu , x.d . ,
davidson , j.a . ,
werner , m.w . , novak , g. , & morris , m. 1990 , , 362 , 114 ( h90 ) hildebrand , r.h . , davidson , j.a . , dotson , j. , figer , d.f . , novak , g. , platt , s.r . , & tao , l. 1993 , , 417 , 565 ( h93 ) hildebrand , r.h . , &
dragovan , m. 1995 , , 450 , 663 jackson , j.m . ,
1993 , , 402 , 173 jones , t.j .
1989 , , 346 , 728 killeen , n.e.b , lo , k.y . , & crutcher , r. 1992 , , 385 , 585 lazarian , a. 1994 , , 268 , 713 1995 , , 453 , 229 lazarian , a. , & roberge , w.g .
1996 , , submitted lee , h.m . , & draine , b.t .
1985 , , 290 , 211 morris , m. , & serabyn , e. 1996 , , 34 , 645 mouschovias , t. ch .
, & morton , s.a . 1985 , , 298 , 190 myers , p.c . , &
goodman , a.a .
1991 , , 373 , 509 purcell , e.m .
1969 , physica , 41 , 100 1979 , , 231 , 404 purcell , e.m .
, & spitzer , l. , jr .
1971 , , 167 , 31 roberge , w.g . ,
hanany , s. , & messinger , d.w .
1995 , , 453 , 238 ( rhm ) schwarz , u.l . , &
lazenby , j. 1990 , in galactic and extragalactic magnetic fields , eds .
r. beck , p. kronberg , and r. wielebinski , p. 383 wardle , m. , & knigl , a. 1990 , , 362 , 120 ( wk ) | we describe an implicit prediction of the accretion disk models constructed by wardle and ( 1990 ) for the circumnuclear disk ( cnd ) of gas and dust near the galactic center : supersonic ambipolar diffusion , an essential dynamical ingredient of the wardle- disks , will cause the alignment of dust grains due to a process described by roberge , hanany , & messinger ( 1995 ) .
we calculate synthetic maps of the polarized thermal emission which would be caused by ambipolar alignment in the preferred wardle- model .
our maps are in reasonable agreement with 100 polarimetry of the cnd if we assume that the grains have shapes similar to those of grains in nearby molecular clouds and that the cnd contains a disordered magnetic field in energy equipartition with its ordered field . |
the study of flavor - changing neutral - current ( fcnc ) interactions has been playing an important role in testing the standard model ( sm ) and probing new physics beyond the sm . as the most massive fermion in the sm
, the top quark is naturally regarded to be more sensitive to new physics than other fermions . in the sm , due to the gim mechanism
, the top quark fcnc interactions are absent at tree level and extremely small at loop levels . in some new physics models beyond the sm
the top quark fcnc may be significantly enhanced .
searching for the top quark fcnc would be a good probe for new physics .
intensive activities to explore the top quark fcnc couplings have been undertaken in recent years .
on the experimental side , the cdf and d0 collaborations have reported interesting bounds on the fcnc top quark decays @xcite from run 1 experiment and will tighten the bounds from the on - going run 2 experiment . on the theoretical side , various fcnc top quark decays and top - charm associated productions at high energy colliders were extensively studied in the sm @xcite , the minimal supersymmetric model ( mssm ) @xcite and other new physics models @xcite .
these studies showed that the sm predictions for such top quark fcnc processes are far below the detectable level and the mssm can enhance them by several orders to make them potentially accessible at future collider experiments @xcite . due to its rather clean environment
, the next generation linear collider ( nlc ) will be an ideal machine to probe new physics .
in such a collider , in addition to @xmath0 collision , we can also realize @xmath2 collision and @xmath1 collision with the photon beams generated by the backward compton scattering of incident electron- and laser - beams @xcite . the susy induced
top - charm fcnc will give rise to various processes at the nlc , i.e. , the top - charm associated productions via @xmath0 , @xmath1 and @xmath2 collisions as well as the top quark rare decays @xmath3 ( @xmath4 , @xmath5 or @xmath6 ) .
it is noticable that some of these processes , like the top - charm associated productions in @xmath2 or @xmath1 collision , have not been studied in the framework of the mssm .
the production in @xmath2 collision may be more important than in @xmath0 collision studied in the literature @xcite .
the reason is twofold .
firstly , the process @xmath8 is a good probe of new physics because it is essentially free of any sm irreducible background @xcite .
secondly , unlike the process @xmath9 , which is @xmath10-channel suppressed in high energy collisions , there are @xmath11- and @xmath12-channel contributions to @xmath8 and thus its cross section may be much larger at the nlc .
it is also noticable that all these fcnc processes at the nlc involve the same part of the parameter space of the mssm .
involves more parameters , we do not include it in our analysis .
] therefore , it is necessary to perform a comparative analysis for all these processes to find out which process is best to probe the top quark fcnc .
this is the aim of this article .
this paper is organized as follows . in sec .
ii , we discuss the mixing between top - squark and charm - squark and derive the fcnc interaction lagrangian in susy - qcd . in sec .
iii , we calculate the @xmath13 productions in @xmath2 and @xmath1 colllisions induced by such fcnc susy - qcd interactions .
numerical results for these productions at the nlc are given in sec .
iv , with the comparison to @xmath13 production in @xmath0 collision and various fcnc top quark decays .
finally , a brief discussion and conclusion can be found in sec . v.
many popular susy models predict the flavor mixings of sfermions . for the squark sector , despite of the possible strong constraints on the down type squark flavor mixings from the low - energy experimental data , limit could enhance the fcnc @xmath14 decays by several orders @xcite and will face tests in the on - going @xmath14-factory experiments . ]
the mixings between top - squark and charm - squark are subject to no strong low - energy constraints @xcite .
such a mixing between top - squark and charm - squark is well motivated in low - energy supergravity models ( sugra ) @xcite . in these models
squark mass matrices are diagonalized simultaneously as with the quark matrices at the planck scale .
but when the terms evolve down to low energy , this diagonality is violated by radiative corrections .
due to large top quark mass , at low energy it is found @xcite that the mixing between top - squark and charm - squark may be significant .
note that only @xmath15 mixes with top - squark while @xmath16 does not in the approximation of neglecting charm quark mass .
as shown in @xcite , the mixing between @xmath15 and @xmath17 is most likely to be large , which is proportional to a sum of some soft masses . motivated by the above arguments , in our analysis we assume the existence of the mixing between @xmath15 and @xmath17 and parameterize the mixing as @xmath18 , where @xmath19 ( @xmath20 ) is the soft mass parameter for left - handed squark of third ( second ) generation and @xmath21 is a dimensionless parameter representing the mixing strength .
@xmath21 can be calculated in terms of other parameters in a given model like msugra @xcite .
but in our calculation we retain it as a free parameter in the range of @xmath22 @xcite . considering the mixing between @xmath17 and @xmath23 and neglecting the mixing between @xmath15 and @xmath16 , we obtain the @xmath24 mass - square matrix in the basis @xmath25 : @xmath26 where @xmath27
, @xmath28 , @xmath29 and @xmath30 is the soft - breaking mass term for right - handed top - squark .
@xmath31 is the coefficient of the trilinear term @xmath32 in soft - breaking terms , @xmath33 is the mixing mass parameter between @xmath34 and @xmath35 in the superpotential and @xmath36 is ratio of the vacuum expectation values of the two higgs doublets . the mass - square matrix in eq.([smass1 ] )
can be diagonalized by a @xmath37 unitary matrix @xmath38 , which rotates the interaction eigenstates ( @xmath17 , @xmath23 , @xmath39 ) into mass eigenstates @xmath40 .
such a rotation results in the fcnc in both the weak interaction sector and the strong susy - qcd sector . of course
, the latter will be dominant .
so in our analysis we only consider the fcnc in susy - qcd sector , given by @xmath41 , \end{aligned}\ ] ] where @xmath42 .
the susy - qcd fcnc in eq.([fcnc ] ) will induce the top - charm associated production in @xmath2 collision , as shown in fig . 1 .
throughout our calculations the charge conjugate production channel , i.e. , the production of @xmath43 , has also been included . neglecting the charm quark mass , the amplitude of this process
is given by @xmath44 with @xmath45 , @xmath46 and @xmath47 here @xmath48 denote the momentum of incoming photons and @xmath49 the momentum of outgoing top and charm quarks .
the coefficients @xmath50 can be obtained by a straightforward calculation of the diagrams shown in fig . [
fig : feyman1 ] and their expressions are presented in the appendix .
we checked that our results satisfy the ward identity , @xmath51 and @xmath52 .
we also checked that all ultraviolet divergences canceled in our results due to the unitary of @xmath38 , which is essentially guaranteed by the renormalizability of the mssm .
note that when @xmath11 or @xmath12 approaches 0 , which is physically permitted , @xmath50 tend to be very large .
indeed , this is the advantage of such a process over @xmath53 which has only @xmath10-channel contribution and thus is suppressed at high energy colliders .
the @xmath54 production in @xmath1 collision proceeds through the process @xmath55 , where the @xmath56-beam is generated by the backward compton scattering of incident electron- and laser - beam and the @xmath57 is radiated from @xmath58 beam .
the subprocess @xmath59 has the same feynman diagrams shown in fig .
[ fig : feyman1 ] . in our calculation
we use the weizs@xmath60cker - williams approximation @xcite which treats @xmath57 from @xmath58 beam as a real photon .
thus the process can be approximated by the simpler fusion reaction @xmath61 and its cross section is given by @xmath62 where @xmath63 is the probability of finding a photon with a fraction @xmath64 of energy @xmath65 in an ultrarelativistic electron and is given by @xcite @xmath66 note that there are also intermediate @xmath6-boson contribution for the process @xmath67 .
however , such contributions are suppressed by the probability function of finding a @xmath6-boson in an ultrarelativistic electron @xcite and can be safely neglected . for both @xmath2 collider and @xmath68 collider ,
the photon beams are generated by the backward compton scattering of incident electron- and laser - beams just before the interaction point .
the events number is obtained by convoluting the cross section with the photon beam luminosity distribution . for @xmath2 collider the events number
is obtained by @xmath69 where @xmath70 is the photon beam luminosity distribution and @xmath71 , with @xmath72 being the energy - square of @xmath73 collison , is defined as the effective cross section of @xmath74 . in optimum case
, it can be written as@xcite @xmath75 where @xmath76 denotes the energy spectrum of the back - scattered photon for unpolarized initial electrons and laser photon beams given by @xmath77 the definitions of parameters @xmath78 , @xmath79 and @xmath80 can be found in @xcite . in our numerical calculation , we choose @xmath81 , @xmath82 and @xmath83 .
for the @xmath84 collider the effective cross section of @xmath85 is defined as @xmath86 the process @xmath9 and the top quark rare decays @xmath87 ( @xmath88 ) were already calculated in the literature . here for comparison we recalculated all of them .
the lengthy expressions are not presented here .
the relevant susy parameters all the top - charm productions and top quark rare decays are @xmath21 , @xmath89 , @xmath90 , @xmath19 , @xmath91 , @xmath92 and @xmath93 . to find out typical magnitudes of these processes , we scan over these susy parameters by requiring @xmath94 , @xmath95 gev@xcite , @xmath96 gev @xcite , @xmath97 tev and restricting other soft mass parameters to be of sub - tev scale .
our scan results are plotted in figs .
[ fig : scan1 ] and [ fig : scan2 ] . for the productions one sees that the typical values are @xmath98 fb and @xmath99 fb and in optimum cases @xmath100 fb and @xmath101 fb . while for the process @xmath9
, the cross section can only reach @xmath102 fb , one order of magnitude lower than that of @xmath8 and comparable with that of @xmath103 .
the reason for this , as we pointed out before , is due to the @xmath11-channel enhancement for @xmath8 and @xmath10-channel suppression for @xmath9 . for the top quark
rare decays we see that the optimum values are @xmath104 , @xmath105 and @xmath106 , which agree with previous results @xcite . comparing with predictions in other models , such as r - parity violating model @xcite and the type iii two higgs doublet model@xcite
, we find that the optimum value of @xmath107 in susy - qcd is larger .
note that our results for @xmath108 are quite different from those in ref.@xcite where the optimum value of @xmath109 can reach @xmath110 fb .
the reason for the difference is that in ref .
@xcite a non - unitary flavor changing matrix is used . during our scan
, we find that , only in case of a light top - squark with mass of @xmath111 gev and a large mass splitting of a few hundreds gev in squark spectrum can relative large cross sections of these processes be obtained .
this is due to a gim - like cancellation which occurs in the limit of degenerate squark masses as a result of the unitarity of the rotation matrix @xmath38 .
the sm and mssm predictions for the top - charm fcnc processes are summarized in table i. note that since the sm prediction for the process @xmath112 has not been done in the literature , the sm value shown in the table is calculated by us .
this is an arduous work .
there are 66 diagrams contributing @xmath113 in @xmath114 gauge and often the contribution from a single box diagram can be decomposed into dozens of terms . in order to avoid artificial error , we use formcalc@xcite in our calculations . from table
i , we see that the mssm can enhance the cross section of top - charm associated production processes by several orders .
table 1 : theoretical predictions for the top - quark fcnc processes .
susy - qcd predictions are the maximum values .
the collider energy is @xmath115 gev for productions . [
cols="<,<,<",options="header " , ] the behaviours of the sm cross sections versus the collider energy are shown in fig .
[ sm ] for three different production processes .
we see that the cross section of @xmath116 drops quickly with the increase of collider energy due to the @xmath10-channel suppression .
let us focus on the most important process @xmath117 and study some of its features .
for example we fix @xmath118 and @xmath119 gev , and choose @xmath120 gev , @xmath121 gev for scenario i ( relatively light squarks for the third - family ) and @xmath122 gev , @xmath123 gev for scenario ii ( relatively heavy squarks for the third - family ) .
the cross section of @xmath8 as a function of @xmath21 is shown in fig .
[ fig : sigma - del ] .
this figure shows that the cross section is enhanced dramatically with the increase of @xmath21 .
clearly this is due to the fact that large @xmath21 will not only enhance the mixing between top - squark and charm - squark but also enlarge the mass splitting between the squarks . in fig .
[ fig : sigma - x ] we show the dependence of @xmath124 on @xmath125 .
we see that @xmath124 increases as @xmath125 gets large .
the reason is that large @xmath125 enhances the mass splitting between the top - squarks and thus lead to a weak cancellation between different feynman diagrams .
the dependence of @xmath124 on @xmath126 and @xmath127 is shown in fig .
[ fig : sigma - mg ] and fig .
[ fig : sigma - s ] , respectively . fig .
[ fig : sigma - mg ] shows that the cross section drops when gluino becomes heavier , showing the decoupling effect of the mssm . from figs .
[ fig : sigma - del]-[fig : sigma - s ] one finds that the cross section in scenario i is always larger than that in scenario ii .
the reason is that scenario i has the relative light squarks for the third - family which lead to a larger mass splitting in squark spectrum .
we have performed a comparative analysis for the susy - induced top - charm fcnc processes at the nlc .
these processes include the top - charm associated productions via @xmath0 , @xmath1 and @xmath2 collisions as well as the top quark rare decays @xmath3 ( @xmath4 , @xmath5 or @xmath6 ) , all of which involve the same susy parameters . in the production channels the @xmath13 production in @xmath2 collision
was found to occur at a much higher rate than @xmath0 or @xmath1 collision . in some part of parameter space ,
the production rate of @xmath8 can reach @xmath128 fb .
this means we may have @xmath129 events each year for the designed luminosity of @xmath130 fb@xmath131/year at the nlc . since the sm value of the production rate is completely negligible , the observation of such @xmath132 events would be a robust indirect evidence of susy .
note that in practical experimental searches of such @xmath132 events , a careful study of backgrounds is needed . to efficiently suppress the backgrounds ,
the signal events may be hurt by @xmath133 or more @xcite .
using polarization at a linear collider can help to suppress the background , as analyzed in the second reference of @xcite . in the rare decay channels
the @xmath134 was found to have the largest branching ratio which can reach @xmath135 , in agreement with previous studies .
although @xmath136 @xmath137 events could be produced at the nlc each year , studies @xcite showed that the sensitivity to such rare decays can only reach @xmath138 .
so the mssm prediction for the branching ratio of @xmath135 is too low to be accessible at the nlc unless the designed luminosity can be further upgraded .
therefore , we conclude that , to probe the susy - induced fcnc top quark interactions at the nlc , @xmath13 production in @xmath2 collision is the best channel . note that in our analysis we assumed flavor mixing occurs between @xmath15 and @xmath17 , which is favored in sugra models .
other kinds of mixings like the mixing between @xmath16 and @xmath17 @xcite may also be of phenomenological interest .
however , we believe our conclusion will be quite model - independent since our results reflect the basic features of the processes .
we thank kaoru hagiwara and tao han for very helpful discussions . this work is supported by young outstanding foundation of academia sinica .
before presenting the explicit form of @xmath139 , we define the following abbreviations @xmath140 where @xmath38 is squark mixing matrix and b , c and d are loop functions defined in @xcite
. then @xmath141 is given by @xmath142 with @xmath143 and @xmath144 .
@xmath145 and @xmath146 in eq.([amp ] ) can be expressed in the form of eq.([m - e ] ) and their corresponding coefficients @xmath50 are given as @xmath147 m. paulini , hep - ex/9701019 ; j. incandela ( cdf ) , fermilab - conf-95/237-e(1995 ) ; d. gerdes , hep - ex/9706001 ; t. j. lecompte ( cdf ) , fermilab - conf-96/021-e ( 1996 ) ; a. p. heinson ( d0 ) , hep - ex/9605010 .
for the fcnc top quark decays in the sm , see , g. eilam , j. l. hewett and a. soni , 44 , 1473 ( 1991 ) ; b. mele , s. petrarca , a. soddu , 435 , 401 ( 1998 ) .
chang , x .- q .
li , j .- x .
wang , m .- z .
yang , 313 , 389 ( 1993 ) c .- s .
huang , x .- h .
wu and s .- h .
zhu , 452 , 14(1999 ) . for a review
see , e.g. , h. e. haber and g. l. kane , phys .
rep.*117 * , 75 ( 1985 ) . for @xmath148 in the mssm ,
see , c. s. li , r. j. oakes and j. m. yang , 49 , 293 ( 1994 ) ; g. couture , c. hamzaoui and h. konig , 52 , 1713 ( 1995 ) ; j. l. lopez , d. v. nanopoulos and r. rangarajan , 56 , 3100 ( 1997 ) ; g. m. de divitiis , r. petronzio and l. silvestrini , 504 , 45 ( 1997 ) .
j. m. yang , b .-
young and x. zhang , 58 , 055001 ( 1998 ) . for @xmath149 in the mssm , see , j. m. yang and c. s. li , 49 , 3412 ( 1994 ) ; j. guasch , and j. sola , 562 , 3 ( 1999 ) ; g. eilam , a. gemintern , t. han , j.m .
yang , x. zhang , 510 , 227 ( 2001 ) . c. s. li , x. zhang and s. h. zhu , 60 , 077702 ( 1999 )
. j. l. diaz - cruz , _ et al .
_ 41 , 891(1990 ) ; g. eilam , j. l. hewett and a. soni , 44 , 1473 ( 1991 ) ; x.l .
wang et al . , 50 , 5781 ( 1994 )
; 20 , l91 ( 1994 ) ; 24 , 359 ( 1995 ) ; david atwood , l. reina and a. soni , 53 , 1199 ( 1996 ) ; s. bar - shalom , _ et al . _ , 79 , 1217(1997 ) ; w. s. hou , g .-
lin and c .- y .
ma , 56 , 7434(1997 ) ; s. bejar , j. guasch and j. sola , 600 , 21 ( 2001 ) ; r. a. diaz , r. martinez , j .- a .
rodriguez , hep - ph/0103307 ; t. han and j. hewett , 60 , 074015 ( 1999 ) ; f. del aguila , j. a. aguilar - saavedra , r. miquel , 82 , 1628 ( 1999 ) ; j. a. aguilar - saavedra , b. m. nobre , hep - ph/0210360 . j. a. aguilar - saavedra , g. c. branco , 495 , 347 ( 2000 ) ; j. a. aguilar - saavedra , 502 , 115 ( 2001 ) ; j. a. aguilar - saavedra , t. riemann , hep - ph/0102197 .
koh abe et al .
, acfa linear collider working group , hep - ph/0109166 . s. bar - sharlom and j. wudka , 86 , 3722(2001 ) ; 60 , 094016 ( 1999 ) . see , e.g , z. xiong and j. m. yang , 628 , 193 ( 2002 ) ; c. bobeth , t. ewerth , f. kruger , j. urban , hep - ph/0204225 ; s.rai choudhury , naveen gaur , hep - ph/0207353 ; hep - ph/0206128 ; hep - ph/0205076 ; hep - ph/0203041 ; see , e.g. , f. gabbiani , e. gabrielli , a. masiero , l. silverstrini , 477 , 321 ( 1996 ) ; m. misiak , s. pokorski , j. rosiek , adv . ser
high energy phys .
15 , 795 ( 1998 ) ; f. borzumati , c. greub and t. hurth , 62 , 075005 ( 2000 ) ; t. besmer , c. greub and t. hurth , 609 , 359 ( 2001 ) .
m. j. duncan , 221 , 285 ( 1983 ) ; 31 , 1139 ( 1985 ) . k. hikasa and m. kobayashi , 36 , 724 ( 1987 ) . s. j. brodsky , t. kinoshita and h. terazawa , 4 , 1532 ( 1971 ) ; h. terazawa , rev
. mod .
phys.*45 * , 615(1973 ) ; b. a. kniehl , 254 , 267(1991 ) .
r. cahn and s. dawson , 136 , 196 ( 1984 ) ; m. chanowitz and m. k. gaillard , 142 , 85 ( 1984 ) ; g. l. kane , _ et .
148 , 367(1984 ) .
i. f. ginzburg , _ et al .
_ , nucl .
instrum*219 * , 5 ( 1984 ) ; v. i. telnov , nucl .
instrum .
meth.*294 * , 72 ( 1990 ) . particle physics group .
15 , 274 ( 2000 ) .
y. zeng - hui , _ et .
16 , 541 ( 2000 ) .
j. yi , _ et .
57 , 4343 ( 1998 ) .
t. hahn , nucl .
suppl.*89 * , 231 ( 2000 ) ; acta phys .
b*30 * , 3469 ( 1999 ) .
k. j. abraham , k. whisnant and b .-
young , 419 , 381 ( 1998 ) .
r. frey _ et al .
_ , fermilab - conf-97 - 085 ( 1997 ) ; hep - ph/9704243 . j. l. diaz - cruz , h. j. he and c. p. yuan , 530 , 179 ( 2002 ) ; h .- j .
he , s. kanemura , c .- p .
yuan , hep - ph/0203090 .
d. atwood , s. bar - shalom , g. eilam and a. soni , hep - ph/0203200 .
g t hooft and m. veltman , 44 , 189 ( 1972 ) ; a. denner , fortschr . phys . * 41 * , 307 ( 1993 ) . | in the minimal supersymmetric model ( mssm ) the hitherto unconstrained flavor mixing between top - squark and charm - squark will induce the flavor - changing neutral - current ( fcnc ) interaction between top quark and charm quark , which then lead to various processes at the next generation linear collider ( nlc ) , i.e. , the top - charm associated productions via @xmath0 , @xmath1 and @xmath2 collisions as well as the top quark rare decays @xmath3 ( @xmath4 , @xmath5 or @xmath6 ) .
all these processes involve the same part of the parameter space of the mssm . through a comparative analysis for all these processes at the nlc
, we found that the best channel to probe such susy - induced top quark fcnc is the top - charm associated production in @xmath2 collision , which occurs at a much higher rate than @xmath0 or @xmath1 collision and may reach the detectable level for some part of the parameter space .
since the rates predicted by the standard model are far below the detectable level , the observation of such fcnc events would be a robust evidence of new physics .
= # 1,nucl .
phys .
b * # 1 * , # 1,phys .
lett .
b * # 1 * , # 1,euro .
phys .
j. c * # 1 * , # 1,phys .
rev .
d * # 1 * , # 1,phys .
rev .
lett .
* # 1 * , # 1,mod .
phys .
lett . a * # 1 * , # 1,j
. phys .
g*#1 * , # 1,commun .
theor .
phys .
* # 1 * , # 1tt@xmath7#1 |
small angle scattering ( sas ) is sensitive to the presence of structural and chemical inhomogeneities inside the systems , which behave as scattering entities . here and in the program
, we shall refer to these entities as _ scattering particles _ whenever they do not bear a charge and _ macroions _ otherwise . in the latter case a set of smaller scattering cores ( _ counterions _ with opposite charge ) are clearly present to satisfy electroneutrality .
polydispersity means that the particles are not all identical but their size , charge or other properties exhibit a large variety of values ( a phenomenon commonly found , for instance , in colloidal and micellar suspensions ) .
hence in presence of polydispersity sas intensity may greatly differ from that of monodisperse systems .
the aim of the present paper is to present a code ( flac ) which can be used in the analysis of experimental data of small - angle neutron scattering from polydisperse systems .
the flac program provides the scattering functions from polydisperse fluids of neutral @xcite or charged hard spheres @xcite , based on the analytical solutions of the ornstein - zernike ( oz ) integral equations within the percus - yevick ( py ) or the mean spherical spproximation ( msa ) closures , respectively @xcite . for the sake of clarity and in order to be self - contained , general formulas appearing in the sas theories will now be briefly recalled . the most general expression of the sas is @xcite @xmath3 here @xmath4 is the macroscopic differential coherent scattering cross section ( scs ) as a function of the exchanged wave vector @xmath5
, whose magnitude is defined by @xmath6 where @xmath7 is the neutron wavelength and @xmath8 the full scattering angle . in eq .
[ sas1 ] , the integral is extended over the total sample volume @xmath9 , @xmath10 is the position vector and @xmath11 the local scattering length density . for x - rays ,
the scattering length density is replaced by the electron density , leaving the general formalism unchanged .
the angular brackets @xmath12 represent an ensemble average over all possible positions and orientations of the particles in the system . in the absence of long range order
, the scattering length density can be thought as having a uniform value @xmath13 on which fluctuations @xmath14 are superimposed .
the distribution of scattering material inside a particle of species @xmath15 is then defined by its form factor @xmath16 where @xmath17 is the scattering amplitude at zero - angle , and @xmath18 is the volume of the scattering core ( i.e the subregion of a macroparticle which contains the scattering material : nuclei , etc . ) .
often it proves convenient to combine eq .
[ sas8 ] and [ sas9 ] as @xmath19 .
let us assume that the polydisperse system can be regarded as a mixture with @xmath20 components . by introducing an effective form factor @xmath21 as @xmath22
one can also define an effective structure factor @xmath23 of the mixture by rewriting eq .
[ sas1 ] in the form @xmath24 here , @xmath21 accounts for the distribution of scattering matter inside the particles , while @xmath23 is related to the particle - particle interactions and the resultant equilibrium structure of the system , and it is defined through eq .
( [ sas12 ] ) .
if the particle - particle interactions are spherically symmetric ( homogeneous and isotropic fluids ) , then the scs can be cast into the fournet - vrij s form @xcite : @xmath25 \label{sas11 } \\
& + & \sum_{i , j=1}^p\sqrt{n_in_j}f_if_j~\left\langle b_i(\bf{q } ) \right\rangle _ \omega \left\langle b_j^{*}(\bf{q})\right\rangle _ \omega ~s_{ij}(q ) , \nonumber\end{aligned}\ ] ] where @xmath26 is the number density of species @xmath15 and the brackets @xmath27 represent angular averages ( this formula being still valid in case of non - spherically symmetric particles ) .
the @xmath28 are the ashcroft - langreth partial structure factors @xcite , defined by
@xmath29 where @xmath30 is the kronecker delta and @xmath31 is the three - dimensional fourier transform of the total correlation function @xmath32 , which can be obtained by solving the oz integral equations of the liquid state @xcite . the first term on the rhs of eq .
[ sas11 ] is called the `` incoherent '' part , and the remainder the `` coherent '' one .
the incoherent part vanishes if the distribution of scattering matter is spherosymmetric and in this case @xmath33 fortunately , for most of the investigated systems the particles are formed by a single type of uniform scattering material ( first phase = solute ) and are embedded in a uniform medium ( second phase = solvent for a solution , or matrix for an alloy ) . in this case , @xmath34 in eq .
[ sas8 ] can be replaced by @xmath35 , where @xmath36 and @xmath37 are the scattering length density of the scattering particles and the embedding medium , respectively .
thus the zero - angle scattering amplitude ( eq.[sas9 ] ) is given by @xmath38 and the form factor @xmath39 ( eq . [ sas8 ] ) becomes @xmath40 in this case , eq . [ sas11 ] takes the simple form @xmath41 then the effective structure factor can be expressed in terms of partial structure factors as @xmath42 the @xmath39 can be calculated analytically with the result @xmath43 where @xmath44 is the scattering core radius of a particle of species @xmath45 and @xmath46 is the first - order spherical bessel function . in general , @xmath47 , with @xmath48 being the hard sphere radius of the @xmath15-th particle .
clearly , @xmath49 in the case of micelles or in the presence of a depletion zone around a precipitate , the scattering particles can be thought as a two - shell spheres , characterized by an inner core and an outer shell with different scattering length ( or electron ) densities . even in this case
the form factor @xmath50 has a relatively simple form @xmath51 where @xmath52 and @xmath53 are the radii of the inner and outer shell , respectively , @xmath54 ( @xmath55 ) and @xmath13 , @xmath56 and @xmath57 are the scattering length ( or electron ) density of the solvent , of the inner and the outer shell , respectively .
experimental data are in general plagued by several effects which might affect the scattering intensity , as compared to the theoretical one , and hence may lead to an incorrect interpretation .
most of them have been well studied and the way of controlling them is now well established . the instrumental resolution ( i.e. the incident radiation wavelength spread , the finite collimation and the detector resolution ) together with the radial average of the sas data performed when isotropic data are collected on a 2d detector can be computed following the recipe given in ref .
@xcite , where each effect is described by a gaussian contribution , which is also used to model the primary beam for the beam width effect @xcite .
the vertical slit effect is also taken into account in the calculation , as described in ref .
@xcite ; this correction becomes relevant for slit systems and double crystal diffractometers . finally , possible corrections due to multiple scattering are also calculated using the mux routine @xcite , which is based on the theory developed in ref . @xcite .
since sas experiments are performed with more instrumental configurations in order to scan the largest possible q range , up to three different setups are considered during smearing of the scattering intensities .
two different types of background are considered in flac . a flat incoherent one , like a plateau on which the small angle scattering intensity is superimposed , and the porod one .
the latter is due to the asymptotic scattering of large inhomogeneities and can be approximated as @xcite @xmath58 where @xmath59 is the porod constant .
as often reported in the literature ( see e.g. ref .
@xcite ) , the hard sphere radius @xmath48 of a particle need not coincide with the radius @xmath60 of its scattering core . to include this possibility
, the ratio @xmath61 can be tuned in the program . for neutral hard spheres ,
the interparticle potential @xmath62 is described as @xmath63 where @xmath64 .
the oz equations for an hard sphere mixture can be solved by using the py closure .
a closed - form for the sas intensity for polydisperse hard sphere fluids was obtained by vrij @xcite .
our code for this part is constructed along the lines of that reference . in the case of charged hard spheres ,
the interparticle potential @xmath62 takes the following form @xmath65 where coulombic ( attractive and repulsive ) interactions are added to the previous neutral case , @xmath66 being the elementary charge , @xmath67 the valence of the @xmath15-th species and @xmath68 the dielectric constant of the solvent .
this potential defines the primitive model of ionic fluids @xcite .
the msa analytical solution of the oz equations for this model allows to write a closed - form for the sas intensity of polydisperse ionic fluids @xcite .
we assume that polydispersity is associated to macroions , whereas counterions are monodisperse .
the electroneutrality condition , necessary to have a stable system , then reads : @xmath69 where @xmath70 and @xmath71 are the number density and charge of the counterions , respectively .
the flac program is able to calculate the scs of neutral or charged polydisperse hard spheres . in both cases flac
also provides the result corresponding to non - interacting spheres ( i.e. @xmath72 ) .
the scattering density can correspond either to a filled or to a two - shell hollow sphere .
these models are suitable for a large variety of physical systems , among which colloids and mycelles .
flac can work in two different modes : 1 ) it can provide the theoretical calculation of the scattering function , given a set of @xmath73 ( experimental or simulated ) values ; 2 ) it can fit experimental data by using the two models described above .
the first mode is particularly useful to gauge the input parameters as a starting point for the fit .
this is governed by the _
ifit _ flag and it will described in detail below .
the filenames to be used are handled in the subroutine _
file_name _ and the proper input data are read by using the subroutine _
read_data_. subsequently , the subroutine _ control _ checks the consistency of the input values .
if required , the definition of the smearing effects is worked out before the start of the calculation of the scs in the _ fres_prep _ subroutine .
the va05ad routine of the harwell subroutine library ( hsl ) @xcite needs the definition of the functional to be minimized in the _ calfnf _ subroutine and the statistical errors are estimated in the _ kovafd _ routine developed at the institut fr festkrperforschung of the forschungszentrum jlich , germany . in input flac requires a set of physical parameters ( average size of the particles , volume fraction , etc ) plus the parameters of the distribution .
in addition to the two parameters of the distributions considered here ( gaussian , schulz , weibull and log - normal ) a parameter for the normalization is also required .
we remark , however , that these three parameters are actually computed in terms of the above physical parameters . for each @xmath73 value ,
the intensity is calculated by using the subroutine _ intensity _ , where the proper model is selected . at this stage ,
the porod background is considered . in sequence ,
the corrections due to instrumental smearing effects , vertical slit and multiple scattering are performed .
since a flat background does not alter any of the previous corrections , the incoherent background is added at the final stage . during the fit , a file called *
.int * ( see below ) is output at each iteration , for a clearer control of the convergence of the numerical procedure . upon convergence , flac estimates the errors in the physical quantities and reorganizes the data for the output files . at the initial stage flac requires a filename interactively and then it sets up input and output filenames by appending the proper extension to this filename . a short description of each extension
is reported below along with the meaning of all the files .
we note that only the * .dat * ( containing the various parameters ) is strictly necessary for running flac , the experimental data in * .ex1 , * .ex2 * and * .ex3 * being necessary only in the case of fit . in this case
the three files refer to experimental data obtained at three different experimental configurations that are are then assembled together into a single array .
all the other output files are created by flac .
the quantities contained in each file are organized in columns whose labels are reported in brackets ( used here for clarity and not appearing in the file ) and have their meaning written underneath .
: + * general input data + 0.1 in * experimental data ( q i e ) of data collected in different instrumental configurations . this file is required only in case of fit . * * scattering vector @xmath73 ( @xmath74 ) * * experimental intensity ( cm@xmath74 ) * * experimental uncertainty ( cm@xmath74 ) : + * general results + 0.1 in * calculated quantities ( q e d c b a ) * * theoretical model + porod background + instrumental resolution ( cm@xmath74 ) * * a + slit effect ( cm@xmath74 ) * * b + incoherent background ( cm@xmath74 ) * * c + multiple scattering ( cm@xmath74 ) * * d + random error ( simulation ) / experimental data ( fit ) ( cm@xmath74 ) * * scattering vector @xmath73 ( @xmath74 ) + 0.1 in * calculated quantities ( q d c b a ) * * effective structure factor * * effective form factor ( cm@xmath74 ) * * theoretical model ( cm@xmath74 ) * * c + porod background ( cm@xmath74 ) * * scattering vector @xmath73 ( @xmath74 ) + 0.1 in * calculated quantities on different @xmath73 range than the experimental one ( q@xmath76 c b a ) .
this file is generated only in case of fit .
* * effective structure factor * * effective form factor ( cm@xmath74 ) * * theoretical model ( cm@xmath74 ) * * scattering vector @xmath77 ( @xmath74 ) + 0.1 in * backgrounds ( q p f ) * * flat incoherent background ( cm@xmath74 ) * * porod background ( cm@xmath74 ) * * scattering vector @xmath73 ( @xmath74 ) + 0.1 in * scattering particles ( or scattering cores ) and hard sphere size distributions ( r a b d m n ) * * scattering particle radius ( ) * * scattering particle size distribution ( @xmath78 ) * * error of a ( @xmath78 ) * * hard sphere diameter ( ) * * hard sphere size distribution ( @xmath78 ) * * error of m ( @xmath78 ) + 0.1 in * pair correlation function ( r g ) * * distance from the center of a particle ( ) * * pair correlation function -0.2 cm we note that at the beginning of the calculation flac transforms all dimensional quantities in suitable powers of so that all calculations are performed in such units . in the final output
all quantities are then restored to their original more convenient units .
all relevant parameters are to be specified in * .dat * file .
first , the spatial range ( _ r_min _ and _ r_max _ ) of the size distribution function associated to polydispersity and its number of points ( _ num_points_int ) are required . for simplicity
, the number of input points is given in powers of 2 throughout the code .
hence , for instance , @xmath79 is the actual number of points of the distribution function .
the mesh size in reciprocal space @xmath80 is computed from the range extension ( _ q_min _ and _ q_max _ ) and the corresponding number of points ( _ num_points _ ) .
_ the instrumental resolution ( for wavelength band and finite collimation ) , detector , radial average and beam width effects are controlled by separate flags : _ ires1 _ , _ ires2 _ , _ ires3 _ , _ ires4 _ and _ ires5 _ , respectively . the number of points for the convolution ( _ nfold _ ) , the radii of source ( _ r1 _ ) and the sample ( _ r2 _ ) slits , source - sample ( _ rl1 _ ) and sample - detector ( _ rl2 _ ) distances , the wavelength ( _ rlam _ ) and its band ( _ dlam _ ) , the spatial resolution of the detector ( _ rdet _ ) , the sampling used for radial average ( _ rrav _ ) , the fwhm of the primary beam _ sbeam _ and the total number of experimental configurations _
nconf _ have to be provided and they will be used only if these corrections are considered in the calculation . note that _ rl1 _ and _ rl2 _ are arrays because of the different experimental configurations ( maximum 3 ) considered in this analysis . for the vertical slit , the flag ( _ isl _ ) , the number of points ( _ nsl _ ) used and the maximum @xmath73 value ( _ tmax _ ) are considered as input parameters . the correction for multiple scattering is controlled by a flag ( _ ims _ ) , and it requires the sample thickness ( _ thick _ ) and the number of points to be used for the 2d fourier transform ( _ nft _ ) ; for this calculation , flac is exploiting the imsl routines @xcite .
the data for the fitting routine ( vao5ad ) of the hsl are also input parameters : the step increment ( _ h _ ) , the maximum excursion ( _ dmax _ ) and the accuracy ( _ acc _ ) of the numerical calculation can be gauged at will . moreover
, the maximum number of iterations ( _ maxfun _ ) as well as the output form ( _ iprint _ ) can be also varied ( in this last option , a customized option for a better visualization of the evolution of the fit parameters , is also possible ) . in the general part of the input data ,
one finds the main flag ( _ ifit _ ) , which discriminates between the simulation of a theoretical curve or the fit of an actual set of experimental data ( _ ifit = 1 _ ) . in the first case
the theoretical scattering functions are provided either for a generic range of @xmath73 ( _ ifit = 2 _ ) or for a specific set of experimental @xmath73 values ( _ ifit = 0 _ ) .
this latter case is particularly useful to find a proper starting set of parameters for the best fit analysis . for both the fit and the simulation of experimental data
, one should define the first ( _ n1 _ ) and the last point ( _ n2 _ ) of each experimental scattering intensity file to be used and another flag indicating whether the experimental data also include the experimental uncertainty ( _ ierr _ ) .
note that also _ n1 _ and _ n2 _ are arrays like _
rl1 _ and _
rl2_. for simulated data , a random error can be introduced in the calculated intensity and _ err_int _ is the band width where the intensity can be randomly shifted using a uniform random number generator controlled by the input variable _ iseed_. a weighted fit can be chosen with _
ipes _ as well as the backgrounds ( flat and porod ones ) , which are regulated by two flags : _ ifin _ and _ ipor _ , respectively .
the flat background can be considered as it stands or it may take into account the volume fraction of the scattering particles ; this case is particularly useful for highly concentrated systems .
the choice of the size distribution function ( _ idist _ ) allows four different functions : gaussian , schulz , weibull and log - normal .
calculations for concentrated or dilute systems can be controlled by the _
inter _ flag .
the form factors of the scattering particles can be chosen between the one- and two - shell spheres by _
iff_. for charged hard spheres , the counterions charge ( _ zc _ ) and the bjerrum length ( _ lb _ ) defined by @xmath81 are to be given . here
@xmath82 is the boltzmann constant , @xmath83 the absolute temperature , @xmath66 the unit charge and @xmath68 the dielectric constant .
finally a set of @xmath84 physical parameters are provided for the calculation , each one with its own flag , describing whether its value is kept pinned to that given in the input data file ( corresponding flag=0 ) , left free for the optimization ( corresponding flag = 1 ) or fixed to a value given in an external file * .res * produced as output by flac ( corresponding flag = 2 ) .
these parameters are the volume fraction _ xf _ , the average radius _ xc _ of the scattering particles ( in the case of a two - shell sphere , this correspond to the outer shell radius ) and the dispersion _ xd _ of the particle size ; for monodisperse systems _ xd _ must be set to 0 .
a condition we have used in our calculation is that the hard sphere radius must be larger or equal to the outer dimension of the scattering particle ; @xmath85 then represents the ratio between the hard sphere radius and the outer dimension of the scattering particles . the porod constant ( _ xp _ ) and the flat background ( _ xfo _ ) are required and might also be optimized .
the differences of scattering length densities between the outer and inner shell ( _ xb _ ) , between the outer shell and the solvent for the scattering particles ( macroions ) ( _ xcon _ ) and between the counterions and the solvent ( _ xcon1 _ ) are also necessary for the calculation . the thickness of the outer shell for two - shell spheres is set by the variable _ xa _ and _ xz _ represents the charge associated to the macroions with sizes equal to the average of the distribution . finally , a scaling factor _
xk _ is necessary when data are not available in absolute units . at the end of the input file ,
the name of the external file where some data have to be used is also requested .
all the input data are tested for non - appropriate values .
all adopted distribution functions depend upon three parameters . for three distributions ( gaussan , schulz and log - normal ) ,
they can be analytically computed from the knowledge of the volume fraction , the mean radius and the polydispersity index ( dispersion ) which is a measure of the strength of polydispersity and whose definition can be found in ref @xcite . in the weibull case ,
the lack of a one - to - one correspondence in the analytical expression enforces a numerical self - consistent procedure .
all the @xmath86 physical parameters to be optimized are constraint to be positive , apart from _ xs _ which must be larger than @xmath87 and _ xb _ which can assume all real values . in the case of charged hard spheres ,
the population of counterions is determined by the electroneutrality condition and the value of the screening parameter @xmath88 is computed iteratively using eq .
50 of ref .
@xcite . the coherent transmission factor @xmath89 is calculated by estimating the theoretical scattering cross section over a very wide interval in the @xmath73 space @xmath90 \right\ } , \label{coh}\ ] ] where @xmath91 is the sample thickness and the term between square brackets represents the macroscopic coherent scattering cross section @xmath92 under the approximation of isotropic scattering , i.e. the integral of the differential one over the solid angle .
this transmission factor is an indicator of the multiple scattering probability : the higher @xmath89 , the smaller is the multiple scattering probability .
this effect is incorporated in flac , by the use of the mux routine @xcite .
the output of the routine is adapted to the experimental @xmath73 grid by using linear interpolation , with the exclusion of the last point which is obtained by extrapolation .
the structure factor for a system containing monodisperse counterions and macroions can also be computed with the charged monodisperse option . in this case
the scattering is reduced to the sum of 4 terms , corresponding to a @xmath93 matrix .
the average pair correlation function @xmath94 is calculated by using the real space mesh size @xmath95 determined directly from the one in reciprocal space by using the standard relation @xmath96 , where @xmath97 is the number of points _
num_points _ , and it exploits a standard fast fourier transform ( fft ) algorithm .
flac must be placed in the same directory of the files containing the parameters ( _ flac.par _ ) and the common definitions ( _ distribution.com , flag.com , general.com , ggc.com , intensity.com , library.com , names.com , parameters.com , points.com , quantities.com_ and _ smearing.com_ ) and it must be linked with imsl library @xcite . 0.5 cm
in order to illustrate the practical use of flac , we provide , along with the code , a set of so far unpublished experimental data collected , by us , at the point geometry kwsi diffractometer at the fzj in jlich , germany , on silica particles in hydrogenated water .
samples have been prepared from commercial ludox hs30 at the nominal volume concentration of 16.5% .
the measurements have been performed by using 4.56 neutrons with a wavelength band of about 20% and the data have been collected with three different experimental configurations , i.e. sample - to - detector and collimation distance equal to 14 , 4 and 1.2 m , in order to investigate the largest possible q range @xcite .
the samples were contained in quartz cell of 2 mm thickness , the multiple scattering being then not negligible .
hence , we have performed the data analysis by using the neutral one - level hard sphere model with 6 free parameters , i.e. the volume fraction , the average radius , the dispersion of the size distribution function , the hard sphere factor , the incoherent background and the contrast .
the weibull size distribution function has been selected in the present case .
we have optimized the free parameters by including in the calculation also the smearing due to the wavelength band , different finite collimations and multiple scattering . in fig.1
, experimental data collected at three different instrumental configurations are depicted together with the theoretical scattering cross section , the incoherent scattering , the effective form factor and the smeared scattering cross section .
error bars are smaller than the size of the symbols and hence are not displayed .
the agreement between experimental data and calculated cross section is excellent and the physical parameters shown in tab .
[ tab1 ] are well representative of the system .
the coherent transmission @xmath98 is relatively high , showing that the coherent multiple scattering does not play a major role in this sample .
a nominal value of the scattering contrast of commercial ludox hs30 has been estimated to be about 4@xmath9910@xmath100 @xmath101 , in good agreement with our optimized value .
the italian murst ( ministero delluniversit e della ricerca scientifica e tecnologica ) , the infm ( istituto nazionale per la fisica della materia ) are gratefully acknowledged for financial support .
we thank raffaele della valle for a critical reading of the manuscript .
the authors are indebted to the hsl management for the free use of the routines , and to the computing group of the institut fr festkrpeforschung of the forschungszentrum jlich , germany , for their assistence during the preparation of this work .
the dupont company is also gratefully acknowledged for providing the ludox hs30 material .
fig.1 : best fit of commercial ludox hs30 with neutral one - level hard sphere . the experimental data collected at 14 , 4 and 1.2 m are shown together with the theoretical intensity , the effective form factor , the incoherent background and the best fit obtained by smearing for the wavelength band , the finite collimation at different instrumental configurations and the multiple scattering .
.[tab1]_physical parameters obtained by the best fit with a weibull size distribution function of the commercial ludox hs30 .
the parameters optimized by the best fit procedures are : c@xmath102 the volume fraction , @xmath103 the average radius , @xmath104 the dispersion of the size distribution function , @xmath105 the ratio of the radii of the hard sphere and the scattering particle , @xmath106 the incoherent background , @xmath107 the scattering contrast . moreover , n@xmath108 the scattering particle density and s@xmath108 the specific surface are then calculated .
the coherent transmission t@xmath109 is also estimated . _ [ cols="<,^,^ " , ] | flac is a program to calculate the small - angle neutron scattering intensity of highly packed polydisperse systems of neutral or charged hard spheres within the percus - yevick and the mean spherical approximation closures , respectively .
the polydisperse system is defined by a size distribution function and the macro - particles have hard sphere radii which may differ from the size of their scattering cores . with flac
, one can either simulate scattering intensities or fit experimental small angle neutron scattering data . in output scattering intensities , structure factors and pair correlation functions
are provided .
smearing effects due to instrumental resolution , vertical slit , primary beam width and multiple scattering effects are also included on the basis of the existing theories .
possible form factors are those of filled or two - shell spheres .
0.5 cm 0.5 cm program summary 0.3 cm _ title of program : _ flac 0.3 cm _ catalogue identifier : _ 0.3 cm _ program obtainable from : _ cpc program library , queen s university of belfast , n. ireland 0.3 cm _ computer : _ digital workstation au 433 ( 128 mb ram ) , pentium i mmx 200 mhz ( 64 mb ram ) , macintosh powerbook g3 400 mhz ( 192 mb ram ) 0.3 cm _ operating systems : _ digital unix 4.0e , windows nt service pack 4 , mac 0s 9.0.2 0.3 cm _ programming language : _ fortran 77 0.3 cm _ memory required to execute with typical data : _ 640 kwords 0.3 cm _ no .
of bits in a word : _
32 0.3 cm _ no . of lines in distributed program , included test data , etc . : _
6873 ; of which 1304 are routines of the harwell subroutine library ( hsl ) library and their use must be acknowledged in any paper publishing results obtained by flac .
the entire code must be linked with the international mathematical statistical libraries ( imsl ) library .
an example of the input file _
flac.dat _ is also available ( 94 lines ) together with 11 files * .com where common definitions are found ( 331 lines ) and a file ( _ flac.par _ ) containing the parameters used for dimensioning the arrays ( 7 lines ) . 0.3 cm _ additional keywords : _ small - angle scattering , hard spheres , polydispersity , percus - yevick , mean spherical approximation . 0.3 cm _ nature of physical problem : _ the problem is the calculation of the scattering cross section in small - angle scattering of polydisperse neutral and charged hard spheres . both dense or dilute systems are considered .
0.3 cm _ method of solution : _ the algorithms implemented here are obtained by solving the orstein - zernike integral equations within the percus - yevick or mean spherical approximation closures for neutral or charged hard spheres , respectively @xcite . 0.3 cm _
restriction on the complexity of the problem : _ only hard sphere ( neutral or charged ) potentials are used . 0.3 cm _ typical running time : _ a test run considers 200 points for defining the size distribution function and calculates the scattering cross section over @xmath0 points , without any smearing effects . for neutral hard spheres , on a digital workstation au 433 this test takes 7.7 s , while on the macintosh and on the pc 26.1 and 49.6 s , respectively . on including instrumental smearing considering only one experimental configuration ( by convoluting over 11 points ) the cpu time on the digital workstation au 433 increases up to 46.8 s ;
if multiple scattering correction is also taken into account ( by using 2@xmath1 points for the 2d fourier transform ) , the cpu time is 47.3 s. furthermore , the addition of a vertical slit ( sampled by 2@xmath2 points ) requires 270.1 s of cpu time . for charged hard spheres ,
the calculation without any correction takes 12.1 s of cpu time .
0.3 cm _ references _ * a.vrij , j. chem
. phys . * 69 * , 1742 ( 1978 ) ; * 71 * , 3267 ( 1979 ) . * d. gazzillo , a. giacometti and f. carsughi , j. chem .
phys . * 107 * , 10141 ( 1997 ) . 0.3 cm long write - up |
it is commonly accepted that most quasars contain energetic outflows of ionized gas emerging from their accretion disks at speeds ranging from @xmath11 5,00030,000 km s@xmath12 ( e.g. , turnshek et al .
1988 ; weymann et al .
these outflows imprint broad absorption features bluewards of the resonant uv emission lines of c iv , si iv , n v , and o vi .
broad absorption features are expected to be observed only for lines of sight that traverse the outflow . the outflow is thought to be driven by radiation pressure on spectral lines from uv photons of the central source ( e.g. , arav et al .
1995 ; murray et al . 1995 ; proga , stone , & kallman 2000 ; srianand et al .
an estimate of the mass - outflow rate may result from the study of the properties of the outflowing winds of quasars .
this quantity may be used to evaluate the contribution of outflowing winds in distributing accretion - disk material into the vicinity of the quasar central engine and into the host galaxy over a typical life time of a quasar . constraining this rate is also important for understanding the connection between black hole and bulge growth in the host galaxy ( e.g. , fabian 1999 ) .
an estimate of the mass - outflow rate requires knowledge of the velocities and locations of the various ions that contribute to the wind .
broad absorption features in the uv band often show multiple detached components with different velocities , column densities , and ionization states .
one needs to include the contributions of all components to obtain an accurate value of the total mass - outflow rate .
present estimates of mass - outflow rates are based mostly on the contributions from ions absorbing in the rest - frame uv band .
the present x - ray data for balqsos are sparse , and only poor to moderate signal - to - noise ratio ( s / n ) spectra are available ( e.g. , chartas et al . 2001 ; gallagher et al . 2001 ; green et al . 2001 ; oshima et al .
2001 ; brinkmann et al .
2002 ; gallagher et al .
the few moderate s / n x - ray spectra of balqsos available show that their x - ray faintness is due to absorption with typical hydrogen column densities ranging from @xmath0 10@xmath1310@xmath14 @xmath15 .
the ionization , kinematic and spatial properties of the x - ray absorbing material are not well constrained .
the physical relationship between the uv and x - ray absorbers is unclear .
is the x - ray absorber part of the outflow ?
another unresolved issue is how moderately ionized species can survive the extreme uv and soft x - rays produced by the central source . to account for this
, theoretical studies have postulated the presence of an optically thick layer of shielding gas between the central source and the outflow that prevents the outflow from becoming completely ionized ( e.g. , murray et al .
recent simulations by proga et al .
( 2000 ) indicate that the outflow is self - shielding , i.e. , the shielding gas is an integral component of the outflow .
the observed x - ray absorbing gas in balqsos has been suggested as a candidate for the shielding gas ; however , there has been little direct observational evidence for this association .
recently , narrow absorption lines ( nals ) in the x - ray band have been detected in _
chandra _ and _ xmm - newton _ observations of bright seyfert 1 galaxies ( e.g. , kaspi et al .
the nals are blueshifted relative to the systemic velocity suggesting that the nal material is part of an outflow with a mean outflow velocity of a few hundred km s@xmath12 . to improve our understanding of the x - ray absorption in balqsos
, we performed a long _ chandra _ observation of the bright gravitationally lensed bal quasar apm 08279 + 5255 .
the flux magnification of apm 08279 + 5255 , estimated to be @xmath0 100 ( egami et al .
2000 ; mu~ noz et al .
2001 ) in the x - ray band , and its high redshift of @xmath1 allowed us to study the kinematic and ionization properties of a bal quasar in the x - ray band .
specifically , the high redshift of apm 08279 + 5255 places the strong fe k features at energies where the telescope effective area is much larger . here
we present the results of these observations .
apm 08279 + 5255 was observed as part of the guaranteed time observer program with the advanced ccd imaging spectrometer ( acis ) instrument ( g. p.
garmire et al .
2002 , in preparation ) onboard the _ chandra x - ray observatory _ ( cxo ) on 2000 october 11 and on 2002 february 24 for 9.2 ks and 88.8 ks , respectively .
the pointing of the telescope placed apm 08279 + 5255 on the back - illuminated s3 chip of acis . in table 1
we list the observation epochs , exposure times and detected event rates for the lensed images ( see @xmath16 3 ) . for preparing the _ chandra _ data for analysis we used the ciao 2.2 and caldb 2.12 products provided by the _ chandra _ x - ray center ( cxc ) . to improve the spatial resolution we removed a @xmath17 025 randomization applied to the event positions in the cxc processing and employed a sub - pixel resolution technique developed by tsunemi et al .
( 2001 ) . to account for the recently observed quantum efficiency decay of acis , possibly caused by molecular contamination of the acis filters , we have applied a time - dependent correction to the acis quantum efficiency based on the presently available information from the cxc .
the acis quantum efficiency decay is insignificant for energies above 1 kev and does not affect the main results of our analysis .
the _ chandra _ image of apm 08279 + 5255 obtained from the 88.8 ks observation is shown in figure 1 . to improve the spatial resolution we have applied the lucy - richardson ( l - r ) maximum likelihood deconvolution technique ( richardson 1972 ; lucy 1974 ) . for the deconvolution we supplied a point spread function ( psf ) created by the simulation tool ` marx ` ( wise et al .
the x - ray spectrum used to generate the psf was that determined from our spectral analysis ( see @xmath16 4 ) .
we find a separation between the x - ray images a and b of 038 @xmath17 001 .
recent observations of apm 08279 + 5255 with the nicmos camera onboard _ hst _
( ibata et al . 1999 ) imply the presence of three images with a separation between the two brightest images a and b of @xmath18=038 and between a and the fainter image c of @xmath19=01 . to estimate the x - ray flux ratios we modeled the _
chandra _ images of a , b and c with point spread functions generated by the simulation tool ` marx ` .
the x - ray event locations were binned with a bin - size of 00246 .
the simulated psfs were fit to the _ chandra _ data by minimizing the cash @xmath20 statistic formed between the observed and simulated images of apm 08279 + 5255 .
the relative positions of the images were fixed to the observed nicmos values .
we find x - ray flux ratios of @xmath21_{x } = 0.21 \pm 0.02 $ ] and @xmath22_{x } = 0.88 \pm 0.03 $ ] .
the x - ray flux ratios are close to the flux ratios measured in the _ hst _ f160 band of @xmath23_{f160 } = 0.22 \pm 0.01 $ ] and @xmath24_{f160 } = 0.78 \pm 0.01 $ ] .
we briefly comment that x - ray flux variability is detected in both images of apm 08279 + 5255 , and this issue will be the focus of a future paper .
spectra of apm 08279 + 5255 from the combined images for the two epochs were first extracted and analyzed separately to search for any significant long - term variability .
the source spectra were extracted from circular regions centered on the midpoint between images a and b with radii of 3 .
the background was determined by extracting events within a source - free annulus centered on the midpoint between images a and b with inner and outer radii of 10 and 30 , respectively .
all derived errors below are at the 90% confidence level unless quoted otherwise .
results from spectral fits to the 9.1 ks observation of apm 08279 + 5255 were reported in gallagher et al .
the x - ray flux of apm 08279 + 5255 appears to have decreased by a factor of @xmath0 2 between the first and second epochs ( see table 1 ) .
unfortunately , there was a moderate degree of photon pile - up in the first observation of apm 08279 + 5255 which can lead to spectral distortions and loss of events .
we therefore did not combine the spectra of the first and second epochs .
the second observation of apm 08279 + 5255 was performed in a sub - array mode to reduce the effects of pile - up .
the reduced x - ray flux of apm 08279 + 5255 during the second epoch and the reduced ccd frame - time of 2.5 s used for this observation significantly reduced the effects of pile - up , which is negligible in the subsequent analysis . in the spectral analysis that follows we focus on the data obtained from the 88.8 ks observation of apm 08279 + 5255 .
a variety of models were fit to the 88.8 ks spectrum of apm 08279 + 5255employing the software tool xspec v11 ( arnaud 1996 ) .
the spectrum was initially fit with a model consisting of a simple power law with galactic absorption due to neutral cold gas with a column density of @xmath25 = 3.9 @xmath7 10@xmath26 @xmath15 ( stark et al .
the model also included neutral intrinsic absorption at @xmath1 .
our fits support the presence of an intrinsic absorber with a column density of @xmath25 = ( 6.0@xmath27 ) @xmath7
10@xmath13 @xmath15 ( see table 2 ) .
the fit is not acceptable in a statistical sense with @xmath28 = 182.1 for 109 degrees of freedom ( dof ) .
the fit residuals show several absorption features between 1.53 kev that contribute to the unacceptable fit .
to illustrate better the presence of these absorption features and the absorption at energies below 1 kev , we fit the apm 08279 + 5255 data above the rest frame energy of 10.8 kev with a simple power - law model modified by galactic absorption and extrapolated this model to lower energies .
the spectrum above 10.8 kev ( rest - frame ) is less susceptible to intrinsic absorption and excludes the absorption features .
following this analysis strategy , the low - energy residuals indicate that intrinsic absorption is present in the spectrum .
in addition , the spectrum shows two strong absorption lines from 1.53 kev ; these are presented in figure 2a .
we attempt to model the residual features by considering a range of models of increasing complexity .
we first add to our absorbed power - law model a redshifted gaussian component near the most obvious absorption feature appearing at an energy of @xmath0 1.65 kev ( observed frame ) .
including one gaussian component in our model leads to a significant improvement in fit quality at the @xmath2 99.9% confidence level ( according to the @xmath3-test ) with @xmath28 = 146.9 for 106 dof .
the best - fit energy of the absorption feature is @xmath29 = 8.05@xmath30 kev ( rest frame ) .
this absorption feature is in a well - calibrated energy region , and the combined effective area of the _ chandra_mirrors and acis is known to vary smoothly near this energy .
the absorption feature is not resolved by acis ; we place an upper limit on its width of 140 ev ( corresponding to a fwhm of @xmath31 12,300 km s@xmath12 ) at the 90% confidence level .
we next add a second gaussian component near 1.9 kev to model the remaining residuals near this energy .
this fit is significantly improved compared to the previous fit at the @xmath2 99.9% confidence level and yields @xmath28 = 106.7 for 103 dof ; the fit is now statistically acceptable .
the best - fit value for the photon index is @xmath32 , consistent with the range of @xmath9 measured for large samples of radio - quiet quasars at lower redshifts ( e.g. , george et al . 2000 ; reeves & turner 2000 ) .
the best - fit energy for the second absorption line is @xmath33 = 9.79@xmath34 kev with a width of @xmath35 = 0.41@xmath36 kev .
the second absorption feature falls near the instrumental iridium edge produced by the _
chandra _
mirrors . to ascertain any systematic calibration uncertainties near the mirror edge we have fit the spectra of several test - sources with expected smooth power - law spectra near the 2 kev iridium edge .
since @xmath28 residuals depend on the statistics , we filtered the test - source data in time to produce spectra with a total number of counts equal to that observed in the second observation of apm 08279 + 5255 . typical @xmath28 residuals for these test - sources near the mirror edge are less than @xmath0 1@xmath37 indicating that the observed 5@xmath37 residuals near 2 kev are real and not due to systematic errors in the calibration of the effective area of the _ chandra _ mirrors .
we find a ratio of @xmath0 0.5 between data and model near the two absorption features for the second observation of apm 08279 + 5255 .
this ratio is significantly less than the observed ratio of @xmath0 0.95 between data and model for our test - sources .
the fit that includes two gaussian absorption - line components shows positive residuals near 8.4 kev and 11.2 kev ( rest frame ) .
these residuals may at least partially be the result of the simplistic absorption - line models that we have adopted in the present analysis .
uv bals usually show multiple absorption components with non - gaussian profiles , and such complex profiles may also be present in x - ray bals .
to test further the robustness of our modeling we also attempted to fit the 1.53 kev residuals in figure 2a with a model that included an absorption edge . specifically , for a model consisting of a simple power law with galactic absorption , intrinsic absorption , and one absorption edge we find a best fit energy and optical depth of the edge of @xmath38 = 7.68@xmath39 kev and @xmath40 = 0.37@xmath41 , respectively .
this model did not provide an acceptable fit with @xmath42 = 146 for 107 dof . in figure 2b
we show that the fit of the apm 08279 + 5255 spectrum with a one - edge model produces significant residuals .
the reason for these large residuals is that the absorption edge is too broad to fit the narrow absorption lines .
the addition of a second absorption edge to the model did not result in a significant improvement of the fit .
our analysis of the _ chandra _ spectrum of apm 08279 + 5255 shows strong evidence for the presence of absorption lines at rest - frame energies of 8.05@xmath43 kev and 9.79@xmath34 kev ( fit 3 of table 2 ) .
of all the abundant elements , iron absorption lines would be the closest in energy to the observed features . in this sense ,
our interpretation that the absorption lines are associated with fe k absorption is the most conservative one possible ( e.g. , absorption lines from relativistic oxygen would require much larger blueshifts ) . at higher atomic numbers than iron ( corresponding to higher absorption - line energies ) , there are no abundant elements .
observationally , fe k absorption lines have been seen from other objects , such as the x - ray binaries grs 1915 + 105 ( kotani et al .
2000 ) , circinus x-1 ( brandt & schulz 2000 ) , and gx 13 + 1 ( ueda et al .
2001 ) and possibly the agn ngc 3516 ( nandra et al .
we note that the expected ratios of the energies of the iron k@xmath44 to k@xmath45 transitions of fe xxv and fe xxvi are just outside the 2@xmath37 confidence limits of the ratio of the energies of the two observed absorption lines .
because of the possible presence of multiple components within each of the observed absorption lines , we expect that future x - ray observations of apm 08279 + 5255 with higher energy resolution will unambiguously show whether the observed absorption lines correspond to the k@xmath45 and k@xmath44 transitions of fe xxv or fe xxvi . a plausible site that may be producing the absorption features is the outflowing disk wind .
complex absorption features are also observed in the uv spectrum of apm 08279 + 5255 . in particular , recent high resolution spectroscopy of apm 08279 + 5255 reveals multiple velocity components of c iv @xmath46(1548 , 1551 ) , at @xmath0 4670 km s@xmath12 , @xmath0 9670 km s@xmath12 , and @xmath0 12,400 km s@xmath12 ( srianand & petitjean 2000 ) .
broad absorption due to o vi @xmath46(1032 , 1037 ) , n v@xmath46(1238 , 1242 ) and si iv @xmath46(1393 , 1402 ) is also detected .
the wide range of ionization levels , velocities and column densities inferred from the uv bals of apm 08279 + 5255 imply that the wind may be composed of multiple regions of different densities , with large ionization gradients ( srianand & petitjean 2000 ) . for our estimate of the ejection velocities of the absorbers we assume as argued above that the x - ray absorption features are due to absorption from fe ions in the flow .
the rest energies of the most likely resonant absorption lines of fe are 6.70 kev ( fe xxv k@xmath45 ) , 7.88 kev ( fe xxv k@xmath44 ) , 6.97 kev ( fe xxvi k@xmath45 ) and 8.25 kev ( fe xxvi k@xmath44 ) .
we estimate that the 8.05 kev and 9.79 kev absorption features correspond to wind velocities ( depending on the ionization state of fe ) of @xmath47 ( fe xxv k@xmath45 ) , @xmath48 ( fe xxvi k@xmath45 ) and @xmath49 ( fe xxv k@xmath45 ) , @xmath50 ( fe xxvi k@xmath45 ) respectively , relative to the systemic redshift of @xmath1 . for the velocity calculations we considered the special relativistic velocity correction and assumed that the angle between the wind velocity and our line of sight is 20@xmath51 . this angle is not constrained with the present data
; however , hydrodynamical simulations indicate that the bal wind opening angle may range between 10@xmath5130@xmath51 depending on the location of the inner radius of the disk .
we examined the low - resolution optical spectrum of apm 08279 + 5255presented in ellison et al .
( 1999 ) for evidence of notable absorption in c iv , si iv , and n v with velocities consistent with the x - ray absorption lines . in the spectral region
covered , no strong systems were apparent , though a detailed analysis of a high - resolution spectrum ( such as that presented in srianand & petitjean 2000 for lower velocities ) remains to be done .
the non - detection of these ions is not a serious problem as they probably are not present in such a highly ionized gas . to obtain some insight into the kinematics of the wind flow and the relative locations of the uv and x - ray absorbers , we calculated the wind velocity as a function of radius .
more realistic velocity profiles of disk outflows have been obtained from hydrodynamical calculations ( e.g. , proga et al .
2000 ) that predict radial and azimuthal variations of the wind velocity . for the purposes of our simple analysis
we assume that the velocity of an outflow produced by radiation pressure from a central source with a uv luminosity of @xmath52 and a mass of @xmath53 is given by : @xmath54^{1/2}},\ ] ] where @xmath55 is the eddington luminosity , @xmath56 is the force multiplier ( see the discussion of @xmath56 in laor & brandt 2002 ) , @xmath57 is the radius at which the wind is launched from the disk , and @xmath58 is the distance of the accelerated portion of the flow from the central source . in figure 3
we plot wind velocity versus radius for material launched at radii of 2 @xmath7 10@xmath8 cm , 5 @xmath7 10@xmath8 cm and 1 @xmath7 10@xmath59 cm .
we have assumed @xmath60 = 100 ( e.g. , arav , li , & begelman 1994 ) , @xmath52 = 4 @xmath7 10@xmath61 erg s@xmath12 , @xmath62 = 2 @xmath7 10@xmath63 erg s@xmath12 , and @xmath62/@xmath55 = 0.1 .
simulations performed by proga et al .
( 2000 ) for a disk accreting onto a 10@xmath64 @xmath65 black hole yield a minimum launching radius for the uv absorber of @xmath0 10@xmath66 cm . scaling this result for the likely black hole mass of apm 08279 + 5255 of @xmath0 2 @xmath7 10@xmath67 @xmath65
, we estimate a minimum launching radius for the uv absorber of @xmath0 2 @xmath7 10@xmath8 cm for this luminous quasar . from a qualitative perspective
our velocity curves indicate that the wind reaches nearly its terminal velocity close to the launching radius .
a possible explanation of the relatively large velocities of the x - ray bals compared to the uv bals is that the material creating the x - ray bals is launched at smaller radii ; the relatively small launching radii implied are consistent with the higher degree of ionization needed to produce the observed absorption lines if they are due to fe xxv and/or fe xxvi .
a rigorous calculation of the velocity of the x - ray bal wind is beyond the scope of this letter .
we note , however , that the largest unknown in this estimation is the value of @xmath68 , which will depend on the ionization state and velocity structure of the gas .
if the observed x - ray absorption is due to fe xxv , the high ionization parameter of this gas indicates that only heavier metals will still have any electrons . in this case
, the dominant radiation pressure may be provided by x - rays rather than uv photons .
specifically , such highly ionized gas will be driven primarily by bound - free absorption and compton scattering , though considerable non - thermal broadening may cause bound - bound absorption to contribute significantly as well ( e.g. , chelouche & netzer 2001 ; d. chelouche 2002 , private communication ) . a comparison between the recombination time - scale of the ionized wind , @xmath69 , and the travel - time
, @xmath70 , for the gas to reach a certain distance from the launching radius can constrain the location of the material making the x - ray bals . for @xmath69 @xmath71 @xmath72
we expect the x - ray bal material to be at small radii ( near the launching radius ) to account for the high ionization .
if @xmath69 @xmath73 @xmath72 the x - ray bal can be located significantly away from the launching radius .
the recombination time - scale for fe xxv is @xmath69 @xmath74 s , where @xmath75 k is the electron temperature and @xmath76 @xmath77 is the electron number density ( allen 1973 ) . for a range of electron densities of 1 @xmath7 10@xmath78 @xmath77 to 1 @xmath7 10@xmath67 @xmath77 ( hydrodynamical models of line - driven disk winds by proga et al .
2000 predict electron number densities at the base of the wind of 5 @xmath7 10@xmath78 @xmath77 to 5 @xmath7 10@xmath79 @xmath77 ) we find the recombination time - scales to range between 4.4 @xmath7 10@xmath80 s and 4.4 s. for a launching radius of 2 @xmath7 10@xmath66 cm the amount of time needed for the radiatively driven gas to reach a distance of 5 @xmath7 10@xmath66 cm from the launching radius is @xmath0 4 @xmath7 10@xmath81 s. we conclude that @xmath69 @xmath71 @xmath72 implying that the x - ray bal material is located at relatively small radii ( @xmath82 2 @xmath7 10@xmath8 cm ) near the launching radius of the wind .
again this geometry explains both the high velocity of the observed x - ray bals and the high ionization needed to obtain fe xxv or fe xxvi . using a curve of growth analysis ( spitzer 1978 ) we estimated the hydrogen column densities implied by the observed equivalent widths of the two absorption lines at 8.05 and 9.79 kev .
assuming the ion species responsible for the x - ray absorption in both lines is fe xxv and @xmath83 parameters of the order of the observed widths of the lines ( @xmath84 , where @xmath85 is the velocity width of the line ) , we calculate that the column densities of the absorbers are @xmath86 @xmath11 ( 3.4@xmath87 ) @xmath7 10@xmath59 @xmath15 and @xmath88 @xmath11 ( 3.8@xmath89 ) @xmath7 10@xmath59 @xmath15 , respectively . for no ionization correction and
assuming solar abundances , the implied total hydrogen column densities of these absorbers are @xmath90 @xmath11 ( 1.0@xmath91 ) @xmath7 10@xmath92 @xmath15 and @xmath93 @xmath11 ( 1.1@xmath94 ) @xmath7 10@xmath92 @xmath15 , respectively .
we emphasize that there are significant limitations with the present curve of growth analysis ; the absorption lines may contain multiple unresolved components implying that the equivalent widths and @xmath83 parameters used should be considered as upper limits .
in addition , the velocity widths estimated from fits of gaussian lines to observed absorption features can only be used to derive @xmath83 parameters when the absorber is optically thin . it is not clear from the present data if the intrinsic absorber which attenuates the low - energy continuum and the absorber responsible for the absorption lines are the same .
the models adopted in table 2 assume neutral absorption at @xmath1 .
for an ionized absorber or for the case where there is partial covering , one expects the estimated hydrogen column density to be even larger than the value estimated assuming a neutral absorber ( i.e. , a fit to the spectrum of apm 08279 + 5255 with an ionized absorber having an ionization parameter of @xmath95 yields a column density of @xmath25 = 7.5 @xmath7 10@xmath13 @xmath15 ) .
we note that the low - energy absorption could include overlapping bals from h - like and he - like ions of mg , si , and s that perhaps could still survive along with the highly ionized h - like and he - like ions of fe ( eg . , kallman & bautista 2001 ) . based on our estimated column densities of the x - ray bals , we calculated the strengths of the associated absorption edges of fe xxv ( fe xxvi ) to be @xmath96 @xmath0 0.07(@xmath97 @xmath0 0.04 ) .
we added these absorption edges to our spectral model for apm 08279 + 5255 ( fit 3 of table 2 ) and found that we do not expect them to be detectable , consistent with the data . the observed relativistic velocities , crude estimates of the column densities , and locations of the x - ray bals allow us to place constraints on the mass - outflow rate for apm 08279 + 5255 .
we estimated the mass - outflow rate as a function of the distance @xmath58 to the absorber for @xmath98 ranging from 110 , where @xmath99 is the thickness of the absorber .
we assumed a hydrogen column density of @xmath25 = 1 @xmath7 10@xmath92 @xmath15 , a covering fraction of 0.2 , and a wind velocity of @xmath5 . to obtain a reasonable mass - outflow rate ( e.g. , proga et al .
2000 ) of @xmath82 1 m@xmath100 year@xmath12 the radial location of the gas creating the x - ray bals must be less than @xmath0 10@xmath8 cm which is consistent with the radius implied from the observed high velocities and ionization states of the x - ray bals . to summarize
, we have reported on the discovery of the first x - ray bals in the gravitationally lensed quasar apm 08279 + 5255 .
the energies of the two observed absorption features of 8.05 kev and 9.79 kev ( rest frame ) suggest the presence of two distinct absorption systems with velocities of @xmath0 @xmath5 and @xmath0 @xmath6 , respectively .
the combination of estimated short recombination time - scales for fe xxv and/or fe xxvi and the observed relativistic velocities suggest that the x - ray absorbers are launched at radii of @xmath82 2 @xmath7 10@xmath8 cm .
curve of growth estimates imply column densities of the two x - ray bals in apm 08279 + 5255 of the order of 1 @xmath7 10@xmath92 @xmath15 .
one of the key implications of our results is that x - ray bals appear to be located within the uv bal region and may therefore represent the shielding gas proposed in several theoretical studies of line - driven disk winds ( e.g. , murray et .
al . 1995 ; proga et al . 2000 ) .
we expect future follow - up observations of apm 08279 + 5255with the _ chandra _ high - energy transmission grating and the _ xmm - newton _ rgs to identify the absorbing ions and resolve the x - ray absorption features into multiple components allowing for tighter constraints of the properties of x - ray bals .
we also plausibly expect variability of the absorption - line profiles over short time - scales of @xmath0 days based on our estimates of the launching radii of the x - ray absorbers . during the review of this paper ,
a related paper was posted on the apjl web site by hasinger , schartel , & komossa ( 2002 ) .
they report the detection of an ionized fe k edge in apm 08279 + 5255 with _ xmm - newton_. a visual comparison between the _ xmm - newton _ and _ chandra _ spectra of apm 08279 + 5255suggests the presence of the two absorption features in both cases ; however , the equivalent widths of the absorption features appear to be reduced during the deep _ xmm - newton _ observation .
the edge model adopted in the analysis of the _ xmm - newton _ spectrum of apm 08279 + 5255 does not provide an acceptable fit to the _ chandra_spectrum of this object .
this may imply significant variability of the x - ray bals as suggested in our paper .
+ we thank michael eracleous and doron chelouche for useful discussions and suggestions .
we acknowledge financial support from nasa grants nas 8 - 38252 and nas 8 - 01128 .
wnb acknowledges financial support from nasa ltsa grant nag5 - 8107 and nasa grant nag5 - 9932 .
scg acknowledges financial support from nasa gsrp grant ngt5 - 50277 and from the pennsylvania space grant consortium .
arav , n. , li , z. , & begelman , m. c. 1994 , , 432 , 62 + arav , n. , korista , k. t. , barlow , t. a. , & begelman , m. c. 1995 , , 376 , 576 + arnaud , k. a. 1996 , asp conf .
ser . 101 : astronomical data analysis software and systems v , 5 , 17 + allen , c. w. , 1973 , astrophysical quantities . athlone press , london + brandt , w. n. & schulz , n. s. 2000 , , 544 , l123 + brinkmann , w. , ferrero , e. , & gliozzi , m. 2002 , , 385 , l31 + chartas , g. , dai , x. , gallagher , s. c. , garmire , g. p. , bautz , m. w. , schechter , p. l. , & morgan , n. d. 2001 , , 558 , 119 + chelouche , d. & netzer , h. 2001 , , 326 , 916 + ellison , s. l. , lewis , g. f. , pettini , m. , sargent , w. l. w. , chaffee , f. h. , foltz , c. b. , rauch , m. , & irwin , m. j. 1999 , , 111 , 946 + egami , e. , neugebauer , g. , soifer , b. t. , matthews , k. , ressler , m. , becklin , e. e. , murphy , t. w. , & dale , d. a. 2000 , , 535 , 561 + fabian , a. c. 1999 , , 308 , l39 + gallagher , s. c. , brandt , w. n. , laor , a. , elvis , m. , mathur , s. , wills , b. j. , & iyomoto , n. 2001 , , 546 , 795 + gallagher , s. c. , brandt , w. n. , chartas , g. , & garmire , g. p. 2002
, , 567 , 37 + george , i. m. , turner , t. j. , yaqoob , t. , netzer , h. , laor , a. , mushotzky , r. f. , nandra , k. , & takahashi , t. 2000 , , 531 , 52 + green , p. j. , aldcroft , t. l. , mathur , s. , wilkes , b. j. , & elvis , m. 2001 , , 558 , 109 + hasinger , g. , schartel , n. , & komossa , s. 2002 , , 573 , l77 + ibata , r. a. , lewis , g. f. , irwin , m. j. , leh ' ar , j. , & totten , e. j. 1999 , , 118 , 1922 + kaspi , s. , brandt , w. n. , george , i. m. , netzer , h. , crenshaw , d. m. , gabel , j. r. , hamann , f. w. , kaiser , m. e. , koratkar , a. , kraemer , s. b. , kriss , g. a. , mathur , s. , mushotzky , r. f. , nandra , k. , peterson , b. , shields , j. c. , turner , t. j. , & zheng w. 2002 , astro - ph/0203263 + kallman , t. & bautista , m. 2001 , , 133 , 221 + kotani , t. , ebisawa , k. , dotani , t. , inoue , h. , nagase , f. , tanaka , y. , & ueda , y. 2000 , , 539 , 413 + laor , a. & brandt , w. n. 2002 , , 569 , 641 + lucy , l. b. 1974 , , 79 , 745 + muoz , j. a. , kochanek , c. s. , & keeton , c. r. 2001 , , 558 , 657 + murray , n. , chiang , j. , grossman , s. a. , & voit , g. m. 1995 , , 451 , 498 + nandra , k. , george , i. m. , mushotzky , r. f. , turner , t. j. , & yaqoob , t. 1999 , , 523 , l17 + oshima , t. , mitsuda , k. , fujimoto , r. , iyomoto , n. , futamoto , k. , hattori , m. , ota , n. , mori , k. , ikebe , y. , miralles , j. m. , & kneib , j .-
, 2001 , , 563 , l103 + proga , d. , stone , j. m. , & kallman , t. r. 2000 , , 543 , 686 + reeves , j. n. & turner , m. j. l. 2000 , , 316 , 234 + richardson , w. h. 1972 , optical society of america journal , 62 , 55 + spitzer , l. 1978 , physical processes in the interstellar medium ( new york : wiley ) + srianand , r. & petitjean , p. 2000 , , 357 , 414 + srianand , r. , petitjean , p. , ledoux , c. , & hazard , c. 2002 , mnras , in press , astro - ph/0205524 + stark , a. a. , gammie , c. f. , wilson , r. w. , bally , j. , linke , r. a. , heiles , c. , & hurwitz , m. 1992 , , 79 , 77 + turnshek , d. a. , grillmair , c. j. , foltz , c. b. , & weymann , r. j. 1988 , , 325 , 651 + tsunemi , h. , mori , k. , miyata , e. , baluta , c. , burrows , d. n. , garmire , g. p. , & chartas , g. 2001 , , 554 , 496 + ueda , y. , asai , k. , yamaoka , k. , dotani , t. , & inoue , h. 2001 , , 556 , l87 + weymann , r. j. , morris , s. l. , foltz , c. b. , & hewett , p. c. 1991 , , 373
, 23 + wise , m. w. , davis , j. e. , huenemoerder , houck , j. c. , dewey , d. flanagan , k. a. , and baluta , c. 1997 , _ the marx 3.0 user guide , cxc internal document _ available at http://space.mit.edu/asc/marx/ + clclllll + + & & & & & & & + & & & & & & & + & & & & & & & + & & ( s ) & 10@xmath101 cnts s@xmath12 & 10@xmath101 cnts s@xmath12 & 10@xmath101 cnts s@xmath12 & 10@xmath101 cnts s@xmath12&10@xmath102 cnts
s@xmath12 pixel@xmath12 + & & & & & & & + 2000 oct 11 & 1643 & 9,137 & @xmath103@xmath104 & @xmath103@xmath104 & @xmath103@xmath104 & 10.96 @xmath17 0.34 & 1.68 @xmath17 0.22 + 2002 feb 24 & 2979 & 88,817 & 3.1 @xmath17 0.07 & 2.8 @xmath17 0.06 & 0.66 @xmath17 0.07 & 6.56 @xmath17 0.09 & 0.99 @xmath17 0.05 + & & & & & & & + notes : + @xmath105 @xmath106 , @xmath107 and @xmath108 are the event rates for images a , b and c estimated from a two - dimensional fit to the _ chandra _ image of apm 08279 + 5255 .
only events with standard _ asca _ grades 0,2,3,4,6 and with energies from 0.210 kev were used in the binned image . + @xmath109 @xmath110 is the detected event rate from apm 08279 + 5255extracted from a circular region centered on the midpoint of images a and b with a radius of 3 .
only events with standard _ asca _ grades 0,2,3,4,6 and with energies from 0.210 kev were extracted .
+ @xmath111 @xmath112 is the detected background event rate per acis pixel ( 1 acis pixel = 0492 ) extracted from an annulus centered on apm 08279 + 5255 with inner and outer radii of 10 and 30 , respectively .
only events with standard _ asca _ grades 0,2,3,4,6 and with energies from 0.210 kev were extracted .
+ @xmath104 it was not possible to determine the individual count rates for the first observation of apm 08279 + 5255 due to the low s / n and the presence of moderate photon pile - up .
+ clcc + + + & & & + & & & + & & & + & & & + 1 & power - law ( pl ) and & @xmath113 & 1.72@xmath114 + & neutral absorption at & @xmath115 & ( 6.0@xmath27 ) @xmath7 10@xmath13 @xmath15 + & source . &
@xmath116 & 182.1/109 + & & @xmath117@xmath104 & 1.4 @xmath7 10@xmath118 + & & & + & & & + 2 & pl , neutral absorption & @xmath113 & 1.73@xmath119 + & at source , and one & @xmath115 & ( 6.4@xmath120 ) @xmath7 10@xmath13 @xmath15 + & gaussian absorption line & e@xmath121 & 8.05@xmath30 kev + & at source . &
@xmath122 & @xmath82 0.140 kev + & & ew@xmath121 & 0.23@xmath123 kev + & & @xmath116 & 146.9/106 + & & @xmath117@xmath104 & 5.3 @xmath7 10@xmath118 + & & & + 3 & pl , neutral absorption & @xmath113 & 1.72@xmath10 + & at source , and two & @xmath115 & ( 6.7@xmath124)@xmath7 10@xmath13 @xmath15 + & gaussian absorption lines & e@xmath121 & 8.05@xmath43 kev + & at source . &
@xmath122 & @xmath82 0.140 kev + & & ew@xmath121 & 0.24@xmath123 kev + & & e@xmath125 & 9.79@xmath34 kev + & & @xmath35 & 0.41@xmath36 kev + & & ew@xmath125 & 0.43@xmath126 kev + & & @xmath116 & 106.7/103 + & & @xmath117@xmath104 & 0.41 + & & & + notes- + @xmath127 all model fits include fixed , galactic absorption of @xmath128 @xmath15 ( stark et al .
+ @xmath129 all absorption - line parameters are calculated for the rest frame .
+ @xmath130 all errors are for 90% confidence with all parameters taken to be of interest except absolute normalization . + @xmath131 @xmath132 is the probability of exceeding @xmath28 for @xmath133 degrees of freedom . | we report the discovery of x - ray broad absorption lines ( bals ) from the balqso apm 08279 + 5255 originating from material moving at relativistic velocities with respect to the central source .
the large flux magnification by a factor of @xmath0 100 provided by the gravitational lens effect combined with the large redshift ( @xmath1 ) of the quasar have facilitated the acquisition of the first high signal - to - noise x - ray spectrum of a quasar containing x - ray bals .
our analysis of the x - ray spectrum of apm 08279 + 5255 places the rest - frame energies of the two observed absorption lines at 8.1 and 9.8 kev .
the detection of each of these lines is significant at the @xmath2 99.9% confidence level based on the @xmath3-test .
assuming that the absorption lines are from fe xxv k@xmath4 , the implied bulk velocities of the x - ray bals are @xmath0 @xmath5 and @xmath0 @xmath6 , respectively .
the observed high bulk velocities of the x - ray bals combined with the relatively short recombination time - scales of the x - ray absorbing gas imply that the absorbers responsible for the x - ray bals are located at radii of 2 @xmath7 10@xmath8 cm , within the expected location of the uv absorber . with this implied geometry
the x - ray gas could provide the necessary shielding to prevent the uv absorber from being completely ionized by the central x - ray source , consistent with hydrodynamical simulations of line - driven disk winds .
estimated mass - outflow rates for the gas creating the x - ray bals are typically less than a solar mass per year .
our spectral analysis also indicates that the continuum x - ray emission of apm 08279 + 5255 is consistent with that of a typical radio - quiet quasar with a spectral slope of @xmath9 = 1.72@xmath10 .
# 1 # 1 |
the standard slow - roll inflationary paradigm , which can give rise to the primordial perturbations , has been emerged as a solution of cosmological problems , namely , flatness , homogeneity , and isotropy problems @xcite .
observational precision has been dramatically improved recently @xcite and will be further improved in near future @xcite .
thus the refined understanding of the origin of inflation would be hoped even though there exists degeneracy problem @xcite .
indeed , some well known inflationary models are started to be disfavored by recent data @xcite .
some of the key measurements toward the origin of the inflation are the scalar spectral index , @xmath2 , the tensor - to - scalar ratio , @xmath3 , and the primordial non - gaussianity ( ng ) , @xmath4 .
the final nine - year wmap data , in conjunction with ecmb , bao and @xmath5 data @xcite , suggest that these observables are within a window : @xmath6 combination of these measurements provide a powerful guidance to distinguish specific inflationary models .
for instance , it is known that canonical single field models can not produce a sizable ng @xcite and thus the observation of the large ng will strongly disfavors single field models but favors multi - field inflationary models . recently
, ` higgs - inflation ' scenarios where the inflaton , @xmath7 , is identified with the electroweak higgs boson in jordan frame with non - minimal ( nm ) coupling @xmath8 with the ricci scalar , @xmath9 , @xcite and its generalization to arbitrary monotonic function , @xmath10 , have been studied with one - field model @xcite .
the most interesting observation would be the fact that an asymptotically flat potential in einstein frame is obtained when the ratio between the scalar potential @xmath11 and the square of the nm coupling term @xmath12 approaches a constant , @xmath13 where @xmath14 is a positive constant .
the rather extensive study on ng in single - field inflation model with nm coupling term was recently performed @xcite , in particular , for the @xmath15 potential with @xmath16 term @xcite . in this paper
we consider two inflaton fields aiming toward even further generic multi - field cases with nm couplings , namely @xmath17 and @xmath18 .
indeed multi scalar fields and their nm couplings with ricci scalar are naturally introduced in supergravity theories taking generic khler potential into account @xcite .
particularly , we address whether a large ng could be accompanied with nm coupling term in the presence of multi - field inflation .
to be specific , we focus on the case where one of the two fields has a significant nm coupling in this paper and reserve more generic case for the future .
the rest of this paper is organized as follows : in section [ sec:2fields ] , we develop a set of tools in @xmath19n - formalism to analyze the two - field inflation with nm coupling term .
the analytic expressions for physical quantities including power spectrum , spectral index , tensor - to - scalar ratio , and non - linearity parameter are derived for a generic setup . in the subsequent section [ sec : model ] , as a concrete example , we apply our results to a specific model of separable potential , @xmath20 with the nm term @xmath21 . for the future experiment
, we show the parameter space which is consistent with the current observation but still provides large ng .
finally , the conclusion is given in section [ sec : conclusion ] . in appendices ,
we provide useful analytic expressions for the number of e - foldings in various two - field cases with and without nm couplings , which are extensively used in the text .
having two scalar fields , @xmath22 , we write the generic ( super)gravity action in jordan frame as @xmath23 where the nm coupling term , in general , is expanded as @xmath24 and @xmath25 is the inflaton potential in the jordan frame . units . ]
the constant term @xmath26 can be absorbed in the definition of @xmath27 so is neglected in below .
the specific forms of @xmath28 and @xmath29 depend on their origins . for the given nm coupling term
, we always find the conformal transformation leading to the einstein frame : @xmath30 where the conformal factor , @xmath31 , is given by @xmath32 the resultant action in the einstein frame is , then , @xmath33 -u(\phi_{\rm{j}},\chi_{\rm{j } } ) , \label{eq : le}\end{aligned}\ ] ] where @xmath34 and @xmath35 is the potential in the einstein frame , which is related to @xmath29 by @xmath36 note that the subscript comma ` , ' denotes the partial derivative , _
e.g. _ , @xmath37 .
now a few comments on the general properties of the action are in order : * up to this point , the form of the lagrangian is symmetric about @xmath38 if @xmath39 and @xmath40
. the symmetry can be ( explicitly ) broken either by the nm coupling term and/or the potential . *
suppose that @xmath41 and @xmath42 in the limit of large fields , @xmath43 .
a generic condition for large field inflation can be easily noticed since the potential in the einstein frame , eq . , looks @xmath44 if the asymptotic value is constant , as in eq .
, the potential involves a large plateau which may be responsible for large field inflation .
even when one of the two directions is asymptotically flat , it is still possible to have a slow - roll inflation along that direction . in this case , the condition for the flat potential at the particular field direction , @xmath45 here , becomes @xmath46 these results are analogous to those in @xcite .
the kinetic term in eq . is not separable in general and thus is not canonically normalizable as far as @xmath48 , @xmath49 , and @xmath50 in eq . are mixed functions of @xmath45 and @xmath51 .
we restrict our interest to the cases of canonically normalizable two fields . to be specific , considering that the nm coupling is only a function of one of the two fields , @xmath52 or the terms depending on @xmath51 are negligibly smaller than the terms with @xmath45 , we have @xmath53 , @xmath54 , and @xmath55 . more concretely , @xmath56 then the canonically normalized field in the einstein frame is obtained by solving the equation : @xmath57 the action in the einstein frame , eq . ,
is reduced to a well known form @xcite : @xmath58 where @xmath59 is canonically normalized field of @xmath45 and @xmath60 .
one should notice that the ordinary minimally coupled case is recovered by letting @xmath61 ( or @xmath62 ) .
if the potential is a product - separable in jordan frame , as @xmath63 , the potential in einstein frame is also separable as @xmath64 where @xmath65 .
this type of model has been extensively studied in @xcite and we will closely follow the existing studies in subsequent sections .
in addition , for the brevity of the notation , we will drop subscript ` @xmath66 ' in the einstein frame below . in the background friedmann - robertson - walker metric , the equations of motion for the scalar fields are @xmath67 where @xmath68 is the hubble parameter . from the einstein equations we get @xmath69 under the assumption of slow - roll , eqs .
become @xmath70 in slow - roll inflationary models , it is convenient to introduce slow - roll parameters @xmath71 the slow - roll condition requires @xmath72 , @xmath73 , @xmath74 , and @xmath75 .
it is also useful to introduce a dimensionless angle , @xmath76 , defined by @xmath77 and , in the slow - roll approximation , @xmath78 in terms of the dimensionless angles , the cosmological observables are expressed as @xmath79,\nonumber\end{aligned}\ ] ] where @xmath80,\end{aligned}\ ] ] and @xmath81 note that the sub(super)script @xmath82 indicates the value at the horizon exit ( end of inflation ) and the prime index denotes the derivative with respect to the corresponding fields as usual . .
[ fig : x00 ] ] to see the numerical significance of the non - linearity parameters , @xmath83 , @xmath84 , @xmath85 , @xmath86 , and @xmath87 , we plotted them in fig .
[ fig : x00 ] with fixed @xmath88 for numerical evaluation .
similar results are found for @xmath89 , too .
the general tendency is summarized as follows : * since the @xmath83 is always less than unity and smaller than other parameters , we will ignore @xmath83 in evaluating cosmological observables .
* the parameters , @xmath84 , @xmath86 , and @xmath87 can be significant when @xmath90 or @xmath91 . on the other hand , when @xmath92 or @xmath93 , @xmath85 and @xmath87 are significant .
* due to the large @xmath86-dependent term in eq .
, @xmath94 case produces large ng when @xmath95 , respectively .
details of the cosmological observables depend on the explicit realization of the model . in the next section
, we will more closely examine a particular case with @xmath21 and @xmath96 , which leads an interesting phenomenology .
in this section , we consider a phenomenological potential of the form in the jordan frame , @xmath98.\ ] ] this potential could be regarded as a hybrid of the chaotic inflation with @xmath99 term @xcite and the natural inflation with cosine term along @xmath100 direction @xcite .
interestingly , the potential in the einstein frame involves a flat plateau along @xmath7 direction ( or natural inflation along the orthogonal @xmath100 direction ) . when the nm coupling term is of the form of @xmath101 , the condition , @xmath102 , @xmath103 in eq .
is satisfied .
the potential and the typical paths of inflation are depicted in fig .
[ fig : pot ] for minimal ( @xmath104 , left ) and non - minimal ( @xmath105 , right ) case , respectively . ) and nm ( @xmath106 ) case , respectively .
we fix @xmath107 and @xmath108 .
the two dotted lines in the lower two figures stand for the typical trajectories for minimal ( @xmath104 ) and nm ( @xmath109 ) cases with different initial field values .
[ fig : pot ] ] in the minimally coupled case ( @xmath104 or equivalently @xmath61 in eq . ) , the potential in the jordan frame and the einstein frame become identical with @xmath110 and @xmath111 : @xmath112.\end{aligned}\ ] ] here we drop the the subscript , ` @xmath113 ' .
the shape of the potential and the typical trajectories of the inflaton are depicted in the left two figures in fig .
[ fig : pot ] . the potential may be understood as the mass term along the @xmath7 direction , which varies periodically along the @xmath100 direction . for each trajectory , the number of e - foldings is found in a rather simple form ( see appendix [ appendix : ne minimal ] for derivation ) : @xmath114,\end{aligned}\ ] ] that determines the evolving field values of @xmath7 and @xmath100 for a given number of e - foldings : @xmath115,\end{aligned}\ ] ] where @xmath116 and @xmath117 are the initial values of inflaton .
here we only consider ` large - to - small ' evolution along the @xmath7 direction and @xmath118 for reality of @xmath7 during inflation . in general , @xmath100 is a periodic function sitting in the cosine potential but we only restrict ourselves to the case where the inflation takes place within one period of oscillation . taking @xmath119 , we find @xmath120 and @xmath121 as the values of fields at the end of inflation @xmath122 , @xmath123 , @xmath3 , and @xmath124 are depicted for @xmath104 case .
the large ng above @xmath125 is hardly obtainable in this scenario when the observational constraints from wmap9 are taken into account .
[ fig : m ] ] the slow - roll parameters are obtained following the eq . with @xmath61 :
\label{eq : min}\end{aligned}\ ] ] the slow - roll conditions , @xmath127 , lead to the domain of the field values in @xmath128 having all the slow - roll parameters in eq . , we can find the analytic expressions of the power spectrum , spectral index , tensor - to - scalar ratio , and the non - linearity parameter @xcite .
the results are presented in appendix [ appendix : min ] .
one immediately notices that the power spectrum @xmath129 sets the value of @xmath130 parameter as @xmath131 but the other parameter @xmath132 is still to be determined .
thus we will treat @xmath132 as a free parameter . in fig .
[ fig : m ] , we plotted the cosmological observables in @xmath133 and @xmath134 with @xmath108 . in the parameter space we typically found @xmath135 .
though we did not present numerical results for a different domain where the inflation takes place with @xmath136 , we sometimes found larger values of @xmath137 .
however , this domain turns out to be largely in conflict with the measured cosmological observable : @xmath138 .
we summarize some representative results in table .
[ table : m ] .
.results for the minimally coupled case with @xmath104 and @xmath139 .
[ table : m ] [ cols="^,^,^,^,^,^ " , ]
when the nm coupling term , @xmath140 , is introduced , the phenomenology of inflation shows several distinctive features which may be observationally probed in coming years . in this paper , we focus on inflationary models with two fields , one of which includes a significant nm coupling .
interestingly , a large ng may arise differently from the case without nm coupling . in doing so
, we have re - derived analytic expressions for power spectrum , spectral index , tensor - to - scalar ratio , and non - linearity parameter for minimal ( @xmath141 , @xmath62 ) and also for nm ( @xmath141 , @xmath142 ) cases .
the results confirm the previous works on minimal case but we also found some corrections .
for the non - minimally coupled case , we explicitly derive the expressions using the @xmath143n - formalism when the inflaton potential is product - separable , @xmath144 as well as other cases . the analytic results are summarized in appendices and the parameter regime where a large ng arises with nm coupling term is clarified in sec .
ii and iii . for concreteness , a specific two - field inflationary model with @xmath96 is studied in detail in sec .
vi . within the parameter space
consistent with the observational data , it turned out that a large ng is hardly obtained . on the other hand , as soon as we introduce a nm coupling term , @xmath101 , to this model , we could read out significantly large non - linearity parameter with the desirable values of spectral index and tensor - to - scalar ratio . in near future
we expect that more accurate observations will be reported @xcite and a large ng may be actually observed . in that case , as we learned in this paper , we should take the inflationary models with nm coupling term more seriously and we even may hope to discriminate among various inflationary scenarios .
in this appendix , we estimate the primordial perturbations by using the @xmath143n - formalism @xcite . we provide the results for the minimally coupled case ( @xmath61 ) and for the non - minimally coupled case ( @xmath145 ) . in the case of minimally coupled case with product - separable potential ,
the analytic forms of the power spectrum , spectral index , tensor - to - scalar ratio , and the non - linearity parameter are known respectively @xcite @xmath146,\end{aligned}\ ] ] where @xmath147 in this subsection , we provide detail calculation for @xmath143n - formalism for the case of product potential , @xmath148 .
we recapitulate in this subsection the content derived in @xcite with some corrections of errors and typos .
the number of e - foldings is given by @xmath149 using the constant of motion along the trajectory @xcite , @xmath150 we could read out first derivatives of @xmath151 as follows : @xmath152 once we find the first derivatives of @xmath151 , it is straightforward to obtain second derivatives of @xmath151 , @xmath153 where @xmath154,\nonumber \\ \eta^{ss}&\equiv&\frac{1}{\epsilon}\left(\eta^{\phi}\epsilon^{\chi}+\eta^{\chi}\epsilon^{\phi}\right).\end{aligned}\ ] ] now that we have first and second derivatives of the number of e - foldings , we are able to obtain power spectrum , spectral index , tensor - to - scalar ratio , and the non - linearity parameter .
the power spectrum of curvature perturbation is given by @xmath155 where @xmath156 and @xmath157 is defined in such a way that the multi - field action in the einstein frame is written as of the form of @xmath158 with @xmath159 s as inflaton fields .
substituting eq . into eq .
, one obtains @xmath160 where @xmath161.\end{aligned}\ ] ] the spectral index is defined by @xmath162 noticing that @xmath163 and inserting eq .
, we could read out @xmath164 where @xmath165 .
it is straightforward to get @xmath166 \nonumber .\end{aligned}\ ] ] the definition of the tensor - to - scalar ratio is @xmath167 where @xmath168 is the power spectrum of gravitational waves .
some computation gives @xmath169 from the @xmath143n - formalism , the scale - independent part of the non - linearity parameter in the slow - roll approximation is given by @xmath170 since the scale - dependent part of the non - linearity parameter is too small to be detectable @xcite , we just take into account the scale - independent part only . using eq . and eq . , one could obtain @xmath171.\end{aligned}\ ] ] if we let @xmath61 , then , , , and reduce to the results of minimally coupled case in eq . . in terms of the dimensionless angle @xmath76 ,
the results of non - minimally coupled case in eqs .
, , , and are those in eqs . .
note that the results are slightly different from @xcite .
in this appendix , we provide some useful expressions for the number of e - foldings in various slow - roll potentials with two inflaton fields , which are extensively used in the text .
let us assume general potential of two scalar fields @xmath172 , without nm coupling , and compute the number of e - foldings , @xmath151 , defined by @xmath173 since we are interested in the slow - roll inflation , the equations of motion are approximated by the slow - roll equations , @xmath174 from the second equation , we express the integrand of eq . in terms of that of a scalar field @xmath7 ,
thus , the number of e - foldings is given by @xmath175 some separable cases are of our interest , @xmath176 , @xmath177 , @xmath178^p$ ] , and @xmath179^p$ ] with @xmath180 because these cases make evaluation of the number of e - foldings tractable and span wide range of two fields scalar potential for slow - roll inflation .
* case 1 : @xmath181 + for the sum - separable potential case , we read a relation @xmath182 from the slow - roll equations , @xmath183 . inserting this relation into eq .
with @xmath176 , we obtain @xmath184 * case 2 : @xmath185 + for a product - separable potential , the slow - roll equations give a relation @xmath186 . inserting this relation with @xmath177 into eq . , we obtain @xmath187 * case 3 : @xmath188^{p}$ ] + for a more general case with @xmath180 , we still have the relation @xmath182 as in the case with @xmath189 . using this relation in eq . , we find @xmath190 * case 4 : @xmath191^{p}$ ] + since the potential is product - separable , @xmath192 , we have the same relation as in case 2 , @xmath186 .
therefore , we have @xmath193 we turn on nm coupling with one of the two fields , @xmath194 , @xmath195 . then , in eq .
, the third slow - roll equation of @xmath100 is modified , @xmath196 which allows us to use the previously obtained formula of the number of e - foldings in eq . .
* case 1 : @xmath181 + for this sum - separable potential case , the slow - roll equation in eq .
provides @xmath197 .
thus , @xmath198 * case 2 : @xmath185 + for the product - separable potential , we have @xmath199 and thus @xmath200 * case 3 : @xmath188^{p}$ ] + for the general cases with @xmath180 , we have a relation @xmath201 and thus we have @xmath202 * case 4 : @xmath191^{p}$ ] + similar to the fourth case in the previous subsection ( @xmath62 ) , we have the relation @xmath199 and thus the number of e - foldings is @xmath203
this work is supported by the project of global ph.d .
fellowship which the national research foundation of korea conducts from 2011 with grant number 2011 - 0008792 ( j. k. ) and by basic science research program through the national research foundation of korea funded by the ministry of education , science and technology with grant numbers 2011 - 0011660 ( y. k. ) , and also 2011 - 0010294 and 2011 - 0029758 ( s. c. p. ) .
a. h. guth , `` the inflationary universe : a possible solution to the horizon and flatness problems , '' phys .
d * 23 * , 347 ( 1981 ) ; a. d. linde , `` a new inflationary universe scenario : a possible solution of the horizon , flatness , homogeneity , isotropy and primordial monopole problems , '' phys .
b * 108 * , 389 ( 1982 ) ; a. albrecht and p. j. steinhardt , `` cosmology for grand unified theories with radiatively induced symmetry breaking , '' phys .
* 48 * , 1220 ( 1982 ) .
e. komatsu _ et al .
_ [ wmap collaboration ] , `` seven - year wilkinson microwave anisotropy probe ( wmap ) observations : cosmological interpretation , '' astrophys .
j. suppl .
* 192 * , 18 ( 2011 ) [ arxiv:1001.4538 [ astro-ph.co ] ] . c. l. bennett , d. larson , j. l. weiland , n. jarosik , g. hinshaw , n. odegard , k. m. smith and r. s. hill _
et al . _ ,
`` nine - year wilkinson microwave anisotropy probe ( wmap ) observations : final maps and results , '' arxiv:1212.5225 [ astro-ph.co ] ; g. hinshaw , d. larson , e. komatsu , d. n. spergel , c. l. bennett , j. dunkley , m. r. nolta and m. halpern _ et al .
_ , `` nine - year wilkinson microwave anisotropy probe ( wmap ) observations : cosmological parameter results , '' arxiv:1212.5226 [ astro-ph.co ] .
v. acquaviva , n. bartolo , s. matarrese and a. riotto , `` second order cosmological perturbations from inflation , '' nucl .
b * 667 * , 119 ( 2003 ) [ astro - ph/0209156 ] ; j. m. maldacena , `` non - gaussian features of primordial fluctuations in single field inflationary models , '' jhep * 0305 * , 013 ( 2003 ) [ astro - ph/0210603 ] ; t. tanaka , t. suyama and s. yokoyama , `` use of delta n formalism : difficulty of generating large local - type non - gaussianity during inflation , '' prog .
* 190 * , 135 ( 2011 ) .
f. l. bezrukov and m. shaposhnikov , `` the standard model higgs boson as the inflaton , '' phys .
b * 659 * , 703 ( 2008 ) [ arxiv:0710.3755 [ hep - th ] ] .
s. c. park and s. yamaguchi , `` inflation by non - minimal coupling , '' jcap * 0808 * , 009 ( 2008 ) [ arxiv:0801.1722 [ hep - ph ] ] .
y. kim and s. c. park , `` hyperbolic inflation , '' phys .
d * 83 * , 066009 ( 2011 ) [ arxiv:1010.6021 [ hep - ph ] ] . e. komatsu and t. futamase , `` constraints on the chaotic inflationary scenario with a nonminimally coupled inflaton field from the cosmic microwave background radiation anisotropy , '' phys .
d * 58 * , 023004 ( 1998 ) [ astro - ph/9711340 ] .
e. komatsu and t. futamase , `` complete constraints on a nonminimally coupled chaotic inflationary scenario from the cosmic microwave background , '' phys .
d * 59 * , 064029 ( 1999 ) [ astro - ph/9901127 ] .
f. vernizzi and d. wands , `` non - gaussianities in two - field inflation , '' jcap * 0605 * , 019 ( 2006 ) [ astro - ph/0603799 ] ; k. -y .
choi , l. m. h. hall and c. van de bruck , `` spectral running and non - gaussianity from slow - roll inflation in generalised two - field models , '' jcap * 0702 * , 029 ( 2007 ) [ astro - ph/0701247 ] .
m. sasaki and e. d. stewart , `` a general analytic formula for the spectral index of the density perturbations produced during inflation , '' prog .
phys . * 95 * , 71 ( 1996 ) [ astro - ph/9507001 ] .
c. t. byrnes , k. -y .
choi and l. m. h. hall , `` conditions for large non - gaussianity in two - field slow - roll inflation , '' jcap * 0810 * , 008 ( 2008 ) [ arxiv:0807.1101 [ astro - ph ] ] . j. garcia - bellido and d. wands , `` metric perturbations in two field inflation , '' phys . rev .
d * 53 * , 5437 ( 1996 ) [ astro - ph/9511029 ] .
j. -o .
gong and h. m. lee , `` large non - gaussianity in non - minimal inflation , '' jcap * 1111 * , 040 ( 2011 ) [ arxiv:1105.0073 [ hep - ph ] ] . | we consider a model of inflation with non - minimal coupling with two inflaton fields .
it is noticed that a large class of inflaton potentials in product form as well as additively separable form can generate large non - gaussianities especially when non - minimal coupling term is introduced . a potentially realistic hybrid inflationary model with a potential , @xmath0 , and a non - minimal coupling term , @xmath1 ,
is analyzed in detail .
the model explicitly demonstrate the appearance of a large non - gaussianity . |
this work was supported by cnpq and fapesp .
navarra , m. nielsen , m.e .
bracco , m. chiapparini and c.l .
schat , phys . lett . *
b489 * , 319 ( 2000 ) .
navarra , m. nielsen , m.e .
bracco , phys . rev . *
d65 * , 037502 ( 2002 ) .
matheus , f.s .
navarra , m. nielsen and r. rodrigues da silva , phys . lett . *
b541 * , 265 ( 2002 ) . | we employ qcd sum rules to calculate the @xmath0 form factors and coupling constant by studying the three - point @xmath1 correlation function .
we find that the momentum dependence of the form factor depends on the off - shell meson .
we get a value for the coupling which is in agreement with estimates based on constituent quark model .
hadrons are composites of the underlying quarks whose effective fields describe point - like physics only when all the interacting particles are on mass - shell . when at least one of the particles in a vertex is off - shell , the finite size effects of the hadrons become important .
therefore , the knowledge of the form factors in hadronic vertices is of crucial importance to estimate any hadronic amplitude using hadronic degrees of freedom .
this work is devoted to the study of the @xmath1 form factor , which is important , for instance , in the evaluation of the dissociation cross section of @xmath2 by pions and @xmath3 mesons using effective lagrangians @xcite . since a decrease of @xmath2 production in heavy ions collisions might signal the formation of a quark - gluon plasma ( qgp ) @xcite , a precise evaluation of the background , i.e. , conventional @xmath2 absorption by co - moving pions and @xmath3 mesons , is of fundamental importance .
the @xmath1 coupling has been studied by some authors using different approaches : vector meson dominance model plus relativistic potential model @xcite and constituent quark meson model @xcite .
unfortunately , the numerical results from these calculations may differ by almost a factor two .
the relevance of this difference can not be underestimated since the cross section is proportional to the square of the coupling constants . in ref . @xcite it was shown that the use of different coupling constants and form factors can lead to cross sections that differ by more than one order of magnitude , and that can even have a different behavior as a function of @xmath4 . in previous works
we have used the qcd sum rules ( qcdsr ) to study the @xmath5 @xcite , @xmath6 @xcite and @xmath7 @xcite form factors , considering two different mesons off mass - shell . in these
works the qcdsr results for the form factors were parametrized by analytical forms such that the respective extrapolations at the off - shell meson poles provided consistent values for the corresponding coupling constant . in this work
we use the qcdsr approach to evaluate the @xmath1 form factors and use the same procedure described above to estimate the @xmath1 coupling constant . the three - point function associated with a @xmath8 vertex ( see fig . 1 ) , where @xmath9 and @xmath10 are the incoming and outgoing external mesons respectively and @xmath11 is the off - shell meson , is given by _
( p,)&=&d^4x d^4y e^ip^.x e^-i(-p).y + & & 0|t\{j_3(x ) j^_2(y)j^_1(0)}|0 , [ cor ] where @xmath12 is the interpolating field for @xmath13 . for @xmath2 , @xmath14 and @xmath15 mesons
the interpolating fields are respectively @xmath16 , @xmath17 and @xmath18 with @xmath19 and @xmath20 being a light quark and the charm quark fields .
the phenomenological side of the vertex function , @xmath21 , is obtained by the consideration of @xmath9 and @xmath10 state contribution to the matrix element in eq .
( [ cor ] ) : _ ^(phen)(p,)=1p^2-m_1 ^ 21 ^ 2-m_3 ^ 2 0|j_3|h_3 ( ) + h_3()|j^_2|h_1(p)h_1(p)| j^_1|0 + , [ phe1 ] where h. r. means higher resonances .
the matrix element of the current @xmath22 defines the vertex function @xmath23 : h_3()|j^_2|h_1(p)=h_2(q)|j^_2|0 , [ for ] where @xmath24 .
calling @xmath25 and @xmath26 the four momentum of @xmath27 and @xmath15 respectively one has v^(p_1,p_2,p_3)=g_dd*(q^2 ) ^_^(p_1 ) _ ^(p_2)p_3p_2 .
[ ver ] the vacuum to meson transition amplitudes appearing in eqs .
( [ phe1 ] ) and ( [ for ] ) are given in terms of the corresponding meson decay constants @xmath28 by 0|j^(d)|d = m_d^2f_dm_c , [ fd ] and v(p,)|j^_|0=m_vf_v^ * _ , [ fv ] for the vector meson @xmath29 or @xmath30 .
therefore , using eqs .
( [ for ] ) , ( [ ver ] ) , ( [ fd ] ) and ( [ fv ] ) in eq .
( [ phe1 ] ) we get _
^(phen)(p,)&=&cg_dd*(q^2 ) _ p^^(q^2-m_2
^ 2)(p^2-m_1 ^ 2)(^2-m_3 ^ 2 ) + & + & , [ phen ] where c = m_d^2m_d^*m_f_df_d^*f_m_c .
[ chh ] the contribution of higher resonances and continuum in eq .
( [ phen ] ) will be taken into account as usual in the standard form of ref .
@xcite .
the qcd side , or theoretical side , of the vertex function is evaluated by performing wilson s operator product expansion ( ope ) of the operator in eq .
( [ cor ] ) .
writing @xmath31 in terms of the invariant amplitude : _
( p,)=(p^2,^2,q^2)_p^^ , we can write a double dispersion relation for @xmath32 , over the virtualities @xmath33 and @xmath34 holding @xmath35 fixed : ( p^2,^2,q^2)=-14 ^ 2ds du ( s , u , q^2)(s - p^2)(u-^2 ) , [ dis ] where @xmath36 equals the double discontinuity of the amplitude @xmath37 on the cuts @xmath38 , @xmath39 , with @xmath40 in the case of off - shell @xmath14 or @xmath15 and @xmath41 in the case of off - shell @xmath2 . we consider diagrams up to dimension three which include the perturbative diagram and the quark condensate . to improve the matching between the two sides of the sum rules ,
we perform a double borel transformation in both variables @xmath42 and @xmath43 we get one sum rule for each meson considered off - shell . calling @xmath44 the @xmath45 form factor for the off - shell meson @xmath46 , we get the following sum rules : c e^- e^- = _ 4m^2^s_0 _ u_min^u_0 dsdu ^(d)(u , s , t)e^-e^-(u_max - u ) , [ ffd ] c e^- e^- = _ 4m^2^s_0 _ u_min^u_0 dsdu ^(d^*)(u , s , t)e^-e^-(u_max - u ) , [ ffds ] and c e^- e^- = _ m^2^s_0 _ u_min^u_0 dsdu ^(j/)(u , s , t)e^-e^-(u_max - u ) , [ ffpsi ] with @xmath47 , ^(d)(u , s , t)=^(d^*)(u , s , t)= ( 1 + ) , @xmath48 , @xmath49 and u^max_min= , in the case of off - shell @xmath15 or @xmath14 . in the case of an off - shell @xmath2 we get : ^(j/)(u , s , t)= -4 ^ 2<|qq>(s - m_c^2)(u - m_c^2 ) , and u^max_min= . in the eqs .
( [ ffd ] ) , ( [ ffds ] ) and ( [ ffpsi ] ) we have transferred to the qcd side the higher resonances contributions through the introduction of the continuum thresholds @xmath50 and @xmath51 .
the parameter values used in all calculations are @xmath52 , @xmath53 , @xmath54 , @xmath55 , @xmath56 , @xmath57 , @xmath58 , @xmath59 .
the continuum thresholds for the sum rules are @xmath60 and @xmath61 with @xmath62 .
we first discuss the @xmath0 form factor with an off - shell @xmath15 meson . in fig .
2 we show the behavior of the form factor @xmath63 at @xmath64 and @xmath65 , as a function of the borel mass @xmath66 using @xmath67 .
we can see that the qcdsr results are rather stable in the interval @xmath68 . in fig .
3 we show @xmath69 as a function of @xmath66 and @xmath70 . from fig .
3 we see that the stability is still good even considering the two independent borel parameters .
the same kind of stability is obtained for other values of @xmath71 and for the other two form factors .
fixing @xmath72 and @xmath73 we show , in fig . 4 , the momentum dependence of the qcdsr results for the form factors @xmath74 , @xmath75 and @xmath76 through the circles , squares and triangles respectively . since the present approach can not be used at @xmath77 , to extract the @xmath78 coupling from the form factors we need to extrapolate the curve to @xmath79 : the mass of the off - shell meson .
in order to do this extrapolation we fit the qcdsr results with an analytical expression .
we tried to fit our results with a mono - pole form , since this is very often used for form factors , but the fit was only good for @xmath80 . for @xmath74 and @xmath75
we obtained good fits using a gaussian form .
we get : g^(j/)_dd^*(q^2)= , [ papsi ] g^(d^*)_dd^*(q^2)= 19.9 , [ pads ] g^(d)_dd^*(q^2)= 12.7 .
[ pad ] these fits are also shown in fig . 4 through the dotted , dashed and solid lines respectively .
from fig . 4 we see that all three form factors lead to compatible values for the coupling constant when the form factors are extrapolated to the off - shell meson mass ( shown as open circles in fig . 4 )
considering the uncertainties in the continuum threshold , and the difference in the values of the coupling constants extracted when the @xmath15 , @xmath14 or @xmath2 mesons are off - shell , our result for the @xmath0 coupling constant is : g_dd^*=(3.48 0.76)^-1 .
[ fi ] in table i we show the results obtained for the same coupling constant using different approaches.in refs .
@xcite and @xcite .
[ cols="^,^,^ " , ] values of the coupling constant @xmath81 in @xmath82 evaluated using different approaches .
while our result is compatible with the coupling obtained using constituent quark meson model @xcite , it is half of the value obtained with the vector meson dominance model plus relativistic potential model @xcite .
to summarize : we have used the method of qcd sum rules to compute form factors and coupling constant in the @xmath83 vertex .
our results for the coupling show once more that this method is robust , yielding numbers which are approximately the same regardless of which particle we choose to be off - shell and depending weakly on the choice of the continuum threshold . as for the form factors ,
we obtain a harder ( softer ) form factor when the off - shell particle is heavier ( lighter ) . |
for most purposes , the sun can be assumed in hydrostatic and thermal equilibrium , neglecting the effects of rotation and magnetic field .
however models built with these simplifying assumptions do not agree with the helioseismic data , in particular with those obtained by the satellite soho , and it appears that macroscopic mixing processes must be taken into account not only in the convection zone , but also in the radiative interior .
turbulent mixing may be introduced in the models by adding a turbulent term @xmath2 in the diffusive part of the equation for the time evolution of the chemical abundance @xmath3 .
the equation becomes : @xmath4 where the velocity v@xmath5 of species i with respect to the center of mass is : @xmath6 the velocity @xmath7 is the sum of one term which depends on the concentration gradient @xmath8 , and one which does not , @xmath9 ( proffit and michaud 1991 ) . in this section
, we address the mixing which may occur in the shear layer connecting the differential rotation in the convection zone with the solid rotation of the radiative interior .
spiegel and zahn ( 1992 ) have given a physical interpretation of this tachocline by invoking anisotropic turbulence , with much stronger viscous transport in the horizontal than in the vertical direction .
such turbulence reduces the differential rotation and therefore inhibits the spread of the layer deep inside the radiative zone .
the conservation of mass , momentum and entropy is given by : @xmath10 @xmath11 @xmath12 where @xmath13 is the density , @xmath14 the pressure , @xmath15 the temperature , @xmath16 is the local velocity in the rotating frame , @xmath17 the differential rotation , @xmath18 the specific entropy , @xmath19 the viscous stress tensor , @xmath20 gravitational potential . + in order to solve the system of equations , some simplifying assumptions are made : * the flow field is axisymmetric @xmath21 * anelastic approximation : @xmath22 is negligible * advection is small compared to coriolis acceleration * the tachocline is thin compared to the pressure scale height * viscous forces are small compared to coriolis force . after separating each variable into its mean value on the sphere plus a perturbation , as @xmath23 ,
the linearized form of the set of equations , in the stationnary state ( @xmath24 ) is ( spiegel and zahn 1992 ) : @xmath25 \nonumber \\ \mbox { diffusion and advection of angular momentum } \nonumber \\
\frac{n^2}{g}\frac{t}{\rho r^2}\frac{\p \psi}{\p x}=\frac{1}{\rho c_p r^2}\left[\chi r^2 \frac{\p \hat{t}}{\p r}\right ] \nonumber \\ \mbox { diffusion and advection of heat } \nonumber\end{aligned}\ ] ] in spherical coordinates ( @xmath26 , @xmath27 ) , where @xmath28 is the horizontal turbulent viscosity ( @xmath29 ) and defining a stream function @xmath30 for the meridional flows by : @xmath31 we now project on horizontal eigenfunctions , the variables of the set of equations describing the tachocline : @xmath32 we find that the dominant term for the even eigenfunction is @xmath33 .
the final step is to connect the meridional velocity to the differential rotation . approximating the stream function by @xmath34 and using the conservation of angular momentum ,
we get : @xmath35 with @xmath36 . by introducing the variables @xmath37 and @xmath38 the tachocline thickness , where @xmath39 is the radius , n the brunt - visl frequency and @xmath40 the radiative diffusivity at the base of the convective zone , we thus obtain this radial dependancy for @xmath41 : @xmath42 the anisotropic diffusion invoked to stop the spread of the tachocline will also interfer with the advective transport of chemicals .
chaboyer and zahn ( 1992 ) have shown that the result is a diffusive transport in the vertical direction , with an effective diffusivity given by : @xmath43 where @xmath44 are the coefficients of the expansion of the vertical component of the velocity @xmath45 in legendre polynomials .
the eigenfunctions @xmath46 above may be projected on these legendre polynomials , and to a good approximation one has @xmath47 replacing @xmath41 by ( [ u4 ] ) , one reaches the following expression for the vertical diffusivity @xmath48 we shall treat the tachocline thickness @xmath49 as an adjustable parameter , roughly bounded between 0.03 and 0.08 @xmath50 ( corbard et al . 1997 ) . with the new latitudinal dependence of @xmath51 ( thompson et al .
1996 ) , @xmath52 nhz , the coefficient @xmath53 and @xmath54 . starting from the reference model of brun , turck - chize and morel ( 1998 ) built with the stellar evolution code cesam ( morel 1997 ) , which include the recent opal opacities and eos , alderberger et al .
( 1998 ) nuclear reaction rates and microscopic diffusion ( see also turck - chize et al .
these proceeding ) , we introduce the coefficient @xmath2 in our diffusion equation with different values of @xmath49 and @xmath55 ( fig .
the brunt - visl frequency @xmath55 is taken as constant and @xmath49 is chosen such as to yield a tachocline of @xmath56 , as determined by ( corbard et al .
1997 ) . in table 1 and figures
2,3,4,5 we present models including the coefficients shown in fig .
1 . for one coefficient ( @xmath57 )
we also test the influence on our results of the calibration of @xmath58 .
p3cm*6c parameters & reference & @xmath57 & @xmath57 & @xmath59 & @xmath60 + opacities & opal / a & opal / a & opal / a & opal / a & opal / a + diffusion & yes & yes & yes & yes & yes + age ( gyr ) & 4.6 & 4.6 & 4.6 & 4.6 & 4.6 + @xmath49 ( @xmath61 ) & - & 0.05 & 0.05 & 0.08 & 0.05 + @xmath55 ( @xmath62hz ) & - & 100 & 100 & 100 & 10 + ( z / x)@xmath63 & fixed & fixed & free & fixed & fixed + + @xmath64 & 1.768 & 1.754 & 1.763 & 1.748 & 1.759 + y@xmath65 & 0.2721 & 0.270 & 0.2721 & 0.2693 & 0.2692 + z@xmath65 & 1.96 @xmath66 & 1.913 @xmath66 & 1.96 @xmath66 & 1.904 @xmath66 & 1.898 @xmath66 + ( z / x)@xmath65 & 0.0277 & 0.0269 & 0.0277 & 0.0267 & 0.0267 + + y@xmath63 & 0.2425 & 0.2454 & 0.2475 & 0.2458 & 0.247 + z@xmath63 & 1.803 @xmath66 & 1.796 @xmath66 & 1.841 @xmath66 & 1.79 @xmath66 & 1.795 @xmath66 + ( z / x)@xmath63 & 0.0244 & 0.0244 & 0.0251 & 0.0244 & 0.0244 + @xmath67 & 2 & 2 & 2 & 2 & 2 + + r@xmath68/r@xmath69 & 0.713 & 0.7145 & 0.7134 & 0.7145 & 0.715 + t@xmath70
( k ) & 2.192 & 2.181 & 2.196 & 2.181 & 2.178 + + y@xmath71 & 0.642 & 0.639 & 0.642 & 0.638 & 0.641 + z@xmath71 & 2.095 @xmath66 & 2.046 @xmath66 & 2.096 @xmath66 & 2.035 @xmath66 & 2.03 @xmath66 + t@xmath72 ( k ) & 15.71 & 15.68 & 15.71 & 15.66 & 15.67 + @xmath73 ( g/@xmath74 ) & 153.5 & 153.3 & 153.5 & 153.0 & 153.1 + ( 8,5 ) our standard model has a surface abundance of helium of 0.2425 in mass ( see table 1 ) , which is a bit too low if we compare with the basu and antia ( 1995 ) value for the opal equation of state ( rogers , swenson and iglesias 1996 ) , @xmath75 .
when allowing for turbulent diffusion in the tachocline , the settling of helium is reduced by 16% for @xmath57 up to 25% with @xmath60 , leading to surface abundances @xmath76 or @xmath77 which are in better agreement with the helioseismic value .
note the smooth composition profile in fig .
3 , with its extended plateau below the convection zone . ( 8,5 ) ( 8,5 ) we remark that the present z / x value for a mixed model plays an important role in this study . if we do not calibrate the z / x value of the mixed model , starting with the heavy element composition @xmath78 of the reference model , we obtain different surface abundances ( see for example column 2 and 3 of table 1 for the coeficient @xmath57 ) . also , when comparing the two sound speed profiles ( figs . 4 and 5 ) , we see that in both cases the bump around 0.7 @xmath50 is pratically erased by the introduction of the tachocline mixing , but the calibrated model is modified along the whole structure and not only close to the bump .
this could be understood by the fact that the tachocline mixing inhibits the elements settling and then less helium reaches the center , modifying the central conditions .
the remaining part of the discrepancy between the nuclear core and the convection zone may come from uncertainties on opacities coefficients or knwoledge of the element abundances ( see brun , turck - chize and morel 1998 ) or from some other mixing .
the low value of the tachocline thickness ( @xmath79 ) inferred by elliot et al .
( these proceedings ) could be due to the turbulent coefficient used and also to the fact that they did not calibrate their model in z / x .
( 8,4.8 ) ( 8,4.8 )
we also introduce a turbulent pressure in the total pressure ( i.e. @xmath80 , where @xmath81 is the sum of the gas+radiation+coulombian pressure ) .
the turbulent pressure is neglected in most solar models , in spite of the fact that it represents a substantial fraction of the total pressure in the uppermost part of the convection zone , which has an important weight in the determination of the frequency of the acoustic modes . by definition , the turbulent pressure is given by the horizontal average @xmath82 , @xmath83 being the vertical turbulent velocity , but here we link it to the convective velocity @xmath84 drawn from the mixing lenth treatment ( mlt ) : @xmath85 , where @xmath86 takes into account the geometry of the turbulent motions .
we follow cox and giuli ( 1968 ) who introduce a modified adiabatic gradient .
@xmath87 where @xmath88 is the adiabatic temperature gradient .
doing so we obtain for the convective efficiency @xmath89 and for the velocity @xmath90 a corrected expression : @xmath91 where , @xmath92 is the temperature gradient of the medium , @xmath93 is the temperature gradient of the moving element , @xmath94 is the mixing length , @xmath95 the pressure scale height , g the gravitational acceleration and @xmath96 . after some algebraic manipulations ,
we deduce a new cubic equation for @xmath89 : @xmath97=0\ ] ] where @xmath98 is the gradient if all the energy is carried by radiation , @xmath99 , @xmath100 , @xmath101 the specific heat at constant pressure , a the radiation constant , c the light speed and @xmath102 the opacity coefficient .
the real gradient @xmath92 is thus obtained by : @xmath103 we test different values for @xmath86 from 0.5 to 1.5 ( see fig .
7 ) , and find that the best agreement with direct numerical simulations ( see rosenthal et al .
( these proceedings ) ) are obtained for @xmath104 ( see fig .
6 ) . for the other values ,
the turbulent contribution to the total pressure is either too low ( @xmath105 ) or too high ( @xmath106 ) .
this study confirms the important role played by the turbulent pressure in the upper layers of the solar convection zone ( see baturin and mironova and bhmer and rdiger these proceedings ) .
its introduction in the model removes a large part of the discrepancy between observed and predicted acoustic frequencies , as shown in fig .
again the best fit is obtained with @xmath104 . but this improvement can not hide the fundamental shortcomings of the mixing length treatment , in particular that of being a local theory which does not allow for convective penetration .
serious progress will only be achieved by realistic hydrodynamical modelization , as illustrated by the contributions of rosenthal et al .
( these proceedings ) .
adelberger , e. et al .
1998 , to appear in rev .
alexander , d. r. and ferguson , j. w. 1994 , , 437 , 849 .
basu , s. and antia , h. m. 1995 , mnras , 276 , 1402 .
baturin , v. a. and mironova , i. v. 1998 , `` turbulent convection and p - mode oscillations '' , these proceedings .
bhmer , s. and rdiger , g. 1998 , `` influence of turbulence on the solar p - mode frequencies '' , these proceedings .
brun , a. s. , turck - chize , s. and morel , p. 1998
, astro - ph 9806272 , to appear in 506/2 .
chaboyer , b. and zahn , j. p. 1992
, , 253 , 173 .
corbard , t. , berthomieu , g. , provost , j. and morel , p. 1997
, , 330 , 1149 .
cox , j. p. and giuli , r. t. 1968 .
_ principles of stellar structure _ , gordon&breach , new york .
elliot , j. r. , gough , d. o. and sekii , t. 1998 , `` helioseismic determination of the solar tachocline thickness '' , these proceedings .
iglesias , c. and rogers , f. j. 1996 , , 464 , 943 .
morel , p. 1997 , a&a sup .
, 124 , 597 .
proffit , c. r. and michaud , g. 1991 , , 380 , 238 .
rogers , f. j. , swenson , j. and iglesias , c. 1996 , , 456 , 902 .
rosenthal , c. s. , christensen - dalsgaard , j. , kosovichev , a. g. , nordlund , . ,
reiter , j. , rhodes jr , e. j. , schou , j. , stein , r. f. and trampedach , r. 1998 , `` tests of convective frequency effects with soi / mdi high degree data '' , these proceedings .
spiegel , e. a. and zahn , j. p. 1992
, , 265 , 106 .
thompson , m. j. , toomre , j. and the gong dynamics inversion team 1996 , sci , 272 , 1300 .
turck - chize , s. , brun , a. s. , chize , j. p. and garci , r. 1998 , `` predictions of the solar neutrino fluxes and the solar gravity mode frequencies from the solar sound speed profile '' , these proceedings . | with the recent results of helioseismology aboard soho , solar models are more and more constrained ( brun , turck - chize and morel 1998 ) .
new physical processes , mainly connected to macroscopic motions , must be introduced to understand these new observations . in this poster , we present solar models including a turbulent pressure in the outer layers and mixing due to the tachocline ( spiegel and zahn 1992 ) . +
our results lead us to conclude that : + - turbulent pressure improves the absolute value of the acoustic mode frequencies ( @xmath0 at 4 mhz ) + - mixing in a tachocline thickness of @xmath1 ( corbard et al .
1997 ) looks promising .
key words : sun : interior ; sun : turbulent processes . |
in proton beam therapy treatment planning , analytical descriptions of the treatment beams are commonly used to determine the dose deposited in clinical patient models .
although monte carlo based methods have become faster during the last few years , there still is a distinct advantage of using more efficient , analytical models when having to perform multiple calculations such as in the process of 4d robust optimization and adaptive therapy .
this advantage has greater significance in the case of pencil beam based proton therapy , where multiple small beams need to be tracked and calculated . despite all these advantages , analytical algorithms
have been shown to be less reliable in clinically more complex treatments like lung and breast treatments@xcite , which prompted this effort to provide a more accurate description of the dose deposition by a scanned proton beam .
in addition , a more analytical description provides greater insight in the macroscopic process of how a pencil beam behaves physically in a medium as issues such as energy and medium vary . a pencil beam entering
a medium will generate secondary particles such as scattered neutrons , generated photons , @xmath4rays and large angle scattered protons that produce a _ nuclear halo _ of dose around the central beam axis .
although , in the region away from the central axis , the contribution from a single pencil beam is small , a complex treatment plan is made of many pencil beams and the summation of the lateral contributions could be significant .
the lateral extent of a beam at different depths is illustrated in figure [ depth_change ] .
monte carlo based calculation showing changes in lateral dose deposition for a pencil beam of nominal energy of 230 mev .
a logarithmic scale is used to better illustrate the difference in contributions from the nuclear halo and the primary particles . ] the description and treatment of the nuclear halo has been the focus of research by a number of groups who have proposed various methodologies describing the effects in an analytical way .
gottschalk et al .
provided an in - depth analysis of all the physical processes contributing to the nuclear halo , subdividing a pencil beam into a combination of four distinct regions ; core , halo , aura and ( possibly ) spray ; in the pencil beam@xcite .
this approach requires up to 25 different physical parameters to characterise the beam . in most implementations ,
the contribution from the nuclear halo is solved by adding different distributions to a central gaussian distribution describing the core of the pencil beam . the first proposed solution for the nuclear halo by pedroni et al .
added another , broader gaussian to the core ; a methodology that is implemented in the varian eclipse ( varian medical systems , palo alto , ca ) calculation algorithm@xcite .
the simplicity of this calculation also allowed for faster gpu implementation@xcite . in further refinements of this approach , other groups attempted combinations of gaussian , lorentz ( also known as cauchy ) and lvy distributions@xcite , increasing the complexity of the fitting procedure and necessitating look - up tables for the various parameters . a key insight that enables our novel approach is that each of these methods combine two or more stable functions in their analytical representation . a further clinically used algorithm for pencil beam calculation
is used in raystationtps ( raysearch laboratories , stockholm , sweden ) . in this system
, each spot is modelled as a superposition of 19 gaussian distributions ( 19 sub - spots : 1 at the center , and 6 and 12 positioned at two concentric circles around the center)@xcite .
in this paper we review the concept of stable distributions and show that they can be used to represent the evolution of a proton pencil beam in a medium .
we demonstrate that this approach provides a more accurate description of the pencil beam and is more efficient than the use of normal distributions , or a sum thereof .
furthermore , we show that this parametrization allows interpolation of non measured energies from measured ( or calculated using monte carlo ) depth profiles .
finally , we implement this algorithm in an open source treatment planning toolkit , matrad@xcite .
stable distributions are a class of distributions which generalize a property of the normal distribution .
namely , they extend the central limit theorem which says that if the number of samples drawn from random variables , _ with or without _ finite variance , tends to infinity , then the measured distribution tends to a stable distribution
. if the variance is finite , the resultant distribution tends towards the normal distribution , a member of the class of stable distributions .
other than for specific cases , these distributions do not possess an analytical representation .
it is therefore necessary to describe them in terms of their characteristic function which always exists for a given stable distribution .
more generally , the characteristic function , @xmath5 , of a distribution is the fourier transform of the probability function , @xmath6 , of that distribution , e.g. : @xmath7 it can be shown that all stable distributions can be characterised as having the same characteristic function , @xmath5 , barring a change in the parameters @xmath8 .
@xmath9\ ] ] with @xmath10 except for @xmath11 , in which case @xmath12 .
the parameter @xmath13 $ ] determines the shape of the distribution , @xmath14 $ ] is a measure for symmetry , @xmath15 $ ] is a scale factor and @xmath4 a position , or the most probable value@xcite . for a symmetric , zero centred distribution the equation [ cf_long ] reduces to : @xmath16 in an appendix
we show from first principles that this equation represents all symmetric zero centered distributions that follow the central limit theorem
. as @xmath17 and @xmath18 can vary continuously there are an infinite number of stable distributions , most of which do not have an analytical representation in real space .
indeed , only for @xmath19 , @xmath20 , and @xmath21 ( @xmath22 ) is a closed form known .
these correspond to , respectively , the gauss , lorenz , and lvy distributions . using this generalization
it is possible to define a class of uni modal distributions whose properties can be exploited to describe physical random walk processes which combine different physical properties@xcite .
the provision center for proton therapy is currently operational using an iba cyclotron which provides proton beam scanning technique up to a maximum energy of 230 mev . for this work ,
the provision beams from 98 to 230 mev were accurately reproduced by the monte carlo code fluka by adapting the simulated beams to the commissioning experimental beam data@xcite . at each energy , the beam is defined at the surface of the phantom by a two - dimensional normal distribution characterized by position and standard deviation , @xmath23 . using fluka ,
the dose distribution in medium ( water ) is calculated in a 200x200x350 mm@xmath24 cube with 1 mm@xmath24 tally volumes . the calculated dose distribution in each 200x200 slice perpendicular to the beam axis is then considered to be a two dimensional dose distribution .
generation of secondary particles under the form of gamma s , neutrons and @xmath4rays was enabled .
the simplest approach of analytically predicting the behavior of the dose deposited by a proton beam is based on a gaussian parametrization with the width of the beam defined by a variable standard deviation , @xmath23@xcite .
this implies that at any given depth the lateral dose deposition can be described as : @xmath25 in commercially available algorithms it was seen that this approach did not predict the contributions of pencil beams further away from the beam axis in an adequate way .
indeed in the eclipsetreatment planning software , developed by varian ( varian medical systems , palo alto , ca ) , an additional gaussian term is used to describe this , as proposed by pedroni _
et al._@xcite .
raystation ( raysearch , stockholm sweden ) also employs such a strategy introducing multiple off axis gaussian contributions . in the remainder of the paper we will only concentrate on the two gaussian solution ,
which is parametrized as follows : @xmath26 with @xmath27 $ ] .
this parametrization implies that to fit the behaviour of a proton pencil beam at a given depth we need to determine three parameters : @xmath28 , @xmath29 , and @xmath30 .
because the majority of stable distributions do not possess an analytic form ; it is difficult to use the classical approach to fit the data .
indeed , the fitting of stable distributions is the subject of scientific research by itself .
we opted to use a maximum likelihood estimation based on pre computed spline approximations@xcite .
in essence it selects the distributions that match the pre computed ones the best .
once the parameters are determined , the characteristic function is calculated in complex space and using an inverse fourier transform the actual stable distribution was generated , a straightforward methodology also proposed by mittnik et al.@xcite . the resulting curve
could then be compared with the monte carlo simulation . to fit the normal distribution based algorithms , a classical methodology using a least square fit of the analytical function based on a levenberg
marquardt algorithm@xcite was used .
both the stable and gaussian fit were compared to the monte carlo simulation using pearson s @xmath31 measure .
@xmath32 the @xmath31 value for @xmath33 degrees of freedom then yields the probability that the fitted distribution is different from the simulated one , with @xmath34 being the number of parameters in the fit . for stable parametrization @xmath35 ( @xmath17 and @xmath18 ) , for a double gaussian @xmath36 ( @xmath28 , @xmath28 , and @xmath30 ) .
we denote the @xmath31 value for normal and stable distribution as respectively @xmath37 and @xmath38 . for all pencil beams with nominal energy ( @xmath39 )
the centrally located transversal distribution is extracted at all available depths ( @xmath1 ) .
subsequently , the normalised stable distribution parameters @xmath40 and @xmath41 are determined using the above - mentioned fitting procedure . in a first approximation
we consider the pencil beams to be circularly symmetric .
finally , the total integral dose at each depth @xmath42 is also calculated .
this procedure yields three parameters which vary as a function of depth and nominal beam energy allowing us to calculate the dose distribution at any depth in a homogeneous medium .
we propose a methodology to determine the beam characteristics of intermediate energies from two provided beam characterisations .
let @xmath43 , @xmath44 , and @xmath45 be the parameters fully describing a proton pencil beam with nominal energy @xmath46 .
then it is possible to calculate the parametrization of an intermediate energy @xmath47 by interpolation of the parametrization of energies @xmath46 and @xmath48 disregarding a scaling in the depth parameter which depends on the range of the given energy .
let @xmath49 , : @xmath50 where : @xmath51 and @xmath52 with @xmath53 being the range of a proton in the medium under consideration . using the methodology of intermediate morphing
, we determine the minimal amount of beams we need to fully characterize in order to obtain a full set of data across all nominal energies .
we choose a threshold of 1% error determining the width of the beam ( @xmath18 ) and 3% for the shape ( tailedness ) ( @xmath17 ) .
the total deposited energy needs to be correct to 1% dose and 1 mm position . to allow testing of our parametrised beam model with clinical patient plans
, we implemented the stable distribution dose calculation algorithm in an open source treatment planning system , matrad ( dkfz , heidelberg , de ) .
matrad is written in matlab ( mathworks , natick , ma ) and provides functionality for importing patient data , ray tracing , inverse planning and treatment plan visualisation .
the proton dose calculation component was extended to support a beam model described by a stable distribution , in addition to single and double gaussian models . to provide radial symmetry of the beam ,
a 2d normalisation is required when computing the lateral profile in a plane of distance @xmath1 into a medium .
if @xmath54 is the value of the stable distribution that describes the 1d beam profile at a distance @xmath55 from the central axis , the 2d beam profile is described by : @xmath56 where , @xmath57 is the distance from the pencil beam central axis and @xmath17 and @xmath18 are the parametrization at depth @xmath1 .
@xmath58 is the normalisation required such at the volume under the 2d distribution is unity and is calculated using the shell formula as : @xmath59 as there is no analytical representation in real space for stable distributions except when @xmath17 equals specific values , numerical computation of this integral is required to provide the normalisation . to increase efficiency ,
the integral @xmath58 is pre calculated for each combination of @xmath17 and @xmath18 in the discrete beam parametrization and is interpolated as required within the dose calculation engine .
figure [ v_variation ] shows that @xmath58 varies smoothly within the relevant range of parameter space .
the 2d normalisation parameter , @xmath58 varies smoothly over the relevant range of the parameter space defined by @xmath17 and @xmath18 and is therefore interpolated as required within the dose calculation engine . ]
the parameters required to fully characterise the knoxville beam at all energies were implemented in : namely , @xmath17 , @xmath18 and the integrated dose at a distance @xmath1 along the beam path .
this implementation allows complex treatment plans to be generated .
figure [ fit1]a ) shows the lateral dose deposition at a depth of 20 cm for a proton pencil beam with nominal energy of 226.08 mev with two parameterizations : the appropriate stable distribution and a double gaussian . from visual inspection of the graph ,
it is clear that the stable distribution provides a better fit to the beam profile at this energy and depth .
the fit of each parameterization is quantified by calculating the @xmath31-values , yielding @xmath60 for a single gaussian distribution ( not plotted ) , @xmath61 for the stable distribution , and @xmath62 for a double gaussian .
this corresponds to a probability of 0 , 1 , and 1 , for the lateral dose distributions to be represented by the respective fits .
figure [ fit1]b ) plots the @xmath38value at each calculated lateral point in the beam profile where a value of 0 is a perfect fit at that point .
this graph shows that each of the three distributions provides an adequate representation of the dose close to the centre of the beam . for the gaussian distribution ,
the @xmath31values for each point quickly deviates substantially , showing that this distribution does not represent the range behaviour adequately beyond @xmath6310 mm .
the double - gaussian fit provides a good estimation of the beam profile to a distance of @xmath6360 mm , however it becomes clear that there is a systematic underestimation of the dose faraway which increases the @xmath31value .
furthermore , this parameterization requires the most variables to describe the system .
the stable distribution provides the best fit of the profile of a proton pencil beam at this energy and depth .
the effectiveness of these systems increases the number of variables needed to describe the system and depends on the region of interest ( i.e. the size of the region taken into account to measure the tail contributions ) .
the behaviour of a proton pencil beam , as commissioned at the provision facility can be parametrised at a given depth and for a specific nominal energy using two parameters from the stable distribution fit : @xmath17 , describing the tail of the distribution and @xmath18 providing the width .
these parameters provide a normalised distribution .
a final parameter is the integral dose @xmath64 deposited at that depth ( fig .
[ parametrization]c ) ) .
the @xmath17 parameter reflects the increased contribution of interactions with longer range , most likely from scattered protons .
the contribution diminishes due to two factors : 1 ) the decrease of generated secondary protons due to the lower energy of the primary protons , and 2 ) the decrease in energy of the secondary protons .
figure [ interpol ] shows the methodology interpolating the data from two energies to generate data for a third beamlet . in the remaining figures we calculate the maximal error of the parametric representation . due to the non linearity of the parameters behavior as a function of energy
we expect that linear interpolation is useful only in a limited energy range .
indeed , figures [ interpol]b ) and c ) , show that the parameter @xmath18 is most sensitive and increases deviates to more than 1% if the interpolated energies are more than 20mev apart in nominal energy .
the @xmath17parameter is not very sensitive to interpolating distance as the curve is relatively noisy .
top figure : predicting the next @xmath17graph using two outer source curves . left bottom : variation of @xmath18 is the limiting parameter , 1% error equates and error of 0.01 mm in estimation of the width .
right bottom : the @xmath17 parameter does not change greatly as it is dominated by the curve noise.,title="fig:",scaledwidth=45.0% ] + top figure : predicting the next @xmath17graph using two outer source curves .
left bottom : variation of @xmath18 is the limiting parameter , 1% error equates and error of 0.01 mm in estimation of the width .
right bottom : the @xmath17 parameter does not change greatly as it is dominated by the curve noise.,title="fig:",scaledwidth=45.0% ] top figure : predicting the next @xmath17graph using two outer source curves . left bottom : variation of @xmath18 is the limiting parameter , 1% error equates and error of 0.01 mm in estimation of the width .
right bottom : the @xmath17 parameter does not change greatly as it is dominated by the curve noise.,title="fig:",scaledwidth=45.0% ] the alpha - stable parametrization has been successfully incorporated into the matrad open source treatment planning system
. calculation of alpha - stable distribution is performed using either a fast , parallel c / c++ library _ _
libstable__@xcite , if available , or a native matlab implementation in other situations@xcite . using the c
/ c++ library , calculation of a complete pencil beam on a 200x200x200 3 mm@xmath24 cube takes 20 - 45 seconds , depending on beam energy , on a intel xeon e5 - 2670 based workstation with ten 2.5 ghz cores .
the dose distribution from a single proton beam spot of 120 mev was calculated onto a homogeneous water phantom in matrad , raystation and fluka .
the depth dose curve and beam profile across the bragg peak are shown in figure [ raystation_comparison ] .
comparison of matrad and raystation calculations of a 120 mev proton pencil beam dose distribution.,title="fig : " ] comparison of matrad and raystation calculations of a 120 mev proton pencil beam dose distribution.,title="fig : " ] figure [ matrad_pdd ] shows the central axis depth - dose depositions for selected energies from 100.32 to 226.08 mev form the fluka simulations compared to the distributions re - calculated using matrad .
original fluka simulated central axis beam dose depositions for selected energies ( blue ) and re - calculated in matrad ( red ) . ]
the matrad calculated dose distributions use a different grid size and spacing to the fluka data , specifically 200x200x200 3 mm@xmath24 voxels , demonstrating that appropriate lookups as well as interpolation of data are being performed .
the small differences seen are due to the fitting of the alpha - stable distributions to fluka data .
the use of stable distributions provides a way of calculating the dose in a medium in a scanned pencil beam proton therapy machine that lends itself to implementation on gpu type architectures .
the calculation in the fourier space can be done fast and the libraries are available to do this fast on such processors .
alternatively , it is possible to directly estimate the integral provided by the inverse transform yielding : @xmath65 this can be numerically evaluated using a gaussian quadrature method , which is computationally faster than a fast fourier transform@xcite .
although monte carlo simulation type dose calculators are becoming increasingly available , the use of an analytical alternative is interesting if exhaustive searches in treatment plans are being used . providing a parametrization of this type reduces the number of parameters to a more manageable level , allowing a better insight in the physics of proton therapy planning using scanned pencil beams .
it becomes clear , for instance , that the scattering properties of combined scanned beams are different depending on the depth of the treated volume , and therefore different dose characteristics and maybe even biological effects can be expected .
this because there might be variations of let depending on the contributions of the halo at various depths .
it also provides a method to describe issues like changes in medium in terms of the used parameters . in a forthcoming study
we have already established that not all parameters behave in the same manner as a function of depth combined with changes in material ( data not shown ) . in this current study
we considered the pencil beam to be isotropic .
in practise it is possible that that is not the case depending on the geometric properties of the machine used to generate the pencil beams .
for instance , many proton therapy facilities will use spatially sequential magnets to bend the beams in the directions perpendicular to the beam axis in two perpendicular directions to each other .
this results in an ellipsoid spot size due to a different virtual source positions .
the implication is that we have to find a way to combine different stable distributions . in the case of the normal distribution
this is well understood , i.e. combining the variations depending on the mixing angle .
combining generalized stable distributions is less straightforward , but still fairly trivial in the case where the @xmath17 parameter is constant . investigating eq .
[ cf_short ] shows that the combination of two distributions with the same @xmath17 and scale parameters @xmath66 , and @xmath67 yields a new stable distribution with scale parameter @xmath18 : @xmath68 fortunately , we have seen that the parameter @xmath17 depends only on the amount of material that has been passed . as a result
the value of @xmath17 is the same in every direction of the plane .
combining stable distributions with different @xmath17 is not straightforward because as far as we know the resulting distribution is not stable and is still an area of mathematical research .
we have also limited this study to that of symmetric zero centered pencil beams . while the zero centering is easily resolved by a well chosen coordinate transformation ,
the asymmetry of a pencil beam is not resolvable in an easy way . indeed , in some cases the treatment beams are not symmetric , specifically if collimation is used and pencil beams near the collimator jaws need to be considered . in that case
the parameter @xmath69 is not zero and the full expression as outlined in eq .
[ cf_long ] needs to be evaluated .
this is subject of further research by our group . in theory ,
the methodology we have shown here could be extended to other charged particles and photons .
this is an area of further research by our group .
we have demonstrated that alpha
stable distributions are suited to describe charged particle pencil beams in a medium because they provide an accurate and efficient parameterization .
we have shown how this parameterization of the pencil beam allows dose distributions from intermediate energies to be interpolated through intermediate morphing .
furthermore , we have implemented the alpha - stable parameterization into a treatment planning system .
frank van den heuvel , francesca fiorini , and ben george greatfully acknowledge core support by cancer research uk ( cruk ) and the medical research council ( mrc ) .
frank van den heuvel : concept , fitting the stable distributions and editing the article , francesca fiorini : monte carlo simulations and co writing the article , niek schreuder : measurements to verify pencil beam monte carlo calculations and co writing the article , ben george : implementation of the algorithm in matrad and co writing the article 19ifxundefined [ 1 ] ifx#1 ifnum [ 1 ] # 1firstoftwo secondoftwo ifx [ 1 ] # 1firstoftwo secondoftwo `` `` # 1''''@noop [ 0]secondoftwosanitize@url [ 0 ]
+ 12$12 & 12#1212_12%12@startlink[1]@endlink[0]@bib@innerbibempty @noop ( ) @noop ( ) , http://stacks.iop.org/0031-9155/60/i=14/a=5627 [ * * , ( ) ] http://stacks.iop.org/0031-9155/50/i=3/a=011 [ * * , ( ) ] link:\doibase 10.3389/fonc.2015.00281 [ * * ( ) , 10.3389/fonc.2015.00281 ] link:\doibase 10.1016/j.ejmp.2015.05.004 [ * * , ( ) ] @noop _ _ , ( , ) in @noop _ _ ( , ) pp .
`` , '' in @noop _ _ ( , , ) chap . , pp . http://dx.doi.org/10.1038/nature09116 [ * * , ( ) ] , @noop ( ) @noop * * , ( ) `` , '' ( , , ) chap . , pp . \doibase http://dx.doi.org/10.1016/s0895-7177(99)00110-7 [ * * , ( ) ] @noop `` , '' ( ) link:\doibase 10.1007/s11222 - 016 - 9691 - 9 [ ( ) ] @noop * * , ( ) @noop * * , ( )
in this appendix we outline the notion of stable distributions , provide some definitions and show that the characteristic representation parameterizing the quantities ( @xmath17 and @xmath18 ) indeed represents all stable distributions .
the text is extensively based on the treatise by uchaikin and zolotarev and is provided as a synthesis and guideline rather than an original scientific contribution , the original work is much more extensive and dense @xcite .
we start out by quoting the law of large numbers which states that the difference between the estimated mean of a sample from a random variable tends to the mean of the distribution when enough samples are taken .
it is best known in the form as proposed by bernouilli in the 18th century : let @xmath70 , @xmath71, @xmath72 be independent , identically distributed random variables with
mean @xmath73 then : @xmath74 since @xmath75 i.e. it converges to zero in probability as @xmath76 , which provides the reformulation of bernoulli s law of large numbers : @xmath77 a more sophisticated approach considers the random variables as functions on an interval [ 0,1 ] where @xmath78 is an instantiation of the experiment yielding 0 or 1 .
in that case the strong law of large numbers can be replaced by a weaker version : @xmath79 which then becomes a degenerate function if infinite samples are taken , but more importantly , before reaching the degenerate condition the sum tends to the normal distribution , which is the classical form of the the central limit theorem .
we also provide the notion of _ equivalent _ distributions @xmath85 and @xmath86 : @xmath87 as well as the notion of similar distributions which provides the possibility of introducing a linear transformation of the given distribution .
@xmath88 using expression [ equival ] and [ similar ] it is clear that for similar distributions @xmath85 and @xmath86 on an infinitesimal interval @xmath89 : @xmath90 therefore , the same applies to the distribution functions ( cumulative of the density function ) : @xmath91 for example the normal distribution @xmath92 provides the following expression : @xmath93 an interesting property arises when , instead of looking at the variables themselves , we now investigate how sums of these variables ( summands ) behave .
if we have two normal distributions @xmath94 and @xmath95 with variances @xmath28 and @xmath29 then it is easy to see that , using expression [ scaling ] , we get : @xmath96 by setting @xmath97 and with an arbitrary summands , @xmath34 , we obtain a well known result : @xmath98 or more interestingly expressed as : @xmath99 it is the generalization of this property that leads to the notion of stable distributions by allowing arbitrary values of @xmath100 and @xmath101 .
let @xmath70 , @xmath71, be independent , identically distributed random variables , and let there exist constants @xmath105 and @xmath106 such that @xmath107 for some function @xmath108 which is not degenerate is not a step function ]
. then @xmath108 follows the stable law . in this section ,
we show how we move from the expression in [ clt_stable ] to the parameterization that we have been using denoting the type of stable distribution based on the single parameter @xmath17 .
before we move on , we narrow the definition of stable distributions to that of _ strictly _ stable distribution by setting @xmath111 , thus : @xmath112 defining @xmath113 the distribution of the sum . by calculating the variance of the distribution [ strictly ]
we obtain : @xmath114 if @xmath115 and @xmath116 then there is only one possibility @xmath117 which reverts to the result obtained for the normal distribution . with the notion of summands
, we can now use them to further extend the properties to the general case of summing strictly stable random variables .
@xmath118 we now choose to limit ourselves to summands with @xmath119 terms @xmath120 keeping in mind that @xmath121 , which we can generalize to any pair , we can rewrite the expression in ( [ 2ksum ] ) to read : @xmath122 or in much shorter notation @xmath123 and using the expression in defining strictly stable random variables[strictly ] : @xmath124 , we obtain : @xmath125 transforming into @xmath126 \ln b_2 = \ln n^{(\ln b_2)/\ln 2}\ ] ] or @xmath127 we can now repeat this process with @xmath128 , @xmath129 and using induction @xmath130 , yielding @xmath131 , @xmath132 , and @xmath133 , bunching respectively 3 , 4 and @xmath134 summands . in general the following expression is valid : @xmath135 for arbitrary values of @xmath134 . from these
last two formula we conclude that @xmath140 . by induction
we see that there is a single @xmath17 for all these stable distributions , and that the scaling factors @xmath141 follow the following law : @xmath142 [ [ characteristic - function - of - symmetrical - zerocentered - stable - distribution ] ] characteristic function of symmetrical zero centered stable distribution ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ the characteristic function , c.f .
, of a distribution can be defined as the fourier transform of the probability density function , p.d.f .
, of that distribution .
let @xmath143 be a p.d.f .
of a set of random variables @xmath85 , then the c.f . , @xmath144 , is defined as it s expectation value , @xmath145 : @xmath146 from this definition some properties follow immediately : 1 .
@xmath147 2 .
@xmath148 3 .
@xmath149 , where @xmath150 denotes the complex conjugate .
4 . if @xmath85 is symmetric about zero , @xmath151 5 . if @xmath152 then the continuous @xmath34th derivative of the c.f . exists and : @xmath153 6 . if @xmath113 is the sum of independent random variables @xmath70 , @xmath70 , , @xmath72 , then : @xmath154 7 .
any @xmath155 is a uniformly continuous function .
( inversion theorem ) any distribution function @xmath156 is uniquely defined by its c.f .
if @xmath100 and @xmath101 are some continuity points if @xmath156 , then the inversion formula states that @xmath158 the principle advantage of using the c.f .
is that the c.f . of a sum of independent random variables
is equal to the product of the c.f.s of the summands : @xmath159 if we take the logarithm ( obtaining the second characteristic : @xmath160 ) , we find that : @xmath161 this is important as it allows us to assess the summation of a large number of independent random variables without evaluating multiple integrals .
for this reason we cite the continuity theorem : ( continuity theorem ) let @xmath162 be a sequence of c.f.s and let @xmath163 be a sequence of corresponding distribution functions . if @xmath164 as @xmath82 , for all @xmath165 and @xmath157 is continuous at @xmath166 , then @xmath157 is the c.f . of a cumulative distribution function @xmath156 , and the sequence
@xmath163 weakly converges to @xmath156 , @xmath167 .
the inverse is also true : if @xmath167 and @xmath168 is a distribution function , then @xmath164 , where @xmath157 is the c.f .
of the distribution function @xmath168 . to gain some insight in how to perform this
, we can look at two well known stable distributions to find a way forward .
the distributions under consideration are the normal distribution and the cauchy distribution .
the calculation of the characteristic function for these distributions is well known and they also form part of the stable distribution , in the form @xmath169 representing the stable distribution density : note that the traditional form of the density function for the normal distribution is slightly different : @xmath175 this observation allows us to generalize for any symmetric distributions .
let @xmath176 be such a distribution with an arbitrary parameter @xmath17 .
we remind that @xmath177 keeping in mind the property as elucidated in equation [ conv_prop ] @xmath178 using the expression for the second characteristic @xmath179 the results obtained earlier when investigating the normal- and the cauchy distribution lead us to propose a solution of the form @xmath180 this satisfies the previous expression with @xmath181 and an arbitrary complex
valued @xmath182 , which we choose to be @xmath183,\quad \lambda , c_1 , \text { real numbers}\ ] ] to find @xmath184 we use property ( 3 ) of the chacaracteristic function , the link between conjugate characteristic function and negative estimates : @xmath185 , \quad -\infty < k<\infty , 0<\alpha\leq 2\ ] ] with @xmath186 which is a trick to rewrite the equation as @xmath187 , in such a way that we are explicitly splitting the expression in a real and an imaginary part by a good choice of the constant @xmath182 .
we also have not specified what form the function @xmath188 takes , we do know that it will depend on the parameter @xmath17 as well as provide a measure for the asymmetry of the distribution , should that be present .
later we will attribute that to a parameter @xmath69 .
the constant @xmath189 is an arbitrary real number and can serve as a scaling factor , which can be renormalised to @xmath20 , without loss of generality .
this implies that the full expression of a stable distributions characteristic function is of the form : @xmath190 taking into account that the characteristic function of a symmetric stable function is real valued due to property ( 3 ) of the characteristic functions : @xmath191 and the characteristic function for any stable function becomes : @xmath192 with @xmath18 as a scaling factor . | * purpose : * to introduce and evaluate the use of stable distributions as a means of describing the behavior of charged particle pencil beams in a medium , with specific emphasis on proton beam scanning ( pbs ) .
* methods : * the proton pencil beams of a clinically commissioned proton treatment facility are replicated in a monte carlo simulation system ( fluka ) . for each available energy
the beam deposition in water medium is characterized by the dose deposition . using an alpha stable distribution methodology each beam with a nominal energy @xmath0
is characterized by the lateral spread at depth @xmath1 : @xmath2 and a total energy deposition @xmath3 .
the beams are then described as a function of the variation of the parameters at depth .
finally , an implementation in a freely available open source dose calculation suite ( matrad , dkfz , heidelberg , germany ) is proposed .
* results : * quantitatively , the fit of the stable distributions , compared to those implemented in standard treatment planning systems , are equivalent .
the efficiency of the representation is better ( 2 compared to 3 and more parameters needed ) .
the meta parametrization ( i.e. the description of the dose deposition by only providing the fitted parameters ) allows for interpolation of non measured data . in the case of the clinical data used in this paper , it was possible to only commission 1 out of 5 nominal energies to obtain a viable data set .
* conclusions : * alpha stable distributions are intrinsically suited to describe charged particle pencil beams in a medium and can be easily implemented in existing treatment planning systems .
the use of alpha - distributions can easily be extended to other particles . |
m. mendoza would like to acknowledge the brazilian agency capes for financial support , while p.a.s .
is grateful to the continuous support provided by fapesp . | we propose a controllable way of tuning fano resonances in open quantum dots in the absence of magnetic field .
a quantum dot can be modified by changing the gate voltages that define the dot itself .
an extra degree of freedom can be introduced by means of a controllable repulsive perturbation , such as the one induced by scanning a atomic force microscope tip on a real sample .
we numerically investigate the coupling between localized states in the dot and the continuum in the leads .
the advantage of such position controllable perturbation is the selective manipulation of the quantum dot states .
we show that this could be a feasible alternative to quantum dots in aharonov - bohm interferometers as a fano resonance tuning device .
pacs number(s ) 73.40.gk,73.21.fg,73.21.la in the past few years , mesoscopic systems became appropriate tools for investigating wave function interference effects .
the nanometer - scale technology turned into reality structures with dimensions that are smaller than the phase coherence length .
the prototype are systems created in a two - dimensional high - mobility electron gas ( 2deg ) at the interface of a gaas / algaas heterostructure structured by means of the split gates technique @xcite .
two important examples should be mentioned here : the quantum point contact ( qpc ) , where conductance quantization has been seen @xcite ; and the quantum dot ( qd ) , with a wide range of interesting transport properties @xcite .
starting from these two devices , systems of greater complexity have been conceived and fabricated for controling electronic wave function interference effects .
a recent example of such effort is the imaging of the coherent electron flow through qpc , where the interference effects can be controlled by scanning an atomic force microscope ( afm ) tip over the sample @xcite .
another example , related to the present work is the tunable fano resonance in an aharonov - bohm interferometer with an embebed qd @xcite .
we propose that fano resonances could be completely tunable in open quantum dot ( oqd ) structures by moving a local perturbation across the system , like a tip of an afm . indeed
such a procedure enables a selective manipulation of the qd states . up to now , in absence of magnetic fields , fano resonances have been only partially tuned by changing gate voltages in single - electron transistors .
@xcite . in the present approach
, a continuous system is discretized into a lattice , considering a single @xmath0-like orbital per site and only nearest - neighbour hopping elements .
the oqd structure , emulated by a tight - binding lattice model is depicted in fig.1(a ) .
the black circle represent a potential columm simulating the perturbation induced , for instance , by an afm tip located on the sample at that position . in what follows we consider perturbations of a single host lattice site ( @xmath1 ) , which corresponds to a extension relative to the quantum dot of @xmath1/@xmath2=1/49 for oqd with dimensions @xmath3 , where @xmath4 is the host lattice parameter ( fig.1(a ) ) .
the repulsive perturbation will be characterized by a height @xmath5 in the range between @xmath6 mev and @xmath7 ev .
the tight - binding hopping parameters are chosen in order to emulate the electronic effective mass for the gaas bottom of the conduction band , @xmath8 .
hence , @xmath9 , @xmath10 ev for a lattice parameter of @xmath11 .
such parametrization represents quantum dots with lateral sizes up to @xmath12 , at least one order of magnitude lower than the typical dimensions of actual quantum dots constructed by litographic methods .
however , the present work has the proposal of illustrating a selective manipulation of confined states . here
the relevant energy scale is @xmath13 ev . for systems ten times larger , the new energy scale would be of the order of @xmath14 mev , still accessible for available temperatures .
another relevant scale is the ratio between the extension of the perturbative spike and the dot dimension , @xmath15 .
a previous work @xcite suggests that the energy shifts are qualitatively insensitive to this ratio if the perturbation height , @xmath5 , is less than @xmath16 .
the afm tip can also be seen as a controllable impurity in a quantum dot and therefore a simple tunable experimental realization of a multiply connected nanostructure @xcite .
we will interested in the transmission probabilities and local density of states ( ldos ) , which are obtained by means of green s functions approaches . for the ldos we follow a self - energy method
@xcite , the total hamiltonian , @xmath17 , is a sum of four terms : the dot connected to the two point contacts regions , described by the @xmath18 , the left and rigth contact regions , @xmath19 and @xmath20 , respectively ; as well as the coupling term between the contacts and the dot structure , @xmath21 : @xmath22 where @xmath23 describes the interaction between device and contacts .
the corresponding green s function is given by @xmath24 we define the green s functions for the isolated contacts : @xmath25 using equation ( 1 ) in ( 2 ) we obtain : @xmath26 the greens function of the device , @xmath27 , result in @xmath28 where @xmath29 and @xmath30 are the so called self - energies which describe the effects of the leads on the device .
the propagators in the transverse mode representation for the isolated contacts , @xmath31 are : @xmath32 with @xmath33\ ] ] the matrix elements of the unitary transformation matrix , @xmath34 , which transforms the greens function from site to transverse mode representation ( @xmath35 ) are : @xmath36 where n ( transverse site ) and @xmath37 ( transverse mode ) are row and columm indexes of the @xmath34 matrix , respectively .
the @xmath38 , @xmath23 , @xmath39 and @xmath34 are @xmath40x@xmath40 matrices , where @xmath40 is the number of transverse sites . using equation ( 5 ) we obtain the ldos : @xmath41\ ] ] where m is the longitudinal site .
this ldos is proportional to the probability density of a state of energy e. finally , a green s function is related to the conductance by the landauer - bttiker formula @xcite , by means of a recursive solution of the dyson equation as discussed elsewhere @xcite .
we can control the coupling of oqd bound states to the two dimensional reservoirs by changing the width , @xmath42 , of the left and right qpcs , fig .
1(a ) transmission probabilities as a function of the incident electron energy are shown in fig .
1(b ) . for @xmath43 , ( continuous line )
we clearly see resonances in the transmission probability due to the bound states of the oqd , as well as the first conductance plateau . for wider leads , @xmath44 , ( dashed line ) we see the resonances due to oqd states shifted to lower energies , since the effective confinement is diminished with the widening of the contacts . moreover the resonance peaks are broadened because the coupling to the reservoirs is enhanced .
as expected , the conductance plateaus are also shifted to lower energies and the onset of a second plateau can be seen in the same energy range .
two aspects related to widening the leads should be kept in mind ; ( i ) the shift in energy and broadening of the resonances are not a monotonous function of the energy ; and ( ii ) , even the shape of a resonance may the qualitatively changed , as can be seen for the antiresonance for @xmath43 changing into a peak for @xmath44 .
the previous paragraph summarizes the effects on the conductance that can be achieved by varying ( symmetrically ) the leads , i.e , the voltages of the gate electrodes that actually define the quantum dot structure . in
what follows we will be interested in effects due to a new degree of freedom in the manipulation of the dot states , as well as the coupling between them and the continua of states in the leads .
such new mechanism is introduced by a local potential perturbation that can be moved across the sample , like an afm tip , as suggested above .
the novelty of this local perturbation is that the resonances may be selectively tuned , exploring the wave function symmetries of the states . in order to explore the wave function symmetries , one has to look systematically along three symmetry axis for the present case of a square qd : along the axis connecting the leads ( x axis ) , the axis perpendicular to the line joining the leads and crossing the center of the dot ( y axis ) , and a diagonal , also crossing the center , as indicated in the sketch shown in fig .
1(a ) .
the shift in energy of a certain resonance peak depends on the position of the local perturbation , since such shift is proportional to the wave function amplitude at this position @xcite : @xmath45 where @xmath46 , is the unperturbed energy of state @xmath47 , @xmath21 is the effective perturbative potential , and @xmath48 is the probability density of the unperturbed state .
hence , a mapping of the probability density related to a resonance peak can be obtained by a contour plot of the energy shifts as a function of the perturbation position @xcite .
here we are interested not necessarily in wave function mapping , but in the related selective tuning of the resonances .
fig .
1(c ) shows contour plots of the ldos at the energies of the 4 resonances and the anti - resonance shown in fig .
1(b ) for the @xmath43 wide leads case ( continuous line ) .
the last pannel is for an energy sligthly above the anti - resonance .
having these contour plots in mind it is straightforward to see that the width of the resonance peak is proportional to the wave function amplitude ( ldos ) leaking into the leads .
of particular interest is the strong localization of the fifth state , which correspond to the anti - resonance in fig .
1(b ) , as well as the clear 1d channel character at off resonance energies in the conductance plateau , @xmath49 , fig .
1(c ) . within this framework illustrated by fig .
1 we are able to understand the effects of the position dependent perturbation . in this sense ,
two important effects are shown in fig.2 .
first , for a perturbative spike positioned at the center of the quantum dot , fig.2(a ) , we see the selective shift of the resonances in the perturbed system with @xmath50 ev(continuous line ) compared to the unperturbed one ( dashed line ) .
we clearly identify that important shifts ( for the lowest resonance and the anti - resonance ) occur only for those states with significant amplitude at the center of the dot .
next , for a perturbation @xmath51 ev placed at one of the leads ( continuous line ) , fig .
2(b ) , no energy shifts of the resonances are observed . on the other hand ,
a strong suppression of the conductance plateau can be observed , an effect which has been used recently for imaging the electron flow through quantum point contacts @xcite .
the dependence of the selective resonance shifts on the wave function symmetries for moving the perturbative spike along the y axis can be analyzed by comparying fig .
2(a ) and fig .
2(c ) .
now , for @xmath52 ev ( continuous line ) and @xmath53ev ( dashed line ) , the second resonance remains almost unchanged and the lowest one is shifted , as expected from the ldos in fig .
1(c ) .
having the ldos at the resonance energies in mind , it is also expected that the third resonance would be perturbed in this situation and , indeed , this resonance is not only shifted in energy but the line shape changes into a fano - like form . here
the shift in energy affects dramatically the coupling with the continua in the leads and therefore the line shape change in the transmission probability . moving the perturbation further away from the center in the y direction , fig .
2(d ) , this third resonance evolves towards a symmetric peak again ( continuous line for @xmath54ev ) , illustrating the tunability of a specific resonance with a controllable perturbation .
the symmetry of the localized states in the qd give rise to two new consequences when the perturbation is moved along the diagonal , fig .
3 . here dashed and continuous line are for perturbative spikes at the center and at the diagonal , respectively .
both situations are for @xmath55 ev .
first , we observe the inversion of the fano resonance associated to @xmath56 @xcite : from fig .
2 ( c ) to fig .
3 , the peak position is reversed .
moreover , only an off center perturbation along the diagonal breaks the symmetry of the @xmath57 state , allowing the coupling with the 1d channel and leading to a fano resonance .
this symmetry is not broken with the perturbation positioned along the other symmetry axis ( x and y ) , since the ldos related to @xmath57 is negligible there . in summary
, we propose that scanning an afm across an open quantum dot could be an efficient tool for a selective manipulation of quantum states .
the local perturbation induced by the tip is an important new degree of freedom in order to explore and modify resonant states . as an example , we show that the coupling of a specific level with the continuum can be tuned over a wide range in which the line shape of the associated resonance in the conductance changes from a symmetric peak into a asymmetric fano resonance , which can be inverted or tuned back into a single peak .
this effect suggests an interesting alternative to quantum dots in aharonov - bohm interferometers @xcite as a fano resonance tuning device . |
the discovery of @xmath7-electron compounds that display a unique combination of competing ground states promotes advances in the field of kondo physics by challenging our understanding of the underlying many - body interactions.@xcite the flurry of research activity associated with the ce@xmath11t@xmath12in@xmath13 ( t = co , ir , or rh ; @xmath14 = 1 or 2 ; @xmath15 = 0 or 1 ) compounds certainly fits this description.@xcite the compounds that reside within this `` family '' exhibit essentially all of the many ground states that have been observed in @xmath7-electron systems , including paramagnetism , antiferromagnetism , and exotic ambient - pressure and pressure - induced superconductivity .
these compounds also exhibit a mixture of essentially all known @xmath7-electron phenomena , including a kondo - renormalized ground state , fermi - liquid behavior , and both pressure - induced and ambient - pressure non - fermi - liquid ( nfl ) behavior in the vicinity of a antiferromagnetic quantum - critical point ( qcp).@xcite to date , our understanding of @xmath7-electron kondo systems has rested , in part , on the competition between kondo and ruderman - kittel - kasuya - yosida ( rkky ) interactions.@xcite the complex phase diagram defined by the ce@xmath11t@xmath12in@xmath13 family challenges this understanding .
the ce@xmath11t@xmath12in@xmath13 family contains three distinct subgroups , each with differing degrees of anisotropy . the first subgroup ( @xmath15 = 0 ) contains a single member , the cubic antiferromagnet ( afm ) cein@xmath16 . at ambient pressure cein@xmath16 orders
magnetically at @xmath17 = 10 k and has a slightly enhanced sommerfeld coefficient @xmath18 ( @xmath19 as @xmath20 of 100 mj / mole k@xmath21.@xcite an applied pressure of @xmath22 25 kbar drives @xmath17 to zero,@xcite leads to nfl behavior,@xcite and produces a superconducting state with a maximum transition temperature @xmath23 = 0.2 k. the second subgroup ( @xmath14 = @xmath15 = 1 ) contains three known ce - based members , cecoin@xmath1 , ceirin@xmath1 , and cerhin@xmath1 .
these tetragonal ce-115 compounds have a quasi - two - dimensional structure that is composed of a cubic cein@xmath16 element separated by a tin@xmath24 layer .
cerhin@xmath1 is an ambient pressure antiferromagnet ( @xmath17 = 3.8 k ) with an enhanced @xmath18 @xmath25 400 mj / mole k@xmath21.@xcite superconductivity occurs at @xmath23 = 2.1 k in a pressure of 16 kbar.@xcite in contrast ambient - pressure unconventional@xcite superconductivity occurs in both ceirin@xmath1 ( t@xmath26 = 0.4 k)@xcite and cecoin@xmath1 ( @xmath23 = 2.3 k).@xcite in both superconductors many - body interactions produce a highly renormalized mass state ( @xmath18(ceirin@xmath1 ) = 0.75 j / mole k@xmath21 , @xmath18(cecoin@xmath1 ) @xmath25 1 j / mole k@xmath21 ) that becomes evident in the specific heat @xmath27(t ) below 10 k. cecoin@xmath1 exhibits clear nfl behavior in resistivity @xmath28(t ) and @xmath27(t ) data , indicating that this compound resides near a antiferromagnetic quantum - critical point.@xcite while the specific - heat of ceirin@xmath1 exhibits fermi - liquid behavior just above @xmath23 , @xmath29 , thermal - expansion , and @xmath30 data suggest that this compound may be close to a qcp as well.@xcite the third subgroup ( @xmath14 = 2 , @xmath15 = 1 ) is composed of three double - layer members ce@xmath24tin@xmath31 ( t = rh , ir , co ) .
these ce-218 compounds exhibit the same phenomena seen in the ce-115 materials , including paramagnetic@xcite and afm@xcite ground - states , pressure - induced and ambient pressure superconductivity,@xcite and nfl behavior.@xcite anisotropy is a feature common to nearly all of the properties exhibited by the ce-115 compounds . in part , these anisotropies stem from the tetragonal 115 lattice structure . band structure calculations and de haas - van alphen ( dhva ) measurements indicate that the 115 materials have a complicated quasi-2d fermi surface ( fs ) composed of multiple electron and hole orbits.@xcite when considering the novel properties of the 115 compounds it is critical to differentiate between phenomena that are controlled by prosaic single - electron physics and those determined by correlated electron interactions .
the hall effect provides a useful means of elucidating the relative importance of single - electron ( i.e. , conventional electronic structure ) and many - body interactions in @xmath7-electron systems .
the hall effect in these systems is strongly influenced by the scattering of charge carriers via the orbital angular momenta of localized electrons.@xcite in most @xmath7-electron compounds this orbital skew scattering effect dominates the hall coefficient @xmath32(t ) at temperatures greater than the characteristic kondo temperature @xmath33 .
in contrast , the electronic - structure component is generally a small , temperature - independent contributor to @xmath32(t ) .
many body correlations frequently dominate the hall coefficient as the system approaches the kondo temperature from above.@xcite at and below @xmath33 , @xmath32(t ) is influenced by the onset of coherence and the development of the abrikosov - suhl resonance in the electronic density of states near the fermi energy.@xcite for these reasons t and h - dependent hall measurements provide insights into the relative importance of electronic structure and many - body interactions in the physical properties of @xmath7-electron compounds .
we have measured the in - plane hall coefficients of cerhin@xmath1 , ceirin@xmath1 , and cecoin@xmath1 and their respective non - magnetic lanthanum analogs in fields to 90 koe and at temperatures from 2 k to 325 k. the hall effect is dominated by skew - scattering above @xmath2 50 k in the ce-115 compounds , and a precipitous negative drop is present in the in - plane hall coefficient below 50 k that is inconsistent with incoherent orbital scattering . this same temperature dependence , but with a more modest t - dependent amplitude , is evident in hall measurements on the non - magnetic analogs larhin@xmath1 , lairin@xmath1 , and lacoin@xmath1 signifying that the hall effect in the ce-115 materials is dominated by the conventional hall carrier contribution .
measurements on the ce-115 compounds indicate that the hall anomaly present below 50 k can be suppressed significantly by a 90 koe field , suggesting that field - dependent many - body kondo interactions influence the unusual hall effect intrinsic to the 115 electronic structure .
lastly , hall angle ( @xmath4 ) measurements indicate that the ratio @xmath34 ) varies as @xmath6 below 20 k in all three ce-115 compounds ; la-115 @xmath35 ) data between 30 k and 100 k vary quadratically with temperature too .
this hall - angle temperature dependence has been observed in high - t@xmath26 cuprates@xcite as well , and some have speculated that this behavior is linked to antiferromagnetic spin fluctuations due to a nearby qcp.@xcite while it is tempting to make the same connection for these ce heavy - fermion compounds , the fact that @xmath36 in all six 115 compounds suggests that this temperature dependence is unconnected with qcp - related spin - fluctuations .
single crystal ce-115 and la-115 transport samples were produced with an indium flux - growth technique .
x - ray diffraction on powdered crystals indicates that the crystals are single - phase and form in the primitive tetragonal hocoga@xmath1 structure . while the excess in flux is critical for growing the crystals , residual indium has the negative side effect of contaminating essentially all transport measurement on as - grown samples . to eliminate this problem
all samples employed in this study were polished and then pre - screened via @xmath28(t ) and magnetic susceptibility @xmath37(t ) measurements to ensure that no extrinsic ( in ) superconductivity was evident at 3.4 k. hall measurements were made on single - crystal samples that were cut and polished into thin hall bars with sample thicknesses ranging from 0.1 to 0.4 mm .
all hall coefficients reported here are in - plane measurements with the magnetic field applied along the c axis
. the standard four - terminal contact arrangement was used , and the contacts were made with silver conductive epoxy .
the longitudinal resistivity ( @xmath38 ) and the hall voltage v@xmath39 were measured with a low - frequency resistance bridge .
systematic errors associated with hall contact misalignment were eliminated by ascribing the asymmetric component of the field - reversed transverse voltage to the hall voltage , @xmath40/2$].@xcite the hall coefficient was calculated from the standard expression r@xmath41 , where @xmath42 is the applied current flowing perpendicular to the applied field @xmath43 , and @xmath44 is the sample thickness .
the hall measurements were made in fields from 1 koe to 90 koe in order to examine the @xmath32 field dependence and to determine the hall coefficient in the low - field limit , @xmath3 .
the hall coefficient is extremely field - dependent in the ce-115 compounds below 20 k ; this strong field - dependence coupled with the fact that @xmath45 = 0 in the zero field limit means that some care is required to accurately determine @xmath3 .
we employed two methods in determining r@xmath46 from the measured hall voltages . in method
one @xmath47 was determined from the measured @xmath48 in fields ranging from 1 to 90 koe , and @xmath3 was calculated by extrapolating the data to @xmath43 = 0 . in method
two @xmath49 was determined from the zero - field limit of the hall resistivity field derivative , @xmath50 = @xmath51 , where @xmath52 .
both methods produce @xmath3 values that are identical within experimental error .
in contrast to the strongly field - dependent hall voltage in the ce-115 compounds , the la-115 materials exhibit a field - independent hall coefficient ( i.e. , @xmath53 ) .
as such , the t - dependent la-115 hall data reported in sec .
[ results ] were measured in 10 koe .
in - plane hall coefficients ( h = 10 koe ) plotted as a function of temperature for larhin@xmath1 , lairin@xmath1 , and lacoin@xmath1 .
low - temperature data are highlighted in the inset .
the data are normalized by the absolute value of each compound s 300 k hall coefficient .
the solid line is a fit to the lacoin@xmath1 data utilizing eq .
( [ eqn : lahall ] ) . ]
the la-115 compounds are isostructural analogs of the ce-115 compounds . as such ,
the la-115 transport properties are indicative of the non - magnetic contributions to transport in the ce compounds .
the characteristics of these la-115 compounds are , for the most part , unremarkable and typical of a non - magnetic intermetallic system .
larhin@xmath1 , lairin@xmath1 , and lacoin@xmath1 exhibit a temperature - independent pauli paramagnetic susceptibility , and the resistivity varies linearly with temperature above @xmath2 50 k. room - temperature resistivities range from 10 to 20 @xmath54 cm , while anisotropic resistivity measurements on larhin@xmath1 find that the nonmagnetic electronic anisotropy inherent in the tetragonal 115 structure is relatively small.@xcite the low - temperature ( 10 k ) magnetoresistance ( mr ) of larhin@xmath1 and lairin@xmath1 is positive and grows quadratically with field .
these properties are consistent with a simple metallic system
. the temperature - dependent in - plane hall coefficients of the three la-115 compounds are displayed in fig .
[ fig : hallla ] .
all three la-115 compounds exhibit a negative hall coefficient , with 300 k values of -7.5 x 10@xmath55 m@xmath56/c , -5.1 x 10@xmath55 m@xmath56/c , and -4.2 x 10@xmath55 m@xmath56/c for the co , rh , and ir compounds , respectively .
r@xmath39 is nearly temperature - independent above 100 k , drops monotonically below 100 k , and begins to saturate below 20 k. the inset to fig .
[ fig : hallla ] shows that @xmath32 is essentially t - independent below 10 k for the three la-115 compounds . while the data displayed in fig .
[ fig : hallla ] were measured in a 10 koe field , measurements made in fields from 1 to 90 koe show that r@xmath39 is field - independent ( @xmath57 ) at 5 , 10 , 50 , 100 , and 300 k. the data in fig .
[ fig : hallla ] are normalized by the absolute value of each compound s 300 k hall coefficient to underscore the fact that @xmath32 follows essentially the same temperature - dependence in all three la-115 compounds .
the data above 20 k are well described by the expression @xmath58 where @xmath59 , @xmath60 , and @xmath61 are fitting parameters .
( [ eqn : lahall ] ) was used to produce the fit to the normalized lacoin@xmath1 data that is shown in fig .
[ fig : hallla ] ; the fitting parameters are @xmath60 = 0.07 , @xmath61 = -0.06 k@xmath62 , and @xmath59 = -0.93 .
surprisingly , the same expression ( with @xmath59 = 0 ) describes the normal - state hall response in both yba@xmath24cu@xmath16o@xmath63 and mgb@xmath24.@xcite the substantial t - dependence present below 100 k in the la-115 data is quite unusual since metals typically have a nearly constant hall coefficient .
a t - dependent hall coefficient is usually a sign that the underlying electronic structure is composed of multiple electron and hole bands with differing mobility temperature dependencies .
band - structure calculations and dhva measurements do find that the la-115 electronic structure is very complex,@xcite so the la-115 @xmath32 temperature dependence appears to be a reflection of the 115 electronic structure .
the in - plane hall angle @xmath35 ) plotted as a function of @xmath6 for larhin@xmath1 , lairin@xmath1 , and lacoin@xmath1 .
the dashed lines are linear fits to the data . for clarity the ir and rh data have been vertically offset by 5 and 10 , respectively . ]
we next consider the la-115 hall angle ( @xmath4 ) , which is plotted as @xmath64 vs @xmath6 in fig .
[ fig : cotla ] .
@xmath4 is defined as the angle between the applied current and the resulting electric field , and it is determined experimentally via @xmath65 = @xmath66 .
the data fall on a straight line from roughly 30 to 100 k for all samples , indicating that @xmath65 varies with temperature as @xmath67 this quadratic temperature dependence occurs over precisely the same temperature range where @xmath32 is extremely temperature dependent .
the @xmath6 behavior results from the combination of a resistivity that varies linearly with @xmath68 and a hall coefficient that varies inversely with @xmath68 .
the hall angle of both yba@xmath24cu@xmath16o@xmath63 [ ] and mgb@xmath24 [ ] follow the same anomalous temperature dependence .
the la-115 compounds exhibit a conventional hall angle temperature dependence ( @xmath69 ) only above 100 k where @xmath32 is nearly constant .
magnetic - scattering contributions to the in - plane resistivity plotted as a function of temperature for cerhin@xmath1 , ceirin@xmath1 , and cecoin@xmath1 ; data below 10 k are highlighted in the inset . ]
figure [ fig : rhomag ] shows the magnetic resistivity @xmath70 of cerhin@xmath1 , ceirin@xmath1 , and cecoin@xmath1 plotted as a function of temperature from 1 to 325 k. the resistivity contribution from magnetic scattering is calculated by subtracting the electron - phonon contribution ( the resistivity of the non - magnetic la analog ) from the ce-115 resistivity , @xmath71 .
@xmath72 exhibits a broad maximum located below 50 k in all three compounds ; the coherence temperature @xmath73 is defined by the temperature where the resistivity peaks .
although the resistivity maximum in the cerhin@xmath1 data is not as pronounced as those in the ceirin@xmath1 and cecoin@xmath1 data , @xmath70 varies roughly as @xmath74 in each of the ce-115 materials for @xmath75 .
the inset to fig .
[ fig : rhomag ] shows the low - temperature behavior of the magnetic resistivity .
the superconducting transitions in ceirin@xmath1 and cecoin@xmath1 are clearly evident as abrupt drops in @xmath70 at their respective transport @xmath23 s , while the onset of magnetic order at @xmath17 ( 3.8 k ) in cerhin@xmath1 leads to a more subtle inflection - point anomaly . just above their superconductivity transitions ,
@xmath29 for ceirin@xmath1 and cecoin@xmath1 varies with temperature as @xmath76 and @xmath77 , respectively.@xcite these power - laws differ from that of a fermi - liquid ( @xmath78 ) , and are suggestive of non - fermi - liquid behavior .
low - field ( @xmath79 ) in - plane ( @xmath80 c ) hall coefficients plotted as a function of temperature for cerhin@xmath1 ( @xmath81 ) , ceirin@xmath1 ( @xmath82 ) , and cecoin@xmath1 ( @xmath83 ) .
low - temperature data are highlighted in the inset . ]
figure [ fig : hallce ] shows the low - field ( @xmath84 0 ) hall coefficients of cerhin@xmath1 , ceirin@xmath1 , and cecoin@xmath1 plotted as a function of temperature between 2 and 300 k ; the low - temperature behavior is highlighted in the figure s inset .
@xmath0 is electron - like at all temperatures , and the precipitous drop that occurs in the hall response of all three materials below 50 k is certainly the most prominent feature in the data .
the data displayed in the inset indicate that the hall response saturates below 4 k , and that the drop in @xmath0 occurs at roughly the same temperature ( @xmath240 k ) in the three compounds .
while the three ce-115 compounds all show signs of the same drop in @xmath32 below 40 k , the feature is far more prominent for cecoin@xmath1 and ceirin@xmath1 than for cerhin@xmath1 . the hall voltage in cecoin@xmath1 drops to zero below 2.3 k for fields less than @xmath85 due to the onset of superconductivity (
this data is omitted for clarity in fig .
[ fig : hallce ] ) . with regard to the afm transition in cerhin@xmath1 ,
no discernable anomaly is present in @xmath32 at or below @xmath17 . at higher temperatures
@xmath32 is weakly temperature dependent in all three ce-115 compounds .
the hall coefficients of cerhin@xmath1 and ceirin@xmath1 have a positive slope above 50 k ; between 50 k and 300 k @xmath86 grows by 25% for cerhin@xmath1 , and 9% for ceirin@xmath1 .
the cecoin@xmath1 hall coefficient shows no such positive slope above 100 k ; instead , the cecoin@xmath1 hall response drops by roughly 10% between 100 k and 300 k. the 300 k hall coefficient is nearly the same in the three ce-115 compounds , and the room temperature value ( -3.5 x 10@xmath55 m@xmath56/c ) corresponds to an _ effective _ carrier concentration of 2.9 @xmath87 per formula unit .
figure [ fig : hallh ] displays the ce-115 hall coefficients for temperatures below 60 k when measured in four different fields ( @xmath88 0 , 10 , 50 , and 90 koe ) . increasing field strength qualitatively influences the hall response in the three ce-115 compounds in the same manner .
the data indicate that the hall response is extremely field - dependent in the temperature region where it is also very temperature - dependent .
the most important feature of the data is that the large negative drop present below 40 k in the zero - field ce-115 data is progressively diminished as the applied field is increased .
while a field of 90 koe does not eliminate the drop in @xmath32 that occurs below 40 k , the magnitude of the effect is diminished to the point where the hall response in the ce-115 compounds is comparable to that seen in the la-115 non - magnetic analogs .
increasing the field above roughly 5 koe also produces a shallow minimum in @xmath32 that is centered between 5 and 10 k. below 15 k the @xmath32 field dependence decreases monotonically with increasing field : changing the field from the zero - field limit to 10 koe reduces @xmath32 by roughly 50% , while the hall response in 50 and 90 koe differ by less than 10% relative to r@xmath89 0 ) .
the extreme field - dependence exhibited by the ce-115 compounds is in stark contrast to the field - independent hall response exhibited by the la-115 compounds in the same temperature range .
measurements on cerhin@xmath1 , ceirin@xmath1 , and cecoin@xmath1 at 50 , 100 , and 300 k in fields between 1 and 90 koe indicate that the hall response becomes field - independent at these elevated temperatures . the field - dependent hall data shown in fig .
[ fig : hallh ] satisfy a scaling relationship that is similar to one used in analyzing single - impurity magnetoresistance data.@xcite following the standard definition of the relative magnetoresistance , @xmath90 , we define the relative change in the hall coefficient as @xmath91 isotherms of the field - dependent hall response can be superimposed by plotting @xmath92 as a function of the transformed field parameter @xmath93 defined by @xmath94 the insets in fig .
[ fig : hallh ] show @xmath92 plotted vs @xmath93 for cerhin@xmath1 ( a ) , ceirin@xmath1 ( b ) , and cecoin@xmath1 ( c ) . in constructing these plots the scaling parameter @xmath95
was set to 2 , and the values used for the thermal scaling parameter @xmath96 are listed in the insets .
each inset contains ten superimposed isotherms that were measured at temperatures ranging from 4 to 25 k ( 8 to 30 k for cecoin@xmath1 ) , and each isotherm is composed of hall data that was measured in fields ranging from 1 to 90 koe .
scaling works best for @xmath95 = 2 @xmath97 0.1 , and the @xmath96 values can be varied by @xmath97 0.3 k without adversely effecting the analysis ; attempts to use the spin-1/2 mr exponent ( @xmath95 = 1)@xcite were unsuccessful regardless of the value used for @xmath96 .
qualitatively , @xmath92 varies with @xmath93 in roughly the same manner for all three ce-115 compounds .
quantitatively , the cecoin@xmath1 data are a stronger function of @xmath93 than are the cerhin@xmath1 and ceirin@xmath1 data .
in - plane hall angle @xmath35 ) plotted as a function of @xmath6 for cerhin@xmath1 , ceirin@xmath1 , and cecoin@xmath1 .
the dashed lines are linear fits to the data . the ceirin@xmath1 data has been vertically offset by 50 for clarity . ]
we turn now to the ce-115 hall angle , which is plotted as @xmath64 vs @xmath6 in fig .
[ fig : cotce ] .
the data fall on straight lines for temperatures less than 30 k , indicating that @xmath65 varies with temperature in a manner consistent with eq .
( [ eqn : lacot ] ) . while 10 koe hall data were used in constructing fig .
[ fig : cotce ] , essentially the same temperature dependence results if data at other fields are used ; all that changes is the overall magnitude of @xmath65 .
the temperature range over which the data follow a quadratic temperature dependence are as follows : 8 to 20 k for cerhin@xmath1 , 4 to 30 k for ceirin@xmath1 , and 3 to 25 k for cecoin@xmath1 .
@xmath65 is nearly constant at higher temperatures because @xmath32 and @xmath38 both become weakly t - dependent above 50 k. as with the la-115 compounds , the ce-115 data vary quadratically with temperature in the same temperature range where @xmath32 exhibits considerably temperature dependence .
the one important difference between the ce and la hall angle data concerns the temperature range over which @xmath65 varies quadratically with temperature .
for the la-115 compounds quadratic behavior is evident from 30 to 100 k , while the ce-115 compounds show this behavior over a more limited temperature range .
the ce-115 hall response is very different from what is observed in most ce kondo - lattice systems .
the canonical ce heavy - fermion hall effect is dominated by a positive skew - scattering contribution that dwarfs the conventional charge - carrier contribution.@xcite in contrast , the hall effect in the ce-115 compounds appears to be governed by the conventional charge - carrier contribution . after disentangling the relative contributions from these two mechanisms it will become clear that kondo interactions and @xmath7-electron effects play an important part in determining the temperature and field dependence exhibited by the hall effect in the 115 compounds .
the hall response in heavy - fermion compounds is produced by a contribution from skew - scattering @xmath98 and the ordinary hall effect @xmath99 , @xmath100 the skew scattering term stems from interactions between the large ce spin state and the applied field that produce a left - right asymmetry in charge - carrier scattering . as first expressed by fert and levy,@xcite the skew - scattering terms takes the form @xmath101 where @xmath102 = @xmath103 is the reduced magnetic susceptibility and @xmath104 is the curie constant . for @xmath105
the parameter @xmath106 becomes @xmath107 where @xmath108 is phase - shift produced by incoherent kondo scattering , @xmath109 is the magnetic ion s land@xmath110 g - factor , @xmath111 is the bohr magneton , and k@xmath112 is boltzmann s constant .
as the system is cooled below @xmath73 incoherent kondo scattering dies off ; calculations based on the periodic anderson hamiltonian suggest that @xmath113 in the coherent regime.@xcite these theoretical results indicate that , starting from @xmath68 = 0 , @xmath98 should increase rapidly from zero , achieve a broad maximum at the temperature where @xmath70 peaks , and gradually decrease at higher temperatures .
the hall effect in many ce , u , and yb heavy - electron systems behave in this manner.@xcite these compounds exhibit a positive skew - scattering hall response because the repulsive single - impurity scattering potential leads to a negative phase shift .
r@xmath39 plotted versus @xmath114 for cerhin@xmath1 , ceirin@xmath1 , and cecoin@xmath1 .
temperature is an implicit parameter in this figure ; 300 k data appear on the left side of the plot and 40 k data appear to the right
. the solid lines are linear fits to the rh and ir data . ]
careful analysis of the hall data indicates that skew - scattering is a minor contributor to the overall hall response in the ce-115 compounds .
this is particularly true below 50 k where the large negative drop in the hall coefficient is inconsistent with skew - scattering .
the most obvious inconsistency is that the low - t anomaly occurs below the coherence temperature where any skew scattering contribution should be dropping to zero .
additionally , the sign of the hall feature is inconsistent with skew scattering , @xmath32 does not peak at @xmath73 , and the sharp drop in the hall response can not be fit to @xmath115 .
skew scattering only becomes a significant contributor to the hall effect above 50 k. this is shown in fig .
[ fig : skew ] where @xmath0 is plotted vs @xmath115 for temperatures from 40 to 300k .
the cerhin@xmath1 and ceirin@xmath1 hall data vary linearly with @xmath115 , as predicted by eq .
( [ eqn : rhskew ] ) .
the parameters determined from the figure are @xmath106 = 0.016 k / t and @xmath99 = -2.9 x 10@xmath55 m@xmath56/c for cerhin@xmath1 , and @xmath106 = 0.010 k / t and @xmath99 = -4.1 x 10@xmath55 m@xmath56/c for ceirin@xmath1 . the resulting phase - shifts ( @xmath116 = -0.042 radians and @xmath117 = -0.025 radians ) are consistent with those reported for other ce heavy - fermion systems.@xcite for cecoin@xmath1 @xmath32 does not scale with @xmath115 .
this suggests that any skew - scattering contribution present in the co material is overwhelmed by the conventional hall term .
the skew - scattering contribution evident in the ce-115 hall data is small , in part , because @xmath70 is five to ten smaller than in other heavy - fermion systems .
similarities in the hall response of the ce and la 115 compounds below 100 k imply that the second term in eq .
( [ eqn : rhtot ] ) is responsible for the anomalous drop present in the ce-115 data .
dhva measurements@xcite indicate that the la and ce compounds share essentially the same electronic structure , so the conventional ce-115 hall term should mimic that of the la-115 materials .
band - structure calculations and dhva measurements indicate that the 115 s are compensated materials ( @xmath118 ) with multiple electron and hole fermi surfaces that form complex 2d and 3d structures.@xcite the hall effect of a multiband system is determined by the weighted sum of the contributions from each band .
qualitatively , a temperature - dependent hall coefficient can occur when multiple electron and hole band cross the fermi energy and the bands have mobilities with different temperature dependencies.@xcite the situation for the 115 s is even more complicated since the electron and hole conduction bands give rise to highly anisotropic fermi surfaces , and , presumably , anisotropic relaxation times .
this description also applies to the electronic structures of mgb@xmath24 and yba@xmath24cu@xmath16o@xmath63 , both of which also have a hall coefficient described by eq .
( [ eqn : lahall]).@xcite while the presence of a complex fs qualitatively accounts for the temperature - dependent la-115 hall coefficient , it does not address the question of why @xmath32 becomes increasingly negative below 100 k. the electron - like hall response may occur because the electron extremal orbits seen in larhin@xmath1 dhva spectra tend to have lighter masses , and hence larger mobilities , than the extremal hole orbits.@xcite despite their similarities , there are also significant difference in the ce-115 and la-115 hall response .
the most obvious difference is that the hall anomalies are much larger in the ce-115 compounds .
the hall response of cecoin@xmath1 changes by a factor of 20 between 100 and 4 k ; the hall coefficients of cerhin@xmath1 and ceirin@xmath1 change by a factor of 4 and 10 , respectively , over the same temperature range . in comparison ,
the hall response at 100 k and 4 k differ by only a factor of two in the la-115 compounds .
these differences presumably stem from the influence of the @xmath7-electrons that are present in the ce compounds .
if we assume that the ce-115 hall response @xmath119 is the sum of a skew - scattering term and a term proportional to the la-115 coefficient , @xmath119 can be expressed as @xmath120 where @xmath121 is defined as the @xmath7-electron hall weighting function .
figure [ fig : hallf ] shows @xmath122 ( symbols ) plotted vs temperature for the three ce-115 compounds ; the solid line in the figure shows @xmath121(90 koe ) for cecoin@xmath1 .
@xmath121 was determined from eq .
( [ rh : ce ] ) by combining ce and la hall data with the skew scattering contribution determined from the analysis associated with fig .
[ fig : skew ] ( we assume @xmath98 = 0 for cecoin@xmath1 ) . @xmath121 is unity from roughly 50 to 300 k , indicating that the conventional hall contribution in the ce compounds matches what is measured for the la-115 s . below 50
k @xmath121 monotonically increases with decreasing temperature , and saturates below 3 k ; the rise in @xmath121 begins at 45 k for cecoin@xmath1 and 25 k for both cerhin@xmath1 and ceirin@xmath1 .
these onset temperatures indicate that the @xmath7-electron contribution to the hall effect begins to grow at roughly the same temperature where the resistivity peaks . as such
, the growth in @xmath121 evident below 50 k appears to be correlated with the commencement of kondo coherence .
the extremal orbit masses observed in dhva spectra can provide a simple explanation for the connection between the ce-115 hall response and kondo interactions below 50 k. those measurements@xcite on cecoin@xmath1 detect a heavy hole orbit with a mass ( 87@xmath123 ) that is consistent with the large electronic specific - heat coefficient observed experimentally.@xcite in comparison , the cecoin@xmath1 electron orbits are significantly lighter ( @xmath215@xmath123 ) .
this extreme electron - hole mass asymmetry would lead to a mobility asymmetry , and , ultimately , an even larger negative hall anomaly than is seen in lacoin@xmath1 .
although the differences are not as extreme , ceirin@xmath1 and cerhin@xmath1 also exhibit asymmetries in their respective hole and electron masses.@xcite of the three ce-115 compounds , cerhin@xmath1 has the smallest electron - hole mass asymmetry , and the least amount of 4f character in its fermi surface;@xcite these fs features are consistent with the fact that cerhin@xmath1 also has the smallest low - t hall anomaly in the ce-115 series .
the enhanced low - t ce-115 carrier masses are a direct result of the kondo interactions that produce the large electronic contribution to @xmath124 .
the kondo resonance gradually develops with decreasing temperature , so that the carrier masses will not be heavy for @xmath105 .
the temperature - dependence shown by the @xmath7-electron hall weighting function is then simply a reflection of the carrier mass enhancement that gradually develops as the system is cooled below @xmath73 .
the low - field @xmath7-electron hall weighting function @xmath121 plotted as a function of temperature for cerhin@xmath1 , ceirin@xmath1 , and cecoin@xmath1 .
the solid line shows @xmath121 in a field of 90 koe for cecoin@xmath1 . ]
the substantial field dependence present in the ce-115 hall response offers confirming evidence that field - dependent many - body interactions are responsible for the sizable difference between the ce and la low - t hall anomalies .
an applied magnetic field has a deleterious effect on the heavy - fermion state because it tends to broaden the kondo resonance and shift it below the fermi energy.@xcite the @xmath7-electron contribution to the fs in - turn drops , and the sommerfeld coefficient and the large zero - field effective mass are reduced.@xcite the solid line in fig .
[ fig : hallf ] shows that the @xmath7-electron contribution to the cecoin@xmath1 hall effect is substantially reduced in 90 koe ; the same is true for cerhin@xmath1 and ceirin@xmath1 .
these results are consistent with the large field - induced reduction of the cecoin@xmath1 carrier mass observed in dhva measurements.@xcite the analysis used in sec .
[ ce115 ] to parameterize @xmath32(h , t ) data also shows a link between the ce-115 hall response and kondo interactions .
the h - t scaling analysis used on the ce-115 hall data is similar to the parametrization that can be applied to single - impurity mr data.@xcite the mr of a spin-1/2 kondo system typically follows the h - t scaling expressed by eq .
( [ eqn : hscale ] ) with @xmath95 = 1 , and @xmath96 = t@xmath125 .
the ce-115 @xmath96 values listed in fig .
[ fig : hallf ] are roughly consistent with the kondo temperatures estimated from low - t sommerfeld coefficients.@xcite .
hence , the @xmath126 data and @xmath127 values are consistent with a field - induced suppression in the @xmath7-electron character of the ce-115 fermi - surface states ; further , h - t hall data scaling is consistent with kondo energy scales of a few degrees kelvin .
our analysis bears some resemblance to the two - fluid kondo lattice model proposed by nakatsuji , _ _ et al.__@xcite this model divides a kondo lattice system into a kondo - gas component ( analogous to a kondo - impurity phase ) and a kondo - liquid component ( analogous to a coherent heavy - fermion phase ) , with the temperature - dependent evolution of kondo - lattice properties controlled by the mixing parameter @xmath128 .
the gas phase , characterized by a single - ion kondo scale @xmath33 , dominates the system s properties at high temperatures .
the liquid phase , characterized by the intersite coupling energy scale @xmath129 , begins to influence the system s physical properties at temperatures less than @xmath129 .
a two - fluid analysis@xcite of @xmath37(t ) and @xmath124 data finds that the energy scales in cecoin@xmath1 are @xmath33 = 1.7 k and @xmath129 = 45 k. the model s crossover from local - moment behavior at high temperatures to itinerant heavy - fermion behavior at low temperatures , and the energy scales derived for cecoin@xmath1 , accurately describe the ce-115 @xmath0 data .
it is particularly noteworthy that the increasing importance of the heavy kondo - liquid phase below @xmath129 in the two - fluid model provides a simple explanation for the rise in @xmath130 that occurs in the ce-115 hall data below 50 k.
lastly , we consider the significance of the @xmath131 behavior present in the 115 hall data .
the same t - dependence is present in high - t@xmath26 cuprate data@xcite and some have suggested that this is linked@xcite to a qcp .
the nfl behavior evident in the physical properties of cecoin@xmath1 ( and possibly ceirin@xmath1 ) below @xmath2 5 k might also be related to a qcp . as such , it is tempting to ascribe the novel @xmath65 temperature dependence in these two ce-115 materials to critical spin fluctuations associated with a nearby qcp .
this interpretation appears untenable since the hall angle data of cerhin@xmath1 and the three la-115 compounds all of which show no qcp - related phenomena also exhibit the @xmath132 behavior .
a more credible conclusion is that the quadratic hall - angle temperature - dependence results from the peculiar `` conventional '' hall response intrinsic to the 115 electronic structure .
the hall response of the ce-115 compounds differs markedly from that of most kondo systems .
@xmath119 is dominated by the conventional hall effect rather than that due to skew - scattering .
this comes about because of the complex electronic structure intrinsic to the 115 system .
the field and temperature - dependent variation of the ce-115 hall coefficients below 50 k are consistent with the same kondo interactions that also influence other transport and thermodynamic properties .
these results indicate that conventional transport mechanisms can not always be ignored in interpreting the physical properties of @xmath7-electron systems .
we thank e. d. bauer , j. m. lawrence , and z. fisk for encouragement and useful discussions .
we also thank n. o. moreno for assistance with ppms hall measurements .
work at los alamos was performed under the auspices of the us department of energy . | the in - plane hall coefficient @xmath0 of cerhin@xmath1 , ceirin@xmath1 , and cecoin@xmath1 and their respective non - magnetic lanthanum analogs are reported in fields to 90 koe and at temperatures from 2 k to 325 k. @xmath0 is negative , field - independent , and dominated by skew - scattering above @xmath2 50 k in the ce compounds .
@xmath3 becomes increasingly negative below 50 k and varies with temperature in a manner that is inconsistent with skew scattering .
field - dependent measurements show that the low - t anomaly is strongly suppressed when the applied field is increased to 90 koe .
measurements on larhin@xmath1 , lairin@xmath1 , and lacoin@xmath1 indicate that the same anomalous temperature dependence is present in the hall coefficient of these non - magnetic analogs , albeit with a reduced amplitude and no field dependence .
hall angle ( @xmath4 ) measurements find that the ratio @xmath5 varies as @xmath6 below 20 k for all three ce-115 compounds .
the hall angle of the la-115 compounds follow this t - dependence as well .
these data suggest that the electronic - structure contribution dominates the hall effect in the 115 compounds , with @xmath7-electron and kondo interactions acting to magnify the influence of the underlying complex band structure .
this is in stark contrast to the situation in most @xmath8 and @xmath9 heavy - fermion compounds where the normal carrier contribution to the hall effect provides only a small , t - independent background to @xmath10 |
management of the environment and inventory of natural resources often requires appropriate remote sensing data acquired at specific times on earth locations .
but quite often , good resolution optical images have large damaged areas with lost information due to clouds or shadows produced by clouds . the general statistical approaches for imputation needs to consider the data loss in one of three categories : missing at random data , completely missing at random data , ( meaning that the missing data is independent of its value ) , and non - ignorable missingness , allison(2000 ) .
the last case is the most problematic form , existing when missing values are not randomly distributed across observations , but the probability of missingness can not be predicted from the variables in the model .
one approach to non - ignorable missingness is to impute values based on data otherwise external to the research design , as older images , little et al ( 2002 ) .
merging information acquired at the same time from two or more sensors ( with possible different resolution ) is the core of data fusion techniques .
their historical goals are to increase image resolution , sharpening or enhancement of the output , tsuda et al ( 2003 ) ; ling ( 2007 ) and pohl ( 1998 ) , and their major problem to overcome is the co - registration of the different sources to merge , blum et al ( 2005 ) . in the last decade several adaptations of data fusion techniques for information recovery have been devised to mitigate the effect of clouds on optical images , a classical problem , since 50% of the sky is usually covered by light clouds . in le hegarat - mascle et al .
( 1998 ) contextual modeling of information was introduced in order to merge data from sar ( synthetic aperture radar ) images into optical images , to correct dark pixels due to clouds .
arellano ( 2003 ) used wavelet transforms first for clouds detection , and then to correct the information of the located clouds pixels by merging older image information with a wavelet multiresolution fusion technique .
rossi et al ( 2003 ) , introduced a spatial technique , krigging interpolation , for correction of shadows due to clouds .
shadows and light clouds do not destroy all information , only distort it , but dark clouds or rain clouds produce non - ignorable missingness , a harder problem for interpolation techniques .
another interesting example of incomplete data is the damaged imagery provided by the landsat 7 etm+ satellite after its failure on may 2003 .
a failure of the scan line corrector , a mirror that compensates for the forward motion of the satellite during data acquisition , resulted on overlaps of some parts of the images acquired thereafter , leaving large gaps , as large as 14 pixels , in others .
about 22% of the total image is missing data in each scene . in figure [ zcent ]
we can see two parts of the same landsat 7 image , one with missing information and another almost without loss .
[ zcent ] the landsat 7 science team ( 2003 ) , developed and tested composite products made with the damaged landsat 7 image and a database of older landsat images , using local linear histogram matching , a method that merges pieces of older images that best matches the surroundings of the missing information region , scaramuzza et al 2004 , usgs / nasa 2004 , ( united states geological survey/ national aeronautics and space administration ) . commonwealth of australia ( 2006 ) reported that the composite products appear similar in quality to non damaged landsat 7 images , but masked environment changes , an usual problem in data fusion when one of the sources is temporally inaccurate .
zhang and travis ( 2006 ) developed a spatial interpolation method based on krigging , the krigging geostatistical technique , for filling the data gaps in landsat 7 etm+ imagery , without need of extra information .
they compare their method with the standard local linear histogram matching technique chosen by usgs / nasa .
they show in their case study that krigging is better than histogram matching in targets with radical differences in radiance , since it produces an imputation value closer to the values of the actual neighbors .
a drawback of spatial techniques is that they rely on a neighborhood that could be completely empty of real information .
the algorithms start imputation in the gap contour , where there are many good pixels , and use imputed and good values to create the next pixel value . in center pixels of large gaps , algorithms based only on damaged images will use previously imputed values to generate the next imputation , degrading visual quality and increasing interpolation error . in this paper , we propose three methods based on data fusion techniques for imputation of missing values in images with non - ignorable missingness .
we suppose there is available temporally accurate extra information for the gap scenes , produced by a lower resolution sensor .
this is not a very restrictive hypothesis , since there are many satellite constellations that can provide temporally accurate data with different cameras at different resolutions .
this study proposes three different alternatives for filling the data gaps .
the first method involve merging information from a co - registered older image and a lower resolution image acquired at the same time , in the fourier domain ( method a ) .
the second used a linear regression model with the lower resolution image as extra data ( method b ) . the third method consider segmentation as the main target of processing , and
propose a method to fill the gaps in the map of classes , avoiding direct imputation for the segmentation task ( method c ) .
radiometric imputation is later made assigning a random value from the convex hull made by the neighbor pixels within its class .
all the methods were compared by means of a large simulation study , evaluating performance with a multivariate response vector with four measures : q , rmse , kappa and overall accuracy coefficients .
two of these performance measures ( kappa and overall accuracy ) are designed for the assessment of segmentation accuracy .
the other two , q and rmse , measure radiometric interpolation accuracy .
difference in performance were tested with a manova mixed model design with two main effects , imputation method and type of lower resolution extra data , and a blocking third factor with a nested sub - factor , introduced by the real landsat image and the sub - images that were used .
method b proved to be the best for all criteria .
we are considering images acquired from the same geographic target , with fixed number of bands @xmath0 , and 256 levels of gray per band . damaged and older images have the same support , lower resolution images have different supports , and we suppose that there are @xmath1 pixels in the damaged image for each pixel in the lower resolution image .
let @xmath2 be the damaged image , @xmath3 a lower spatial resolution image , and @xmath4 , an older image acquired with the same sensor that @xmath2 .
let @xmath5 be the gap to be filled , i.e. the set of pixels of @xmath2 with missing values .
the goal is to input values on @xmath5 using available good data from @xmath2 in a small neighborhood of each pixel , the values of @xmath3 and eventually , the values of @xmath4 .
the first step in processing is the re - sampling of the lower resolution image in order to match the support of the three images .
this is done replacing each pixel @xmath6 of @xmath3 by a matrix of size @xmath7 with constant value @xmath8 .
imputation is done in each band separately with the same algorithm , thus the methods descriptions consider the images as one band only .
the image @xmath4 does not need to be co - registered with reference to the damaged image @xmath2 , since they have been acquired by the same sensor , but calibration is indeed necessary to increase merging accuracy .
we suppose a gaussian calibration have been made on @xmath4 per column , taking as pivots the values in the matched column of @xmath2 .
let @xmath9 be the composite image , output from the imputation method a. we define @xmath9 values in the gap as a mixture of high frequencies of the older image and low frequencies taken from the actual but lower resolution image .
@xmath10 with @xmath11 and @xmath12 the ideal high pass and low pass fourier filters .
the filters are regulated by the @xmath13 coefficient , the bigger @xmath13 is the larger the influence of @xmath3 in the imputation .
high pass filters are related to detail , edges and random noise , and low pass filters to structural information .
therefore , method a takes structural information from the low resolution image and detail from the older one .
imputation will be made with the only help of @xmath3 , a temporally accurate image with lower resolution .
we have expanded the lower resolution image to match the support of @xmath2 , replacing each cross grained pixel by a block of @xmath7 pixels with the same radiometric value .
each one of the constant blocks of the expanded @xmath3 image has a matched block in the damaged image @xmath2 , that could be valid ( having all the information ) , or non valid , ( with some loss ) .
we will follow a time series approach now .
let think we have data collected month to month along several years , and we have a missing january data .
it is reasonably to impute that value using a regression model that only involves other january data , and extra data collected on summer that year . in our case
, we have missing information on a location @xmath6 inside a block @xmath14 , we may impute that value using a regression model that only involves data in the same position inside the other blocks of the image ( january data ) , using as regressor variable the values of @xmath3 ( summer data ) .
@xmath15 with @xmath16 .
the coefficients @xmath17 and @xmath18 are estimated by ordinary least squares using only valid blocks
. then @xmath19 @xmath20 where @xmath21 and @xmath22 we define the imputed image @xmath23 as @xmath2 in the pixels with no missing information and the value predicted by the regression in each damaged pixel .
@xmath24 remote sense images are used in a wide range of applied sciences . for many of them , as agricultural and experimental biology , or environmental science , all useful information is contained on a class map , where the classes are characterized by special features under study .
kind of crop , forested or deforested areas , regions with high , medium and low soil humidity , are examples of such features . in this section we will change our point of view by thinking on class maps instead on radiometric images
let suppose we have our damaged image @xmath2 , and a temporally accurate image @xmath3 with lower resolution , and possible different number of radiometric bands .
this is an advantage over radiometric imputation , which need the same spectral properties in both images .
we also suppose that a map of classes @xmath25 has been drawn from the damaged image @xmath2 , using a non supervised method like @xmath0 means , with @xmath0 different classes .
the class of missing information pixels is another class , called class 0 .
the goal is to assign the pixels without radiometric information to one of the @xmath0 classes , and generate a radiometric value for them randomly from the selected class .
there are two main steps in this method 1 .
an enhancement of the initial classification @xmath25 , that could be made or not 2 .
imputation of pixels in the zero class .
[ [ initial - enhancement ] ] initial enhancement + + + + + + + + + + + + + + + + + + + automatic classification is a difficult task , and the imputation method under study heavily depends of the accuracy of the initial map of classes .
so it is important to pay special attention to the coherence of the classes .
we suggest a set of steps to improve class homogeneity as follows 1 .
given @xmath2 construct a map of classes @xmath25 using @xmath0 means , and assign the zero label to the missing information regions .
2 . given the class image @xmath25 , detect the pixels with non homogeneous neighborhood , i.e , pixels that have no neighbor pixels on the same class .
call this set @xmath26 .
3 . for each pixel in @xmath26 , verify its label using a forward mode filter , and a backwards mode filter .
the filters with give a label that is consistent with the mode of the labels in a ( forward or backward ) neighborhood .
1 . if both filters give the same label , update the label of the pixel to match this one .
if the labels are not the same , maintain the original label , and put the pixel in a new set call @xmath27 .
4 . for each pixel in @xmath27 that do not has label zero , we use again radiometric information for updating the label .
1 . take the one step neighbors of the pixels , and compute the arithmetic mean for each class present in the neighborhood .
2 . update the label of the pixel by the label of the class whose arithmetic mean is closest ( in euclidean distance ) to the radiometric value of the pixel .
it is important to note that after this process , some of the pixels of the zero class may have been classified into a real class , only using contextual information .
the other classes should have more defined borderlines . [
[ class - zero - imputation ] ] class zero imputation + + + + + + + + + + + + + + + + + + + + + we suppose now that we have a stable map of @xmath0 classes @xmath25 , with missing information in class zero , and auxiliary information provided by a temporally accurate lower resolution radiometric image @xmath3 .
if @xmath6 is a pixel in class zero , it does not have radiometric information .
we will impute a label class on it with the following algorithm 1 .
let @xmath28 be the smallest neighborhood of @xmath6 that have a pixel with non zero label .
2 . let @xmath29 be the arithmetic mean of the @xmath3 pixels that have positions in class @xmath30 in @xmath28 .
the label of @xmath6 is the label of the class @xmath30 that makes the smallest euclidean distance from @xmath8 to each @xmath29 . @xmath31 after this process , all pixels have a class label , but pixels in the gaps have not been imputed with radiometric information yet .
a pixel s radiometric value is assigned randomly from the values of the convex hull generated by the pixels of a small neighborhood within its class .
impartial imputation assessment is only possible when ground truth is available .
complete simulation of the three types of images involved in the methods ( old , damaged and lower resolution ) will introduce errors beyond the ones produced by the imputation methods , degrading the quality of the assessment .
for this reason , good quality landsat 7 etm+ imagery that have older matched imagery available were selected , and strips similar to the gaps in the slc - off etm+ imagery were cut manually , guaranteeing ground truth to compare with and co - registration between the images .
the four landsat 7 etm+ images selected were quit large , having many different textures in them , like crop fields , mountains and cities , that challenge the imputation methods differently .
landsat 7 etm+ had a lower resolution sensor in its constellation , the mmrs camera from the sac c satellite , whose imagery could be used as extra data . but to control also co - registration problems , lower resolution imagery were simulated with three resolution reduction methods ( rrm ) , by block averaging the ground truth , the congrid method from envi software and shifted block averaging .
block averaging is a crude way of reducing resolution , the congrid method reduces the blocking effect smoothing the output , giving better visual appearance , but block averaging allows to shift the blocks in a controlled fashion , simulating lack of co - registration . in figure [ examples ] we see an example of four matched images , good , damaged , older and lower resolution by block averaging .
we can see a bright spot in the center of the image , which is still present in the damaged one and the lower resolution one , but it is not present in the older one .
database construction details are giving in the subsection [ database ] .
the performance measures ( rmse , q , kappa and overall accuracy ) were modeled as a response vector in a manova mixed model with main effects ( imputation method , resolution reduction method ) , and random nested effects ( image and sub - image ) . subsection [ measures ] introduces the performance measures and subsection [ anova ] includes a detailed description of the manova design . following the manova rejection ,
simultaneous multivariate comparisons were made by ( bonferroni corrected ) two means hotelling tests , and individual anova results with simultaneous comparisons via fisher lsd tests were studied to determine the best method , and the influence in performance of co - registration problems .
we selected four landsat 7 etm+ images with six bands each , acquired before the failure of the slc mirror .
these images have also companion good quality landsat images acquired approximately an year before . in the table [ tab ] , we see some information about the images selected , courtesy of conae - argentina . .landsat
7 etm+ images selected for simulation study , courtesy of conae - argentina .
[ cols="^,^,^,^",options="header " , ] we would like now to study how the location parameters cluster , to identify the methods that are significantly different from each other .
we have some information form the multivariate bonferroni corrected set of hotelling test we made , but now we will see how the different variables cluster the methods .
we will use fisher lsd multiple comparison test as clustering criteria .
other multiple comparison test like tukey lsd and duncan test gave the same type of clustering output .
we will study now the information contained on table [ table6 ] .
c & 0,59 & 192 & a & & & & + a2 & 0,79 & 192 & & b & & & + a1 & 0,79 & 192 & & b & & & + a3 & 0,80 & 192 & & b & & & + b & 0,83 & 192 & & & c & & + c1 & 0,83 & 192 & & & c & & + method & mean kappa & n & & & & & + c & 0,36 & 192 & a & & & & + a2 & 0,65 & 192 & & b & & & + a1 & 0,66 & 192 & & b & & & + a3 & 0,66 & 192 & & b & & & + b & 0,71 & 192 & & & c & & + c1 & 0,72 & 192 & & & c & & + method & mean rmse&n & & & & & + b & 18,09 & 192 & a & & & & + a3 & 21,62 & 192 & & b & & & + a2 & 23,05 & 192 & & b & c & & + a1 & 23,93 & 192 & & & c & & + c1 & 39,30 & 192 & & & & d & + c & 42,67&192 & & & & & e + method & mean q & n & & & & & + c & 0,53 & 192 & a & & & & + c1 & 0,59 & 192 & & b & & & + a2 & 0,82 & 192 & & & c & & + a1 & 0,83 & 192 & & & c & & + a3 & 0,83 & 192 & & & c & & + b & 0,85 & 192 & & & & d & + when considering imputation method , method b and c1 are grouped together by the two measures of good classification , kappa and overall accuracy .
the other two measures , q and rmse separate them into single clusters .
the three versions of method a have been grouped together by all the measures but rmse , which considered method a1 and a3 significantly different .
the two versions of method c , c and c1 , have been distinguished by all the measures .
the multivariate simultaneous comparison distinguished all methods but the versions of method a. and even though , method a1 was considered significantly different from method a3 .
it is interesting to see that this is the way of grouping of the two radiometric measures .
the classification based measures do not distinguish between method b and c1 , or a1 , a2 and a3 . to the naked eye ,
method b reconstruction of figure [ method ] is very different from method c1 reconstruction , and method a1 is sharper than method a3 .
the multivariate simultaneous comparison agreed with this statement , with an experiment - wise type i error @xmath32=0.05 .
now we can come back to the profile analysis of section [ profiles ] . in that section ,
the performance measures means , computed over the sub - images of each image , were plotted as profiles , and the highest value of q , kappa and overall accuracy indicated the best imputation method , and the lowest value of rmse back up the same statement .
method b seemed to be the best of all them .
but the analysis did not have any statistical confidence .
the simultaneous comparisons made reported method b as significantly different in mean performance from all the others , and table [ table6 ] show the mean value of measures q , overall accuracy and kappa the highest of all , and the lowest of the error measure rmse .
these results give confidence to the previous profile analysis .
now we are concerned with the fact that imputation may have reduced performance when the lower resolution image used as extra information is not co - registered accurately .
when simulating the lower resolution image , it was chosen chose block averaging and congrid method as basic resolution reduction methods , and as a third method , the block averaged image was moved slightly ( method rrm2 ) , simulating lack of co - registration .
simultaneous comparisons using fisher lsd ( see table [ table7 ] ) report that the two versions of block averaging are indistinguishably , but rrm2 produced a significantly different mean in all measures , making them worse . in the case of kappa ,
q and overall accuracy , the means are smaller , and in the case of rmse the mean is higher . 2 & 0,74&0,57&30,71 & 0,69 & 384 & a & + 0 & 0,79 & 0,66&27,05 & 0,76 & 384 & & b + 1 & 0,79 & 0,66&26,58 & 0,77 & 384 & & b +
regression models are considered successful models for imputation in a wide range of situations in all applied sciences . in this paper
, we introduced a simple regression model for imputation of spacial data in large regions of a remote sensed image , method b , that had a statistically significant better performance than two other main methods also adapted from the literature , in the frame of a careful simulation study .
multivariate simultaneous comparisons of all imputation methods agreed with the visual inspection of the reconstructed images .
method b reconstruction shows less contrast between the imputed stripes and the non imputed regions , but looses sharpness , and appear slightly blurred .
method a1 and method a3 reconstructions are different , a1 produced more fine detail , but also a lot more contrast between imputed and non imputed regions , and a3 has a smoother appearance , closer to method b reconstruction , also with less detail .
method c and c1 were designed to give good segmentations , despite the large regions without informations , and method c1 does .
pixels radiometric imputation is made choosing at random a a value form the convex hull of pixels of its class in a small neighborhood . without the help of radiometric extra data , radiometric imputation become quite poor .
method c has a preprocessing step that makes the initial segmentation sharper , which increases the difference with the original image , not only on the imputed regions but in all regions .
one of the hypothesis of all methods was the existence of good , co - registered , temporally accurate imagery with possible lower resolution . to test the dependence of the methods of co - registration ,
the block averaged image that act as lower resolution extra data was shifted slightly and the performance of all methods diminished .
this reduction was observed statistically significant with fisher lsd test of simultaneous comparisons , for each performance measure , and globally with a series of hotelling tests ( bonferroni adjusted ) .
the results introduced in this paper are part of the masters thesis in applied statistics of valeria rulloni , at universidad nacional de cordoba .
this work was partially supported by pid secyt 69/08 .
we would like to thank s. ojeda , j. izarraulde , m. lamfri , and m. scavuzzo for interesting conversations leading to the design of the methods .
the imagery used in the simulation section was kindly provided by conae , argentina .
imputation methods and performance measures were computed with envi software .
statistical analysis was made with infostat , unc , provided by prof .
julio di rienzo
. little , r. and rubin , d .
( 2002 ) . statistical analysis with missing data .
john wiley , new york .
allison , p .
multiple imputation for missing data : a cautionary tale .
methods res .
28(3 ) ( 2000 ) 301309 .
le hegarat - mascle , s. , bloch , i. y vidal madjar , d. ( 1998 ) , introduction of neighborhood information in evidence theory and application to data fusion of radar and optical images with partial cloud cover .
pattern recognition , vol 31 : 11 , pp18111823 .
rencher , a. ( 2002 ) .
methods of multivariate analysis.2d edition .
wiley series in probability and statistics .
richards , j. a. y jia , x. ( 1999 ) , remote sensing digital image analysis . an introduction . springer . | imputation of missing data in large regions of satellite imagery is necessary when the acquired image has been damaged by shadows due to clouds , or information gaps produced by sensor failure .
the general approach for imputation of missing data , that could not be considered missed at random , suggests the use of other available data .
previous work , like local linear histogram matching , take advantage of a co - registered older image obtained by the same sensor , yielding good results in filling homogeneous regions , but poor results if the scenes being combined have radical differences in target radiance due , for example , to the presence of sun glint or snow .
this study proposes three different alternatives for filling the data gaps .
the first two involves merging radiometric information from a lower resolution image acquired at the same time , in the fourier domain ( method a ) , and using linear regression ( method b ) . the third method consider segmentation as the main target of processing , and propose a method to fill the gaps in the map of classes , avoiding direct imputation ( method c ) .
all the methods were compared by means of a large simulation study , evaluating performance with a multivariate response vector with four measures : q , rmse , kappa and overall accuracy coefficients . difference in performance
were tested with a manova mixed model design with two main effects , imputation method and type of lower resolution extra data , and a blocking third factor with a nested sub - factor , introduced by the real landsat image and the sub - images that were used .
method b proved to be the best for all criteria . |
it is known that every irrational number @xmath0 can admit an infinite continued fraction ( cf ) induced by gauss transformation @xmath1 given by @xmath2 for @xmath3 and @xmath4 where @xmath5 denotes the integer part of a real number @xmath6 , that is , @xmath7 , if @xmath8 for some @xmath9 .
we set @xmath10 and have the following cf expansion of @xmath11 : @xmath12.\ ] ] the numbers @xmath13 ( @xmath14 ) are called the partial quotients of @xmath11 . in 1845 , gauss observed that @xmath15 preserves the probability measure given by @xmath16 where @xmath17 is any borel measurable set and @xmath18 is the lebesgue measure on @xmath19 $ ] .
the measure @xmath20 is called gauss measure and equivalent with @xmath18 .
continued fractions is a kind of representation of real numbers and an important tool to study the diophantine approximation in number theory .
many metric and dimensional results on diophantine approximation have been obtained with the help of continued fractions such as good @xcite , jarnik @xcite , khintchine @xcite etc , and also the dimensional properties of continued fractions were considered , for example wang and wu @xcite , xu @xcite etc .
the behaviors of the continued fraction dynamical systems have been widely investigated , for example , the shrinking target problems @xcite , mixing property @xcite , thermodynamic formalism @xcite , limit theorems @xcite etc .
hu and yu @xcite were concerned with the set @xmath21 for any @xmath22 and proved that it is 1/2-winning and thus it has full hausdorff dimension .
asymptotic behaviors of the orbits are one of the main theme in dynamical systems . in this paper , we focus on the continued fraction dynamical system @xmath23 and study firstly the denseness of the orbits , and secondly the size of the scrambled set from the sense of hausdorff dimension .
let @xmath24 the orbit of @xmath11 under @xmath15 is dense in @xmath19~~\},$ ] and @xmath25 be the complement of the set @xmath26 in @xmath19 $ ] .
[ thm1 ] let @xmath15 be the gauss map on @xmath27 .
then + @xmath28 @xmath26 is of full lebesgue measure in @xmath19 $ ] .
+ @xmath29 @xmath25 is of full hausdorff dimension .
+ @xmath30 @xmath25 is of the first category in @xmath19 $ ] .
+ @xmath31 @xmath25 and @xmath26 are dense in @xmath19 $ ] .
the second statement @xmath29 of the theorem is a direct corollary of the result in @xcite since the set studied in @xcite is a subset of @xmath25 .
similar properties with this theorem for @xmath32-transformations have been given by @xcite .
for some other dynamical systems , the set of the points with nondense orbits is usually of full hausdorff dimension , see also @xcite , @xcite and references therein .
the chaotic property is a characterization of the asymptotic behaviors between different orbits in dynamical system .
@xcite let @xmath33 be a metric space .
for two points @xmath34 , @xmath35 is a scrambled pair for the map @xmath36 , if @xmath37 and @xmath38 . a subset @xmath39 , containing at least two points , is a scrambled set of @xmath40 , if for any @xmath41 , @xmath42 , @xmath35 is a scrambled pair for @xmath40 .
if a scrambled set @xmath43 for @xmath40 is uncountable , we say that @xmath40 is chaotic in the sense of li - yorke . it is well known that @xcite the surjective continuous transformation on compact metric space with positive topological entropy is chaotic in the sense of li - yorke . even though the topological entropy of the gauss map is infinite , the result above in @xcite can not be applied since the gauss map is not continuous .
we prove that the gauss map is chaotic as the following , indeed , the scrambled set can be very large from the dimensional sense , not just uncountable .
[ thm2 ] let @xmath15 be the gauss map on @xmath27 .
then there exists a scrambled set in @xmath27 whose hausdorff dimension is 1 . as a consequence
, we have the gauss transformation on @xmath27 is chaotic in the sense of li - yorke . the sizes of the scrambled sets have been considered for many dynamical systems from the sense of the measure , dimension , and topology , see also @xcite , @xcite , @xcite , @xcite , @xcite etc . in 1995
, xiong @xcite proved that there exists a scrambled set of @xmath44 of full hausdorff dimension .
the paper is organized as follows . in section 2 ,
we collect and establish some elementary properties of continued fractions which will be used later .
section 3 is devoted to proving theorem [ thm1 ] and theorem [ thm2 ] is proved in section 4 .
in this section , we present some elementary results in the theory of symbolic dynamics and continued fractions which will be used later , see @xcite . let @xmath45 with @xmath46 or @xmath47 .
denote the symbolic space of one - sided infinite sequences over @xmath48 by @xmath49 the symbol @xmath50 is called the @xmath51-th coordinate of @xmath11 .
we assign the discrete topology to @xmath48 and the product topology to @xmath52 . for @xmath53 , the distance @xmath54 is defined by @xmath55 , where @xmath56 denote by @xmath57 the set of all @xmath58-letter words and @xmath59 for the set of all finite words over @xmath48 . for any @xmath60 , @xmath61
, @xmath62 denotes the concatenation of @xmath60 and @xmath63 . the symbol
@xmath64 " means the diameter , the length and the absolute value with respect to a set , a word , and a real number respectively . for @xmath65 with @xmath66 , we write @xmath67 . the shift map @xmath68 is defined by @xmath69 in this subsection , we collect and establish some elementary properties on continued fractions .
we can refer to @xcite for more details .
we denote by @xmath70 $ ] the @xmath58-th convergent of @xmath11 . with the conventions @xmath71
, we have @xmath72 @xmath73 for any @xmath14 and @xmath74 , let @xmath75 be the denominator of finite continued fraction @xmath76 $ ] .
if there is no confusion , we write @xmath77 instead of @xmath75 for simplicity . for any @xmath14 and @xmath74
, we define the @xmath58-th basic intervals @xmath78 by @xmath79 in the case of @xmath80 , @xmath81 for @xmath82 .
the following lemmata are well known .
@xcite[lem0 ] for any @xmath14 and @xmath74 , we have : + @xmath83 @xmath84 ; + @xmath85 @xmath86 @xcite [ lem01 ] for any @xmath14 and @xmath87 , @xmath88 for any @xmath89 , let @xmath90 and @xmath91 jarnik @xcite showed the following : [ lem1 ] for any @xmath92 , @xmath93 and @xmath94 where @xmath95 " denotes the hausdorff dimension .
* lemma 4.1)[lem4 ] let @xmath96 be the gauss transformation .
then there exists a constant @xmath97 such that for any @xmath98 , @xmath99
[ lem2 ] let @xmath100 $ ] be the cf expansion of @xmath101
. then @xmath102 if and only if there exists some word @xmath103 which does not appear in @xmath104 .
on the one hand , if there exists @xmath103 which does not appear in @xmath104 , let @xmath6 be the middle point of the interval @xmath105 , then @xmath106 for any @xmath107
. hence @xmath102 by the definition of the set @xmath25 . on the other hand ,
suppose that @xmath102 .
when @xmath108 , since the cf expansion of every rational number is finite , then there exists @xmath109 such that @xmath110 $ ] , thus the finite word @xmath111 is the required .
now it left to check the case of @xmath112 . by contradiction , suppose that every @xmath103 appears in @xmath104 , then , for any @xmath113 and any @xmath14 , the finite word @xmath114 will appear in @xmath115 .
hence there exists @xmath116 such that @xmath117 , then we obtain @xmath118 by @xmath83 of lemma [ lem0 ] , it follows that @xmath119 therefore @xmath120= \overline{\mathbb{q}^{c}\cap[0,1 ) } \subset \overline{\{t^{k}(x ) : k\geq 0\}}.\ ] ] that is to say , @xmath121 by the definition of the set @xmath26 , which contradicts @xmath102
. then the lemma holds .
\(1 ) in view of lemma [ lem2 ] , we have @xmath122 since the cardinality of @xmath123 is countable , we just need to prove @xmath124 , where @xmath18 is the lebesgue measure on @xmath19 $ ] .
the fact that @xmath15 is ergodic with respect to the gauss measure @xmath20 is equivalent to that @xmath125 for any borel measurable subset @xmath126 of positive measure . since @xmath127
, we have @xmath128 since @xmath129 , the assertion @xmath130 follows , which implies @xmath124 by the equivalence between the lebesgue and gauss measures .
\(3 ) according to ( [ eq.1 ] ) , we only need to prove that @xmath135 is nowhere dense since @xmath123 is countable .
recall that the gauss measure @xmath20 is an ergodic measure with respect to @xmath15 .
let @xmath136 and @xmath137 be the endpoints of the interval @xmath105 and @xmath138 . since the open interval @xmath139 , we have @xmath140 then @xmath141 let @xmath142 be the closure of @xmath135 . since @xmath143 is an open set , thus @xmath144 then @xmath145 by the equivalence between the lebesgue and gauss measures .
hence @xmath135 is nowhere dense .
\(4 ) since @xmath146 by ( 1 ) , we know that @xmath26 is dense in @xmath19 $ ] . it suffices to prove that @xmath25 is dense in @xmath19 $ ] , which can be obtained by the facts that @xmath147 and that the set @xmath148 is dense in @xmath19 $ ] .
fix @xmath46 , we define four mappings @xmath151 , @xmath152 , @xmath153 and @xmath154 which are dependent on the symbol @xmath155 .
actually , the idea for constructing these mappings is inspired by xiong @xcite . in the following , we write @xmath156 to emphasize the dependence of the number @xmath155 of letters .
define a positive integers sequence @xmath164 , that is , @xmath165 .
define @xmath166 by @xmath167 for @xmath168 and @xmath169 .
( 4).define @xmath170 such that for @xmath171 , @xmath172 that is to say , the form of @xmath173 is as follows : @xmath174 suppose that @xmath185 .
note that @xmath186 and @xmath187 by the definitions of @xmath188 and @xmath154 and @xmath189 , the symbol @xmath190 appears infinitely often in any @xmath191 , but it does not appear in any @xmath192 . hence @xmath184 . for any @xmath193 with @xmath194 , denoted by @xmath195 and @xmath196
, we shall prove that @xmath197 is a scrambled pair for the shift .
the proof is divided into two cases according to @xmath198 or @xmath199 .
assume that @xmath200 for some @xmath46 , then there exist @xmath201 with @xmath42 such that @xmath202 and @xmath203 .
since @xmath42 , there exists @xmath204 such that @xmath205 . by the definitions of @xmath151 and @xmath152 ,
the symbol @xmath206 and @xmath207 appear infinitely often in the same location of @xmath208 and @xmath209 respectively . by the definition of @xmath154 ,
hence @xmath206 and @xmath207 appear infinitely often in the same location of @xmath173 and @xmath210 respectively .
that is to say , there exists an increasing sequence @xmath211 such that it follows that @xmath213 on the other hand , according to the maps @xmath151 and @xmath152 , there exists an increasing sequence @xmath214 such that @xmath215 by the definition of @xmath154 , there exists an increasing sequence @xmath216 such that @xmath217 that is , @xmath218 , which implies @xmath219 ( 2 ) . suppose @xmath195 and @xmath196 such that @xmath220 and @xmath199 , then there exist @xmath171 and @xmath221 such that @xmath202 and @xmath222 . by the definitions of @xmath151 , @xmath152 , @xmath223 and @xmath224 ,
we have @xmath247 suppose that @xmath198 , that is , @xmath248 for some @xmath46 , then @xmath249 for any @xmath82 . according to lemma [ lem4.4 ] , @xmath250 there exists an increasing sequence @xmath251 such that @xmath252
. then @xmath253 where the first inequality holds since the interval @xmath254 or @xmath255 is contained in the gap between the point @xmath256 and @xmath257 and @xmath249 for any @xmath82 and the second inequality holds by lemma [ lem4 ] . therefore @xmath258
@xmath259 suppose that @xmath195 , @xmath196 for some @xmath220 and @xmath199 . according to ( 2 ) of remark [ rem4.1 ] and constructions of @xmath260 and @xmath261 , there exists an increasing sequence @xmath262 such that @xmath263 then @xmath264 where the first inequality holds since the interval @xmath265 or @xmath266 is contained in the gap between the point @xmath267 and @xmath268 and the second inequality holds by lemma [ lem4 ] .
therefore @xmath269 recall that @xmath271 for any @xmath46 . since the hausdorff dimension of rational number is zero
, we only need to consider the set @xmath272 .
note that @xmath273 . by the construction of @xmath260 , @xmath274 , we know that @xmath275 for any @xmath46 , then @xmath276 consider a map @xmath278 defined by @xmath279 for @xmath280 . since @xmath231 is continuous and bijective from @xmath149 to @xmath281 and @xmath154 is continuous and bijective from @xmath175 to @xmath282 ,
the map @xmath283 is a continuous bijection on @xmath284 .
proposition [ prop4.1 ] is the corollary of the following lemma , thus we omit the proof of proposition [ prop4.1 ] .
recall that the map @xmath288 @xmath289 satisfies locally @xmath290-h@xmath291lder condition , if there exist a real number @xmath292 and a constant @xmath293 such that , for any @xmath34 with @xmath294 , @xmath295 for a fixed integer @xmath46 and any @xmath14 , we can check that @xmath296 where @xmath297 recall that @xmath164 . for any @xmath14 ,
let @xmath298 , we obtain @xmath299 similarly , @xmath300 for any @xmath301 , let @xmath302 be the finite word by eliminating the terms @xmath303 .
then @xmath304 suppose that @xmath317 .
the interval @xmath318 is contained in the gap between @xmath136 and @xmath137 whether @xmath58 is even or odd . similarly to the case @xmath319 , the interval @xmath320 is contained in the gap between @xmath136 and @xmath137 whether @xmath58 is even or odd . by lemma [ lem4 ] and @xmath321 , @xmath322 * proof of lemma [ lem4.6 ] : * from lemma [ lem0 ] , for any @xmath301 , we have @xmath323 let @xmath285 , there exists @xmath324 such that for any @xmath325 , we have @xmath326 by lemma [ lem0 ] and lemma [ lem01 ] , @xmath327
let @xmath328 .
for any @xmath329 , there exists @xmath58 such that @xmath309 and @xmath310 , where @xmath306,c=[c_1,c_2,\cdots]\in\phi(s_n)\subset [ 0,1)$ ] , then @xmath330 where the last inequality holds by lemma [ lem8 ] .
* acknowledgement . *
the authors thank professor ying xiong for the helpful suggestions .
the work was supported by nsfc 11371148 , guangdong natural science foundation 2014a030313230 , and `` fundamental research funds for the central universities '' scut 2015zz055 and 2015zz127 .
10 f. blanchard , e. glasner and s. kolyada , on li - yorke pairs .
math . , ( 2002 ) , 547:51 - 68 .
h. bruin and v. jimnez lpez , on the lebesgue measure of li - yorke pairs for interval maps .
( 2010 ) , no . 2 , 523 - 560 .
c. faivre , a central limit theorem related to decimal and continued fraction expansion , arch . math .
70(6 ) ( 1998 ) 455c463
. k. j. falconer , fractal geometry , mathematical foundations and applications .
new york : wiley , 1990 . i. j. good , the fractional dimensional theory of continued fractions , proc
. cambridge philos . soc . 37 ( 1941 ) , 199 - 228 . h. hu and y .- l .
yu , on schmidt s game and the set of points with non - dense orbits under a class of expanding maps , j. math . anal .
418(2 ) ( 2014 ) , 906 - 920 . i. kan , a chaotic function possessing a scrambled set with positive lebesgue measure .
92 ( 1984 ) , 45 - 49 .
khintchine , continued fractions .
university of chicago press,1964 .
kleinbock , nondense orbits of flows on homogeneous spaces , ergodic theory dynam .
systems 18 ( 2 ) ( 1998 ) 373c396 .
wang and j. wu , hausdorff dimension of certain sets arising in continued fraction expansions . adv .
218 ( 2008 ) , no . 5 , 1319 - 1339 .
h. wang and j. xiong , chaos for subshifts of finite type .
acta math . sin .
( engl . ser . ) 21 ( 2005 ) , no . 6 , 1407 - 1414 .
h. wang and j. xiong , the hausdorff dimension of chaotic sets for interval self - maps .
topology appl .
153 ( 2006 ) , no .
12 , 2096 - 2103 .
j. wu , a remark on the growth of the denominators of convergents .
147(3)(2006 ) , 259 - 264 . | we prove that there exists a scrambled set for the gauss map with full hausdorff dimension .
meanwhile , we also investigate the topological properties of the sets of points with dense or non - dense orbits . |
krawtchouk polynomials are the simplest finite - discrete polynomials of the askey scheme of orthogonal polynomials .
they play an essential role in the study of problems coming from different branches of modern science .
image analysis based on their moments @xcite , exactly solvable birth and death processes @xcite , perfect qubit transfer in a spin chain with nearest - neighbour interaction @xcite are only a few of such problems solved by employing krawtchouk polynomials .
also in the `` discretisation '' of the quantum - mechanical formulation of the harmonic oscillator problem they play an important role . the non - relativistic explicit formulation of this problem in the infinite continuous configuration space is well known within the canonical approach .
the wavefunctions of stationary states are expressed in terms of the hermite polynomials both in momentum and position representation @xcite . under the assumption that the quantum world is quite different from the classical one , various approaches and methods
allow to construct a number of exactly solvable mathematical models of the quantum harmonic oscillator , where the wavefunctions are expressed in terms of other known orthogonal polynomials .
also the krawtchouk polynomials allow one to construct a number of interesting quantum harmonic oscillator models in finite - discrete configuration space . at present , the known models using krawtchouk polynomials are the @xmath1 model @xcite and the @xmath6 model @xcite for the finite quantum harmonic oscillator . in the case of @xmath1 dynamical symmetry , the wavefunctions of the model are defined through the special case of krawtchouk polynomials with a fixed parameter @xmath2 . for the ` supersymmetric ' model with @xmath6 dynamical symmetry , the wavefunctions are expressed by means of general krawtchouk polynomials with parameter @xmath7 . in the present work , we present a new oscillator model with wavefunctions of even and odd states also expressed by krawtchouk polynomials with fixed @xmath2 , but in a very different way : the polynomials that appear are @xmath3 and @xmath4 respectively .
the dynamical symmetry is formally called a @xmath0 deformation of the @xmath1 lie algebra .
the current paper is structured as follows ; in section 2 , the @xmath0 deformation of the @xmath1 lie algebra and its representations are introduced . then
, this algebra is used for the construction of a model of the finite - discrete quantum harmonic oscillator . in section 4
, we discuss explicit expression of the wavefunctions of the constructed model both in momentum and position representations .
finally , section 5 covers a brief summary and discussion of the obtained results .
our starting point is the definition of the formal algebra @xmath1 under the @xmath0 deformed symmetry by its basis elements @xmath8 , @xmath9 and @xmath10 .
the algebra @xmath5 is a unital algebra with basis elements @xmath8 , @xmath9 , @xmath10 , @xmath11 and @xmath12 subject to the following relations : * the operator @xmath11 commutes with all basis elements .
* @xmath12 is a parity operator satisfying @xmath13 and @xmath14=\p j_0-j_0\p=0 , \qquad \{\p , j_\pm\}=\p j_\pm + j_\pm \p= 0 . \label{p}\ ] ] * the @xmath1 commutation relations are @xmath0 deformed as follows : @xmath15 = \pm j_\pm , \label{j0j+ } \\ & [ j_+ , j_- ] = 2j_0 \left ( { \c\p-1 } \right ) .
\label{j+j-}\end{aligned}\ ] ] the next step is to describe representations of this algebra , by defining the actions of the basis elements @xmath8 , @xmath9 and @xmath10 in such a way that they are compatible with the commutation relations and .
it is easy to verify that an action exists on the usual standard basis vectors @xmath16 ( with @xmath17 being an integer and @xmath18 ) , which is reminiscent of but different from the common @xmath1 algebra action .
this is given by : @xmath19 note that @xmath17 must be integer ( and not half - integer ) in order to be a representation ( one way to see it is that one should still have @xmath20 and @xmath21 ) .
the algebra representation is completed by the action of the extra operators @xmath11 and @xmath12 : @xmath22 despite the fact that this algebra is at first sight very close to @xmath1 , its behaviour is quite different .
for example , the representations matrices of @xmath23 in the @xmath5 algebra never coincide with the representation matrices of these elements in the @xmath1 algebra , except for the trivial representation @xmath24 .
note also that there is no `` deformation parameter '' that takes the @xmath5 algebra to the @xmath1 algebra for some limit .
we now consider an oscillator model based on the algebra introduced , and choose the hamiltonian , position and momentum operators in a way that is similar to the non - deformed @xmath1 one - dimensional oscillator case @xcite , i.e. @xmath25 one can easily check that these three operators satisfy the heisenberg equations : @xmath26 = - \mi\hat p,\quad \left [ { \hat h,\hat p } \right ] = \mi\hat q.\ ] ] then , one can also observe that @xmath27 so the spectrum of @xmath28 is equidistant but finite : @xmath29 from the actions and one obtains for the position operator @xmath30 that @xmath31 when @xmath32 is even , and @xmath33 when @xmath34 is odd .
one can do the same computations for the operator @xmath35 . in matrix form ,
both position and momentum operators @xmath36 and @xmath37 take the following form : @xmath38 and @xmath39 with matrix elements @xmath40 the purpose is now to study the eigenvalues of the position ( and momentum ) operator in this model . in other words : we need to find eigenvalues and eigenvectors of these matrices @xmath41 and @xmath42 .
it turns out that these eigenvectors can be constructed in terms of symmetric krawtchouk polynomials . in general ,
krawtchouk polynomials @xmath43 of degree @xmath44 ( @xmath45 ) in the variable @xmath46 , with parameter @xmath7 , are defined in terms of the @xmath47 hypergeometric function of argument @xmath48 as follows @xcite : @xmath49 their orthogonality holds for discrete values of @xmath46 , and is given by : @xmath50 where , @xmath51 here , @xmath52 are pochhammer symbols .
we can introduce the following orthonormal krawtchouk functions : @xmath53 let us also mention that for @xmath2 , the corresponding krawtchouk polynomials are called `` symmetric '' , since @xmath54 in previous examples , the construction of a quantum oscillator model in term of orthogonal polynomials is based on the differential ( or difference ) equation , the solution of which is expressed in terms of that polynomial . in this sense , the @xmath1 one - dimensional krawtchouk oscillator model @xcite is based on the difference equation for krawtchouk polynomials ( * ? ? ?
* ( 9.11.5 ) ) , whereas the @xmath6 one - dimensional krawtchouk oscillator model @xcite is based on the forward and backward shift operator equations for them ( * ? ? ?
* ( 9.11.6 ) , ( 9.11.8 ) ) . for our current construction ,
the essential point here is a set of new difference equations for symmetric krawtchouk polynomials ( @xmath55 ) .
these new relations involve a pair of symmetric krawtchouk polynomials shifted in variables @xmath56 or @xmath57 and parameters @xmath58 and @xmath59 .
the symmetric krawtchouk polynomials satisfy the following difference equations : @xmath60 in these equations , the integer @xmath44 is arbitrary but should be taken appropriately ( e.g. for the first krawtchouk polynomial , the degree @xmath61 should belong to @xmath62 )
. * proof . *
the validity of both equations is related to a higher level hypergeometric series than @xmath63 , namely @xmath64 . for this purpose
, we can use the following relation ( * ? ? ? * ( 39 ) ) : @xmath65 using this equation for the @xmath63 expressions in leads to the following identity to verify : @xmath66 now , to prove the correctness of it is sufficient to simplify the pochhammer symbols in the expansion of the @xmath64 s in the right hand side , collect terms , and the left hand side appears automatically .
the proof of is similar . in this case , using leads to the following identity : @xmath67 the verification of this again depends on elementary manipulations of pochhammer symbols .
now , our aim is to construct eigenvalues and eigenvectors of @xmath41 .
this is rather technical , but is essentially based on and . for this purpose , we construct the matrix @xmath68 in the following proposition .
let @xmath69 be the tridiagonal @xmath70-matrix and let @xmath71 be the @xmath70-matrix with matrix elements as follows .
for an even row index @xmath72 ( @xmath73 ) , let @xmath74 for an odd row index @xmath75 ( @xmath76 ) , let @xmath77 then @xmath68 is an orthogonal matrix : @xmath78 furthermore , the columns of @xmath68 are the eigenvectors of @xmath41 , i.e. @xmath79 where @xmath80 is a diagonal matrix containing the eigenvalues @xmath81 of @xmath41 : @xmath82 [ propu ] * proof . *
the proof of this proposition is quite similar to ( * ? ? ?
* proposition 3 ) . the matrix relations follow directly from the two difference equations and .
the orthogonality of the matrix @xmath68 is slightly tricky , and does not follow ( as one would expect ) directly from the orthogonality of symmetric krawtchouk polynomials . instead , just as in the proof of proposition 2
we need to go to higher level hypergeometric series . reexpressing the @xmath63 s in terms of @xmath64 s by means of
, the corresponding @xmath64 s can be related to hahn polynomials @xcite with parameters @xmath83 .
the orthogonality of these hahn polynomials leads to : @xmath84 2\delta_{n , n ' } , & \hbox{if } n = n , \end{array } \right .
\label{korth}\ ] ] for @xmath85 .
the extra factor 2 for @xmath86 appears because for that case one has to take a limit of the squared norm of the hahn polynomial , so one gets another factor than in the other cases where one can just substitute @xmath83 .
equation is the underlying equation proving the orthogonality of the columns of @xmath68 .
the extra factor 2 is also responsible for the extra factors @xmath87 when @xmath88 .
this proposition gives us essentially the eigenvalues @xmath89 of the position operator @xmath30 .
there are @xmath90 distinct eigenvalues and they are given by : @xmath91 or in general : @xmath92 in figure 1 , we present a plot of the spectrum of the position operator for the @xmath1 , @xmath6 and @xmath5 oscillator models .
one observes that the spectrum of the position operator is quite different compared to the finite - discrete oscillator models already known .
the main difference is the distribution of positions . for the @xmath1 oscillator model ,
the positions are equidistant ( @xmath93 ) .
the distribution of positions for the @xmath6 model is not equidistant , due to @xmath94 . in that case , the interval covering the positions is smaller than the corresponding one for the @xmath1 case .
comparing the @xmath5 case with two others , we observe that some of its properties overlap with @xmath1 and others with @xmath6 .
the interval covering the positions is the same for the @xmath1 and the @xmath5 model : @xmath95 .
both the @xmath6 and @xmath5 oscillator models have a non - equidistant distribution of the position eigenvalues .
the main feature of the @xmath5 oscillator model is that the position values are concentrated near the border of the inteval @xmath96 $ ] , whereas near the middle of the interval the distribution is sparse . , ( a ) in the case of the @xmath1 model , ( b ) in the case of the @xmath6 model and ( c ) in the case of the @xmath5 model.,scaledwidth=70.0% ]
the position wavefunctions of the @xmath5 finite - discrete oscillator model are defined in a similar way as in @xcite .
their explicit expression is computed through the overlap between the @xmath30-eigenvectors @xmath97 and the @xmath28-eigenvectors @xmath98 . let us denote them by @xmath99 , where @xmath100 , and where @xmath89 assumes one of the discrete values of @xmath101 with @xmath102 .
simple computations show that @xmath103 where @xmath104 are the matrix elements of @xmath41 .
then , one can find that for the even case @xmath32 and positive values of position , the wavefunctions have the following expression : @xmath105 whereas , for the odd case @xmath34 and positive values of position , we have @xmath106 one can extend these computations and obtain similar expressions for zero and negative values of the positions . in figure 2 we plot the discrete wavefunctions and for some large @xmath17-value , and for some @xmath44-values . due to the unique property of the position eigenvalues to be located near the border of the interval @xmath96 $ ] , the behaviour of the wavefunctions also reflects this peculair distrubution . [
cols="^,^ " , ] we can define momentum wavefunctions @xmath107 in a similar way . their explicit expression is computed through the overlap between the @xmath108-eigenvectors @xmath109 and the @xmath28-eigenvectors @xmath98 . in this case , the simple relation between @xmath42 and @xmath41 shows that @xmath110 then , in accordance with equations and , one finds that for the even case @xmath32 and for positive values of momentum , the wavefunctions have the following expression : @xmath111 whereas for the odd case @xmath34 and positive values of momentum , we have @xmath112 these computations can easily be extended for zero and negative values of the momentum .
both wavefunctions @xmath113 and @xmath114 are related by the so called @xmath0 symmetric krawtchouk transform : @xmath115 where , @xmath116 .
the matrix elements @xmath117 can be computed by using .
despite the fact that the non - relativistic one - dimensional quantum oscillator in the canonical approach has a nice explicit solution in terms of hermite polynomials , there are still attempts to construct new oscillator models inspired by different starting points .
one of these starting points is the assumption of discrete space , in which case one is looking for oscillator models with a discrete position spectrum ( which could be finite of infinite discrete ) .
the most popular finite discrete oscillator model @xcite is based on the @xmath1 algebra , and uses krawtchouk polynomials . in current paper ,
we have constructed a new model for the one - dimensional oscillator , with wavefunctions expressed in terms of the krawtchouk polynomials with parameter @xmath2 ( the symmetric krawtchouk polynomials ) . in order to construct a solvable oscillator model ,
we have first defined a new algebra , formally called the @xmath1 algebra under the @xmath0 deformation .
the representations of this algebra were given and are similar to representations of @xmath1 , but the action of the raising and lowering operators @xmath9 and @xmath10 differs considerably . as a consequence , when considering the oscillator model related to this algebra , the matrix representations of position and momentum operators is also different . in the standard @xmath98 bases , the position operator is tridiagonal , and we have shown that its eigenvalues and eigenvectors involve krawtchouk polynomials .
the matrix relation for the eigenvalue problem is equivalent to certain finite difference equations for krawtchouk polynomials .
the resulting model is unique in the sense that it does not generalize any other existing oscillator model .
peculiar properties are the non - equidistant distribution of the position values as well as the behaviour of the wavefunctions around zero and near the border of the maximum and minimum eigenvalues .
eij kindly acknowledges support from research grant * eif-2012 - 2(6)-39/08/1 * of the science development foundation under the president of the republic of azerbaijan . | we define a new algebra , which can formally be considered as a @xmath0 deformed @xmath1 lie algebra . then , we present a one - dimensional quantum oscillator model , of which the wavefunctions of even and odd states are expressed by krawtchouk polynomials with fixed @xmath2 , @xmath3 and @xmath4 . the dynamical symmetry of the model is the newly introduced @xmath5 algebra .
the model itself gives rise to a finite and discrete spectrum for all physical operators ( such as position and momentum ) . among the set of finite oscillator models it is unique in the sense that any specific limit reducing it to a known oscillator models does not exist . |
the present simulations have been performed at various @xmath3 values and lattice volumes at @xmath2 ( table [ tab1 ] ) .
the effective lattice resolution ranges from @xmath4 gev ( @xmath5 ) down to @xmath6 gev ( @xmath7 ) while the ratio @xmath8 varies from 0.83 to 0.55 , corresponding to sea quarks that are slightly heavier than the strange quark and of about one quarter of its mass , respectively .
in addition , results from quenched reference simulations at @xmath9 ( @xmath10 gev ) and @xmath11 ( @xmath12 gev ) are presented .
-.6 cm ? =
? .simulation parameters . [ cols="<,>,>,>,>",options="header " , ] -.7 cm -.15 cm = 7.5 cm -1.2 cm in quenched simulations , finite size effects on the scalar glueball are small @xcite as long as @xmath13 where @xmath14 denotes the torelon mass .
when including sea quarks , the glueball might break up and decay into a single torelon @xcite through an intermediate mesonic state .
the expected torelon masses @xmath15 are included into fig . [ fig4 ] for @xmath16 ( dashed error bars ) .
contrary to the quenched case , we indeed find significant fse at @xmath17 ( and probably @xmath18 ) .
in general , the large volume data is consistent with quenched results .
the @xmath19 mass appears to exceed the @xmath20 reference value but agrees with the continuum extrapolated result .
we acknowledge support by the dfg ( grants schi 257/1 - 4 and schi 257/3 - 2 ) , the eu ( contracts sc1 * -ct91 - 0642 , chrx - ct92 - 0051 and chbg - ct94 - 0665 ) and pparc ( grant gr / k55738 ) . | we present results on the static potential , and torelon and glueball masses from simulations of qcd with two flavours of dynamical wilson fermions on @xmath0 and @xmath1 lattices at @xmath2 . |
sne ia have been considered as thermonuclear explosions of a co wd .
they play important roles in astronomy as a standard candle to determine the cosmological parameters ( e.g. , * ? ? ?
* ; * ? ? ?
* ) and as major sources of iron group elements ( e.g. , * ? ? ?
* ) . however , their progenitor systems and explosion mechanisms are still uncertain ( e.g. , * ? ? ?
* ; * ? ? ?
a merger of co wds has been considered as one of the potential progenitors of sne ia , which is called the double degenerate ( dd ) scenario @xcite .
a binary of co wds loses its angular momentum by emitting gravitational waves and finally merges .
when the total mass exceeds the critical mass(@xmath1 , * ? ? ?
* ) for the carbon ignition in the center , such a system has been proposed to give rise to an sn ia explosion .
some observational studies reported that neither surviving companions ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) nor signatures of them ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) were detected in some well - studied sn ia remnants and nearby sne ia , but recently there are several examples which possibly show these signatures as well both in sne with special properties @xcite and in a normal sn ia @xcite .
these studies suggest that at least a part of sne ia would originate from either the dd scenario or the @xmath2(angular momentum)-loss induced `` delayed '' carbon ignition in the single degenerate ( sd ) scenario ( e.g. , * ? ? ?
after @xcite firstly performed three dimensional ( 3d ) hydrodynamical simulations of co wd mergers with their sph code ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , similar merger simulations have been performed by various groups ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
these studies showed that a co wd merger could lead to an sn ia if several conditions are satisfied .
@xcite simulated mergers of co wds , both of the binary members have masses @xmath3 , with their sph code .
they found that the secondary wd violently accreted onto the primary and carbon detonation would occur during the merger .
they showed that detonation wave propagated through the primary and the primary finally exploded as an sn ia .
they called this explosion mechanism `` the violent merger scenario . '' as the mass ratio @xmath4 , where @xmath5 is the primary mass and @xmath6 is the secondary mass , approaches unity , the mass accretion becomes more violent and the carbon detonation occurs more easily .
therefore , the mass ratio is an important parameter for the violent merger scenario .
@xcite investigated the critical mass ratio @xmath0 above which the violent merger scenario is realized , and found that @xmath7 for @xmath8 .
@xcite studied mergers of wds over a wide range of masses , i.e. , @xmath9 .
they found that dynamical carbon burning did not occur in all their models .
however , their numerical resolution was low ( @xmath10 particles per star ) . @xcite performed sph simulations of co wd mergers over a wide range of masses @xmath11 and with a resolution of 500,000 particles per @xmath12 , i.e. , @xmath13 .
they investigated whether or not dynamical carbon burning would occur during the merger . @xcite and @xcite showed that the numerical resolution is another important factor in studying the violent merger scenario , and @xcite suggested that it should be @xmath14 ( see also * ? ? ? * ) . assuming critical mass ratios for the violent merger scenario based on the result of @xcite
, @xcite estimated the brightness distribution of sne ia and suggested that it could be qualitatively consistent with what is observed @xcite . because the critical mass ratio could have influence on the distribution of sn ia brightness , detailed investigation of the critical mass ratio
is required to verify the violent merger scenario .
recently , @xcite reported that the central system of the bipolar planetary nebula ( pn ) henize 2 - 428 might be a super - chandrasekhar dd pair .
they estimated the combined mass as @xmath15 and the mass ratio as @xmath16 . although there are some negative arguments against their conclusion ( e.g. , * ? ?
* ) , henize 2 - 428 is a candidate progenitor for the violent merger scenario , if the interpretation by @xcite is correct .
while @xcite have already simulated a set of primary and secondary masses which cover a plausible parameter space for henize 2 - 428 , their mass grids were too coarse .
in this paper , we perform some additional ( much finer grids of the masses ) sph simulations of co wd mergers to accurately obtain the critical mass ratio for the violent merger scenario .
we try to derive the relation between @xmath5 and @xmath0 , and compare it with previous studies .
we also use our results to discuss the final fate of henize 2 - 428 .
this paper is organized as follows . in section [ method ]
, we briefly describe our numerical method .
our results are shown in section [ results ] and we discuss them in section [ discussion ] .
we summarize our findings in section [ summary ] . .
vertical dashed lines show the time when the first detonating particle appears .
, width=302 ] figure [ fig3 ] shows the density and temperature of the particle having smallest @xmath17 ratio in each merger simulation , where @xmath18 and @xmath19 are carbon burning and dynamical timescales , respectively , as defined below .
colors and shapes of symbols indicate the total mass , @xmath20 , and primary mass , @xmath5 , respectively . a black solid line shows the line of @xmath21 .
our sph simulations of co wd mergers are essentially the same as those of @xcite .
we used the octree on opencl ( otoo ) code , which was developed for various particle simulations for astrophysical fluid phenomena @xcite .
the equation of state ( eos ) is that of @xcite .
initial setups are also the same as those in @xcite , which refer to @xcite and @xcite .
our co wd models have an uniform composition of @xmath22 carbon and @xmath22 oxygen in mass . @xcite
simulated mergers of co wds whose masses were 0.5 , 0.6 , 0.7 , 0.8 , 0.9 , 1.0 , and @xmath23 . in this paper , we performed additional simulations of co wd mergers to accurately determine the critical mass ratio @xmath0 to realize the violent merger scenario .
we add @xmath24 0.725 , 0.75 , 0.775 , 0.825 , 0.85 , and @xmath25 for each @xmath26 0.75 , 0.8 , 0.9 , 1.0 , and @xmath23 , because @xmath0 seems to be in this mass range ( see figure 5 of * ? ? ?
the numerical resolution is @xmath27 per solar mass @xmath28 .
our sph scheme derives the temperature of a particle from the density and internal energy through the eos .
the estimations of density and internal energy have some numerical noises , and the derived temperature is also fluctuated @xcite . because nuclear burning rate is sensitive to the temperature
, it might strongly enhance the fluctuated temperature .
therefore , we firstly performed our simulations without nuclear reactions , as in our previous work @xcite , and determined the critical mass ratio for the violent merger scenario .
then , we added simulations including nuclear reactions in order to confirm that the inclusion of nuclear burning in our sph simulation increases the temperature high enough to really satisfy the detonation condition for the models above the critical mass ratio ( see section [ nuclear_burning ] ) .
figure [ fig1 ] shows time evolutions of the orbital separation and the maximum temperature in the case of @xmath29 .
they indicate that the temperature increases drastically when the secondary is completely disrupted and accretes onto the primary , which is consistent with previous studies .
black vertical dashed lines show the time when our detonation condition , which is described below in section [ condition ] , is first fulfilled .
figure [ fig2 ] depicts the density and temperature profiles of the equatorial plane at the time which is indicated by the black dashed lines in figure [ fig1 ] .
the profile and morphology are consistent with the similar cases in @xcite . .
a black cross indicates the first detonating particle .
, width=264 ] .
a black cross indicates the first detonating particle . , width=264 ] for the violent merger scenario , it is crucial whether or not carbon detonation occurs during a merger .
although @xcite used the results of @xcite as the detonation condition , the treatment is still controversial . in this paper
, we judge it from the condition to trigger the dynamical carbon burning .
we obtain the dynamical timescale ( e.g. , * ? ? ?
* ) , @xmath30 and the carbon burning timescale , @xmath31 for each sph particle . here ,
@xmath32 and @xmath33 are particle s density and temperature , respectively , @xmath34 is the specific heat at constant pressure of a particle , @xmath35 is the gravitational constant , and @xmath36 is the energy generation rate of carbon burning ( see equation ( 6 ) of * ? ? ?
* ) . when there are any particles that satisfy @xmath37 , carbon ignites dynamically .
we regard this as the detonation condition . if @xmath37 , the temperature increases more rapidly than decreases by adiabatic expansion and accelerates carbon burning .
the temperature would increase high enough to satisfy the detonation condition , such as the results of @xcite .
thus , we consider that the condition of the dynamical carbon burning ( @xmath37 ) is a plausible detonation condition in the case without nuclear reactions .
we confirm this later by including carbon burning in our sph code in section [ nuclear_burning ] .
as described in section [ method ] , temperature is possibly affected by numerical noises in our sph simulations . in order to minimize the influence of temperature noise in judging the condition for carbon detonation
, we examine whether the particle continuously satisfies the above condition at least for a dynamical timescale ( @xmath38 s at @xmath39 g @xmath40 ) .
if so , we assume that dynamical carbon burning is definitely initiated and detonation occurs .
ratio for each co wd merger model .
colors of symbols indicate the total mass of the system .
shapes of symbols do the mass of the primary .
filled squares are the @xmath23 primary , filled circles @xmath41 , filled triangles @xmath42 , filled inverted triangles @xmath43 , filled diamonds @xmath44 , open diamonds @xmath45 , open squares @xmath46 , open circles @xmath47 .
symbols surrounded by black frames indicate @xmath48 ( filled inverted triangle ) and @xmath49 ( filled diamond ) , respectively .
a black solid line indicates @xmath50.,width=302 ] for example , mergers of @xmath48 and @xmath49 , which are surrounded by black frames in figure [ fig3 ] , both satisfy @xmath37 .
however , in the case of @xmath49 , there are no particles which continuously satisfy the above condition for a dynamical timescale .
we thus regard that detonation does not occur in the merger of @xmath49 . in the case of @xmath48 , on the other hand , three particles keep @xmath37 for longer than @xmath19 .
therefore , we regard that carbon detonation does occur in the case of @xmath48 and eventually the system leads to an sn ia in the violent merger scenario .
we plot our numerical results of violent merger scenario in figure [ fig4 ] . here , the filled circles denote models which satisfy our condition for carbon detonation while the crosses do models which do not .
the black dashed line indicates @xmath51 .
figure [ fig4 ] shows that carbon detonation occurs only in the models whose primary and secondary masses are more massive than @xmath52 , which is consistent with the results of @xcite .
we derived approximate formulae of the critical mass ratio , @xmath0 , as a function of the primary mass , @xmath5 , and plot them in figure [ fig5 ] , both of which are on logarithmic scales .
the red solid line is for an upper bound , while the blue solid line is for a lower bound .
they are represented by @xmath53 @xmath54 respectively .
considering the lowest mass ratios in which the model satisfies our condition for carbon detonation , we derive equation ( [ upper_bound ] ) . on the other hand , we derive equation ( [ lower_bound ] ) from the highest mass ratios in which the model does not satisfy our condition for carbon detonation .
if a sizable helium layer exists on the co wd , helium detonation could occur during merger and induce carbon detonation ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) . in this helium - ignited violent merger scenario ,
@xmath0 could be smaller than equations ( [ upper_bound ] ) and ( [ lower_bound ] ) .
but such a study needs much finer resolution because the occurrence of helium detonation depends strongly on the mass of the helium layer ( e.g. , * ? ? ?
we leave such a systematic study to our future work .
@xcite obtained @xmath55 for @xmath56 .
our results show a different value of @xmath57 for @xmath56 .
this discrepancy could arise from the differences in ( 1 ) the initial conditions , ( 2 ) numerical resolution , and ( 3 ) inclusion of nuclear burning .
we discuss these three effects in this order . ) , while a blue solid line shows a lower bound represented by equation ( [ lower_bound ] ) .
we also present the results including carbon burning .
red and blue dashed lines correspond to the upper and lower bounds for the violent merger scenario in the case with carbon burning , respectively .
the critical mass ratios adopted by @xcite are also depicted by a black solid , dashed , and dash - dotted lines .
see text for more details.,width=302 ] @xcite used non - spinning wds and set them at the initial separation where the roche lobe overflow ( rlof ) just started in the sense of eggleton s ( 1983 ) approximation . on the other hand , we assume that the wds spin synchronously with the orbital motion , and set them at the initial separation large enough to avoid the rlof .
we decrease the separation until the rlof starts @xcite . for the initial condition of @xcite , as mentioned in @xcite , the secondary accretes onto the primary more violently than for the case of synchronously spinning wds . as a result ,
carbon detonation would occur easily .
we examined the merger of @xmath58 wds in the similar initial condition to that of @xcite .
we thereby found that there exist 16 particles which satisfy the condition for carbon detonation in this case , while there is no such a particle for our initial condition ( see figure [ fig3 ] ) .
although no definite conclusion has been reached yet , several numerical studies suggest that compact wd binaries would reach synchronization before their merging due to angular momentum dissipation by tidally excited gravity waves ( e.g. * ? ? ?
* and references therein ) . numerical resolution might also cause the difference .
@xcite reported that the maximum temperature and the minimum @xmath59 ratio did not converge in the range of @xmath60 .
this is because smaller hot regions can be resolved as the numerical resolution becomes higher .
several previous studies also indicated that numerical resolution would affect their results ( e.g. , * ? ? ?
* ; * ? ? ?
* ) . in this paper
, we performed several simulations with higher numerical resolutions ( @xmath61 , @xmath62 ) than our standard ones ( @xmath63 ) , for @xmath64 , @xmath65 , and @xmath66 wds .
our numerical results are summarized in figure [ fig6 ] .
the maximum temperature still increases and the minimum @xmath59 ratio decreases except for @xmath67 as the numerical resolution increases .
similar trends were also reported in @xcite , @xcite , and @xcite .
these trends indicate that our numerical results have not been fully converged yet
. therefore , our critical mass ratio could be slightly lower than the red solid line in figure [ fig5 ] .
ratio on the numerical resolution .
the horizontal axis is the number of sph particles per solar mass .
the shapes and colors of symbols have the same meanings as those in figure [ fig3].,width=302 ] inclusion of nuclear reactions could also make the difference between theirs and ours .
@xcite found that particles with high temperature ( @xmath68 k ) did not exist in the case of @xmath69 without nuclear reactions . on the other hand , they existed in the case with nuclear reactions . in order to evaluate the effect of nuclear energy release , we performed additional simulations including nuclear burning for some models . in these simulations , we include only carbon burning , i.e. , @xmath70 reaction , to avoid large computational cost for solving a reaction network .
the formulation of the reaction rate is the same as equation ( 6 ) of our previous work @xcite .
because the initiation of carbon detonation is judged by this reaction , such a simple treatment of nuclear reactions would be sufficient for our purpose .
electron screening effect is negligibly small in the region of @xmath71@xmath72 g @xmath40 and @xmath73 k , where carbon detonation would occur , so we do not include this effect .
figure [ fig7 ] shows the comparison between simulations with and without carbon burning for the case of @xmath74 .
the red lines indicate the case with carbon burning , while the black ones do the case without .
although the orbital evolution ( figure [ fig7](a ) ) is almost the same , the maximum temperature ( figure [ fig7](c ) ) realized in the case with carbon burning is higher than the case without during the merger phase .
this difference comes from the nuclear energy release ( figure [ fig7](b ) ) .
we confirmed that the inclusion of carbon burning increase the maximum temperature .
figure [ fig8 ] summarizes the density and temperature of a particle which has the smallest @xmath59 ratio for all the cases we recalculated including nuclear burning .
the symbols surrounded by black frames are the results of the cases with carbon burning .
red lines indicate the case with carbon burning , while black ones do the case without .
( a ) the orbital evolution , ( b ) the energy release rate of carbon burning , and ( c ) the time evolution of maximum temperature .
, width=302 ] since our simulations include only carbon burning , the effect of nuclear reactions would be underestimated . to check this effect
, we perform post - processing calculations , using an @xmath75-chain reaction network containing 13 species from @xmath76he to @xmath77ni @xcite .
figure [ fig9 ] shows the results for a particle which has the highest temperature in the case of @xmath74 .
this calculation was done for a period in which carbon burning continued .
the red lines indicate the case with the @xmath75-chain reaction network , while green lines do the case with only carbon burning . because we find essentially no difference between them , our simulations including only carbon burning are appropriated to estimate the effect of nuclear burning , and being consistent with the previous studies using @xmath75-chain networks @xcite , at least until the initiation of detonation .
thus , the inclusion of carbon burning lowers @xmath0 to realize violent merger - induced explosion if we use the same criterion as in section [ critical ] . in figure [ fig10](a ) ,
green triangles are the models which satisfy our detonation condition with carbon burning , although they do not satisfy the condition without .
however , our detonation condition in section [ critical ] is considered under the situation without nuclear burning , so it would be no longer valid in the case with carbon burning .
therefore , for comparison with the results of @xcite , we adopt the same detonation condition as theirs , i.e. , the detection of particles having @xmath78 g @xmath40 and @xmath79 k @xcite .
the results are presented in figure [ fig10](b ) .
although the boundary of figure [ fig10](b ) is slightly lower than that of figure [ fig4 ] , the difference between them is small .
thus , we confirm that the detonation condition adopted in section [ condition ] is plausible for the case without nuclear reactions . , but for the case with carbon burning .
the symbols surrounded by black frames are the case with carbon burning while not surrounded are the case without .
black dashed line is the demarcation of detonation condition by @xcite .
, width=302 ] the approximated lines of @xmath80 in the case with carbon burning are also presented in figure [ fig5 ] , where the red ( blue ) dashed line indicates an upper ( lower ) bound .
the value of @xmath0 with carbon burning is slightly lower than that without carbon burning .
the lower bound without carbon burning is almost the same as the upper bound with carbon burning , so we adopt the boundary of equation ( [ lower_bound ] ) .
then , our @xmath0 is more stringent than that derived in @xcite .
this difference mainly comes from the initial condition as discussed in section [ comparison ] .
using the results of @xcite and the binary population synthesis ( bps ) calculation @xcite , @xcite estimated the brightness distribution of sne ia arising from the violent merger scenario .
they adopted the value of @xmath81 for @xmath82 from @xcite , and assumed @xmath83,\ ] ] where they used @xmath84 .
we add these three lines in figure [ fig5 ] .
our @xmath80 is basically more stringent than ruiter et al.s ( 2013 ) assumptions . , we have calculated nuclear energy generation , width=302 ] using the 13 species @xmath75-chain reaction network @xcite .
green lines denote the case of only carbon burning while red ones do those of post - processing results .
( a ) density , ( b ) temperature , ( c ) specific internal energy , and ( d ) pressure . [ fig9 ] in order to examine how our @xmath80 would affect the brightness distribution of the violent merger scenario , we adopt the same assumptions as in @xcite except for @xmath80 .
for the primary at the time of merging , we use the same wd mass distribution as @xcite ( see also * ? ? ? * ) , although it is highly uncertain whether the wd can increase its mass by avoiding the formation of a common envelope during the very rapid accretion from the he star companion .
we also assume the same @xmath85 relation of sne ia as ruiter et al.s ( 2013 ) ( see figure 4 in * ? ? ?
* ; * ? ? ?
* ) and a flat mass ratio distribution of dd systems for simplicity .
using either our critical mass ratio or theirs , we calculated the brightness distribution and compared them each other ( figure [ fig11 ] ) .
we find that there is no significant differences between them qualitatively .
this result implies that the @xmath0 of the violent merger - induced explosion is not so crucial for the brightness distribution of sne ia , as mentioned in @xcite .
our results are closest to the @xmath86 case among ruiter et al.s three cases . in the both cases of our @xmath0 and ruiter et al.s @xmath87
, the brightness distribution concentrates around @xmath88 mag . in figure
[ fig11 ] , we add the observational volume - limited brightness distributions of sne ia .
one in panel ( a ) is derived by the lick observatory supernova search ( loss , * ? ? ?
* ) , and the other in panel ( b ) is obtained by rotse - iiib ( see table 1 in * ? ? ? * ) .
the fraction of faint events in our models is lower than the observation of loss .
this discrepancy might decrease if we consider the viewing angle effects ( see also discussion in * ? ? ?
* ) , which might be increasingly important especially for a large value of @xmath0 . on the other hand , the results of rotse - iiib are consistent with ours , except most luminous ( @xmath89 mag ) events
. it should be noted , however , that there are large uncertainties for the observational brightness distribution ( e.g. , * ? ? ?
* ) and our models are too simple to be compared with the observational results .
more detailed studies are required to reach the definitive conclusion . ,
but for the case with nuclear burning . filled circles and crosses have the same meanings as those in figure [ fig4 ] .
green triangles indicate the mass combinations which satisfy the detonation condition with carbon burning although they do not without carbon burning .
( a ) the results based on our detonation condition in section [ critical ] .
( b ) the results based on the detonation condition of @xcite . ,
title="fig:",width=302 ] , but for the case with nuclear burning .
filled circles and crosses have the same meanings as those in figure [ fig4 ] .
green triangles indicate the mass combinations which satisfy the detonation condition with carbon burning although they do not without carbon burning .
( a ) the results based on our detonation condition in section [ critical ] .
( b ) the results based on the detonation condition of @xcite .
, title="fig:",width=302 ] as the mass ratio is approaching unity , the primary is more strongly deformed by the secondary at merging . as a result ,
the central density of the primary becomes lower .
because the nucleosynthesis in sn ia explosions is so sensitive to the density profile near the center of the primary in the violent merger scenario , the amount of @xmath77ni decreases as the deformation of the primary becomes higher @xcite . since our @xmath0 is higher than that in the previous studies , the amount of @xmath77ni in the violent merger scenario would decrease , while this effects
are not included in our brightness estimates .
we summarize the final fates of co wd mergers in the total mass versus mass ratio ( figure [ fig12 ] ) , as well as in the mass ratio versus primary mass diagram ( figure [ fig13 ] ) .
the black dotted line shows the line of @xmath90 .
if the total mass does not exceed this value , the merger remnant becomes a wd without exploding as an sn ia @xcite .
the direct outcome is a violent merger - induced explosion ( vm ) in the red hatched region .
we adopt equation ( [ lower_bound ] ) as the boundary of this immediate explosion .
in other regions , our condition for the violent merger scenario is not satisfied and the merger remnant reaches a quasi - stationary state .
the merger remnant consists of three parts , a cold core , hot envelope , and outer disk ( e.g. * ? ? ?
* ) . in the blue hatched region ,
off - center carbon burning occurs in the hot envelope ( see , e.g. , * ? ? ?
* ) and the core of the merger remnant would be converted to an oxygen - neon - magnesium ( onemg ) wd @xcite . because the total mass exceeds @xmath91 ,
the merger remnant finally collapses to a neutron star , i.e. , the accretion induced collapse ( aic ) ( e.g. , * ? ? ?
black , cyan , and orange ones are derived from hypothetical @xmath0 of @xcite .
a hatched purple one in panel ( a ) is a volume - limited brightness distribution derived by loss @xcite , while a hatched green one in panel ( b ) is that obtained by rotse - iiib @xcite .
magnitudes of models are bolometric , while those of loss and rotse - iiib are @xmath92 band.,width=302 ] in the green hatched region , the core of the merger remnant remains a co wd because off - center carbon burning does not occur in the early remnant phase ( @xmath93 after the secondary is completely disrupted , e.g. , * ? ? ?
further evolution of the merger remnant depends on the viscosity and the accretion of material from the outer keplerian disk @xcite .
if the viscous heating is sufficiently large in the viscous evolution phase ( @xmath94 , e.g. , * ? ? ?
* ) , or if the accretion rate exceeds the critical rate of @xmath95 in the thermal evolution phase ( @xmath96 , e.g. , * ? ? ?
* ; * ? ? ?
* ) , off - center carbon burning is ignited .
the final result would be aic .
on the other hand , if the viscous heating is too small or if the accretion rate does not exceed the critical rate , the mass of the merger remnant could grow to exceed @xmath91 without off - center carbon burning .
we should note that @xmath91 of the rotating co wd is larger than @xmath97 because of the centrifugal force , being @xmath98 if the rotation is uniform ( or larger if differential rotation ) . eventually , the merger remnant would explode or collapse depending on @xmath99 and the timescale of the angular momentum loss from the merger remnant @xcite .
we call the case of explosion as the accretion induced explosion ( aie ) . because of the above uncertainties , we indicate the green hatched region as aie / aic in figure [ fig12 ] . the boundary between the aic and aie / aic is derived from the results of @xcite . in the magenta hatched region , because off - center carbon burning does not occur and the total mass of the merger remnant does not exceed @xmath91 , it possibly leaves a massive co wd ( mwd ) .
outer boundaries are set from the wd mass range ( @xmath100 ) .
figures [ fig12 ] and [ fig13 ] could not be conclusive .
if higher numerical resolutions are applied , the vm region in figures [ fig12 ] and [ fig13 ] would extend downward . as mentioned in section [ comparison ] ,
the initial condition of simulations also affects the merger outcomes .
if we adopt non - synchronously spinning systems as initial conditions , the vm and aic region might extend ( e.g. , * ? ? ?
* ; * ? ? ?
* ) . in figure [ fig13 ]
, we depict the results of previous works for comparison .
the squares indicate the models calculated by @xcite .
their wd masses are @xmath101 , @xmath102 , @xmath103 , and @xmath104 for @xmath56 @xcite .
the filled squares are the models which satisfy the detonation condition of @xcite , while the open one did not .
their results indicate that @xmath55 for @xmath56 .
note that our boundary is @xmath57 for @xmath56 , which is more stringent than theirs .
we discuss the final fate of henize 2 - 428 on the basis of our results .
henize 2 - 428 is a bipolar pn and its central system was identified as a binary ( e.g. , * ? ? ?
@xcite analyzed the light curve and spectra of the central system . from the depth of the light curve minima , the intensities of two he ii @xmath105 nm absorption lines , and the obtained radial velocity amplitudes , they interpreted the central system of henize 2 - 428 as a dd ( double post - agb core ) system with nearly equal masses and effective temperatures .
they estimated the total mass of the central system as @xmath106 , which clearly exceeds the chandrasekhar limit .
they also derived the merging time of this system as @xmath107myrs from the orbital period of @xmath108 days .
if their interpretation is correct , the central system of henize 2 - 428 is the first detected super - chandrasekhar dd system , which is a candidate of sn ia progenitors in the dd scenario .
because the mass ratio is close to unity , it likely leads to an sn ia in the violent merger scenario . ) as the boundary between the vm and aic or aie / aic .
the boundary between the aie / aic and aic is derived from @xcite .
the blanck ( white ) regions are where the mass of a wd is @xmath109 ( he wd region ) or @xmath110 ( onemg wd region ) .
he and onemg wds are not the subject of the present study .
the black dotted line represents @xmath111 .
if the total mass does not exceed this value , the merger remnant becomes a wd without exploding as an sn ia @xcite .
the black point shows the possible ranges of the mass ratio and total mass of the central system of henize 2 - 428 .
the parameters of henize 2 - 428 are in the region of vm , although we consider the errors .
see text for more details .
, width=302 ] , but in the mass ratio versus primary mass diagram .
black squares denote the models calculated by @xcite .
filled squares indicate the models which satisfy the detonation condition of @xcite , while a open square does the model that does not .
, width=317 ] @xcite fixed the mass ratio of the central system of henize 2 - 428 at unity based on the similar depths of light curve minima and the radial velocity amplitudes . in this paper , considering @xmath112-uncertainties mentioned in @xcite , we derive possible ranges of the mass ratio and primary mass without fixing the mass ratio . in figures
[ fig12 ] and [ fig13 ] , the black point with error bars depicts the possible ranges of the mass ratio and primary mass thus derived . even considering the uncertainties of the mass ratio and primary mass
, we regard that henize 2 - 428 is a possible progenitor of sne ia in the violent merger scenario .
we summarize our main results as follows : \(1 ) based on the sph simulations of merging co wds , we derived the critical mass ratio for the violent merger scenario , i.e. , the @xmath0 versus @xmath5 relation , and compared our result with the previous studies .
our @xmath80 is more stringent than that derived by @xcite .
we conclude that this small difference stems mainly from the differences in the initial condition ( synchronously spinning or not ) .
( 2 ) we confirmed that the difference in the critical mass ratio does not affect the brightness distribution of sne ia significantly , as claimed in @xcite .
our results are close to that of the @xmath86 case in @xcite and consistent with the observational one obtained by rotse - iiib @xcite .
our larger @xmath0 would also decrease the relative rate of higher central density primary wds at the merger , and as a result the total amount of @xmath77ni synthesized in the violent merger would be reduced .
\(3 ) we also summarized the direct outcome of co wd mergers and their final fates in the diagram of @xmath113 versus @xmath114 ( and @xmath114 versus @xmath5 ) .
on the basis of this diagram , we discussed the fate of the central system of the bipolar planetary nebula henize 2 - 428 , which was recently suggested to be a possible super - chandrasekhar dd system merging in a timescale much shorter than the hubble time . even considering uncertainties of the proposed system in the @xmath113 versus @xmath114 diagram
, we identify the final fate of this system as almost certainly an sn ia in the violent merger scenario .
we thank the anonymous referee for many detailed comments that helped to improve the paper . simulations in this paper
were performed by using computational resources of kavli institute for the physics and mathematics of the universe ( ipmu ) , and ha - pacs at the center for computational sciences in university of tsukuba under interdisciplinary computational science program .
this research has been supported in part by grants - in - aid for scientific research ( 23224004 , 23540262 , 23740141 , 24540227 , 26400222 , and 26800100 ) from the japan society for the promotion of science and by the world premier international research center initiative , mext , japan .
this work is partly supported by mext program for the development and improvement for the next generation ultra high - speed computer system under its subsidies for operating the specific advanced large research facilities . | mergers of carbon - oxygen ( co ) white dwarfs ( wds ) are considered as one of the potential progenitors of type ia supernovae ( sne ia ) .
recent hydrodynamical simulations showed that the less massive ( secondary ) wd violently accretes onto the more massive ( primary ) one , carbon detonation occurs , the detonation wave propagates through the primary , and the primary finally explodes as a sub - chandrasekhar mass sn ia .
such an explosion mechanism is called the violent merger scenario .
based on the smoothed particle hydrodynamics ( sph ) simulations of merging co wds , we derived more stringent critical mass ratio ( @xmath0 ) leading to the violent merger scenario than the previous results .
we conclude that this difference mainly comes from the differences in the initial condition , synchronously spinning of wds or not . using our new results , we estimated the brightness distribution of sne ia in the violent merger scenario and compared it with previous studies .
we found that our new @xmath0 does not significantly affect the brightness distribution .
we present the direct outcome immediately following co wd mergers for various primary masses and mass ratios .
we also discussed the final fate of the central system of the bipolar planetary nebula henize 2 - 428 , which was recently suggested to be a double co wd system whose total mass exceeds the chandrasekhar - limiting mass , merging within the hubble time . even considering the uncertainties in the proposed binary parameters
, we concluded that the final fate of this system is almost certainly a sub - chandrasekhar mass sn ia in the violent merger scenario . |
although the determination of the fundamental parameters of rr lyrae ( rr ) stars is interesting in itself , it is also important from a practical point of view because rr stars play a considerable role in establishing galactic and extragalactic distance scales .
the preston index and spectroscopic observations are used to determine their atmospheric metallicity @xmath8 $ ] and interstellar reddening @xmath3 .
since confirmed rr - type components are not known in binary systems , the mass determination is based on both stellar evolution and pulsation theories . due to the uncertainty of parallax data , the baade - wesselink ( bw )
method is mostly used to infer their distance @xmath5 @xcite . in the bw analysis , and to determine @xmath8 $ ] and @xmath3 , the quasi - static atmosphere approximation ( qsaa ) is employed to interpret photometry and spectroscopy .
the qsaa was introduced by @xcite : ` the simplest approach is to assume that at each phase , the atmosphere adjusts itself practically instantaneously to the radiative flux coming from the interior and to the effective gravity @xmath2 @xmath14 where @xmath15 are the instantaneous values of the radius and acceleration , which is supposed uniform throughout the atmosphere ' , @xmath16 is the newtonian gravitational constant , @xmath6 is the stellar mass and dot is a differentiation with respect to time @xmath17 . `
one may then build a series of static model atmospheres , ' and select one of them at each phase by spectroscopic or photometric observations .
its flux , colours , effective temperature @xmath1 and surface gravity @xmath2 are accepted as the atmospheric parameters of that phase providing a basis for the determination of other parameters such as angular radius , mass , distance , etc .
the subject of the this paper is the qsaa and its generalization to a non - uniform atmosphere .
we will investigate the qsaa from the point of view of atmospheric emergent flux and hydrodynamics . by comparing the observed colour indices with those of static model atmospheres ( @xcite , @xcite ) we select the phases when they coincide .
considering hydrodynamics , we do not construct a consistent dynamic model of an rr atmosphere .
however , we find a better description of the pulsating atmosphere if we characterize it by pressure and density stratifications in addition to the two parameters @xmath18 and @xmath19 .
we determine the fundamental parameters @xmath6 and @xmath5 from the hydrodynamic considerations without using the bw method or theories of stellar evolution and pulsation at all .
our method uses photometry as observational input ; spectroscopy and radial velocity observations are not needed . in section 2 , conditions of the qsaa are formulated for a spherically pulsating compressible stellar atmosphere with a velocity gradient .
practical methods are described to determine @xmath8 , e(b - v)$ ] , @xmath20 and @xmath21 from @xmath7 colours of kurucz atmospheric models , @xmath22 being the pulsation phase .
the laws of mass and momentum conservation are used to determine the mass and distance of the pulsating star .
section 3 presents the results obtained from the @xmath7 photometry of su dra .
the discussion and conclusions are given in sections 4 and 5 , respectively .
first , we describe a photometric method to select the phases in which qsaa can be regarded as a good approximation from a point of view of fluxes .
next , we introduce the time - dependent pressure and density stratifications of the photometrically selected static models in the laws of mass and momentum conservation to find the phases in which qsaa can be regarded as a more or less good approximation from a hydrodynamic point of view as well . using the phases of valid qsaa from photometric and hydrodynamic points of view , we determine mass , distance and infer the internal motions of the atmosphere with respect to the stellar radius @xmath18 . in this paper
, we define @xmath18 as where the optical depth is zero , since the emergent fluxes of a theoretical model are given for @xmath23 in any photometric band .
of course , the zero boundary condition is a mathematical idealization @xcite . in the following ,
we tacitly assume that the density is not zero at the boundary , i.e. @xmath24 if @xmath25 .
the functions @xmath26,e(b - v)]$ ] and @xmath27,e(b - v)]$ ] of the kurucz atmospheric models form a suitable grid for interpolation in the ranges @xmath28 and @xmath29 if the colour indices @xmath30 are selected appropriately from @xmath7 photometry . for fixed @xmath8
$ ] and @xmath3 the intersection of the two functions @xmath31,e(b - v))\}_{i=1,2}$ ] as a function of @xmath32 gives a pair of @xmath33 for a given pair of @xmath34 .
although it is more practical to use @xmath35 instead of @xmath36 @xcite , the other common colour indices , e.g. @xmath37 etc .
, will also be kept . using @xmath7 photometry
, we can construct four independent colour indices @xmath38 , i.e. @xmath39 combinations of colour indices resulting in six pairs of @xmath33 for a given @xmath22 .
since the @xmath40 band covers the balmer jump , the indicator of @xmath41 , the combinations must contain at least one colour index containing @xmath40 photometry .
@xmath42 such combinations can be constructed .
the average of the 30 pairs of @xmath33 will be accepted for a phase @xmath22 .
_ condition i. _ if the scatters @xmath43 of @xmath33 are compatible with the expected scatter from the error of the colour indices , qsaa provides a good approximation in this phase from the photometric point of view .
the bolometric corrections and @xmath7 physical fluxes of a model with given @xmath44,e(b - v)$ ] are available in tabular form @xcite and they allow us to determine the angular radius of the star defined by @xmath45 technical details of finding the appropriate model and determining @xmath0 at a phase @xmath22 are given in @xcite and @xcite . , @xmath46 , @xmath0 is available upon request . ] on the scale provided by the kurucz atmospheric models , the above procedure offers a possibility to determine reddening and atmospheric metallicity from photometry without using spectroscopy . in the shock
free phases , the qsaa is expected to reproduce the colours well , therefore , @xmath47 must be minimal as a function of the @xmath22-independent model parameters @xmath48 $ ] @xcite .
of course the method can also be applied for non - variable stars to determine @xmath3 , @xmath8 $ ] , @xmath1 , @xmath46 , @xmath0 . in the original form of the qsaa @xcite
, the atmospheric motion is described by the parameters @xmath49 and the outward or inward accelerations are driven by the variable @xmath50 $ ] in ( [ 1.102 ] ) where @xmath51 .
this is essentially a uniform atmosphere approximation ( uaa ) .
however , the atmosphere of an rr star is a dilute , compressible gaseous system with variable temperatures .
therefore , the introduction of additional parameters containing @xmath20 promises a better description in the frame of the qsaa . considering the geometry of radial pulsation modes @xcite and neglecting rotation will result in all components of the vector field velocity @xmath52 being zero except for the radial component @xmath52 .
the viscosity is negligible at the velocities occurring in an rr atmosphere .
therefore , the momentum conservation is expressed by the euler equation of hydrodynamics : @xmath53 where @xmath54 , @xmath55 is the pulsation period .
the contribution of the neglected tangential motions is the average @xmath56/r$ ] @xcite . in perfect spherical symmetry @xmath57 .
the actual pressure and density stratifications were devided into two parts @xmath58 and @xmath59 , where @xmath60 and @xmath61 denote the values of a static model atmosphere , while @xmath62 , @xmath63 are the dynamical corrections due to pulsation . from the hydrodynamic point of view , the essence of the qsaa is that at a given @xmath17 @xmath64 i.e. the @xmath65-independent effective gravity of a static model atmosphere is introduced in ( [ 1.100 ] ) and @xmath66 , @xmath67 . in what follows
, we drop this latter restriction and define a dynamical correction term of acceleration @xmath68 to account for the difference of a static and dynamical atmosphere .
because of the lack of dynamical model atmospheres , @xmath69 must be determined empirically from the quantities derived from photometry .
equation ( [ 1.100 ] ) is transformed to definition ( [ 1.102 ] ) of qsaa if @xmath70 and @xmath57 , @xmath71 , because these simplifications imply @xmath72 and @xmath73 , i.e. the uaa is valid . as a dynamical equation ,
this simplified equation of motion is the basis of mass determination in bw analyses ( e.g. @xcite , @xcite , etc ) .
if the sound velocity is assumed as an upper limit for @xmath74 and @xmath75 , @xmath76 is some @xmath77 .
it is negligible in comparison with the other components of the acceleration at any @xmath78 .
typical values are @xmath79 during the pulsation cycle of an rr star while @xmath2 can exceed @xmath80 when the atmosphere is in the state of maximal compression . thus , @xmath76 will be neglected and @xmath81 is a periodic function of @xmath17 with zero - points .
if the @xmath65-dependence of the temperature @xmath82 is neglected and a constant density upper boundary condition is assumed , an integration of ( [ 2.100 ] ) over the interval @xmath83 $ ] with the equation of state of a perfect gas gives the approximate density stratification of the model atmosphere : @xmath84\}\ ] ] where @xmath85 are the reciprocal barometric scaleheight at @xmath18 , the universal gas constant and average molecular mass , respectively .
essentially , we use the static model atmospheres to measure @xmath86 . the atmospheric pulsation is driven by @xmath87 and the thermal processes are represented by the variable @xmath88 .
_ condition ii . _
if @xmath89 , i.e. @xmath90 , qsaa provides a good approximation in this phase from the hydrodynamic point of view .
this condition formulates the fact that the dynamical excess of acceleration in the upper photosphere is smaller than the @xmath91 error of @xmath92 and the same error is assumed for @xmath93 .
the combination of ( [ 2.120 ] ) and ( [ 1.100 ] ) allows us to estimate the constant and @xmath94 terms of @xmath69 .
however , the satisfaction of condition ii can be checked afterwards when the terms with other powers of @xmath5 were determined .
the quotient @xmath95 characterizes the degree of excellence of the qsaa .
the qsaa is exact from the hydrodynamic point of view if @xmath71 .
the perfect spherical symmetry means that in a euler picture @xcite we have to use ( [ 1.100])-([2.110 ] ) with @xmath96 and the mass conservation law @xmath97 must be taken into account . with assumption ( [ 2.110 ] ) , the analytic solution to ( [ 1.101 ] ) is @xmath98 where @xmath99 , @xmath100 , @xmath101 , @xmath102 , @xmath103 .
@xmath104 was introduced to incorporate uaa by taking @xmath105 and @xmath106 .
the main term in @xmath107 is @xmath108 and the convergence of ( [ 2.302 ] ) is excellent because @xmath109 .
after differentiations of @xmath107 , @xmath88 and @xmath0 , taking @xmath110 , the following form is convenient for a numerical solution : @xmath111 where @xmath112r^2/ g$ ] .
it can be solved if there exist two or more @xmath17 intervals satisfying condition i and @xmath113 is identical for them .
the assumed differences @xmath114 and the satisfaction of condition ii must be verified afterwards . equation ( [ 107a ] ) must be solved for different phase pairs @xmath115 and the values @xmath5 and @xmath6 must be omitted from the final averaging if @xmath116 . to account for the difference of the local and effective temperatures , the boundary temperature @xmath117 of a gray atmospheric model at @xmath23 @xcite will be used in @xmath118 .
hydrogen and helium are mainly neutral at the boundary temperature of rr stars . consequently , assuming a typical chemical composition means that @xmath119 is the appropriate choice in @xmath118 . , @xmath0 , @xmath32 , @xmath120 of su dra as a function of phase .
filled circles indicate the phases when condition i is satisfied.,width=317 ] , @xmath1 values in bins of size @xmath121 , respectively . dotted : su dra at @xmath122 , solid line : su dra at @xmath123 , dashed : bd + 67708 .
lower panels : standard errors @xmath124 and @xmath125 from 30 possible combinations of @xmath126 as a function of phase . at phases below the dotted lines
the scatter can be attributed to observational errors of the colour indices.,width=317 ]
@xcite derived good quality light curves of the rrab star su dra by homogenizing 45 years of @xmath7 observations ( see his table 7 ) .
the method described in the previous section will now be applied . instead of smoothing the light curves , binned data will be used to keep the results as close as possible to observed quantities .
the 30 colour index pairs containing at least one @xmath40 were obtained from @xmath127 . taking @xmath8=-1.60 $ ] , @xmath128 @xcite the basic quantities for ( [ 107a ] ) are summarized in table [ tab1 ] for different phases .
1 is a plot of @xmath20 , @xmath129 , @xmath21 , @xmath130 for one whole pulsation .
@xmath131 and @xmath132 are plotted in the lower panels of fig .
[ fig2 ] . assuming random errors of @xmath133 for the colour indices will result in @xmath134 and @xmath135 .
these values are indicated by the dotted horizontal lines .
condition i is satisfied in the phase points lying below or close to the dotted lines . at phases lying above the dotted lines ,
the monochromatic flux of static models and su dra differs significantly on a level which has noticeable effect on the broad band colours @xmath7 .
the assumed @xmath8=-1.6 $ ] , @xmath128 were verified by a variation procedure @xcite . in the shock
free phases @xmath136 , minimization of @xmath137 resulted in @xmath8=-1.60\pm 0.10 $ ] , @xmath138 . to demonstrate the difference between good and poor qsaa , the histograms of the 30 @xmath139 values
are plotted in the upper panels of fig .
[ fig2 ] for @xmath140 of su dra and bd + 67 708 .
they show normal distributions with small scatter for the non - variable bd + 67708 and su dra at @xmath122 . at @xmath123 ,
the distribution is almost uniform . at the next phase point @xmath141 ,
merely 14 intersections of @xmath142,e(b - v))\ } _ { i=1,2}$ ] , i.e. 14 pairs of @xmath139 were found instead of 30 pairs .
thus , at @xmath143 the observed colours differ significantly from those of any static model of @xcite . however , it is interesting to note that the small errors @xmath144 , @xmath145 indicate a phase island at @xmath146 with observed and static model colours in agreement for some @xmath147 minutes .
this happens to be the phase of the hump on the light curve when the inward and outward motions encounter and produce a shock @xcite . in the interval @xmath148 the atmosphere is in a state of maximal compression by the shock coming from the sub - photospheric layers and
@xmath18 is nearly minimal .
this is the rising branch and the start of the descending branch in the light curve .
therefore , the values of @xmath149 obtained from qssa , if they can be found at all , must be considered as a first approximation only .
this is reflected in large @xmath150 , @xmath125 except for @xmath151 .
.the basic quantities for ( [ 107a ] ) at various phases .
the units are @xmath152 . [
cols="<,<,<,<,<,<,<",options="header " , ] @xmath153 $ ] indicates the values which we accept as the best ones . in the interval @xmath154 , the acceleration is dominated by @xmath155 and
the atmosphere is practically at a standstill .
the values @xmath156 and @xmath157 indicate @xmath158 and @xmath159 .
if @xmath160 is neglected by setting @xmath161 and the data of table [ tab1 ] are used with @xmath71 , the upper limit @xmath162 is obtained from @xmath163 . if this estimation is improved by @xmath164 obtained for the distance and mass @xmath153 $ ] , the upper limit shifts to @xmath165 . from the functions
@xmath166 the upper mass limit is @xmath167 .
these upper limits are in accordance with the values @xmath153 $ ] .
however , they should be treated as indicative only , because they are sensitive to the actual value of @xmath168 .
these values could be determined by numerical differentiations from @xmath169 in @xmath170 when the qsaa is not optimal to obtain all of them .
hence , the phase island @xmath171 could not be included in the mass and distance determination because a very transient good qsaa is embedded in a strongly shocked interval and the mass @xmath6 appearing in @xmath160 plays a secondary role in the dynamics . for the sake of completeness
, it must be mentioned that the pair @xmath172 and @xmath173 in ( [ 107a ] ) gives @xmath174 pc . adding this value to those involved in @xmath153 $ ] , @xmath175 pc and @xmath176 is obtained .
this distance is closer to @xmath177 pc suggested by the _
parallax @xmath178 mas @xcite .
however , this @xmath179 pc was omitted because it is an outlier point above the @xmath180 level , it exceeds the upper limit @xmath181 pc and the inaccurate knowledge of @xmath182 makes it uncertain . close to the shocked phases ( i.e. for @xmath183 , @xmath184 , @xmath185 , @xmath186 , @xmath187 , @xmath188 , @xmath189 , @xmath190 , @xmath191 ) , negative stellar masses were obtained if @xmath192 .
the condition @xmath193 gives an upper limit for @xmath194 at these phases and shows clearly the limitations of treating an rr atmosphere in the qsaa . in these phases the characteristic value @xmath195
the sharp peaks @xmath196 and @xmath197 are produced by @xmath198 , i.e. the atmosphere is almost completely free of acceleration .
@xmath199 is a result from the poor representation of the density stratification by ( [ 2.110 ] ) , i.e. the dynamical term is large in @xmath200 producing @xmath201 .
this is the start of the descending branch in the light curve , when the atmosphere starts expanding rapidly .
the averaged surface gravity for the whole pulsation cycle is @xmath202 , which is in excellent agreement with @xmath203 from @xmath204 and @xmath7 photometry ( @xcite , @xcite ) , respectively .
the averaged effective temperature of @xmath205 k is significantly higher than @xmath206 k ( @xcite , @xcite , @xcite ) .
the minimal , averaged and maximal angular radii are @xmath207 rad corresponding to @xmath208 , respectively .
the following equilibrium luminosity and effective temperature @xcite give the position of su dra in a theoretical hertzsprung - russell diagram ( hrd ) : @xmath209 , @xmath210 ^ 2\}^{1/4 } = \langle\vartheta^2t_{\rm e}^4\rangle^{1/4 } \langle\vartheta\rangle^{-1/2}=(6813\pm 20)\mbox{k}$ ] .
the magnitude averaged absolute brightness is @xmath211 mag .
from the point of view of radiative transfer , the plane - parallel approximation of the model atmospheres @xcite could be applied beyond doubt , because @xmath212 holds for the whole pulsation cycle of an rr star . in favour of applying qsaa
, semi - quantitative arguments were that the temperature changes are very slow even in a high amplitude rr star such as su dra : @xmath213 , while the characteristic times - cale of the radiative processes is below @xmath214 ( @xcite , @xcite ) .
the effect of the variable effective gravity can be characterized by our condition ii , which provides information on whether qsaa may be assumed .
the limits are @xmath215 .
the upper limit comes from the rising branch of the light curve , when the main shock hits the atmosphere . in general ,
the photometric input of the present method is identical with that of the bw method . in order to obtain the fundamental parameters , radial velocity data and their problematic conversion to pulsation velocities are not necessary .
on the other hand , @xmath0 and @xmath88 must be differentiated numerically ; differential quotients are sensitive to the non - validity of qsaa .
our photometric and hydrodynamic considerations revealed empirical quantitative conditions to find the phase intervals of the pulsation when the static model atmospheres of @xcite are satisfactory to derive the variable and non - variable physical parameters of the pulsating atmosphere .
outside these intervals , dynamical model atmospheres are necessary to refine the parameters from qsaa , which is beyond the scope of this paper .
the fundamental parameters @xmath5 , @xmath6 , @xmath8 $ ] , @xmath3 were determined using photometric quantities only in phases when the qsaa was a good approximation ( i.e. both conditions i and ii were satisfied ) .
the values obtained from averaging over the entire pulsation cycle can be considered as a first approximation only , because qsaa was assumed in all phases regardless of it being a good or poor approximation .
the large error of @xmath216 and @xmath217 originates from @xmath218 of our best value @xmath153 $ ] in table [ tab2 ] . to give a good impression of the accuracy of inverting the @xmath7 photometry to physical parameters ,
the comparison star bd + 67 708 was used , because its colours are similar to those of su dra ( table 3 in @xcite ) .
the results are @xmath8=-0.77\pm .03 $ ] , @xmath219 , @xmath220 rad , @xmath221 , @xmath222 k. the errors are roughly in the same order of magnitude as those of su dra when conditions i and ii are satisfied . figs .
[ fig3](c ) , ( d ) demonstrate that significant corrections must be added to @xmath108 , @xmath19 if the true pulsation velocity and acceleration are required at @xmath223 .
moreover , the triangles show another correction : @xmath224 if @xmath225 , i.e. @xmath226 is the velocity profile . in other words , even in the shock free phases , considerable phase - dependent velocity and acceleration gradients exist in the layers @xmath227 of the line formation .
the problem of @xmath228 is present in all phases , i.e. even in the shock free intervals , which are used in modern bw analyses ( e.g. @xcite ) .
the non - uniform motion of the outermost layers introduces uncertainty when pulsation velocities are determined , because the centre - of - mass velocity @xmath229 must be subtracted from the observed radial velocities . a recent exposition of the problem for cepheid stars
is given in @xcite . by definition @xmath230
refer to @xmath25 , while the variable component of the radial velocity is an average of velocities ( e.g. ( [ 2.302 ] ) ) over the layers @xmath231 i.e. @xmath232 .
the effect on @xmath229 and @xmath5 has not yet been studied at all .
nevertheless , the importance is obvious , since an error of @xmath233 in @xmath229 results in an error @xmath234 @xcite .
there is a considerable uncertainty of @xmath229 in the literature , suggesting an error of @xmath5 as large as a factor of 2 .
it suffices to mention that for su dra @xmath235 are given by @xcite and @xcite from high dispersion spectra and spectral masking method coravel - technique , respectively . a wavy fine structure in the variation of @xmath18
is clearly seen in figs .
[ fig1 ] , [ fig3](b ) .
the outward motion starts at @xmath236 while the outward motion is reversed @xmath237 , i.e. the atmospheric velocity is @xmath238 at these phases .
the wavy fine structure is without doubt real , because the beginnings of the outward motion are connected with significant maxima in @xmath33 .
the well - known bump and hump at @xmath239 @xcite in the light curve are caused by the precursor and main shocks , respectively . at @xmath240 ,
a small change is observable in the slope of the light curve @xcite . here
, the existence of a shock is a new finding .
it is a pre - precursor shock ; we propose the designation _ jump _ for it . by integrating radial velocities ,
authors tend to smooth out the mentioned fine structure of motions [ e.g. @xcite in the case of sw and ] .
this practice seems to be unjustified .
the velocity of the atmosphere is @xmath238 in the interval @xmath241 , the atmosphere is roughly at a standstill , the brightness starts rising , a small depression of @xmath242 is visible at @xmath243 , and @xmath20 is monotonic .
it is not clear whether the small undulation of @xmath129 around @xmath244 is real or not , because the maximal @xmath245 was found just at @xmath246 .
surprisingly , in comparison with previous studies , considerable differences were found only in @xmath247 , @xmath248 .
this is due to the fact that the interpolated @xmath20 depends on @xmath21 and that the information from a five colour photometry was used in a more complex manner : an averaged value from 30 colour index pairs was utilized .
the photometry covers the whole spectrum between 350 and 1000 nm .
the use of only one colour index , e.g. @xmath249 solely ` because of its apparent merits ' @xcite , can result in a systematic error of @xmath20 .
it can explain the @xmath250 k difference , because in that previous work qsaa was assumed in all phases despite violating both conditions i and ii .
an inspection of the functions @xmath142,e(b - v))\ } _
{ i=1,2}$ ] derived from the tables @xcite shows that if merely one colour index is to be used , the optimal choice for determining @xmath1 would be @xmath251 .
this is because @xmath251 is almost independent of @xmath32 for the actual values of @xmath8 $ ] and @xmath3 of su dra .
however , in phases violating condition i , the sole use of @xmath251 would also introduce a systematic error , similar to the use of @xmath249 .
there is a remarkable decrease of @xmath252 , @xmath253 if our compressible qsaa is substituted for uaa , i.e. more physical input is used in the frame of a one - dimensional model in space . since ( [ 107a ] ) was derived from the dynamic equation ( [ 1.100 ] )
, the mass determination is more accurate from this while the distance is more uncertain .
this is reflected in the shape of the curves in fig .
[ fig3](a ) . the large formal error of @xmath5 originates from the non - separability of errors in @xmath254 and @xmath255 . a first attempt to derive @xmath5 and @xmath6 of an rr star from the over - simplified version ( [ 1.102 ] ) of ( [ 1.100 ] )
was described by @xcite .
the present results for the distance and mass of su dra differ only slightly from the previous @xmath256 , @xmath257 .
the most probable reason of the very good coincidence is the existence of the phase island with the good qsaa at @xmath258 , just when the atmosphere is at a standstill ; consequently , ( [ 1.102 ] ) is a good approximation because of @xmath259 .
the values @xmath260 pc , @xmath261 given by @xcite are very close to the present ones . however , some caution is appropriate because of the underestimated uncertainties in their derivation .
their @xmath6 originates from using ( [ 1.102 ] ) , i.e. @xmath106 , the uaa corresponding to line 3 in table [ tab2 ] .
furthermore , because of the uncertain value of @xmath229 , @xmath262 is well possible . in the bw method
, the propagation of the error @xmath263 can be estimated from ( [ 2.202 ] ) as follows .
typical radius changes of an rr star are @xmath264 km within @xmath265 .
the error of the radius change is @xmath266 km , i.e. @xmath267 .
by @xmath268 , the final error will be @xmath269 .
this considerably exceeds the error originating from the projection factor ( e.g. @xcite ) converting the observed radial velocity to pulsation velocity .
furthermore , by ( [ 2.202 ] ) and ( [ 1.102 ] ) , the error @xmath269 propagates to an error @xmath270 . the lower limit of the magnitude averaged visual absolute brightness is @xmath271 mag if @xmath272 pc , while @xmath273 mag belong to the improbable values @xmath274 pc , respectively .
observed colours and magnitudes of a spherically pulsating star have been compared with those of static kurucz model atmospheres to determine fundamental parameters of the star in the frame of the qsaa .
photometric and hydrodynamic conditions have been formulated for the validity qsaa in spherically pulsating stars . *
the qsaa has been generalized for a non - uniform , compressible atmosphere with radial velocity gradient .
this is a step forward because the hitherto available uaa described the motion of the atmosphere by effective gravity and differentiating the sole parameter radius @xmath18 of optical depth zero with respect to time . *
our combined photometric and hydrodynamic method uses the variation of effective gravity , angular radius , velocity , acceleration in the euler equation .
the complete input of the euler equation has been derived from photometry and theoretical model atmospheres .
spectroscopic and radial velocity observations were used neither explicitly nor implicitly .
this is a definite advantage in comparison with the bw method , because less observational efforts are needed to get the fundamental parameters and the problematic conversion of observed radial velocities to pulsation velocities is not necessary . * concerning the interpretation of multicolour photometry , the inputs are identical with those of the bw method .
the present refinements are the quantitative conditions whether photometry can or can not be interpreted by static model atmospheres .
* first , the phases have been selected by the photometric conditions when the qsaa is valid .
secondly , in a number of these phases , the laws of mass and momentum conservation have been applied in the euler formalism of hydrodynamics to determine the mass and distance of the star from the motion of the atmospheric layers in the neighbourhood of zero optical depth .
afterwards , it was checked whether the hydrodynamic condition of the qsaa was satisfied in the used phases ; i.e. phases that satisfied the photometric condition but violated the hydrodynamic condition had to be excluded .
_ atmospheric dynamical mass _ seems to be an appropriate term to indicate that the mass has been derived by a method which is completely different from a mass derived by pulsation or evolution theories @xcite .
* as a by - product , a variation procedure has been given for estimating atmospheric metallicity and interstellar reddening toward a star from @xmath7 photometry .
this method was successfully applied for the non - variable comparison star bd + 67 708 , giving @xmath8=-0.77\pm .03 $ ] , @xmath275 , @xmath220 rad , @xmath276 , @xmath222 k. * using the @xmath7 photometry of the high amplitude rrab star su dra , the following fundamental parameters have been found : @xmath277=-1.60\pm .10,\ ] ] @xmath278 @xmath279 @xmath280 @xmath281 @xmath282 @xmath283 @xmath216 and @xmath248 are approximate values , since they originate from averaging over the whole pulsation cycle containing phases in which the qsaa is merely a first approximation . * the internal motions of the atmosphere with respect to zero optical depth
have been found to be significant in comparison with velocities and accelerations derived from the uaa . from the internal motions of the atmosphere , some constraints have been sketched for converting observed radial velocities to pulsation and centre of mass velocities .
estimations have been given for the error propagation in a baade - wesselink analysis .
the author is grateful to j. m. benk , z. kollth , .
kspl , l. szabados and an anonymous referee for thoughtful reading of the paper and useful comments leading to an improved presentation of the results . | a photometric calibration of kurucz static model atmospheres is used to obtain the following parameters of rr lyrae stars : variation of stellar angular radius @xmath0 , effective temperature @xmath1 , gravity @xmath2 as a function of phase , interstellar reddening @xmath3 towards the star and atmospheric metallicity @xmath4 .
photometric and hydrodynamic conditions are given to find the phases of pulsation when the quasi - static atmosphere approximation ( qsaa ) can be applied .
the qsaa is generalized to a non - uniformly moving spherical atmosphere , and the distance @xmath5 , mass @xmath6 and atmospheric motion are derived from the laws of mass and momentum conservation . to demonstrate the efficiency of the method , the @xmath7 photometry of su dra was used to derive the following parameters : @xmath8=-1.60\pm .10 $ ] dex , @xmath9 , @xmath10 pc , @xmath11 , equilibrium luminosity @xmath12 and @xmath13 k. hydrodynamics stars : atmospheres stars : fundamental parameters stars : individual : su dra stars : oscillations stars : variables : rr lyrae . |
the ( type - i ) seesaw mechanism @xcite is arguably the simplest and best motivated framework not only to give neutrinos mass but also to explain the smallness of said masses compared to the electroweak scale .
the necessary right - handed neutrino partners are assumed to be heavy to suppress the active neutrino masses , which also allows for baryogenesis through leptogenesis @xcite via the decay of the heavy states , thus explaining the baryon asymmetry of the universe . while initially motivated in the context of grand unified theories ,
the seesaw mechanism has since been applied in its own right . the number of right - handed neutrinos , transforming as singlets of the standard model ( sm ) gauge group , is then no longer constrained by anomaly cancellations or the need to fill up an irreducible representation , but rather a free parameter .
at least two right - handed neutrinos are necessary to reproduce the observed neutrino mass - squared differences , as well as leptogenesis @xcite .
there is however no upper bound , and models with @xmath2@xmath3 singlets have been studied in the context of leptogenesis @xcite , lepton flavor violation ( lfv ) @xcite and as a way to explain the large mixing angles of the active neutrinos @xcite .
like for most things in life , there is also a motivation from string theory @xcite .
the number of singlets is formally infinite in extra - dimensional theories , as right - handed neutrinos are not restricted to the sm - brane and will thus lead to a tower of kaluza - klein excitations @xcite .
still lacking , however , is a proper parametrization of the arising mixing and coupling matrices in terms of physical quantities , as needed for example for efficient parameter scans ( as the number of redundant parameters explodes for large @xmath0 ) .
the goal of this work is to provide exactly these parametrizations for arbitrary @xmath0 , both for the most general case ( top - down approach , to some degree discussed in ref .
@xcite ) and the seesaw limit ( bottom - up ) .
the rest of this paper is organized as follows : we first fix our notation in sec . [
sec : framework ] , discuss the top - down parametrization in sec .
[ sec : top - down ] and the bottom - up parametrization in sec .
[ sec : bottom - up ] , also deriving constraints on the involved parameters from perturbativity and lfv . to illustrate possible effects of @xmath0 singlets we discuss basis independence ( neutrino anarchy ) in this framework and present the resulting distributions for the neutrino masses in sec .
[ sec : anarchy ] .
[ sec : thegroupun ] gives a brief introduction to the group @xmath4 and various representations of its elements , as needed for our parametrizations .
[ sec : measures ] collects and extends basis independent measures for miscellaneous matrices necessary for the discussion of anarchy .
finally , app .
[ sec : nequal2 ] is devoted to the somewhat special case @xmath5 , which would interrupt the flow if included in the main text , which will only cover @xmath6 .
introducing @xmath0 sm singlet right - handed neutrinos @xmath7 to the standard model modifies the lagrangian by the following terms : @xmath8 where a sum over @xmath9 ( sometimes @xmath10 in the following ) and @xmath11 is understood and will in the following often be denoted in a vector notation ( e.g. as @xmath12 ) . without loss of generality we work in a basis where the charged lepton mass matrix is diagonal , which can be accomplished by unitary transformations of the lepton fields .
@xmath13 can be diagonalized by an @xmath14 unitary matrix @xmath15 like @xmath16 .
this merely redefines @xmath17 , so from now on we will work in a basis where @xmath13 is diagonal , with positive real entries @xmath18 .
can store @xmath19 parameters ( together with @xmath17 ) , while a non - diagonal @xmath13 has @xmath20 independent entries . ]
after the higgs doublet @xmath21 acquires its vacuum expectation value @xmath22 , the dirac mass matrix for the neutrinos @xmath23 is generated , leading to a @xmath24 majorana mass matrix for the neutral fermions @xmath25 : ( for @xmath26 ) see e.g. refs .
@xmath27 assuming @xmath28 gives the low - energy neutrino mass matrix for the flavor neutrinos @xmath29 of seesaw form : @xmath30 diagonalization of @xmath31 can be performed in the following way @xmath32 with the unitary pontecorvo - maki - nakagawa - sakata matrix ( pmns matrix ) in standard parametrization @xmath33 and the majorana phase matrix @xmath34 . here
we used the abbreviations @xmath35 and @xmath36 for the three mixing angles .
the phase matrix @xmath37 can be absorbed by the lepton fields and is hence unphysical .
the neutrino mass eigenstates are then @xmath38 . at low energies the @xmath39 complex entries of the symmetric majorana matrix @xmath31 decompose into the three eigenvalues @xmath40 ( neutrinos masses ) , three mixing angles @xmath41 , @xmath42 , @xmath43 ( atmospheric , solar and reactor angle ) and three cp violating phases @xmath44 , the latter two being unobservable in neutrino oscillations ( but in principle testable in @xmath45 experiments @xcite ) .
all the information about neutrino mixing is encoded in @xmath17 , the only non - diagonal matrix in the lepton sector . for the general complex @xmath46 matrix @xmath17
, we can write down a singular value decomposition @xmath47 with positive singular values @xmath48 . out of the @xmath49 real parameters in @xmath17 ,
three phases can be absorbed in the lepton fields , so only @xmath50 are physical .
we will confirm this with an explicit parametrization of the unitary matrices @xmath51 below .
for the @xmath52 matrix @xmath53 , we take the pmns parametrization from eq . ,
while the @xmath14 matrix @xmath54 can be written as @xmath55 here , @xmath56 denotes matrices of the form @xmath57 where @xmath58 sits at the @xmath59-th row and @xmath60-th column .
see app .
[ sec : thegroupun ] for a derivation and a proof for the validity of this form of @xmath54 .
it is easy to check that @xmath54 ( @xmath53 ) has @xmath61 ( @xmath62 ) real parameters , @xmath63 ( @xmath64 ) of which are phases . from the form of @xmath17 in eq .
, it is clear that the angles @xmath65 either act on zeros in @xmath66 or can be absorbed by the phase matrix @xmath67 in @xmath68 .
an overall phase and @xmath69 can be absorbed by the lepton fields , so @xmath53 contains only three angles and three phases .
the ordering of the @xmath70 in @xmath54 is of course arbitrary , which allows us to move all the rotations that act on the zeros in @xmath66 to the right , so @xmath71 matrices of the type drop out , leaving @xmath72 real parameters in @xmath54 , half of which are phases .
consequently we can restrict the product in eq . to @xmath73 .
subgroup of @xmath74 , acting on the lower right parts , can be modded out . ] as expected , @xmath17 contains three ( from @xmath48 ) plus six ( @xmath53 ) plus @xmath75 ( @xmath54 ) equals @xmath50 real , physical parameters , @xmath76 of which are phases .
this agrees with the parameter counting in ref .
@xcite .
the full mass matrix can then be written as @xmath77 with the explicit form for @xmath53 : @xmath78 and the shorthand notation @xmath79 etc .
for the mixing angles . in the seesaw limit , we thus find @xmath80 note that the occurring mass matrix @xmath81 is for @xmath82 no longer the most general complex symmetric @xmath14 matrix , because we already restricted @xmath54 to the physical angles . from eq . or eq . one can now extract the physical masses and mixing angles in terms of the parameters @xmath48 , @xmath83 , @xmath84 , @xmath85 and @xmath86 . for completeness ,
let us count the total number of parameters in the lepton sector .
we have @xmath87 charged lepton masses , @xmath0 heavy neutrino masses and @xmath50 real parameters in @xmath17 , summing up to @xmath88 real parameters .
seeing as the seesaw formula contains more parameters than measureable observables , it proves convenient to parametrize our ignorance of the high - energy sector of the seesaw model . in other words , there are infinitely many different @xmath17 and @xmath13 that lead to the same low - energy neutrino parameters via the seesaw formula @xmath89 .
the parametrization of this ambiguity is the goal of the casas - ibarra - parametrization @xcite of the yukawa coupling : @xmath90 this @xmath17 solves @xmath91 as long as the complex @xmath46 matrix @xmath92 satisfies @xmath93 .
correspondingly , any values of @xmath13 and @xmath92 in eq .
lead to the same low - energy model , which means we can fix the pmns mixing angles and masses to the observed ones and study @xmath17 independently in other processes .
we will now give an explicit parametrization of @xmath92 , which has @xmath94 complex entries obeying @xmath95 real constraints ( @xmath96 is symmetric ) ; we therefore expect to find @xmath97 real physical parameters .
this agrees of course with the parameter counting in the last section , as we also have three charged lepton masses , three light neutrino masses , @xmath0 heavy neutrino masses and six parameters in the pmns matrix , summing up to @xmath88 input parameters .
since we only need three orthonormal complex vectors @xmath98 to define @xmath92 , we can write @xmath99 with a complex orthogonal @xmath14 matrix @xmath100 .
similar to the @xmath74 parametrization used for @xmath54 in the previous section , we can write @xmath101 as a product of the @xmath102 matrices from eq . , but with the arguments @xmath103 and @xmath104 in order for @xmath101 to be complex orthogonal : @xmath105 here , @xmath106 is a diagonal matrix with entries @xmath107 , which decides the determinant of @xmath101 and therefore whether it describes a rotation or a reflection .
we can once again mod out the @xmath108 subgroup that acts on the zeros in @xmath109 by reordering the rotations , so the product in eq .
can be restricted to @xmath73 . generally speaking
, @xmath92 is an element of the quotient space @xmath110 . for real angles
@xmath111 , @xmath92 is an element of the compact stiefel manifold @xmath112 , which admits a haar measure and could thus be used for statistical considerations . note that the case @xmath113 is special in the sense that there are no superfluous rotations that need to be modded out .
one can therefore take any complete parametrization for @xmath114 without having to worry about unphysical parameters .
for complex angles , @xmath101 and therefore @xmath92 and @xmath17 can have arbitrarily large elements , because the entries @xmath115 grow exponentially with @xmath116 . as a result ,
the yukawa couplings @xmath17 can be very large even for small @xmath13 , which is somewhat counterintuitive to the seesaw formula and relies on cancellations in the matrix product @xmath117 .
even though these cancellations suggest finetuning of parameters , this is a viable possibility and deserves discussion , especially because a low @xmath13 with large yukawa couplings makes an lhc discovery ( direct or indirect ) possible @xcite .
popular models to obtain this finetuning in a natural way are for example inverse @xcite and linear seesaw @xcite , usually based on the case @xmath118 with an imposed structure in @xmath13 and @xmath17 by some symmetry .
a useful constraint on the complex mixing angles @xmath111 comes from the perturbativity of @xmath17 , on which many calculations involving @xmath17 rely . for this
, we propose yet another , slightly different parametrization of @xmath92 , which we obtain by noting that @xmath119 , and that the ordering of the @xmath70 is arbitrary : @xmath120 the @xmath97 parameters that make up @xmath92 are thus equally divided among the new real parameters @xmath121 and @xmath122 .
the product involving @xmath121 is just an orthogonal matrix , so we can restrict @xmath123 or @xmath124 .
the parametrization can be more useful than , because the @xmath121 drop out of the expression @xmath125 , which arises from ( the often occurring ) @xmath126 for degenerate @xmath13 .
perturbativity of the yukawa couplings gives constraints of the form @xcite @xmath127 this inequality is satisfied if @xmath128{tev}}{m_i } } \sqrt{\frac{\unit[10^{-2}]{ev}}{m_j } } \ , , \label{eq : perturbativitybound}\end{aligned}\ ] ] which constrains each entry in @xmath92 .
note that the upper bound goes down for increasing @xmath0 .
once the smallest masses are specified , the above equation can be used to give a total upper bound on the parameters @xmath122 .
for example , the largest entry in @xmath92 can be estimated as @xmath129 if just one @xmath122 is large , and @xmath130 if all @xmath122 are large .
this typically leads to upper bounds on @xmath131 of @xmath132@xmath133 . in the seesaw limit ,
the neutrino states @xmath134 that couple to the charged current are approximately given by @xcite @xmath135 , with the mass eigenstates @xmath136 and @xmath137 .
this leads to lepton flavor violating decays @xmath138 @xcite , of which @xmath139 gives the strongest constraint : @xmath140 with the diagonal matrices @xmath141 and @xmath142 and the monotonic decreasing loop function @xmath143 here we already omitted the contribution from the light neutrinos , as the corresponding term @xmath144 is suppressed by the tiny mass - squared differences of the active neutrinos @xcite .
the remaining contributions are naively of order @xmath145 , but can be enhanced via @xmath92 . in the approximation @xmath146 we have @xmath147 , so eq . simplifies to @xmath148 since this branching ratio can easily exceed the current limit @xmath149
@xcite for perturbative couplings it gives constraints on the @xmath122 . naively , the relevant product in eq .
is @xmath150 , so the lfv constraints have a different structure than the perturbativity bound .
note that this constraint is connected to the unitarity of the lepton mixing matrix @xcite , because the relevant matrix @xmath151 is no longer unitary at this order .
consequently , the above bound also ensures the validity of our parametrization for @xmath17 from eq .
, as the matrix @xmath152 in @xmath17 only corresponds to the lepton mixing matrix if @xmath153 .
while not particularly nice to look at analytically , the constraints given here should be useful in numerical scans involving @xmath17 for arbitrary many right - handed neutrinos , extending for example the analysis of ref .
the idea that the large lepton mixing angles and small neutrino hierarchy are not due to some flavor symmetry but rather the absence of any distinction between the neutrino generations has been around for over a decade @xcite .
the basis independence of the neutrino mass and mixing matrices leads to a distribution of the mixing angles according to the haar measure of the lie group @xmath154 that diagonalizes @xmath31 ( so @xmath155 for complex , @xmath156 for real @xmath31 ) .
since these distributions prefer large mixing angles , this ansatz seems to fit well to the observed large leptonic mixing angles ( see also , however , a critique of the anarchy approach in refs .
@xcite ) . while the distribution of the mixing angles is uniquely given by the haar measure , the distributions of the masses and yukawa couplings are not unique , and usually the simplest ( linear ) measure is used ( listed in app . [
sec : measures ] ) .
let us now discuss neutrino anarchy , i.e. basis independence , with @xmath0 right - handed neutrinos .
in the seesaw limit we have the low - energy mass matrix for the active neutrinos from eq .
@xmath157 where @xmath53 and @xmath54 are now assumed to be distributed according to the haar measure of @xmath158 and @xmath74 , respectively , while we take the simplest linear measure for the eigenvalues of @xmath13 @xcite @xmath159 we employ the invariant boundary @xmath160 for the scanning region ; see fig . [
fig : massdist ] for a visualization of the distribution .
eigenvalues @xmath161 according to the linear measure from eq . with the boundary constraint @xmath162.,title="fig:",scaledwidth=46.0% ]
eigenvalues @xmath161 according to the linear measure from eq . with the boundary constraint @xmath162.,title="fig:",scaledwidth=46.0% ] for the singular values of @xmath66 we use the linear measure ( see app . [
sec : measures ] for the derivation ) @xmath163 with a similar boundary @xmath164 .
the effect of large @xmath0 will be a reduced hierarchy in the singular values @xmath48 ( see fig .
[ fig : yukawadist ] ) . note that eq .
is only valid for @xmath6 , see app .
[ sec : nequal2 ] for the used measure for @xmath5 .
we will now briefly describe how the random matrices in this paper are generated .
the random unitary matrices @xmath53 and @xmath54 in eq
. following the haar distribution can efficiently be obtained from a qr decomposition , as described in refs .
@xcite . drawing eigenvalues from eq . is a bit more complicated .
the procedure usually employed in the literature is the creation of a symmetric matrix @xmath13 , where @xmath165 and @xmath166 are uniformly distributed in @xmath167 $ ] . to stay rotationally invariant , only matrices with @xmath168
are then used and diagonalized ( roughly speaking , we change the scanning region from a hypercube to a hypersphere ) .
this method works well for @xmath169 , but becomes virtually unusable for @xmath170 , because the volume of the hypercube increases sharply with @xmath0 compared to the hypersphere , making it hard to accumulate statistics .
for this reason , we draw the eigenvalues directly out of the probability distribution using a multivariate metropolis algorithm @xcite .
this markov chain monte carlo method is sufficiently fast even for large @xmath0 and conveniently works without knowing the normalization factor of the distribution .
we use a multivariate gaussian proposal distribution to generate the next steps in the chain and pick only every hundredth ( tenth for @xmath171 ) valid point to reduce correlations in the samples .
the value @xmath172 will fix the overall light neutrino mass scale and can be used to fix one of the mass - squared differences @xmath173 . of the yukawa matrix according to the linear measure from eq . for @xmath26 ( black ) and
@xmath174 ( red / dashed ) with the boundary constraint @xmath175.,scaledwidth=50.0% ] the pmns mixing angles and phases do not change with @xmath0 and are given by the haar measure for @xmath158 @xcite ( ignoring unphysical phases ) : @xmath176 so the only effect of @xmath1 is a change of the three eigenvalues of @xmath31 , i.e. the ratio @xmath177 , where we sorted the masses like @xmath178 . in this notation
, @xmath179 corresponds to normal hierarchy ( nh ) and @xmath180 to inverted hierarchy ( ih ) , with best - fit values @xmath181 and @xmath182 , respectively , taken from a recent global fit @xcite . \equiv \log_{10}\left[(m_2 ^ 2-m_1 ^ 2)/(m_3 ^ 2-m_1 ^ 2)\right]$ ] for various @xmath0 .
the purple / dashed vertical lines indicate the @xmath183 range for @xmath184 @xcite ( for nh ) .
the black vertical line divides the nh and ih solutions.,scaledwidth=98.0% ] we observe from fig .
[ fig : ratios ] that for increasing @xmath0 , the distribution for @xmath185 shifts to larger values , while the width decreases .
this reduction of hierarchy can be tracked back mainly to the yukawa couplings , as the factor @xmath186 in @xmath187 pulls the @xmath48 tightly together .
an analysis without this factor shows that the behavior of @xmath185 goes in the opposite direction , i.e. the maximum shifts toward small values for large @xmath0 while being diluted .
however , the distribution then quickly converges as the only @xmath0-dependend change comes from @xmath188 , for which only the first dozen or so @xmath18 are relevant , due to the hierarchy in @xmath13 .
correspondingly , to shift the maximum of @xmath185 to its observed value , one has not only to omit the factor @xmath186 in @xmath187 , but also to increase the hierarchy by inserting something like @xmath189 , i.e. using a non - linear measure ( see ref .
@xcite for other @xmath48 distributions that accomplish this task ) .
we conclude that the large @xmath0 limit in the anarchy approach worsens the agreement with data .
while the one data point nature provides can obviously not be used to find the overlying distribution , it should be fair to say that anarchy works best for small @xmath0 .
let us finally comment on the limit @xmath190 .
as the yukawa couplings @xmath48 become quasi - degenerate for very large @xmath0 , the neutrino masses are simply given by the upper left @xmath52 submatrix of @xmath188 ( times a prefactor ) . since there is no preferred direction in the @xmath54 rotations , the entries @xmath191 have the same mean magnitude for a given @xmath0 .
the mean entries @xmath192 then take the form of a random walk with decreasing step - size @xmath193 , and due to the hierarchy in the singlet masses @xmath161 , only the first couple of steps are relevant .
the mean magnitudes are again the same for all entries , so the only structure in the neutrino mass matrix comes from random phases @xmath194 : @xmath195 .
so , if the number @xmath0 is high enough , anarchy eventually leads to democracy .
it is easy to show that a matrix of this type predominantly yields nh solutions , so even though ih becomes more and more probable for large @xmath0 , it will never dominate .
one of the simplest ways to explain the masses of the active neutrinos is the introduction of right - handed partners . as the number @xmath0 of these sm singlets is in principle unconstrained
, it is interesting to study implications of varying @xmath0 .
we have given explicit parametrizations for the mixing and coupling matrices connecting the @xmath0 right - handed neutrinos to the sm in terms of physical parameters , both in the top - down and the bottom - up approach ( casas - ibarra - parametrization ) .
for the latter , constraints on the involved parameters from perturbativity of the yukawa couplings as well as charged lepton flavor violation have been discussed . as a novel application of the @xmath0 singlet framework we studied basis independence in the neutrino sector , i.e. anarchy . of the low - energy neutrino parameters , only the neutrino mass distribution changes with @xmath0 , and we showed that anarchy with the simplest linear measure seems to prefer small @xmath0 in view of the observed mass - squared differences .
the author thanks werner rodejohann and mattias blennow for reading the manuscript and offering valuable comments .
this work was supported by the erc under the starting grant manitop , by the max planck society through the strategic innovation fund , and by the imprs - ptfs .
for convenience we provide a short and rather colloquial review of the group @xmath196 .
formally one can define it as the set of complex @xmath197 matrices that leave the inner product of @xmath198 invariant , i.e. @xmath199 . with a little effort one can further show that the entries of @xmath4 form a group under matrix multiplication , actually even a compact connected lie group . counting the number of entries in @xmath200 and constraints
, we find that @xmath4 has @xmath201 real parameters .
one can show that every element in @xmath4 can be written as @xmath202 with skew - hermitian @xmath203 . we can choose a basis for @xmath204 such that the @xmath201 linearly independent @xmath205 satisfy lie algebra relations @xmath206
= i f_{klm } x_m$ ] with structure constants @xmath207 .
the @xmath201 generators @xmath205 can be further separated into @xmath208 symmetric generators @xmath209 and @xmath210 antisymmetric real generators @xmath211 .
it also proves convenient to treat the @xmath212 diagonal ( symmetric ) generators @xmath213 separately , as they are especially easy to exponentiate . to fix the normalization
, we chose @xmath214 , while the @xmath210 non - diagonal @xmath215 and @xmath216 have just two non - vanishing entries , namely @xmath217 in the upper right half and @xmath107 in the lower left ( plus sign for @xmath215 , minus sign for @xmath216 ) . note that this normalization and therefore also @xmath207differs from the common convention , but is of no importance in the following ( see also refs .
@xcite ) . a general element of @xmath4 can then be written as the product of all rotations : @xmath218 the equivalence of this parametrization to the initial @xmath202 can be proven using the baker - campbell - hausdorff formula and the fact that the generators satisfy a lie algebra .
the order of the rotations in eq .
is of course arbitrary @xcite and should be chosen to simplify given expressions .
the three different types of rotations take the form @xmath219 and @xmath220 with real angles @xmath221 .
another useful parametrization for a general unitary matrix along the same lines as above is given by @xmath222 where we combined the generators @xmath216 and @xmath215 that have the same vanishing entries .
this yields a product of matrices of the type @xmath223 with the real parameters @xmath224 , @xmath225 and @xmath226 in our chosen normalization .
once again , the ordering of the matrices in eq .
is arbitrary .
this appendix provides the necessary measures for the anarchy approach in sec .
[ sec : anarchy ] .
we refer to ref .
@xcite for a detailed derivation of the known measures , and simply quote the result for real and complex majorana @xmath197 matrices : @xmath227 where the decompositions @xmath228 and @xmath229 were used . @xmath230 and @xmath231 denote the usual haar measure for @xmath232 and @xmath4 , respectively , while the diagonal matrix @xmath233 contains the eigenvalues of @xmath234 .
we will omit the calculation of the normalization factor for all measures in this appendix , as they are irrelevant for our considerations .
it can in principle be calculated by integration over some volume , e.g. @xmath235 , up to the arbitrary overall mass scale @xmath236 . for dirac matrices , the measures read @xcite @xmath237 with @xmath238 , @xmath239 , and the haar measures for @xmath232 @xmath240 and @xmath4 @xmath241 respectively . for @xmath242 ,
unphysical phases in the overlap of @xmath243 and @xmath244 should be modded out , once a parametrization is specified .
these invariant measures are of course not unique , but rather the simplest consistent ansatz that give the right mass dimension .
contains @xmath245 independent real parameters , so the measure @xmath246 has mass dimension @xmath245 . rewriting @xmath246 in the form
matches this mass dimension , as can be easily verified . ] with the above equations , one can easily write down the measure for an hermitian matrix @xmath247 under @xmath4 : @xmath248 as a cross - check , one can plug the hermitian matrix @xmath249 into eq . to rederive eq . .
we can now give a measure for a non - quadratic @xmath46 complex dirac matrix @xmath250 with @xmath251 , @xmath252 .
considering @xmath253 shows that @xmath254 should at least contain a factor @xmath255 which however has too small of a mass dimension .
the same procedure for the @xmath14 hermitian @xmath256 runs into problems , because @xmath257 of the eigenvalues are zero and would give a vanishing measure .
taking the product only over the non - vanishing factors however yields @xmath258 which has the right mass dimension ( @xmath49 ) and can therefore be viewed as the linear measure for @xmath259 .
the overlap of @xmath243 and @xmath244this time including angles , see sec .
[ sec : top - down]should again be modded out in a given parametrization to avoid overcounting of physically equivalent configurations ( similar to gauge - fixing , as pointed out in ref .
@xcite ) .
the measure can of course be easily extended to a general measure over @xmath260 or @xmath261 with @xmath262 : @xmath263 with the @xmath264 matrices @xmath265 and @xmath266 .
unphysical rotations need to be modded out again .
a simple check shows that the mass dimensions are right and @xmath267 indeed leads back to eq . .
. can actually be obtained from exactly these requirements , without the need to use hermitian matrices in the derivation .
for example , @xmath268 should contain the known @xmath267 factor @xmath269 , which leaves some factor of mass dimension @xmath270 to the power of @xmath271 .
since a factor @xmath272which would increase the hierarchy in the @xmath48has mass dimension @xmath273 , it can only be used in special cases , not for general @xmath274 ( at least for integer @xmath275 and @xmath276 ) .
thus , for general @xmath274 , the only factor with mass dimension @xmath270 is @xmath277 , leading back to eq . .
in this appendix we collect the formulas for the case @xmath5 ( see also refs .
@xcite ) , which is a little different than the @xmath6 cases and would interrupt the flow if included in the main text . with the @xmath278 diagonal matrix @xmath13 we find the top - down parametrization in the singular value decomposition for the yukawa coupling @xmath279
two phases of @xmath54 can be absorbed by @xmath68 , three of @xmath53 in the lepton fields . since one of the diagonal phase rotations of @xmath53 acts on the zeros , there are only two phases in @xmath53 left , so we can write @xmath280 with the shorthand notation @xmath79 etc . for the mixing angles . the linear measure for @xmath281 for the anarchy framework
can be derived along the same lines as in app .
[ sec : measures ] , with the result @xmath282 finally , the bottom - up parametrization is given by eq .
with two different @xmath92 matrices depending on the low - energy neutrino hierarchy ( @xmath283 ) @xcite : @xmath284 p. minkowski , phys .
b * 67 * , 421 ( 1977 ) ; + m. gell - mann , p. ramond , and r. slansky , in _ supergravity _ , p. 315
, edited by p. van nieuwenhuizen and d. freedman , north holland , amsterdam , 1979 ; + t. yanagida , proc . of the _ workshop on unified theory and the baryon number of the universe
_ , kek , japan , 1979 ; + r. n. mohapatra and g. senjanovi , phys . rev .
lett . * 44 * , 912 ( 1980 ) .
see , for example , s. f. king , nucl .
b * 576 * , 85 ( 2000 ) [ http://arxiv.org/abs/hep-ph/9912492[hep-ph/9912492 ] ] ; + p. h. frampton , s. l. glashow and t. yanagida , phys .
b * 548 * , 119 ( 2002 ) [ http://arxiv.org/abs/hep-ph/0208157[hep-ph/0208157 ] ] ; + s. antusch , p. di bari , d. a. jones and s. f. king , phys .
d * 86 * , 023516 ( 2012 ) [ http://arxiv.org/abs/1107.6002[arxiv:1107.6002 ] [ hep - ph ] ] ; + k. harigaya , m. ibe and t. t. yanagida , phys .
d * 86 * , 013002 ( 2012 ) [ http://arxiv.org/abs/1205.2198[arxiv:1205.2198 ] [ hep - ph ] ] . k. r. dienes , e. dudas and t. gherghetta , nucl .
b * 557 * , 25 ( 1999 ) [ http://arxiv.org/abs/hep-ph/9811428[hep-ph/9811428 ] ] ; + g. r. dvali and a. y. smirnov , nucl .
b * 563 * , 63 ( 1999 ) [ http://arxiv.org/abs/hep-ph/9904211[hep-ph/9904211 ] ] ; + n. arkani - hamed , s. dimopoulos , g. r. dvali and j. march - russell , phys .
d * 65 * , 024032 ( 2001 ) [ http://arxiv.org/abs/hep-ph/9811448[hep-ph/9811448 ] ] .
j. schechter and j. w. f. valle , phys .
d * 22 * , 2227 ( 1980 ) ; + g. c. branco , l. lavoura and m. n. rebelo , phys . lett .
b * 180 * , 264 ( 1986 ) ; + y. liao , nucl .
b * 749 * , 153 ( 2006 ) [ http://arxiv.org/abs/hep-ph/0604016[hep-ph/0604016 ] ] .
z. z. xing , phys .
b * 660 * , 515 ( 2008 ) [ http://arxiv.org/abs/0709.2220[arxiv:0709.2220 ] [ hep - ph ] ] ; + z. z. xing , phys .
d * 85 * , 013008 ( 2012 ) [ http://arxiv.org/abs/1110.0083[arxiv:1110.0083 ] [ hep - ph ] ] . for recent reviews ,
see w. rodejohann , int .
j. mod .
e * 20 * , 1833 ( 2011 ) [ http://arxiv.org/abs/1106.1334[arxiv:1106.1334 ] [ hep - ph ] ] ; + j. j. gmez - cadenas , j. martn - albo , m. mezzetto , f. monrabal and m. sorel , riv .
nuovo cim . * 35 * , 29 ( 2012 ) [ http://arxiv.org/abs/1109.5515[arxiv:1109.5515 ] [ hep - ex ] ] .
j. a. casas and a. ibarra , nucl .
b * 618 * , 171 ( 2001 ) [ http://arxiv.org/abs/hep-ph/0103065[hep-ph/0103065 ] ] ; + j. a. casas , a. ibarra and f. jimnez - alburquerque , jhep * 0704 * , 064 ( 2007 ) [ http://arxiv.org/abs/hep-ph/0612289[hep-ph/0612289 ] ] .
a. de gouva and w. c. huang , phys .
d * 85 * , 053006 ( 2012 ) [ http://arxiv.org/abs/1110.6122[arxiv:1110.6122 ] [ hep - ph ] ] ; + p.
s. b. dev , r. franceschini and r. n. mohapatra , phys .
d * 86 * , 093010 ( 2012 ) [ http://arxiv.org/abs/1207.2756[arxiv:1207.2756 ] [ hep - ph ] ] .
e. k. akhmedov , m. lindner , e. schnapka and j. w. f. valle , phys . rev .
d * 53 * , 2752 ( 1996 ) [ http://arxiv.org/abs/hep-ph/9509255[hep-ph/9509255 ] ] ; + s. m. barr , phys .
lett . *
92 * , 101601 ( 2004 ) [ http://arxiv.org/abs/hep-ph/0309152[hep-ph/0309152 ] ] ; + c. h. albright and s. m. barr , phys .
d * 69 * , 073010 ( 2004 ) [ http://arxiv.org/abs/hep-ph/0312224[hep-ph/0312224 ] ] .
thorough discussions can be found in s. antusch , c. biggio , e. fernndez - martnez , m. b. gavela and j. lpez - pavn , jhep * 0610 * , 084 ( 2006 ) [ http://arxiv.org/abs/hep-ph/0607020[hep-ph/0607020 ] ] ; + d. v. forero , s. morisi , m. trtola and j. w. f. valle , jhep * 1109 * , 142 ( 2011 ) [ http://arxiv.org/abs/1107.6009[arxiv:1107.6009 ] [ hep - ph ] ]
. l. j. hall , h. murayama and n. weiner , phys .
lett . * 84 * , 2572 ( 2000 ) [ http://arxiv.org/abs/hep-ph/9911341[hep-ph/9911341 ] ] ; + g. altarelli , f. feruglio and i. masina , jhep * 0301 * , 035 ( 2003 ) [ http://arxiv.org/abs/hep-ph/0210342[hep-ph/0210342 ] ] ; + a. de gouva and h. murayama , phys . lett .
b * 573 * , 94 ( 2003 ) [ http://arxiv.org/abs/hep-ph/0301050[hep-ph/0301050 ] ] ; + a. de gouva and h. murayama , http://arxiv.org/abs/1204.1249[arxiv:1204.1249 ] [ hep - ph ] .
early references include a. y. smirnov , phys .
d * 48 * , 3264 ( 1993 ) [ http://arxiv.org/abs/hep-ph/9304205[hep-ph/9304205 ] ] ; + a. kleppe , in _
lohusalu 1995 , neutrino physics _
118125 ; + e. ma , d. p. roy and u. sarkar , phys . lett .
b * 444 * , 391 ( 1998 ) [ http://arxiv.org/abs/hep-ph/9810309[hep-ph/9810309 ] ] . | introducing @xmath0 right - handed neutrinos to the standard model yields , in general , massive active neutrinos .
we give explicit parametrizations for the involved mixing and coupling matrices in terms of physical parameters for both the top - down and the bottom - up approach for arbitrary @xmath0 .
bounds on the complex mixing angles in the bottom - up approach from perturbativity of the yukawa couplings to charged lepton flavor violation are discussed . as a novel illustration of possible effects from @xmath1 ,
we extend the neutrino anarchy framework to arbitrary @xmath0 ; we show that while the anarchic mixing angles are insensitive to the number of singlets , the observed ratios of neutrino masses prefer small @xmath0 for the simplest linear measure .
# 1 |
the @xmath1-matrix @xcite is conventionally used in the analysis of scattering data , the parameterization of scattering phase shifts and the extraction of resonant energies and widths from them .
the scattering phase shifts can also be analyzed in the @xmath0-matrix formalism of scattering theory @xcite . the inverse scattering oscillator - basis @xmath0-matrix approach was suggested in ref .
@xcite . it was further developed in ref .
@xcite where some useful analytical formulas exploited in this paper , were derived .
the @xmath0-matrix parameterization of scattering phase shifts was shown in ref .
@xcite to be very accurate in describing @xmath6 scattering data .
this parameterization was used to construct high - quality non - local @xmath0-matrix inverse scattering @xmath6 potentials jisp6 @xcite and jisp16 @xcite . in what follows ,
we demonstrate that the @xmath0-matrix can be used for a high - quality parameterization of scattering phase shifts in elastic scattering of nuclear systems using @xmath3 as an example .
the resonance parameters , its energy and width , can be easily extracted from the @xmath0-matrix parameterization .
resonance energies are conventionally associated with eigenstates above reaction thresholds obtained in various nuclear structure models , e. g. , in the shell model .
this is well - justified for narrow resonances , however these eigenstates can differ significantly from the resonance energies in the case of wide enough resonances .
the @xmath0-matrix parameterization naturally provides eigenstates that should be obtained in the shell model or any other many - body nuclear structure theory based on the oscillator basis expansion ( e. g. , in the resonating group model ) to support the experimental nucleon - nucleus scattering phase shifts in any given model space and with any given oscillator spacing @xmath2 .
the shell model eigenstates are provided by the @xmath0-matrix phase shift parameterization not only in the case of resonances , narrow and wide ones , but also in the case of non - resonant scattering as well , for example , in the case of @xmath3 scattering in the @xmath7 partial wave .
we will explore these correspondences between the @xmath0-matrix properties and results from nuclear structure calculations in some detail below .
next , we extend the oscillator - basis @xmath0-matrix inverse scattering approach of ref .
@xcite to the case of charged particles using the formalism developed in ref .
this extended formalism is shown to work well in the description of @xmath4 scattering and the extraction of @xmath4 resonance energies and widths .
the shell model eigenstates desired for the description of the experimental phase shifts , are also provided by the coulomb - extended @xmath0-matrix inverse scattering formalism .
we also carry out no - core shell model @xcite calculations of @xmath5he and @xmath5li nuclei and compare the obtained eigenstates with the ones derived from the @xmath0-matrix parameterizations of @xmath3 and @xmath4 scattering .
the @xmath0-matrix formalism @xcite utilizes either the oscillator basis or the so - called laguerre basis of a sturmian type .
the oscillator basis is of a particular interest for nuclear applications . here
we present a sketch of the oscillator - basis @xmath0-matrix formalism ( more details can be found in refs .
@xcite ) and some details of the inverse scattering @xmath0-matrix approach of ref . @xcite .
the extension of @xmath0-matrix inverse scattering formalism to the case of charged particles is suggested in subsection [ iscoul ] while subsection [ isncsm ] describes how to relate the @xmath0-matrix inverse scattering results to those of the shell model .
scattering in the partial wave with orbital angular momentum @xmath8 is governed by a radial schrdinger equation @xmath9 here @xmath10 , @xmath11 is the relative coordinate of colliding particles and @xmath12 is the energy of their relative motion . within the @xmath0-matrix formalism
, the radial wave function @xmath13 is expanded in the oscillator function series @xmath14 where the oscillator functions @xmath15 @xmath16 is the associated laguerre polynomial , the oscillator radius @xmath17 , and @xmath18 is the reduced mass of the particles with masses @xmath19 and @xmath20 . the wave function in the oscillator representation @xmath21 is a solution of an infinite set of algebraic equations @xmath22 where the hamiltonian matrix elements @xmath23 , the nonzero kinetic energy matrix elements @xmath24 and the potential energy @xmath25 within the @xmath0-matrix formalism is a finite - rank matrix with elements @xmath26 the potential energy matrix truncation ( [ trunc ] ) is the only approximation of the @xmath0-matrix approach .
the kinetic energy matrix is not truncated , the wave functions are eigenvectors of the infinite hamiltonian matrix @xmath27 which is a superposition of the truncated potential energy matrix @xmath28 and the infinite tridiagonal kinetic energy matrix @xmath29 .
note that the hamiltonian matrix , i. e. both the kinetic and potential energy matrices , are truncated in conventional oscillator - basis approaches like the shell model .
hence the @xmath0-matrix formalism can be used for a natural extension of the shell model .
note also that within the inverse scattering @xmath0-matrix approach , when the potential energy is represented by the finite matrix ( [ trunc ] ) , one obtains the exact scattering solutions , phase shifts and other observables in the continuum spectrum ( see @xcite for more details ) .
the phase shift @xmath30 and the @xmath31-matrix are expressed in the @xmath0-matrix formalism as @xmath32 @xmath33 where @xmath34 is the rank of the potential energy matrix ( [ trunc ] ) , the kinetic energy matrix elements @xmath29 are given by eqs .
( [ tnm ] ) , regular @xmath35 and irregular @xmath36 eigenvectors of the infinite kinetic energy matrix are @xmath37 @xmath38 @xmath39 , @xmath40 is a confluent hypergeometric function , the dimensionless momentum @xmath41 .
the matrix elements , @xmath42 are expressed through the eigenvalues @xmath43 and eigenvectors @xmath44 of the truncated hamiltonian matrix , i. e. @xmath43 and @xmath44 are obtained by solving the algebraic problem @xmath45 only one diagonal matrix element @xmath46 , @xmath47 is responsible for the phase shifts and the @xmath31-matrix .
the @xmath0-matrix wave function is given by eq .
( [ eq : row ] ) where @xmath48 in the ` asymptotic region ' of the oscillator model space , @xmath49 .
asymptotic behavior @xcite of functions @xmath50 and @xmath51 defined as infinite series , @xmath52 and @xmath53 [ here @xmath54 and @xmath55 are spherical bessel and neumann functions , and momentum @xmath56 , assures the correct asymptotics of the wave function ( [ eq : row ] ) at positive energies @xmath12 , @xmath57 \\ \mathop{\longrightarrow}\limits_{r\to\infty } \frac{1}{r } \ , \sin[kr+\delta_l-\pi l/2 ] .
\label{asm - gen } \end{gathered}\ ] ] in the ` interaction region ' , @xmath58 , @xmath21 are expressed through matrix elements @xmath59 ( see @xcite for more details ) .
however a limited number of rapidly decreasing with @xmath60 terms with @xmath61 in expansion ( [ eq : row ] ) does not affect asymptotics of the continuum spectrum wave function .
a similarity between the @xmath0-matrix and @xmath1-matrix approaches was discussed in detail in ref .
note that the oscillator function @xmath62 tends to a @xmath63-function in the limit of large @xmath64 @xcite , @xmath65 where @xmath66 is the classical turning point of the harmonic oscillator eigenstate described by the function @xmath62
. therefore expansion ( [ eq : row ] ) describes the wave function @xmath67 at large distances from the origin in a very simple manner : each term with large enough @xmath64 gives the amplitude of @xmath67 at the respective point @xmath68 . within the @xmath0-matrix approach , the oscillator representation wave functions @xmath21 in the ` asymptotic region ' of @xmath49 and in the ` interaction region ' of @xmath69 are matched at @xmath70 @xcite .
this is equivalent to the @xmath1-matrix matching condition at the channel radius @xmath71 the @xmath0-matrix formalism reduces to those of the @xmath1-matrix with channel radius @xmath72 if @xmath73 is asymptotically large . in particular , the function @xmath46 [ see ( [ oscrmnn ] ) ] was shown in ref .
@xcite to be proportional to the @xmath74-matrix ( that is the inverse @xmath1-matrix ) in the limit of @xmath75
. at small enough values of @xmath64 , oscillator functions @xmath62 differ essentially from the @xmath63-function .
therefore the @xmath0-matrix with realistic values of truncation boundary @xmath73 differs essentially from the @xmath1-matrix approach with realistic channel radius values @xmath76 .
it appears that the @xmath0-matrix formalism with its matching condition in the oscillator model space , is somewhat better suited to traditional nuclear structure models like the shell model . in the inverse scattering @xmath0-matrix approach ,
the phase shifts @xmath30 are supposed to be known at any energy @xmath12 and we are parameterizing them by eqs .
( [ osctg ] ) , ( [ eq : snl ] ) , ( [ eq : cnl ] ) , and ( [ oscrmnn ] ) , i. e. one should find the eigenvalues @xmath43 and the eigenvector components @xmath77 providing a good description of the phase shifts . if the set of @xmath43 and @xmath77 values is known , i. e. the function @xmath46 is completely defined , the @xmath31-matrix poles are obtained by solving numerically an obvious equation , @xmath78 where solutions for @xmath79 ( or @xmath80 ) should be searched for in the desired domain of the complex plane . knowing the phase shifts @xmath30 in a large enough energy interval @xmath81 , one gets the set of eigenenergies @xmath43 , @xmath82 1 , ... , @xmath73 by solving numerically the equation @xmath83 where @xmath84 is given by eq .
( [ an1-delta ] ) .
the equation ( [ elam ] ) has exactly @xmath34 solutions .
the last components @xmath85 of the eigenvectors @xmath44 responsible for the phase shifts and the @xmath31-matrix , are obtained as @xmath86 where @xmath87 the physical meaning of the eqs .
( [ elam ] ) , ( [ nlambda ] ) is the following .
the equation ( [ elam ] ) guarantees that the phase shifts @xmath30 exactly reproduce the experimental phase shifts at the energies @xmath88 . the equation ( [ nlambda ] )
fixes the derivatives of the phase shifts @xmath89 at the energies @xmath88 fitting them exactly to the derivatives of the experimental phase shifts at the same energies . the solutions @xmath90 and @xmath85 , @xmath82 1 , ... ,
@xmath73 depend strongly on the values of the oscillator spacing @xmath2 and @xmath73 , the size of the inverse scattering potential matrix .
larger values of @xmath73 and/or @xmath2 , imply a larger energy interval @xmath81 where the phase shifts are reproduced by the @xmath0-matrix parameterization ( [ osctg ] ) .
a hermitian hamiltonian generates a set of normalized eigenvectors @xmath44 fitting the completeness relation , @xmath91 experimental phase shifts generate a set of @xmath85 , @xmath82 1 , ... , @xmath73 that usually does not fit eq .
( [ complete ] ) .
it is likely that the interval of energy values used to find the sets of @xmath90 and @xmath85 , spreads beyond the thresholds where new channels are opened .
thus inelastic channels are present in the system suggesting the hamiltonian should become non - hermitian .
the approach proposed in ref .
@xcite , suggests to fit eq .
( [ complete ] ) by changing the value of the component @xmath92 corresponding to the largest among the energies @xmath90 with @xmath93 .
this energy @xmath94 is usually larger than @xmath95 , the maximal energy in the interval @xmath81 where the experimental phase shifts are available . therefore changing @xmath92
should not spoil the phase shift description in the desired interval of energies below @xmath95 ; more over , one can also vary subsequently the energy @xmath94 to improve the description of the phase shifts in the interval @xmath81 .
we are not discussing here the construction of the inverse scattering potential but point the interested reader to ref .
we note only that if the construction of the @xmath0-matrix inverse scattering potential is desired , one should definitely fit eq .
( [ complete ] ) , otherwise the construction of the hermitian interaction is impossible . in our applications to @xmath3 and @xmath4 scattering we are interested only in the @xmath0-matrix parameterization of scattering phase shifts ; hence we can avoid renormalization of the component @xmath96 .
nevertheless , we found out that this renormalization improves the phase shifts description at energies @xmath12 not close to @xmath43 values .
all the results presented below were obtained with the help of eq .
( [ complete ] ) .
in the case of a charged projectile scattered by a charged target , the interaction between them is a superposition of a short - range nuclear interaction , @xmath97 , and the coulomb interaction , @xmath98 : @xmath99 the coulomb interaction between proton and nucleus is conventionally described as ( see , e. g. , @xcite ) @xmath100 in the case of @xmath4 scattering discussed below , @xmath101 and @xmath102 fm @xcite .
the long - range coulomb interaction ( [ vc1 ] ) requires some modification of the oscillator - basis @xmath0-matrix formalism described in the previous subsection . in the case of charged particle scattering ,
the wave function @xmath103 at asymptotically large distances takes a form : @xmath104 , \label{eq : ascoul}\ ] ] where @xmath105 @xmath106 and @xmath107 are regular and irregular coulomb functions respectively , and sommerfeld parameter @xmath108 . instead of functions @xmath50 and @xmath51
, one can introduce functions @xmath109 and @xmath110 defining them as infinite series , @xmath111 and @xmath112 in order to use @xmath113 and @xmath114 in constructing continuum spectrum wave functions by means of eq .
( [ eq : row ] ) .
such an approach was proposed by the kiev group in ref .
@xcite . within this approach
, the @xmath0-matrix matching condition at @xmath70 becomes much more complicated , resulting in difficulties in designing an inverse scattering approach and in shell model applications . in practical calculations , the approach of ref .
@xcite requires the use of much larger values of @xmath73 , i. e. a huge extension of the model space when solving the algebraic problem ( [ alge ] ) , that makes it incompatible with the shell model applications .
therefore it is desirable to find another way to extend our approach on the case of charged particle scattering .
we use here the formalism of ref .
@xcite to allow for the coulomb interaction in the oscillator - basis @xmath0-matrix theory .
the idea of the approach is very simple .
suppose there are a long - range @xmath115 and a short - range @xmath116 potentials that are indistinguishable at distances @xmath117 . in this case
, the potential @xmath116 generates a wave function fitting exactly ( up to an overall normalization factor ) that of the long - range potential @xmath115 at @xmath118 .
if the only difference between @xmath115 and @xmath116 at distances @xmath119 is the coulomb interaction , then one can equate the logarithmic derivatives of their wave functions at @xmath71 and use the resulting equation to express the long - range potential phase shifts @xmath30 in terms of the short - range potential phase shifts @xmath120 or vice versa .
note that the phase shifts @xmath120 can be obtained within the standard @xmath0-matrix approach discussed in the previous subsection .
the recalculation of the phase shifts @xmath120 into @xmath30 ( or vice versa ) appears to be the only essential addition in formulating such a direct ( or inverse ) coulomb - extended @xmath0-matrix formalism .
to implement this idea , we introduce a channel radius @xmath76 large enough to neglect the nuclear interaction @xmath97 at distances @xmath121 , i. e. @xmath122 , where @xmath123 is the range of the potential @xmath97 . in the asymptotic region @xmath121 , the radial wave function @xmath103 is given by eq .
( [ eq : ascoul ] ) . at short distances @xmath124 ,
the wave function @xmath103 coincides with @xmath125 , the one generated by the auxiliary potential @xmath126 obtained by truncating the coulomb potential @xmath98 at @xmath71 .
the wave function @xmath125 behaves asymptotically as a wave function obtained with a short - range interaction , @xmath127 , \quad b \geq r_{nucl}. % % , \label{eq : asm}\ ] ] the @xmath0-matrix formalism described in the previous subsection , should be used to calculate the function @xmath125 , the auxiliary phase shift @xmath128 and the respective auxiliary @xmath31-matrix @xmath129 . matching the functions @xmath103 and @xmath125 at @xmath71
, the phase shift @xmath30 can be expressed through @xmath128 @xcite : @xmath130 where quasi - wronskian @xmath131\,{f}_{l}(\zeta , kr)\right.\right.\\ \left.\left.-{j}_{l}(kr)\,\frac{d}{dr}{f}_{l}(\zeta , kr)\right\ } \right|_{r = b},\end{gathered}\ ] ] and @xmath132 , @xmath133 and @xmath134 are expressed similarly .
the @xmath31-matrix is given by @xmath135 where @xmath136 , @xmath137 , and the quasi - wronskians @xmath138 are defined by analogy with eq .
( [ wronsk ] ) .
the @xmath31-matrix poles are obtained by solving the equation @xmath139 in the complex energy plane .
this formalism involves a free parameter , the channel radius @xmath76 , used for construction of the auxiliary potential @xmath116 .
as mentioned above , @xmath76 should be taken larger than the range of the short - range nuclear interaction @xmath97 .
on the other hand , the truncated @xmath140 hamiltonian matrix @xmath141 ( @xmath142 , 1 , ... , @xmath73 ) used to calculate the sets of eigenvalues @xmath90 and eigenvectors @xmath143 by solving the algebraic problem ( [ alge ] ) , should carry information about the jump of potential @xmath116 at the point @xmath71 .
therefore @xmath76 should be chosen less than approximately @xmath144 , the classical turning point of the oscillator function @xmath145 , the function with the largest range in the set of oscillator functions @xmath62 , @xmath146 , 1 , ... , @xmath73 used for the construction of the truncated hamiltonian matrix @xmath147 @xmath148 . in a practical calculation
, one should study convergence with a set of @xmath76 values and pick up the @xmath76 value providing the most stable and best - converged results .
as shown in ref .
@xcite , the phase shift @xmath30 calculated at some energy @xmath12 as a function of channel radius @xmath76 , usually has a plateau in the interval @xmath149 that reproduces well the exact values of @xmath30 . in the inverse scattering approach ,
first , we fix a value of the channel radius @xmath76 and transform experimental phase shifts @xmath30 into the set of auxiliary phase shifts @xmath128 : @xmath150 equation ( [ tanret ] ) can be easily obtained by inverting eq . ( [ tantan ] ) .
next , we employ the inverse scattering approach of the previous subsection to calculate the sets of @xmath90 and @xmath85 using auxiliary phase shifts @xmath128 as an input .
the @xmath0-matrix parameterization of the phase shifts @xmath30 is given by eq .
( [ tantan ] ) , the @xmath31-matrix poles can be calculated through eq .
( [ spolecl ] ) . up to this point
we have been discussing the @xmath0-matrix formalism supposing the colliding particles to be structureless . in applications to the @xmath3 and @xmath4 scattering and relating the respective @xmath0-matrix inverse scattering results to the shell model , we should have in mind that the @xmath151 particle consists of 4 nucleons identical to the scattered nucleon and the five - nucleon wave function should be antisymmetrized .
the @xmath0-matrix solutions and the expressions ( [ osctg ] ) for the phase shifts and ( [ smat ] ) for the @xmath31-matrix [ or expressions ( [ tantan ] ) and ( [ smatcl ] ) in the case when both the projectile and the target are charged ] , can be used in the case of scattering of complex systems comprising identical fermions .
the components @xmath152 entering expression ( [ oscrmnn ] ) for the function @xmath46 become , of course , much more complicated : they now appear to be some particular components of the many - body eigenvector .
however , we are not interested here in the microscopic many - body structure of the components @xmath153 ; we shall obtain them by fitting the @xmath3 and @xmath4 phase shifts in the @xmath0-matrix inverse scattering approach .
we focus our attention here on other important ingredients entering expression ( [ oscrmnn ] ) for @xmath46 , the eigenenergies @xmath43 , related to the energies of the states in the combined many - body system , i. e. in the @xmath5he or @xmath5li nucleus in the case of @xmath3 or @xmath4 scattering respectively , obtained in the shell model or any other many - body approach utilizing the oscillator basis .
one should have however in mind that @xmath43 entering eq .
( [ oscrmnn ] ) correspond to the kinetic energy of relative motion , i. e. they are always positive , while many - body microscopic approaches generate eigenstates with absolute energies , e. g. all the states in @xmath5he and @xmath5li with excitation energies below approximately 28 mev ( the @xmath151-particle binding energy ) will be generated negative .
therefore , before comparing with the set of @xmath43 values , one should perform a simple recalculation of the shell model eigenenergies by adding to them the @xmath154he binding energy ; or alternatively one can use the set of @xmath43 values to calculate the respective set of energies defined according to the shell model definitions by subtracting the @xmath154he binding energy from each of @xmath43 .
the physical meaning of transforming these to the shell model scale of values for @xmath43 is to provide the values required from shell model calculations in order to reproduce the desired phase shifts .
the comparison of the inverse scattering @xmath0-matrix analysis with the shell model results is useful , of course , only if the same @xmath2 value is used both in the @xmath0-matrix and in the shell model and model spaces of these approaches are properly correlated .
a traditional notation for the model space within the shell model is @xmath155 where @xmath156 is the excitation oscillator quanta . in the case of the @xmath0-matrix ,
we use , also traditionally , @xmath73 , the principal quantum number of the highest oscillator function @xmath145 included in the ` interaction region ' of the oscillator model space where the potential energy matrix elements are retained .
the following expressions relate @xmath156 and @xmath73 in the cases of @xmath157 and @xmath158 partial waves ( @xmath159 waves ) and @xmath7 partial wave ( @xmath160 wave ) : @xmath161 below we are using shell model type @xmath155 notations for labeling both @xmath0-matrix and shell model results .
the @xmath0-matrix inverse scattering approach was well - tested in the nucleon - nucleon ( @xmath6 ) scattering in ref .
@xcite . in the case of @xmath6-scattering ,
the phase shifts are well established in a wide range of energies up to 350 mev in laboratory system .
the goal of ref .
@xcite was to fit scattering phase shift in the entire interval of energies @xmath162 mev using the smallest possible potential matrices or , equivalently , the smallest possible values of @xmath156 ( or @xmath73 ) . in the case of nucleon-@xmath151 scattering ,
the phase shifts are known in a small energy interval up to @xmath163 mev and in some cases up to 25 mev . on the other hand , to compare the @xmath0-matrix analysis with the shell model results , we are interested in large enough values of @xmath156 and in @xmath2 values reasonable for shell model applications . as a result ,
we face a problem of insufficient data : some solutions @xmath43 of eq .
( [ elam ] ) should be allowed far outside the interval of known phase shifts @xmath30 .
the required phase shifts should be known together with their derivatives at the energies around @xmath164 just to find these solutions @xmath43 [ see eq . ( [ an1-delta ] ) ] and respective eigenvector components @xmath85 [ see eqs .
( [ nlambda ] ) and ( [ allambda ] ) ] .
we address the problem of data insufficiency by an extrapolation of the data outside the energy interval of known phase shifts .
the @xmath0-matrix parameterizations presented in fig .
[ na - ext ] were obtained with @xmath165 mev in various model spaces . in each case
two different extrapolations were used for the phase shifts at energies @xmath166 mev , however the experimental phase shifts below @xmath163 mev are equivalently well described if @xmath156 is large enough .
the deviation of the parameterization from the experiment is seen at energies @xmath167 mev only in the case of the @xmath168 model space , the smallest among all model spaces presented in fig .
[ na - ext ] , and even in this case the deviation is small enough .
this is not surprising since the phase shifts given by eqs .
( [ osctg ] ) and ( [ oscrmnn ] ) in the low - energy interval are governed mostly by the @xmath43 values from the same interval and by the respective eigenvector components @xmath153 .
these @xmath43 and @xmath153 values are determined by eqs .
( [ elam ] ) and ( [ nlambda ] ) locally , i. e. they are independent from the phase shift extrapolation . note that in the case of the @xmath168 model space , both @xmath43 values lie in the energy interval of known phases , hence the parameterization in this model space is completely independent from the extrapolation and the two parameterizations obtained in this model space coincide .
the resonance energy @xmath169 and width @xmath170 calculated by locating the @xmath31-matrix pole by solving eq .
( [ spole ] ) , are seen from table [ tab - extr ] to be very stable and insensitive to the extrapolation of the phase shifts .
.the energy @xmath169 and width @xmath170 ( both in mev ) of the @xmath157 resonance in the @xmath3 scattering obtained with @xmath165 mev in various model spaces @xmath155 with two different extrapolations of the phase shifts . [ cols="^,^,^,^,^ " , ] a very interesting and important output of the @xmath0-matrix inverse scattering analysis of the phase shifts is the set of @xmath43 values which are directly related to the eigenenergies obtained in the shell model or any other model utilizing the oscillator basis , for example , the resonating group model .
the @xmath0-matrix parameterizations provide the energies of the states that should be obtained in the shell model or resonating group model to generate the given phase shifts .
these energies are shown to be model space and @xmath2-dependent and very different from the energies of at least wide enough resonances which are conventionally used to compare with the shell model results .
more , the @xmath0-matrix analysis is shown to provide the shell model energies even in the case of non - resonant scattering such as the @xmath7 nucleon@xmath151 scattering .
our comparison of the lowest @xmath171 with the no - core shell model results shows that the shell model fails to reproduce the phase shifts if small @xmath2 values are employed in the calculations .
when @xmath2 and/or model space size is increased , the shell model predictions approach @xmath171 values obtained in the @xmath0-matrix signaling that the shell model results become more and more consistent with the experimental phase shifts .
however some difference between the no - core shell model predictions and the @xmath0-matrix analysis results is seen even in the largest model spaces used in this study .
this difference is really not large , its possible sources are the following . ( i ) there is an ambiguity in the threshold energies used to relate the absolute negative energies obtained in the shell model and positive @xmath43 values defined relative to the reaction threshold .
( ii ) unfortunately , there is no @xmath6 interaction providing correct energies for , at least , light nuclei .
the jisp16 @xmath6 interaction is good enough and provides reliable predictions for energies of levels in all @xmath160 and @xmath159 shell nuclei @xcite .
however , there are small differences between jisp16 level energy predictions and experiment ; these differences are of the same order as the differences between the @xmath0-matrix @xmath171 values and our no - core shell model results .
probably we shall use the @xmath0-matrix results discussed above while attempting to design a new improved version of the jisp16 interaction by trying to eliminate the discrepancy between the shell model results and the @xmath0-matrix analysis of nucleon-@xmath151 scattering .
of course , the @xmath0-matrix can be used to relate the shell model energies and data on nucleon scattering by other nuclei .
generally , one can also use other elastic scattering data , for example , nucleus - nucleus elastic scattering phase shifts to get the @xmath90 values that should be obtained in the shell model studies of the respective compound nuclear systems : the shell model must generate the states with the same energies in the same model space and with the same @xmath2 value to have a chance to generate the experimental phase shifts .
+ we are thankful to g. m. hale and p. maris for valuable discussions and help in our studies .
this work was supported in part by the russian foundation of basic research , by the us doe grants de - fc02 - 07er41457 and de - fg02 - 87er40371 .
a. m. shirokov , a. i. mazur , s. a. zaytsev , j. p. vary , and t. a. weber , phys . rev .
c * 70 * , 044005 ( 2004 ) ; in _ the @xmath0-matrix method . developments and applications _ , edited by a. d. alhaidari , h. a. yamani , e. j. heller , and m. s. abdelmonem ( springer , 2008 ) , 219 .
a. m. shirokov , yu . f. smirnov , and s. a. zaytsev , in _ modern problems in quantum theory _ , edited by v. i. savrin and o. a. khrustalev , ( moscow state university , moscow , 1998 ) , 184 ; teoret .
fiz . * 117 * , 227 ( 1998 ) [ theor .
* 117 * , 1291 ( 1998 ) ] .
a. lurie and a. m. shirokov , ann .
( ny ) * 312 * , 284 ( 2004 ) ; in _ the @xmath0-matrix method . developments and applications _ , edited by a. d. alhaidari , h. a. yamani , e. j. heller , and m. s. abdelmonem ( springer , 2008 ) , 183 .
the forbidden state in this model is a particular case of so - called isolated states , see a. m. shirokov and s. a. zaytsev , in _ the @xmath0-matrix method .
developments and applications _ , edited by a. d. alhaidari , h. a. yamani , e. j. heller , and m. s. abdelmonem ( springer , 2008 ) , 103 ; quant - ph/0312065 ( 2003 ) ; s. a. zaytsev , yu .
f. smirnov , and a. m. shirokov , izv .
nauk , ser . fiz . * 56 * , 80 ( 1992 ) .
j. p. vary , the many - fermion - dynamics shell - model code , " iowa state university , 1992 ( unpublished ) ; j. p. vary and d. c. zheng , _ ibid _ 1994 ( unpublished ) ; test runs can be performed through http://nuclear.physics.iastate.edu/mfd.php . | the @xmath0-matrix inverse scattering approach can be used as an alternative to a conventional @xmath1-matrix in analyzing scattering phase shifts and extracting resonance energies and widths from experimental data .
a great advantage of the @xmath0-matrix is that it provides eigenstates directly related to the ones obtained in the shell model in a given model space and with a given value of the oscillator spacing @xmath2 .
this relationship is of a particular interest in the cases when a many - body system does not have a resonant state or the resonance is broad and its energy can differ significantly from the shell model eigenstate .
we discuss the @xmath0-matrix inverse scattering technique , extend it for the case of charged colliding particles and apply it to the analysis of @xmath3 and @xmath4 scattering .
the results are compared with the no - core shell model calculations of @xmath5he and @xmath5li . |
during the last years considerable attention has been paid to interaction of @xmath5-mesons with four nucleons @xcite .
analysis of different data is mainly focused on the search for @xmath4he bound states . according to the available experimental results , the rise of the @xmath6he experimental cross section at @xmath7 seems to be not as steep as in the @xmath8he reaction . as discussed , e.g. , in ref.@xcite
the most natural interpretation of this fact is that due to additional attraction caused by one extra nucleon the pole in the @xmath4he scattering matrix is shifted into the region of negative values of @xmath9 and turns out to be farther from the physical region than the @xmath10he pole .
it is therefore concluded that formation of the bound @xmath4he state is highly probable . in view of general complexity of the five - body @xmath0 problem
there are still no rigorous few - body calculations of this system . at the same time , a systematic practical way of handling the @xmath11-body interaction is provided by the quasi - particle formalism in which the kernels of integral equations are represented by series of separable terms .
this method becomes especially efficient if the driving two - particle potentials are governed by the nearly lying resonances or bound ( virtual ) states , like in the @xmath2 and @xmath1 case .
then reasonable accuracy may be achieved with only few separable terms retained in the series .
in particular , the quasi - particle formalism is shown to be very well suited for practical calculation of @xmath12 @xcite as well as @xmath13 @xcite scattering ( in ref.@xcite another method based on the hypospherical function expansion has been developed ) . in this letter
we apply the quasi - particle method to study the five - body system @xmath0 . as a formal basis
we use the alt - grassberger - sandhas @xmath11-body equations derived in ref.@xcite . for the sake of simplicity we neglect influence of the spin and isospin on the interaction between nucleons , treating them as spinless indistinguishable particles .
furthermore , since only the threshold @xmath4he energies are considered , we restrict all interactions to @xmath3-waves only .
as is well known , separable expansion of the kernels allows one to reduce the @xmath11-body integral equations to the @xmath14-body equations , where two of @xmath11 particles in each state are effectively treated as a composite particle ( quasi - particle ) .
therefore , the essence of the method is to approximate the @xmath14-particle interaction obtained in the separable - potential model again by the separable ansatz . in this respect , to simplify presentation of the formalism , we start directly from successive application of the quasi - particle technique to 2- , 3- , and 4-body subamplitudes , occurring when the five - body system is divided into groups of mutually interacting particles . in
what follows , we use the concept of partitions as introduced , e.g. , in ref.@xcite .
different partitions ( as well as the quasi - particles related to these partitions ) are further denoted by the symbols @xmath15 , whereas the latin letters @xmath16 are used for numbering the terms in the separable expansions of the subamplitudes .
the notation @xmath17 refers to the partition obtained by dividing the @xmath0 system into @xmath11 groups .
writing @xmath18 means that the partition @xmath19 is obtained from @xmath20 via further division of the quasi - particle @xmath20 into two groups of particles .
the basic ingredient of the formalism is a separable expansion of the quasi - particle amplitudes @xmath21 then the integral equations for the amplitudes @xmath22 are transformed exactly into the quasi - two - body equations which in the operator form read @xmath23 or more explicitly @xmath24 with @xmath25 being the reduced mass associated with the partition @xmath26 .
the effective potentials @xmath27 are determined as matrix elements of the resolvent @xmath28 between the form factors appearing in the expansion ( [ eq1_10 ] ) @xmath29 the structure of eq.([eq1_25 ] ) is conveniently illustrated in the form of diagrams . in fig.[fig0 ] we show as an example one of the effective potentials @xmath30 , connecting two configurations of the type @xmath31 .
since the nucleons are identical , the condition @xmath32 in eq.([eq1_25 ] ) means that the nucleon lines , entering the quasi - particles @xmath17 and @xmath33 and not included into the quasi - particle @xmath34 should be different . to calculate the form factors @xmath35 and the propagators @xmath36 we employed the energy dependent pole expansion ( edpe ) method of ref.@xcite .
considering nucleons as indistinguishable particles we have only two different types of four - particle partitions : @xmath37 the partitions 1 and 2 and the related two - particle subsystems @xmath2 and @xmath1 will further be labeled by the index @xmath38 . in the present calculation ,
the @xmath2 and @xmath1 @xmath3-wave interactions were approximated by simplest rank - one separable potentials .
for @xmath2 we employed @xmath39 the corresponding @xmath40-matrix has the usual form @xmath41 with the @xmath2 propagator @xmath42^{-1}\,,\ ] ] where @xmath43 is the nucleon mass . the form factors were chosen in the yamaguchi form @xmath44 since we treat nucleons as spinless particles , the strength @xmath45 was taken as an average of the singlet and the triplet strength @xmath46 the singlet and the triplet scattering lengths , @xmath47 and @xmath48 , as well as the cut - off momentum @xmath49 were taken directly from the analysis @xcite of the low - energy @xmath50 scattering @xmath51 it is well known that the yamaguchi @xmath2 potential overestimates attraction at high momenta and yields significant overbinding already in the @xmath52he case ( see table [ tab1 ] ) .
therefore we also adopted the spin - independent @xmath2 potential with exponential form factors @xmath53 which yields the same binding energy @xmath54 of two nucleons .
the form factors ( [ eq2_23 ] ) with parameters listed in table [ tab1 ] give for the three- and four - nucleon binding energies , @xmath55 and @xmath56 , the values which are rather close to those of the @xmath52he and @xmath57he nuclei .
at the same time , with the gauss form factors we have a visibly larger value of the @xmath2 effective range @xmath58 ( see table [ tab1 ] ) .
9.1cm@c@|@c@c@ c@c@c@c@ type & @xmath45 & @xmath49 & @xmath54 & @xmath58 & @xmath55 & @xmath56 + & @xmath59 & @xmath60 & mev & fm & mev & mev + yamaguchi & 4.17 & 1.45 & 0.428 & 1.89 & 12.6 & 54.8 + gauss & 6.51 & 1.24 & 0.428 & 2.33 & 8.05 & 30.3 + [ tab1 ] the @xmath1 @xmath3-wave interaction was reduced to excitation of the resonance @xmath61 only . to include pions we used a conventional coupled channel formalism , where the resulting separable @xmath40-matrix has the matrix form @xmath62 with @xmath63 the propagator @xmath64 with @xmath65 , where @xmath66 is the @xmath5 mass ,
is determined by the @xmath61 self - energies @xmath67 and @xmath68 .
the two - pion channel was included via the @xmath69 decay width @xmath70 parametrized in the form @xmath71 the parameters @xmath72 , @xmath73 , @xmath74 , @xmath75 , @xmath76 , and @xmath77 were chosen in such a way that the scattering amplitude @xmath78 corresponding to our @xmath40-matrix @xmath79 ( [ eq2_40 ] ) is close to that obtained in the coupled - channel analyses in the energy region from 20 mev above the @xmath1 threshold to 100 mev below the threshold .
here we took the results of two works @xcite and @xcite predicting rather different values of @xmath80 ( see fig.[fig1 ] ) .
9.1cm@c@|@c@c@ c@c@c@c@ set [ ref .
] & @xmath72 & @xmath81 & @xmath74 & @xmath82 & @xmath76 & @xmath77 + & & mev & & mev & mev & mev + i @xcite & 1.91 & 636 & 0.651 & 850 & 1577 & 4.0 + ii @xcite & 1.23 & 636 & 1.28 & 350 & 1527 & 1.0 + [ tab2 ] we have four different three - body partitions @xmath83 which in the following are numerated by the index @xmath84 . in the latter two cases
there are two pairs of interacting particles propagating independently .
the effective potentials @xmath85 determined by eq.([eq1_25 ] ) for @xmath86 are matrix elements of the free resolvent @xmath87 between the form factors @xmath88 @xmath89 @xmath90 the functions @xmath91 are given by eqs.([eq2_22 ] ) ( or ( [ eq2_23 ] ) ) and ( [ eq2_45 ] ) with @xmath92 .
here we omit the superfluous indices @xmath93 , since our separable ansatz for @xmath2 and @xmath1 amplitudes contains in both cases only one term ( see eqs.([eq2_21 ] ) and ( [ eq2_40 ] ) ) .
there are four two - body partitions of the @xmath0 system : @xmath94 which will be labeled by @xmath95 .
the effective potentials @xmath96 are matrix elements of the resolvent @xmath97 between the form factors @xmath98 appearing in the separable expansion ( [ eq1_10 ] ) for @xmath99 : @xmath100 the propagators @xmath97 ( @xmath38 ) are given by ( [ eq2_15 ] ) and ( [ eq2_50 ] ) with @xmath101
. the calculation of the @xmath102 ( @xmath103 ) and @xmath104 ( @xmath105 ) amplitudes with separable @xmath2 potentials may be found , e.g. , in refs.@xcite and @xcite , and we refer the reader to these works .
the effective @xmath106 amplitudes ( @xmath107 ) describe propagation of two groups of mutually interacting particles .
the corresponding integral equations are schematically presented in figs.[fig2 ] and [ fig3 ] .
after the separable expansions ( [ eq1_10 ] ) for @xmath108 are calculated we build the effective potentials @xmath109 ( [ eq1_25 ] ) as latexmath:[\[\label{eq4_30 } z_{\alpha_2a,\beta_2b}=\sum_{\gamma_3=1}^4\sum_{k , l=1}^{n_{\gamma_3 } } \langle u^{\alpha_2(a)}_{\gamma_3(k)}|\delta^{\gamma_3}_{kl } the five - body @xmath0 equations is diagrammatically presented in fig.[fig4 ] .
after this system is solved , the @xmath111he scattering amplitude can be calculated as @xmath112 here @xmath113 is the normalization constant of the @xmath57he wave function , @xmath114 is the @xmath115he reduced mass , and the momentum @xmath116 is fixed by the on - mass - shell condition @xmath117 where @xmath118 is the four - nucleon binding energy given in table [ tab1 ] .
8.3cm@c@c@ c@c@ |@c@ @xmath119 & @xmath120 & @xmath121 & @xmath122 & @xmath123 [ fm ] + 2 & 2 & 2 & 2 & @xmath124 + 4 & 4 & 4 & 4 & @xmath125 + 4 & 4 & 6 & 6 & @xmath126 + 6 & 6 & 8 & 8 & @xmath127 + 10 & 10 & 12 & 12 & @xmath127 + 20 & 20 & 20 & 20 & @xmath128 + [ tab3 ] in table[tab3 ] we present the value of the @xmath4he scattering length calculated with different number @xmath129 of terms retained in the separable expansion ( [ eq1_10 ] ) of the amplitudes @xmath130 .
as one can see , satisfactory accuracy is achieved with @xmath131 , @xmath132 . in principle , already with first four terms in each expansion the resulting scattering length is within less than 2@xmath133 of the correct value .
thus , also in the five - body case @xmath0 the quasi - particle approach based on the edpe method of ref.@xcite is very suitable for practical applications .
the minimum number of separable terms @xmath129 only slightly exceeds that for the four - body kernels , where convergence is achieved already with first four - six terms in each subamplitude .
as our main result we present the @xmath4he scattering length @xmath134 . it is given in table [ tab4 ] for two versions of the @xmath2 potential .
for comparison purposes also the @xmath10he scattering length calculated with the same sets of the @xmath2 and @xmath135 parameters is presented .
it is remarkable , that despite the larger number of nucleons in @xmath57he the predicted value of @xmath136 is smaller than @xmath137 .
direct calculation shows that the main reason of this somewhat unexpected result is rather rapid decrease of the @xmath1 scattering amplitude in the subthreshold region ( see fig.[fig1 ] ) . because of essentially stronger binding of @xmath57he in comparison to @xmath52he , in the former case the effective in - medium @xmath1 interaction acts at lower internal @xmath1 energies , thus leading to general reduction of the attractive @xmath1 forces ( this question was addressed in detail in refs.@xcite )
this effective weakening may qualitatively explain why the peculiar slope in the @xmath5 spectrum at low energies seen in the data for @xmath6he @xcite and @xmath138he @xcite becomes less steep , when we turn from @xmath10he to @xmath4he .
8.3 cm @c@c@|@c@ c@ @xmath2 & @xmath135 & @xmath137 [ fm ] & @xmath136 [ fm ] + yamaguchi & i & @xmath139 & @xmath140 + & ii & @xmath141 & @xmath142 + gauss & i & @xmath143 & @xmath144 + & ii & @xmath145 & @xmath146 + summarizing , @xmath4he interaction is calculated for the first time correctly dealing with the few - body aspects of the problem .
applying separable representation firstly to the @xmath147 and @xmath148 and then to the @xmath149 and @xmath106 kernels we have solved the five - body alt - grassberger - sandhas equations reducing them to a coupled set of quasi - two - body equations having lippmann - schwinger structure .
the predicted value of @xmath150 is positive and turns out to be smaller than @xmath151 .
this finding should be attributed to effective weakening of the in - medium @xmath1 interaction .
according to our calculation , increase of the attractive forces due to an extra nucleon in @xmath57he is overwhelmed by stronger suppression of the subthreshold @xmath1 interaction in a more dense nucleus .
the resulting attraction in the @xmath0 system is too weak and does not support existence of the @xmath4he bound state , at least with the @xmath1 parameters , used in the present calculation
. this might be the key reason why no signal of @xmath4he bound state formation is still revealed , e.g. , in the @xmath152he@xmath153 and @xmath152he@xmath154 reactions @xcite .
finally , we note that although our results obviously suffer from oversimplified treatment of the @xmath2 potential , they demonstrate applicability of the quasi - particle formalism to the five - body @xmath4he problem .
the edpe method provides rather rapid convergence of the separable expansion , so that transition from @xmath13 to the @xmath0 case is performed without drastic increase of numerical complexity .
at the same time , more refined treatment requires inclusion of the nucleon spin as well as more sophisticated nucleon - nucleon potential instead of our simple rank - one ansatz . | the alt - grassberger - sandhas equations for the five - body @xmath0 problem are solved for the case of the driving @xmath1 and @xmath2 potentials limited to @xmath3-waves .
the quasi - particle ( schmidt ) method is employed to convert the equations into the effective two - body form .
numerical results are presented for the @xmath4he scattering length . |
it is generally accepted that the outburst of a classical nova is caused by a thermonuclear runaway ( tnr ) in the accreted hydrogen - rich envelope on top of a white dwarf in a close binary system ( @xcite 1989 ; @xcite 1983 ; @xcite 1995 ; @xcite 1982 ; @xcite 1990 ) .
detailed one - dimensional spherically symmetric models of the runaway , based on realistic nuclear reaction rate networks but using the mixing - length theory ( mlt ) of convection , have been studied extensively ( @xcite 1974,1985 ; @xcite 1978 ; @xcite 1980 ) , and the results are in good agreement with observational data , such as the total amount of energy released and the metallicity and the abundances of the expelled envelope material . however , using the mlt to model the runaway is very questionable , as can be seen from simple arguments . in general terms , the physics of the tnr can be described as follows .
for a white dwarf accreting matter from a companion star of roughly solar composition , during the accretion process the temperature at the base of the envelope and the nuclear energy release increase steadily but slowly .
once the temperature exceeds about @xmath1 k heat conduction can no longer transport the energy released by nuclear reactions fast enough and the envelope becomes unstable to convection .
however , because of electron degeneracy the nuclear energy released does not go into lifting off the envelope at this stage , but the temperature increases further and the tnr proceeds until the electron degeneracy is removed and an overall expansion sets in . at the onset of the violent burning phase the temperature at the bottom of the envelope is about 10@xmath2 k and it lasts for several minutes .
energy production is dominated by the @xmath3-limited cno - cycle since the temperatures never exceed 4@xmath410@xmath2 k ( see , e.g. , @xcite ( 1981 ) ) .
therefore , the energy production per unit mass is independent of temperature and density and is just a linear function of metallicity ( @xcite 1981 ) .
the limitation of the nuclear energy generation rate is the reason why hydrogen burning on top of a white dwarf is different from , say , the carbon - burning deflagration wave inside a chandrasekhar - mass white dwarf . in the latter case , due to the temperature sensitivity of the energy generation rate and the small thermal width of the burning region a flame forms and
is passively advected by the turbulent motions .
combustion is said to be in the `` flamelet '' regime .
in contrast , in the phase of cno - limited hydrogen burning on top of a white dwarf burning is always slow compared to convective turn - over and combustion takes place in the so - called `` well - stirred '' regime for which either direct numerical simulations have to be carried out or pdf - methods have to be applied . in either case , the treatment of convection is crucial and requires special care because convection dominates over burning . a second aspect which makes reliable numerical simulations of novae very difficult is the fact that due to diffusion and coupling to hydrodynamic flows ( especially convective overshooting ( @xcite 1986 ) ) there can be a significant increase of the energy production rate by mixing c and o from the white dwarf core into the envelope .
the @xmath3-limited cno - cycle operates no longer in a strict equilibrium , but due to the increase of the c and o abundances in mixed zones the energy generation rate is time - dependent and couples to the mixing processes in a very non - linear manner : violent motions can cause mixing which in turn may make the motions even more violent .
it was this effect that lead to fast self - enrichment and a rapid temperature rise in the two - dimensional direct simulations of glasner , livne & truran ( 1997 ) .
a third question concerns the ignition process .
one dimensional models with spherical symmetry consider a global and simultaneous ignition at the bottom layer of the envelope ( global tnr or gtnr ) .
however , according to shara ( 1981,1982 ) also a local tnr ( ltnr ) is possible , because of temperature inhomogeneities arising from magnetic fields , rotation and non - spherically symmetric accretion . since it would take years to spread the combustion front from a ltnr over the white dwarf s surface diffusively by radiative transport and thermal conduction
( @xcite 1981,1982 ) , fryxell & woosley ( 1982 ) discuss that heat transport by mass motions as a multidimensional effect dominates over diffusion . for typical situations at peak temperature
they find laminar burning velocities of about @xmath5 = 30 cm / s only .
in contrast , estimating the effective burning velocity caused by small - scale turbulence motions to be @xmath6 where @xmath7 is the pressure scale hight , and @xmath8 the typical convective velocity , they obtain @xmath9 @xmath10 10 km / s for a typical nova model .
so again , multi - dimensional effects can be crucial for the tnr . to date ,
not very many multi - dimensional studies exist which deal with the various questions raised above .
shankar , arnett & fryxell ( 1992 ) ; shankar & arnett ( 1994 ) ; glasner & livne ( 1995 ) ; glasner , livne , & truran ( 1995 ) and glasner , livne , & truran ( 1997 ) were the first to carry out two - dimensional direct simulations of tnrs . in the most recent computation ( @xcite 1997 ) no initial enrichment of carbon and oxygen was necessary to initiate a rapid tnr which , likely ,
will lead to an outburst similar to that of a classical nova . in their case
violent large eddies were responsible for considerable mixing of envelope and core matter .
their computations were performed with the ( implicit ) vulcan - code ( @xcite 1993 ) in spherical coordinates for a slice of 0.1 @xmath11 radians with reflecting boundaries at the bottom and on both sides . because their results are so important and because sometimes numerical effects can not be separated from physical ones easily we decided to perform independent simulations based upon a different hydro - code ( prometheus ; @xcite ( 1989 ) ) .
moreover , we decided to ignore curvature effects and to represent the surface layers of the white dwarf as well as its envelope by a plane - parallel sheet .
the advantage of this approach is that we can use periodic boundary conditions and thereby avoid common problems of numerical simulations of free convection , namely that reflecting boundaries may act like a containment and may affect the flow patterns in an unphysical way . in spherical coordinates with the polar axis singled out one can not use periodic boundary conditions .
we will come back to this point later in sec .
4 when we interpret our results and compare them with @xcite ( 1997 ) . in the following
we present the first results of our two dimensional ( 2-d ) simulations starting from the same initial conditions as @xcite ( 1997 ) .
we performed the simulations with two different cartesian grids covering the same domain as in glasner , livne & truran ( 1997 ) .
the first grid has approximately the same numerical resolution as was used by glasner , livne & truran ( 1997 ) , and the second one uses 5 times more zones in each direction . in sec .
2 we give a description of our numerical method , including the hydro - code , the nuclear reaction network , and details of the initial conditions . in sec .
3 we present the results , and sec .
4 is devoted to a discussion and , in particular , to a comparison with the work of @xcite ( 1997 ) .
the fact that according to our work 2-d simulations predict that during most of the tnr large scale motions do not dominate the evolution gives us hope that going to the full 3-d problem may not change all our conclusions .
3-d calculations which will proof or disproof this conjecture are under way and will be published in a subsequent paper .
our calculations were performed with a slightly modified version of prometheus , an eulerian code based on a second order godunov method with piecewise parabolic interpolation of the conserved quantities for solving the hydrodynamic equations , as worked out by colella & woodward ( 1984 ) and @xcite ( 1989 ) . since this code
is used frequently by various groups ( see , e.g. , mller ( 1994,1997 ) @xcite ( 1994 ) and niemeyer & hillebrandt ( 1995 ) ) we do not give a detailed description here or provide any numerical tests .
we want to mention in passing , however , that we had to change a few things in the original version of the code to make it applicable to the problem under consideration .
first , a challenging problem was to keep the initial model stable over several thousand dynamical time scales if nuclear reactions were suppressed .
the second problem was the presence of huge gradients at the core - envelope interface which lead to unacceptable numerical diffusion . in order to get a stable initial model we solved the equations of hydrostatic equilibrium for the pressure with a runge - kutta - solver , using the density and composition profiles of the initial model of @xcite ( 1997 ) , which will be described in more detail later in this chapter , and our equation of state .
next we mapped this density distribution on to the grid required by prometheus .
the deviations from the original model of livne and glasner were less than 1 % in pressure and density .
the huge gradients at the core - envelope interface required special attention since gravity in most versions of prometheus is included by calculating the gravitational acceleration _
after _ each hydro time step . since prometheus is a conservative second order godunov scheme one can not impose hydrostatic equilibrium directly
consequently , pressure gradients are not balanced exactly by gravity .
the accuracy one can reach is about 0.1 % which causes a net acceleration for the core - envelope - interface resulting in a systematic slow but significant outward motion of the interface . to overcome this problem
we have treated the time evolution of the gravitational acceleration in a more symmetric way .
it was included by calculating its action prior to and after the hydro timesteps with weighting factors f and ( 1 - f ) , respectively .
we found that for f = 0.12 the interface stayed in place for several thousand dynamical timescales , after the model had been relaxed , with nuclear reactions being turned off .
in fact , we found that these modifications of the original prometheus scheme were significant but not crucial . as a test
, we performed a 2-d simulation with nuclear reactions turned on for f=0 where the interface was moving outwards with the same speed as in the 1-d case without nuclear burning .
we realized that except for the moving of the transition region and some unphysical strong heating at the inner reflecting boundary the flow fields did not change by much .
the main difference was a slightly higher energy production rate for the f=0 case since numerical diffusion mixed additional core material into the burning region of the envelope .
finally , the code was changed to run efficiently on massively parallel computers such as the cray t3e by means of explicit message passing . in our version of prometheus nuclear reactions
are incorporated by solving together with the hydrodynamics a nuclear reaction network including 12 nuclear species , i.e. , @xmath12h , @xmath13he , @xmath14c , @xmath15c , @xmath15n , @xmath16n , @xmath17n , @xmath16o , @xmath17o , @xmath18o , @xmath19o , and @xmath19f , linked by reactions described in @xcite ( 1981 ) .
the reaction rates are taken from @xcite ( 1996 ) .
following @xcite ( 1986 ) , we solve the network equations and the energy source equation simultaneously to avoid numerical instabilities . for a given density @xmath20 the network equations and the energy source equation have the following form : @xmath21 where @xmath22 are the nuclear abundances
, @xmath23 is the energy production rate per unit mass , i , j = 1 ... n , and n is the number of nuclear species . f and
h denote nonlinear functions of the arguments .
this system of equations is solved implicitly from time @xmath24 to @xmath25 according to @xmath26 by a newton - raphson - solver .
the coupling to the hydrodynamic equations is done by solving this system with @xmath27 after hydro time step n with step size @xmath28 is executed , and incorporating the results , namely @xmath29 into the hydrodynamic equations in the next time step n+1 .
heat conduction , magnetic fields , radiation and a possible rotation of the white dwarf are ignored . here
, we use the same initial model for our two - dimensional calculations as glasner , livne , & truran ( 1997 ) .
since for the problem of simulating subsonic convection in an explicit scheme such as ours the time step ( here : @xmath10 milliseconds ) is limited by the cell size and the local sound speed , it is not possible to calculate by means of such schemes the whole accretion process ( several @xmath30 years ) until the tnr takes place . moreover , even for an implicit code , such as vulcan , solving the full problem in 2-d is nearly impossible .
therefore the accretion phase and the slow stages of the burning were calculated by @xcite ( 1997 ) using a one dimensional implicit hydro code ( @xcite 1996 ) . during this early phase hydrogen - rich matter of solar composition ( z = 0.02 )
is accreted onto the surface of a 1 m@xmath0 c+o white dwarf at a accretion rate of @xmath31m@xmath0yr@xmath32 .
as at the bottom of the envelope the temperature rises convection sets in .
convection is taken into account by the mlt in the one dimensional model .
the model we obtained from a. glasner and e. livne provided the density , temperature , entropy , and composition profiles at the point at which the temperature at the bottom of the envelope had already reached 10@xmath2 k. the mass of the hydrogen shell is about @xmath33 m@xmath0 .
we mapped this model onto a radial row of our 2-d grid and relaxed it for several hundred dynamical time scales and then mapped it onto the full 2-d grid .
the calculations were done with two different resolutions of a domain nearly identical to what has been used by @xcite ( 1997 ) .
this means that the grid covers 100 km of the outer part of the white dwarf and 1000 km of the envelope , radially , and @xmath10 1800 km in the lateral direction , corresponding to the outer part of a 0.1@xmath11 segment in spherical coordinates . but
both grids are cartesian and the curvature of the wd surface is ignored for reasons given earlier .
the coarser computational grid consists of 100 uneven radial zones and 220 lateral zones which are equally spaced .
hence the spatial resolution within the white dwarf and the first layers of the envelope is about 5 km @xmath34 8 km in this case .
the finer computational grid consists of 500 uneven radial zones and 1000 lateral zones which are equally spaced . here
the spatial resolution within the white dwarf and the first few layers of the envelope is about 1 km @xmath34 2 km .
our equation of state consists of the ideal boltzmann gases of the nuclei under consideration , an electron gas component with arbitrary degeneracy and a photon gas component .
finally , the boundary conditions and the initial perturbation have to be specified . here
we use a reflecting boundary at the bottom and an outflow boundary at the top of the grid .
periodic boundary conditions are applied laterally .
the model is initially perturbed by increasing temperature in one zone of the bottom layer of the envelope by 1% in both cases .
the further evolution is then followed for about 1400 physical seconds in the case of the low resolution grid and for about 180 physical seconds for the high resolution run . in the next section
we summarize the results of both calculations and compare them .
after the ignition of one zone the burning remains local for a few seconds ( fig.1b ) .
once the temperature has gone up a bit a first local tnr sets in .
the burning front propagates laterally with a velocity of about 15 km / s which is very close to the lateral spread by small scale turbulence as was estimated by @xcite ( 1982 ) .
this means that it should take over 100 s to ignite the entire bottom layer of our computational domain by small scale turbulence .
however , this argument ignores the capability of sound waves to ignite an isothermal layer which is close to the ignition temperature .
in fact , we find that spreading happens essentially on a sound travel time once a sufficiently large volume burns and sound waves are emitted from this region , and a full horizontal layer of the computational grid burns after about 14 s ( fig .
of course , our initial conditions , namely an isothermal layer very close to ignition temperature , enhance this effect and , in reality , small scale turbulence may turn out to be more important .
already 20 s after ignition the information about the location of the ignition point is largely lost . during the following 500
s the mean energy generation rate in the most violently burning shells is slowly increasing from about @xmath35 erg / g / s to over @xmath36 erg / g / s ( fig .
2 ) , and individual zones reach up to @xmath37 erg / g / s ( fig . 10b ) . in detail ,
the thermal history of the convectively burning shells turns out to be rather complicated . during the first 25 seconds cooling by convective mixing of cold envelope material into the burning region
is slightly more efficient than nuclear energy release .
thus the maximum temperature drops a bit before rising rather moderately , and we do not find a change into a fierce runaway ( fig .
this first result does not come as a surprise .
in fact , it is consistent with 1-d models which do not assume initial enrichment of carbon and oxygen ( see , e.g. , starrfield , truran , sparks , & kutter ( 1972 ) , and starrfield , sparks , & truran ( 1974 ) ) and just reflects the fact that we do not find much mixing by convective overshooting in the early phase of the tnr .
the moderate increase of both maximum temperature and energy generation rate continues until most of the protons are used up after approximately 500 s. = 8.8 cm [ cols="^,^ " , ] + [ 0.5ex ] table 1 : energy production rate contributions in erg / g / s of different @xmath3-unstable isotopes at different times calculated for the envelope shell at 6000 km .
the dominant energy production in the outer parts of the envelope is due to the @xmath3-decay of @xmath38 . at this time
only a few protons are left in the envelope ( fig . 8) and
the violent burning is close to being extinguishing .
the dredge - up of core material yields a gradual removing of the outer parts of the white dwarf ( fig .
6 ) which becomes remarkable after about 1000 s. the energy production by radioactive decays gets low and the temperature increase levels off .
small scale vortices reappear , as is to be expected from the previous discussion .
= 8.8 cm = 8.8 cm = 8.8 cm = 8.8 cm protons which have been mixed into the cool core during previous stages ( fig . 8) and stored there and , due to low core temperatures , have not been fused by the cno - cycle are now mixed into the envelope again . here
they are instantaneously absorbed via proton capture reactions almost exclusively in the hottest envelope shells resulting in a peak of the energy production there ( fig .
2 ) . although the effect of this extra burning on the over - all dynamics is not very big it may have interesting consequences for the nucleosynthesis yields expected from the outburst , because a small fraction of it might be mixed convectively all the way to the surface .
the growing envelope mass and the slowly decreasing nuclear energy release leads to a decrease of outflow velocities , until the envelope almost stops expanding near the end of the computations ( fig .
most of the outer core material covered by the computational grid has been mixed into the envelope by that time ( fig .
the convective velocities decrease steadily ( fig .
1h ) and the envelope approaches a new hydrostatic equilibrium state .
most of the previously stirred up heavy element enriched matter is settling onto the surface of the white dwarf at the end of the simulations ( fig .
the formation of a new core - envelope boundary is evident from the smooth density transition seen in fig .
6 at 1000 and 1400 seconds , respectively , which is also visible in other observables like temperature ( fig .
3b ) or hydrogen mass fraction ( fig . 8) . finally , in fig .
9 we show how the metallicity of the ejected envelope material evolves with time . in order to do
so we estimate the metallicity in the upper five calculated shells of the envelope and assume that they will be ejected .
the values we obtain in this way are z @xmath39 0.2 % for the maximum outflow velocities around 170 s , increasing up to about z = 0.9 % after 500 s when the outflow velocities are still remarkable .
increasing the resolution of the grid by a factor of five in each spatial direction means that computational costs increase by roughly a factor of 125 ( including the effect of a five times smaller cfl - time step ) .
so it is obvious that we could not compute a model with this resolution to the end of the tnr as was done for the previous run .
however , since here we are more interested in testing the numerical accuracy of our attempt to simulate directly turbulent combustion in the accreted envelope of a white dwarf rather then presenting very realistic models for a nova outburst which can be compared with observations ( what in fact could be done and will be done in a forthcoming paper since all the relevant physics is included in the code ) , we decided to go for the high resolution and to follow the highly resolved model as long as necessary and possible . in practice this meant that we stopped the computations after a physical time of about 180 s , for reasons which will be discussed later .
figure 11 gives the velocity field at different times for the case of high resolution and should be compared with fig .
1 of the low resolution run .
we disturbed as before one zone with a temperature amplitude of 1 % .
because now the mass inside a single zone is much smaller , the energy perturbation is about 20 times smaller as compared to the low resolution case .
naively one expects a somewhat slower evolution in the beginning , and that is exactly what happens .
we begin with showing a first snapshot at 13 s ( fig .
9a ) since then the distortion is comparable ( in energy ) to the low resolution case after 3 s. from then on the evolution of the tnr has many similarities but also distinct differences to the previous one .
it is evident from fig . 11
that the high resolution run also yields small coherent vortices , but now on typical length scales of 40 to 50 km which are smaller compared to the low resolution run by almost exactly the same factor by which the resolution was increased , indicating that these structures are still not fully resolved .
therefore , if it should turn out that the over - all properties of the model are affected in a similar manner this would mean that there is no chance to tackle the nova problem by means of direct numerical simulations .
independent of this practical question , it is interesting to see that again the small eddies seem to carry a considerable amount of the turbulent kinetic energy .
this finding gives us some hope that 2-d simulations can capture essential features and that going to 3-d models might not be necessary , but , of course , this conjecture has to be tested .
the ignition phase is characterized by a flow field dominated by these small scale vortices .
their typical velocities rise more rapidly .
for instance the maximum velocities after 70 seconds for the low resolution run are comparable to those of the high resolution run after 30 seconds .
a likely interpretation is that now the characteristic size of these eddies is considerably smaller than the pressure scale height .
therefore they are less effected by potential energy differences .
note that the fastest structures dominate the color coding of fig .
11 since we used a linear color scale .
= 8.8 cm = 8.8 cm according to fig .
11a , b , c the lateral spread of the burning by small scale turbulence along the white dwarf surface proceeds at a velocity of approximately 10 km / s during the ignition phase , very similar to what we have found in the low resolution case , and in accord with the estimates of @xcite ( 1982 ) . from then
on again sound waves ignite the burning in other parts of the bottom shell of the envelope and they can also be seen propagating vertically in fig .
after about 40 s the entire bottom layer is burning .
so the time required to propagate the front is equal to what we have found for lower resolution .
this result is in agreement with the sound wave interpretation because the sound velocity is independent of the resolution .
it also demonstrates that at least during the ignition phase _
lateral _ energy transport is not dominated by the fast moving small structures . however , the _ vertical _ extend of the convective region is affected as can be seen by comparing panel d of fig .
1 with panel e of fig . 11 but the convective energy is mainly locked in small scale motions .
= 8.8 cm = 8.8 cm = 8.8 cm = 8.8 cm as a result , we find that the temperatures in the hottest zones increase a bit faster than previously ( fig .
12a ) , but the over all effect is rather small as one can see by comparing the laterally averaged temperatures given in fig . 3b and fig
this is a clear indication that larger scales are mainly responsible for the energy transport , and they are well resolved in both cases . the small delay of the temperature increase in the low resolution run just reflects the fact that due to the lower velocities in the small eddies more energy is available for the large scale motions and , therefore , convective energy transport is a bit more efficient .
it is also obvious from comparing for example fig . 2 and fig .
12 or fig . 4 and fig . [ fig17 ]
that there are only minor differences between the two calculations for key quantities such as the energy generation rate or the @xmath14c mass fractions once the envelope has become fully convective after about 150 s. this , in fact , was the main reason why we stopped the highly resolved run after about 180 s. however , there is still a bit of uncertainty in this conclusion .
it appears that in the high resolution case the energy generation rate as well as the maximum temperature reached in any of the shells seems to decrease or at least level off after about 150s .
we belief that this is a temporary effect only . as we have shown earlier , once the energy generation rate has reached values around 10@xmath17 erg /
g / s the small vortices begin to be destroyed by the more violent large scale motions .
it is very likely that the same will happen also in the high resolution calculations , maybe with some delay , due to the higher velocities ( and therefore increased stability ) in the smaller eddies .
14 shows the laterally averaged vertical distribution of the vertical velocity component .
not surprisingly , the high resolution run shows much more structure and rapid oscillation , especially still at 170 seconds , whereas in the low resolution case the envelope is expanding more uniformly at the time of peak outflow velocities ( fig .
5 ) , and they compare more with values reached already after 150 s in the latter .
but again , we consider this to be a minor uncertainty , consistent with the general trend that for the better resolved calculations the early phase of the tnr proceeds a bit faster , whereas later , when larger eddies take over , the evolution is somewhat slower in this case . the density profiles ( fig .
15 ) show the decrease of the envelope density .
just as before , consumption of protons also proceeds slowly ( fig . 16 ) . for completeness , we give the @xmath17o mass fraction and metallicity of the ejected shells in figs .
17 and 9 , respectively .
of course , it would have been nice to give all these data also at later times , but because of the enormous computer time required for this run we stopped it .
we have presented first results of numerical simulations of the tnr in the hydrogen - rich envelope of an accreting white dwarf .
these simulations were performed by means of a modified version of a standard ppm - type hydro code which is able to treat reactive hydrodynamics in one , two or three spatial dimensions .
the models outlined in the previous sections were obtained in planar geometry , the surface layers of the white dwarf and its envelope being mapped onto a 2-d grid in cartesian coordinates . the main intention of the simulations was to investigate the question as to whether convective overshooting can or will lead to self - enrichment of the envelope with c and o from the underlying white dwarf s surface .
the same question was recently addressed in a very similar simulation by glasner , livne & truran ( 1997 ) who found in fact a positive result . according to their study ,
considerable dredge - up of white dwarf matter occurs early in the tnr , already after about 100 s , leading to a violent runaway between 150 and 200 s after the computations had been started in 2-d , with a temperature of 10@xmath2k at the bottom of the envelope .
our results agree in certain respects with those of @xcite ( 1997 ) , and disagree in others .
we agree with them on the general outcome , namely that even with an initially solar composition of the accreted matter most of the envelope is ejected during the outburst .
we also agree that the ejected material will be enriched with matter mixed in from the white dwarf . on the other hand ,
the differences are remarkable , given the facts that we started our calculations from exactly the same initial conditions and that , in one case , even the numerical resolution was nearly equal to theirs . in general terms
, the outburst is much less violent in our calculations as compared to theirs .
this shows up in the longer timescale for the tnr , the lower ejection velocities , the considerably lower temperature near peak energy production , etc .
the cause of the deviations can be traced back to large differences in the convective flow patterns we obtain . in their simulations
a few large eddies dominate the flow already during early phases of the tnr , whereas in our calculations small eddies carry a large amount of the convective kinetic energy until the burning proceeds at maximum rate . in fig .
[ fig18 ] we provide energy spectra for the low resolution run , to support this argument . they are not as steep as expected for instance for unforced decaying incompressible turbulence as demonstrated in @xcite ( 1984 ) .
most of these eddies move very little in vertical direction and , therefore , do not penetrate into the white dwarf .
this is the reason why we observe very little mixing until the envelope lift - off sets in and dredge - up happens in the steep density ( and pressure ) gradient near the white dwarf s surface .
however , this late mixing has little effect on the dynamical evolution of the envelope because at that stage most of the hydrogen has already been burnt .
if true , this finding will , nevertheless , affect the nucleosynthesis predictions obtained from the model , but we are not 100@xmath40 sure .
the reason is that the surface layers of the white dwarf are not well resolved in our `` low '' resolution run and , due to the eulerian nature of the code numerical diffusion may in part be responsible for some of the late mixing . in principle
this could be checked by extending the `` high '' resolution run to about 800 s , but this would cost a lot of computer resources , and we thought it might be more important to save those for 3-d simulations which are under way . the question then remains as to why the flow patterns are so different in the two independent sets of calculations .
our interpretation is that they are mainly caused by the different boundary conditions and symmetries imposed .
we think it is dangerous to impose polar symmetry and reflecting boundary conditions in simulations which aim at investigating geometrical properties of free subsonic convection .
in fact , we found in our simulations that the fast and small vortices moved mainly horizontally all across the computational grid , leaving the grid on one side and coming back from the other .
such motions would be prohibited if reflecting boundaries were imposed , and it is likely that such boundary conditions will suppress these modes . moreover , as can be seen in fig . 1 and 2 of @xcite ( 1997 ) in their computations
most of the action appears to be either at the polar axis or near the horizontal boundaries which indicates unphysical symmetries . in conclusion
, we have demonstrated that direct numerical simulations of thermonuclear combustion in the accreted hydrogen - rich envelopes of white dwarfs are feasible .
we have shown that even in 2-d ( and in contrast to common expectations ) small scale structures carry a large amount of the convective kinetic energy .
the smallest of those eddies were not resolved in our models but their actual size seems to be of minor importance for the general behavior of the tnr .
just the fact that they exist makes a big difference , as the comparison with the work of @xcite ( 1997 ) showed .
they were interpreted as stable solutions of the ( time independent ) euler equations which form from small fluctuations in 2-d convective flows .
the fact that they disappeared once the energy generation ( source terms ) began to dominate over the advection strongly supports this interpretation
. it will be interesting to see which of these conclusions remain valid in 3-d simulations .
the small scale vortices might not be as important any more because their stability rests largely on axi - symmetry which is broken in 3-d .
on the other hand side , their main effect was to hinder the formation of dominant large scale eddies which otherwise would show up more clearly in 2-d .
so there is hope that 2-d simulations do capture the main features of the tnr , but this has to be seen .
_ acknowledgements .
_ the authors are grateful for many enlightening discussions with stanford e. woosley , jens c. niemeyer and ulrich kolb .
they thank ami glasner and eli livne for supplying the initial model , ewald mller for an earlier version of the prometheus code , and rudi fischer and jakob pichlmeier for their help in preparing a parallel version of the code .
they also thank the stuff members of the rechenzentrum garching for their support during the computations . this work was supported in part by nsf at the institute for theoretical physics at the university of california ,
santa barbara under grant no .
phy94 - 07194 and by nasa under grant nag 5 - 3076 at the university of chicago .
the computations were performed at the rechenzentrum garching on a cray t3e . | we present the results of two - dimensional calculations of turbulent nuclear burning of hydrogen - rich material accreted onto a white dwarf of 1.0 m@xmath0 .
the main aim of the present paper is to investigate the question as to whether and how the general properties of the burning are affected by numerical resolution effects . in particular , we want to see whether or not convective overshooting into the surface layers of the c+o white dwarf can lead to self - enrichment of the initially solar composition of the hydrogen - rich envelope with carbon and oxygen from the underlying white dwarf core .
our explicit hydrodynamic code is based on the ppm - method and computes the onset of the thermonuclear runaway on a cartesian grid .
only part of the white dwarf s surface is covered by the computational grid and curvature effects are ignored .
in contrast to previous works we do not observe fast mixing of carbon and oxygen from the white dwarf s surface into the envelope by violent overshooting of large eddies .
the main features of the flow fields in our simulations are the appearance of small persistent coherent structures of very high vorticity ( and velocity ) compared to the background flow .
their typical linear scales are about 10 to 20 grid zones and thus their physical size depends on the numerical resolution , i.e , their size decreases with increasing resolution . for the early phase of the thermonuclear runaway ( tnr ) they dominate the flow patterns and result in very little overshoot and mixing . only at late times , after steady slow mixing and with increasing nuclear energy production , do these structures become weak , but show up again once hydrogen has mainly been burnt and the energy generation rate drops . on the other hand ,
there are no big differences between high and low resolution simulations , as far as the overall properties of the tnr are concerned .
the two simulations which are presented here show only moderate differences in spatially integrated quantities such as laterally averaged temperature , energy generation rate , and chemical composition .
we have not expanded both simulations equally long , but for the physical time under consideration the major difference seems to be that the highly resolved simulation is a bit less violent . in conclusion , we do find some self - enrichment , but on time - scales much longer than in previous calculations . |
hilbert s tenth problem @xcite asks for a _ single _ procedure / algorithm to systematically determine if any given diophantine equation has some positive integer solution or not .
this problem has now been proved to be turing noncomputable , indirectly through an equivalence mapping to the noncomputable turing halting problem @xcite . nevertheless , we have proposed and argued ( with more and more details provided ) in a series of papers @xcite for an algorithm for hilbert s tenth problem based on quantum adiabatic computation @xcite . among the many valid concerns about our algorithm there are also some misleading objections which are still being spread despite their falsehood . in this short note
we will give an updated overall picture of the algorithm and go through , with pointers provided for further discussions elsewhere , some of the valid concerns as well as the misleading objections to dispel the entrenched but baseless prejudice against our proposed algorithm .
a precise statement of the algorithm can be found elsewhere @xcite , but in a few words , given a diophantine equation our aim is to obtain ( by physical means or simulations or otherwise ) the ( fock ) ground state of an appropriate ( and bounded from below ) hamiltonian @xmath0 carrying the information of the input diophantine polynomial .
we will achieve that by starting with yet another easily obtained ground state of another hamiltonian @xmath1 and by adiabatically deforming the hamiltonian @xmath1 in time to the one carrying the diophantine input above .
essential ingredients of the algorithm are : + - ( for an unbounded hamiltonian in a dimensionally infinite space ) will ensure that , provided the adiabaticity and other conditions are satisfied , the initial ground state will turn into the sought - after ground state of @xmath0 with _
high probability_. relevant to this point is the paper by tsirelson @xcite criticising our algorithm when it first came out in 2001 .
this reference has not been published but somehow has been selectively cited by some as an evidence against our algorithm ! those citing , intentionally or not , have either ignored or missed out our reply to tsirelson @xcite which was posted only 3 days later on the same arxiv . in that reply ,
we clearly pointed out that tsirelson s arguments were simply wrong : had they been right , the qat , and not just our algorithm , would have been mathematically wrong for all those years !
+ - is necessary to obtain the adiabaticity condition for the qat with a _
finite _ rate of hamiltonian deformation .
we have employed certain mathematical theorems and a gauge - like symmetry for the class of time - dependent hamiltonians of the algorithm to show @xcite that there is indeed no level crossing .
+ - is necessary because the qat is a nonconstructive theorem and does not tell us how the final probability of obtaining the ground state is approached as a function of the inverse of the rate of change of the hamiltonian deformation .
such final probability certainly does not increase monotonically but varies for different diophantine equations in a complicated manner . to be an algorithm for hilbert s tenth
we need a _ single and universal _ criterion , applicable to any diophantine equation , to identify the sought - after ground state .
we have shown @xcite that it is only the final ground state that can be obtained with a measurement probability _ more than 1/2 _ with our algorithm .
such a probability is our identification criterion for the ground state .
( the proof has now been extended to an infinite number of energy levels , not just the two levels of the ground state and the next excited state , and will appear elsewhere . )
these analytical results have also been supported by numerical simulations @xcite .
+ - is inevitable for our quantum algorithm @xcite , either in determining the identification probability through relative frequencies in physical measurements or in extrapolating to zero - size time step in some numerical simulations @xcite .
we will come back to this probabilistic nature in a section below .
further modification and extension of our work has also appeared in the literature @xcite .
the recursive noncomputability of hilbert s tenth problem lies in the fact that a systematic substitution of positive integers ( in some increasing order of magnitude , say ) into a given diophantine equation could only terminate if the equation has a solution ; otherwise , the substitution would go on indefinitely without any termination point .
( for this reason , the problem is sometimes also termed semi - computable , as we do not have a general method to determine when the equation has _ no _ solution . ) on the other hand , our quantum algorithm apparently is somehow able to explore the infinite ( of the whole domain of positive integers ) in a finite time ! " how can that be ? and some have even used this as an indication , which is misleading , that there must be something wrong with our algorithm !
the logic behind all this is that hilbert s tenth problem belongs to the class of _ finitely refutable mathematical problems _ @xcite .
that is , for any given diophantine equation it only requires a substitution up to some positive integer to determine whether it has any solution or not , even in the case of _ no _ solution : if the equation has no solution within a certain _ finite _ domain of positive integers , it will not have a solution anywhere else in the whole infinite domain !
the noncomputability of hilbert s tenth problem is precisely because we do not have any universal recursive method to determine this finite decisive domain for every diophantine equation . in contradistinction to the recursive mathematics , quantum mechanics can give us the means to determine such finite decisive domain ( in order to make some conclusion in the infinite domain ) through the ( energetically ) ground state @xcite .
the finiteness of such a domain is encoded and reflected accordingly in the finiteness of the algorithm evolution time and in the finiteness of the energy and occupation number of the final ground state .
this is how the paradoxical power of our quantum algorithm can be understood .
( in fact , the quantum algorithm provides an alternative proof for the finitely refutable character of hilbert s tenth and related problems . )
our quantum adiabatic algorithm is probabilistic in the sense that it can produce a result with a probability , which can be made arbitrarily high , of being the correct result . as such
, it has a non - zero probability , even though it can be made arbitrarily small , _ of being incorrect ! _ thus , in a way , in order to compute the noncomputable or decide the undecidable with our algorithm , we will have to allow for the possibility of being wrong , even though we can reduce this chance ( at a cost ) .
it is this probabilistic nature of the algorithm that renders it outside the jurisdiction of cantor s diagonal arguments @xcite employed in the noncomputability proof of the turing halting problem ( and hence of hilbert s tenth problem ) @xcite .
indeed , the discoverers of noncomputability and of incompleteness in mathematics were very much aware of this power of ( probabilistic ) flexibility , as reflected in their own statements @xcite , + - * gdel * : _ ... it remains possible that there may exist ( and even be empirically discovered ) a theorem - proving machine which in fact is equivalent to mathematical intuition , but can not be proved to be so , nor can be proved to yield only correct theorems of finitary number theory . " _
+ - * turing * : _ ... if a machine is expected to be infallible , it can not be intelligent .
there are several theorems which say almost exactly that . but these theorems say nothing about how much intelligence may be displayed if a machine makes no pretence at infallibility . "
_ probabilistic computation may also be more powerful than turing deterministic computation in yet a different way @xcite , contrary to the often misquoted statement that the two are equivalent in terms of computability .
pour - el and richards have shown that @xcite , surprisingly , the solution at a finite time of the linear wave equation in 3 dimensions can be _
noncomputable _ , even with some computable initial condition !
this result not only shows the limitation of turing computability even with linear and supposedly simple differential equations like the well - known wave equations , but also implies a hypothetical physical hypercomputation : starting a controlled wave propagation with a precisely prepared initial condition and then measuring the wave configuration at a given time later to obtain ( compute ) the otherwise noncomputable .
this would work provided that ( i ) the physical wave propagation so performed is governed by the mathematical wave equation ; and ( ii ) the measurement of the wave configuration could be done with infinite precision .
our quantum algorithm not only implies the noncomputability of the solution of the schrdinger equation with a special class of time - dependent hamiltonians ( associated with hilbert s tenth problem ) @xcite ; but also provides the procedure , physical or otherwise , for computing the turing noncomputable , albeit in a probabilistic manner . and thanks to the flexibility of this probabilistic character
, we may not require an infinitely precise measurement ( but see below ) but still have to assume that the mathematics of quantum mechanics underlies any physical implementation of the algorithm , unless the algorithm is simulated on turing machines . beside ours , there are also other appeals to quantum mechanics as the possible evidence and/or resource for hypercomputation @xcite . postulated in quantum mechanics
is the inherent and irreducible randomness ; and true randomness is outside the turing computability @xcite and thus belongs to the domain of hypercomputation , and so does probabilistic computation in general @xcite .
turing machines are not capable of generating truly random numbers , but only pseudo - random output with some finite and recursive algorithms .
hypercomputation beyond the turing computability is thus not so mythical .
at least everyone agrees that random number generation is a kind of hypercomputation , albeit a very special kind .
the problem is whether and how randomness can be harnessed for more interesting computation rather than just being random in itself .
each series of random numbers generated by a series of quantum measurements is different from each other and is not reproducible ( being random by the very definition ) .
what reproducible is not the outcomes but more often is the probabilities for the outcomes .
we have pointed out how _ some _ hypercomputation could be performed with the help of a ( quantum ) probability which is a noncomputable real number @xcite .
the kind of problems that can be solved by this hypercomputation depends on the properties of that real number .
can we make direct use of a particular ( irreproducible ) series of random numbers generated quantum mechanically or otherwise ?
we hope to report these findings elsewhere @xcite .
probabilistic computation may be more powerful than turing computation provided a suitable probability measure can be defined . the systematic substitution of positive integers into a diophantine equation , for example
, does not lend itself to a definable probability measure since the cardinality of any subset used in the substitution is always finite while that of all the positive integers is not .
( recall that there is no recursive way to determine the finite decisive subsets , even when we know that the problem is finitely refutable .
were there a way , the problem would not have been turing noncomputable . ) our algorithm , on the other hand , possesses naturally defined probability measures through the use of quantum mechanics . in a physical implementation of the algorithm , the probability comes from the weak law of large numbers in determining some other quantum probabilities through the relative frequencies obtained from repetitive measurements @xcite .
this determination does not require measurements of infinite precision .
however , there have been some concerns @xcite that infinite precision is still required in physically setting up the various integer parameters in the time - dependent quantum hamiltonians . while the issue deserves further investigations as surely any systematic errors in the hamiltonians would be fatal , we still are not convinced that such integer parameters can not be satisfactorily set up . in particular , we would like to understand the effects of statistical ( as opposed to systematic ) errors on the statistical behaviour of the spectrum of our adiabatic hamiltonians . however , physical implementation is not the only way to carry out the algorithm .
the algorithm could also be simulated on turing machines @xcite .
then how does probability come into such simulations ?
it comes in under the necessary extrapolation of the simulation time steps to zero sizes , which is essentially probabilistic @xcite .
the probability measures here are different from those in the physical implementation but , on the other hand , we do not have the above problem associated with the integer parameters in our hamiltonians .
we have listed and dealt with some of the concerns and objections against our quantum adiabatic algorithm for the turing noncomputable tenth problem of hilbert .
most of these objections ( none of which actually appears in print ) simply root in the belief , and no more than a belief , that there must be something wrong with the algorithm because it claims to be able to compute the very problem that has been mathematically ( and recursively ) proven to be noncomputable ! closer inspection , however , shows that such a belief is simply false as all the noncomputability proof is only valid within a certain framework , outside of which the quantum algorithm operates and hence entails no contradiction with the known and proven facts .
so , what can we expect from the algorithm ?
once and if implemented , it could give an answer , _ with any pre - determined probability _ , with respect to the existence of solution of any given diophantine equation . the probability can be raised arbitrarily without bounds ; but the higher it is , the higher the cost of time and resources it takes .
the other caveat is that the time it takes for a _ successful _ application of the algorithm is not known beforehand but only as an end product , even though it is always finite .
( this successful time is not the time we fix _ a priori _ for each run of the algorithm
each run of the algorithm always has an end point .
we must then keep increasing this running time until successful . )
i am indebted to many of my colleagues for the benefit i have derived through the communication and correspondence with them .
i also wish to acknowledge in particular the on - going support from peter hannaford and alan head .
this point emerged from a group discussion at mit with enrico deotto , ed farhi , jeffrey goldstone and sam gutmann ( 2002 ) .
it goes without saying that if there are mistakes in the interpretation herein , they are solely mine . | we give an update on a quantum adiabatic algorithm for the turing noncomputable hilbert s tenth problem , and briefly go over some relevant issues and misleading objections to the algorithm . |
the full scale terrestrial gravitational wave antennae are now in the process of assembling and tuning .
it is expected that in the year 2001 in one of these antennae ( project ligo - i ) the sensitivity in units of amplitude of the perturbation of metric will be at the level @xmath0 @xcite .
this value means that the optical readout system will be able to register small displacement between two test masses ( two heavy optical mirrors placed in vacuum pipe at the distance @xmath1 cm ) with the resolution @xmath2 cm .
this displacement may be produced by a burst of gravitational wave generated in astrophysical catastrophe .
the characteristic frequency in this wave packet and , correspondingly , in the displacement @xmath3 has to be of the order of @xmath4 s@xmath5 .
the improvement of the readout system and better isolation of the mirrors from different sources of noise will permit to achieve higher sensitivity : in the year 2005 the planned level of sensitivity has to be better than @xmath6 and correspondingly it will be necessary to register @xmath7 cm .
it is worth noting that in the prototype ligo ( where @xmath8 cm ) the already achieved resolution was @xmath9 cm @xcite .
the important feature of this antenna is that the optical readout system does not register displacements between the centers of the mirrors masses but between the internal surfaces of the mirrors .
these surfaces are coated by several thin dielectrical layers , which provide high optical reflectivity of the mirrors and high sensitivity to the displacements .
the necessity to measure the displacements between the internal surfaces inevitably determines the requirement for high quality factors @xmath10 ( small loss angle ) of internal mechanical modes of the mirrors , which provide relatively small values of random thermal fluctuations of the mirrors related to the mirrors center of mass because the value of @xmath11 is substantially smaller than the eigen - frequencies of internal modes .
we shall call as brownian the fluctuations which are usually calculated from phenomenological model of structural losses in the material in order to distinguish them from another type of thermodynamical fluctuations which we describe below .
this requirement for high @xmath12 determines the choice of the mirrors material . on the first stage of ligo
the mirrors are made of very pure fused silica ( @xmath13 ) , because the values of @xmath14 exceed @xmath15 @xcite . at present the possibility to make the mirrors for the second stage of ligo from sapphire monocrystal ( @xmath16 )
is widely debated , because more than 20 years ago the achieved value of @xmath17 for this material was @xmath18 @xcite .
we show below that the value of @xmath19 is not the only parameter which determines the choice of material from which the mirror has to be manufactured . till
now the experimentalists interest has neen concentrated on the reduction of the brownian motion of the mirrors surfaces ( as it was described above ) , to the reduction of thermal fluctuations of the centers of mirrors masses due to the friction in the mirrors suspension @xcite and to excess noise in the suspensions @xcite . to our best knowledge
nobody has so far paid attention to nonlinear effects which may be `` responsible '' for additional fluctuations of the mirrors surface , which may exceed traditional brownian noise for low frequencies .
the simplest of these effects is thermal expansion , originated from anharmonisity of interaction potential of atoms in solid state .
the following numerical example illustrates the role of nonzero value of thermal expansion coefficient @xmath20 of the mirrors material .
suppose that the surface layer of the mirror having thickness @xmath21 cm ( it is the characteristic length of the decay of fourier thermal wave at frequency @xmath22 in fused silica ) is heated due to some process on @xmath23 .
then the distance between the surface of the mirror and its center of mass will also change on the value of @xmath24 : @xmath25 the estimate ( [ deltal ] ) shows , that if the mirrors are made of fused silica and if the temperature of the first layer ( @xmath26 cm thick ) of the mirror changes by 2 nk during @xmath27 s then the antenna will register a `` burst '' of gravitational radiation with amplitude of the order of @xmath28 with @xmath29 s. the aim of this article is to present the results of the analysis of the two effects which are `` responsible '' for random fluctuations of temperature in the bulk of the mirrors and which may mimic at certain level the gravitational wave if @xmath30 : photo - thermal shot noise ( due to random absorption of optical photons in the surface layer of the mirror ) and thermodynamical fluctuations of temperature . as in both effects
we are interested in the fluctuations of the coordinate averaged over the spot on mirror surface with radius @xmath31 cm of laser beam , which is much smaller than the radius of mirror @xmath32 cm , we replace the mirror by half - space : @xmath33 .
we consider also for simplicity this half - space to be isotropic ( the anisotropy of different constants for sapphire is not very large ) .
the role of these two effects for the condition to reach the standard quantum limit of frequency stability was analyzed more than 20 years ago @xcite .
as at room temperature the power radiated from the surface ( stephan - boltzmann law ) is much lower than the heat exchange due to thermal conductivity ( fourier law ) , below we use the simplified boundary condition : @xmath34 where @xmath35 is the deviation of temperature from the mean value @xmath36 . for the calculations we use
approximation in which we neglect the effects of limited speed of sound but take into account thermal relaxation .
the multilayer coating of the mirrors absorbs a fraction of power @xmath37 of the optical beam , circulating between the mirrors of fabry - perot resonator .
this fraction for the best existing today coatings has not been measured with good precision yet , but for estimates the usually accepted value of this fraction is approximately equal to @xmath38 @xcite .
the value of @xmath37 sufficient to reach @xmath28 ( which is close to standard quantum limit ( sql ) of sensitivity ) is approximately @xmath39 or @xmath40 photon / s @xcite .
it is reasonable to estimate that each mirror absorbs @xmath41 erg / s or @xmath42 photon / s .
each absorbed optical photon having energy @xmath43 erg gives birth to a bunch of approximately 50 thermal phonons .
each of these phonons has a relatively short free path @xmath44 ( for @xmath13 and @xmath16 the values are @xmath45 cm and @xmath46 cm @xcite ) .
after a very short time interval which is a few times longer than @xmath47 s ( where @xmath48 is the speed of sound ) these phonons will produce a local jump of temperature inside the coating @xmath49 cm thick . as we want to know the variance of temperature during the averaging time @xmath50 s which is several orders longer than @xmath51 and as during this time a large number ( @xmath52 ) of photons are creating bunches of thermal phonons it is reasonable to use the model of shot noise .
as the length @xmath53 is much smaller than the radius of laser beam @xmath54 cm for semi - qualitative analysis one can use one dimensional model , described by thermal conductivity equation : @xmath55 where @xmath56 , @xmath57 is thermal conductivity , @xmath58 is density and @xmath59 is specific heat capacity . assuming that displacement @xmath60 ( which is of interest ) is proportional to temperature @xmath35 averaged along the axis @xmath61 @xmath62 it is not difficult to obtain the spectral density of @xmath60 : using formula @xmath63 ] @xmath64 for the exact calculations for half - space ( see appendix a ) we introduce the displacement @xmath65 averaged over the beam spot on the surface @xmath66 where @xmath67 is @xmath61-component of vector of deformation on the surface , and find the spectral density of @xmath65 : @xmath68 where @xmath69 is the poisson coefficient . here and
below we use the fact that for materials at room temperature the thermal relaxation time of the spot @xmath70 is very large : @xmath71 it is easy to see that the exact solution ( [ sdsn ] ) differs from the approximate one ( [ sdts1d ] ) only by poisson coefficient .
it is known from thermodynamics @xcite that the total variance of fluctuations of temperature in volume @xmath72 is described by the following simple formula : @xmath73 where @xmath74 is the boltzmann constant . in classical solid state thermodynamics these fluctuations
are not correlated with fluctuations of volume ( which are responsible for the brownian noise ) @xcite but this is not the case for nonzero thermal expansion coefficient . to find the influence of the fraction of this type of fluctuations in the vicinity of @xmath11 on the vibration of the surface along the x - axis in direct way we use the langevin approach and introduce fluctuational thermal sources @xmath75 added to the right part of the equation of thermal conductivity : @xmath76 these sources should be normalized in a way to satisfy formula ( [ tvar ] ) . by substituting the solution of this equation for the temperature @xmath35 into the equation of elasticity and averaging displacement ( [ barx ] ) over the spot on the surface , we find spectral density of its fluctuations in frequency range of interest ( see details in appendix b ) : @xmath77 we can also find the spectral density of displacement using fluctuation - dissipation theorem ( fdt ) if we assume that the only dissipation mechanism in the mirror is thermoelastic damping .
elastic deformations in a solid body through thermal expansion lead to inhomogeneous distribution of temperature and hence to fluxes of heat and losses of energy @xcite .
the same approach was used @xcite for the calculation of vibration of the surface due to the brownian fluctuations . to calculate the fluctuations in this approach
we should apply a periodic pressure @xmath78 distributed over the beam spot on the surface : @xmath79 and we should find susceptibility @xmath80 .
after that in accordance with fdt the spectral density of displacement @xmath65 must be proportional to imaginary part of susceptibility @xmath81 : @xmath82 calculations ( see appendix c ) show that fdt gives the same result as formula ( [ sdtd ] ) .
therefore , we can conclude that thermodynamical fluctuations of temperature are the physical source of fluctuations deduced from fdt based on thermoelastic damping .
calculations of @xmath83 are rather bulky but zener @xcite suggested a very simple approach allowing to estimate thermoelastic loss angles and hence the imaginary parts of elastic coefficients .
it allows to find @xmath84 only from formula for @xmath81 in zero order approximation ( see appendix c ) : @xmath85 where @xmath86 is the young s modulus .
the imaginary parts of elastic coefficients due to thermoelastic damping may be calculated in zener s approach according to the formulas
@xcite : @xmath87 where @xmath88 are consequently iso - thermic and adiabatic young s moduluses ( @xmath89 ) and @xmath90 are iso - thermic and adiabatic poisson coefficients ( @xmath91 ) , @xmath92 is the time of thermal relaxation of our spot . by substituting these values in ( [ rigidity ] ) and then in ( [ spectrum ] ) and taking into account that @xmath93 ( this relaxation time may be found relatively easy from the solution of the problem of thermal relaxation of the heated spot ) we finally obtain the same formula ( [ sdtd ] ) .
it is really worth noting that all the three methods give precisely the same result .
now we want to compare the fluctuations of the mirrors surface caused by the described above mechanisms with other known sources of noise in ligo antennae .
the main determining source of noise associated with the mirrors material is usually described by the losses in the model of structural damping ( we denote it as brownian motion of the surface ) . in this model
the angle of losses @xmath94 does not depend on frequency and analogously to @xcite by substituting @xmath95 in ( [ rigidity ] ) and then in ( [ spectrum ] ) we obtain : @xmath96 for simplicity here we neglected the unknown imaginary part of poisson coefficient .
the sensitivity of gravitational wave antenna to the perturbation of metric may be recalculated from noise spectral density of displacement @xmath97 using the following formula : @xmath98 where we used the fact that antenna has two arms ( with length @xmath99 ) and in each arm only one mirror adds to noise ( on end mirrors the fluctuations are averaged over the larger beam spot )
. the ligo - ii antenna will reach the level of sql , so we also compare the noise limited sensitivity to this limit in spectral form @xcite : @xmath100 on figures 1 and 2 we plot the sensitivity limitations for the perturbation of metric caused by the described thermal mechanisms and sql ( formulas [ sdsn],[sdtd],[sdsd],[hsd],[hsql ] ) for fused silica ( fig.1 ) and sapphire ( fig.2 ) mirrors .
we used the following numerical values for the parameters of the two materials fused silica and sapphire .
( we should note that the actual figures , which various sapphire manufactures provide , significantly vary tens of percents , especially for the thermal expansion coefficient ) :
@xmath101 from these two figures it is evident that the effects associated with thermal expansion in no case may be neglected .
especially unfortunate result is obtained for sapphire
the noise from thermodynamical fluctuations of temperature dominates for frequencies of interest and is several times larger than the sql . while for fused silica the situation is the opposite .
the sql is one order larger than td fluctuations .
it is important to note , that unlike structural damping for which angles of losses at low frequencies were not yet directly measured , our mechanism has fundamental nature and is calculated explicitely without model assumptions .
the analysis and numerical estimates presented above for two effects allow to give some recommendations for the strategy of upgrading laser interferometric gravitational wave antennae .
it is important to emphasize that contrastingly to the gravitational wave , both effects do not displace the centers of masses of the mirrors .
thus , in principle it is possible to subtract these effects as well as usual brownian fluctuations ( with certain level of accuracy ) in a way similar to the procedure suggested for the subtraction of thermal noise in suspension fibers @xcite .
however , till now to our knowledge nobody presented the scheme of such subtraction .
potentially there exists a possibility to decrease substantially these effects by choosing materials for the mirrors with substantially smaller values of @xmath20 .
there also exists the possibility to use special cuts of anisotropic monocrystals for which @xmath20 turns to zero .
thermodynamic fluctuations decreese with growing radius of the beam spot as @xmath102 even stronger then for the brownian fluctuations . in this way , some optimization of the geometry of the resonators may help .
the `` brute force '' method for the improvement of the sensitivity by increasing the circulating optical power @xmath37 does not look promising , taking into account photo - thermal shot noise .
it seems that few mega - watts is the upper limit for @xmath37 .
the evident method to suppress the thermodynamical fluctuations by using larger beamspots is unfortunately limited by diffraction losses and technology . if we choose the ligo - i geometry with beam sizes on internal and end mirrors : @xmath103 then the threshold of sensitivity of thermodynamical fluctuations will be only 1.7 times lower than on our graphics still 2 times higher than the sql .
we are pleased to thank for helpful discussions v.p.mitrofanov , f.ya.khalili , a.i.osipov , i.a.kvasnikov , h .- j.kimble and r.weiss .
this research was partially supported by the california institute of technology , us national science foundation and the russian foundation for basic research grant # 96 - 15 - 96780 .
in this appendix we calculate the spectral density ( [ sdsn ] ) of the fluctuations of the averaged displacement @xmath65 caused by photo - thermal shot noise .
we start from thermal conductivity equation : @xmath104 with boundary condition ( [ boundarytc ] ) and equation of elasticity for the field of deformation @xmath105 ( see @xcite ) with boundary conditions for tensions on free surface : @xmath106 \right|_{x=0}=0 , \\ & & \sigma_{xy}=\left.\frac{e}{2(1+\sigma)}\left ( \frac{\partial v_x}{\partial y}+ \frac{\partial v_y}{\partial x}\right ) \right|_{x=0}=0 , \qquad \sigma_{xz}=\left.\frac{e}{2(1+\sigma)}\left ( \frac{\partial v_x}{\partial z}+ \frac{\partial v_z}{\partial x}\right)\right|_{x=0}=0,\end{aligned}\ ] ] the problem ( [ tcsn ] , [ boundarytc ] ) in half - space can be replaced by the problem ( [ tcsn ] ) in full space , in which the function of source is evenly continued for negative @xmath61 . in this case
the condition ( [ boundarytc ] ) is satisfied automatically .
assuming that the absorbed power changes as @xmath107 one can find the spectral component : @xmath108 after substitution of @xmath109 into ( [ edfull ] , [ edboundary ] ) one can find the solution for @xmath110 as a sum @xcite @xmath111 , where @xmath112 satisfies poisson s equation ( it can be deduced from ( [ edfull ] ) ) : @xmath113 without boundary conditions and function @xmath114 satisfies equation ( [ edfull ] ) with zero right part and boundary conditions : @xmath115 \right|_{x=0},\label { boundaryva}\\ \sigma_{xy}&=-&\left.\frac{e}{(1+\sigma)}\ , \frac{\partial^2 \varphi}{\partial x \partial y } \right|_{x=0 } , \qquad \sigma_{xz}=-\left.\frac{e}{(1+\sigma)}\ , \frac{\partial^2 \varphi}{\partial x \partial z } \right|_{x=0}. \nonumber\end{aligned}\ ] ] it is convenient to choose the function @xmath112 in symmetrical form ( by evenly continuing the @xmath109 for negative values of @xmath61 ) : @xmath116 in this case @xmath117 and therefore ( i ) the function @xmath112 gives no contribution into @xmath65 and ( ii ) tangential tensions @xmath118 and @xmath119 in boundary conditions ( [ boundaryva ] ) turn to zero . the solution of the problem for @xmath120 is known ( see for example @xcite ) . using condition ( [ condition ] ) and following table integrals : @xmath121 we substitute @xmath122 into ( [ barx ] ) and obtain : @xmath123 assuming that spectral density of absorbed power is equal to @xmath124 one can find the formula ( [ sdsn ] ) .
in this appendix we deduce the formula ( [ sdtd ] ) using langevin abroach for thermal conductivity equation ( [ tdft ] ) and recalculate the thermo - dynamical fluctuations of temperature into deformations . using time and space fourier transform one can write down the formal solution of ( [ tdft ] ) : @xmath125 we interest in temperature @xmath126 averaged over some volume @xmath127 . assuming that @xmath128 one can calculate @xmath129 @xmath130 here we equate ( [ baru2 ] ) to ( [ tvar ] ) and find @xmath131 . in order to find td temperature fluctuations in half - space with boundary condition ( [ boundarytc ] )
one can solve equivalent problem for full space with source function even continued for negative @xmath61 , so that its correlator is equal to : @xmath132\ , \delta ( k_y - k_y')\,\delta ( k_z - k_z')\ , \delta ( \omega-\omega')\ ] ] after that we must substitute the expression ( [ u ] ) for @xmath35 into elasticity equation ( [ edfull ] , [ edboundary ] ) .
the solution for @xmath110 can be found according to the same scheme as in appendix a : @xmath111 , where @xmath112 satisfies poisson equation ( [ varphipua ] ) .
again @xmath112 can be written in symmetrical form ( [ simvarphi ] ) and the function @xmath120 must satisfy equation ( [ edfull ] ) with zero right part and following boundary conditions : @xmath133 the solution for @xmath120 is known @xcite @xmath134 substituting it into ( [ barx ] ) we find @xmath135 and spectral density @xmath136 : @xmath137 from the last formula , using integration in spherical coordinate system and neglecting the term @xmath138 in denominator according to condition ( [ condition ] ) , we obtain formula ( [ sdtd ] ) asymptotically correct for the frequencies @xmath139 .
the table integrals ( [ int2 ] , [ int3 ] ) was used above in calculations of @xmath140 .
to calculate the fluctuations of the surface using fdt approach with thermoelastic losses we should solve the system : @xmath141 in zero order approximation we simply solve elasticity problem neglecting temperature @xmath35 in the right part of the first equation in ( [ newv ] ) and in boarder conditions .
this solution is well known @xcite : @xmath142 by substituting this solution in ( [ barx ] ) and using ( [ int2 ] ) we obtain susceptibility @xmath143 ( [ rigidity ] ) .
this result totally coincides with the calculations for the half - space in @xcite ( if we note that the stored static strain energy @xmath144 and @xmath145 ) and has minor numerical difference with the approximation in @xcite . to find first order approximation for the distribution of temperature we substitute @xmath146 in ( [ newu ] ) to find that @xmath147 we then substitute this solution in ( [ newv ] ) to find second order approximation in the same way as in the two appendices above by finding appropriate evenly continued @xmath112 function satisfying ( [ varphipua ] , [ boundaryva ] ) .
we find using ( [ int3 ] ) @xmath148 by substituting this expression in @xmath149 and using ( [ barx ] ) we may finally obtain @xmath150 and hence @xmath151 which leads through ( [ spectrum ] ) to precisely the same expression as ( [ sdtd ] ) . 99 a. abramovici _ et al .
_ , science * 256*(1992)325 .
a. abramovici _ et al .
_ , phys.letters.a * 218 * , 157 ( 1996 ) .
i. batov , m. dantchevskaya , proc . of vii
all - russian conf . on silica glasses , st .- petersburg 1991 , p.45 .
a. gillespi and f. raab , _ phys .
d _ * 52 * , 577 ( 1995 ) | thermodynamical fluctuations of temperature in mirrors of gravitational wave antennae are transformed through thermal expansion coefficient into additional noise .
this source of noise , which may also be interpreted as fluctuations due to thermoelastic damping , may not be neglected and leads to the necessity to reexamine the choice of materials for the mirrors .
additional source of noise are fluctuations of the mirrors surfaces caused by optical power absorbed in dielectrical reflective layers . |
this paper is devoted to the study of two equivalent dynamical systems of the grigorchuk group @xmath0 , the action on the space of the marked schreier graphs and the action on the space of subgroups .
the main object of study is going to be the set of the marked schreier graphs of the standard action of the group on the boundary of the binary rooted tree and their limit points in the space of all marked schreier graphs of @xmath0 . given a finitely generated group @xmath1 with a fixed generating set @xmath2 , to each action of @xmath1 we associate its schreier graph , which is a combinatorial object that encodes some information about orbits of the action .
the marked schreier graphs of various actions form a topological space @xmath3 and there is a natural action of @xmath1 on this space .
any action of @xmath1 corresponds to an invariant set in @xmath3 and any action with an invariant measure gives rise to an invariant measure on @xmath3 .
the latter allows to define the notion of a random schreier graph , which is closely related to the notion of a random subgroup of @xmath1 .
a principal problem is to determine how much information about the original action can be learned from the schreier graphs .
the worst case here is a free action , for which nothing beyond its freeness can be recovered .
vershik @xcite introduced the notion of a totally nonfree action .
this is an action such that all points have distinct stabilizers . in this case
the information about the original action can be recovered almost completely .
further extensive development of these ideas was done by grigorchuk @xcite .
the grigorchuk group was introduced in @xcite as a simple example of a finitely generated infinite torsion group .
later it was revealed that this group has intermediate growth and a number of other remarkable properties ( see the survey @xcite ) . in this paper
we are going to use the branching property of the grigorchuk group , which implies that its action on the boundary of the binary rooted tree is totally nonfree in a very strong sense . ]
the main results of the paper are summarized in the following two theorems .
the first theorem contains a detailed description of the invariant set of the schreier graphs .
the second theorem is concerned with the dynamics of the group action restricted to that invariant set .
[ main1 ] let @xmath4 be the mapping that assigns to any point on the boundary of the binary rooted tree the marked schreier graph of its orbit under the action of the grigorchuk group .
then * @xmath5 is injective ; * @xmath5 is measurable ; it is continuous everywhere except for a countable set , the orbit of the point @xmath6 ; * the schreier graph @xmath7 is an isolated point in the closure of @xmath8 ; the other isolated points are graphs obtained by varying the marked vertex of @xmath7 ; * the closure of the set @xmath8 differs from @xmath8 in countably many points ; these are obtained from three graphs @xmath9 choosing the marked vertex arbitrarily ; * as an unmarked graph , @xmath7 is a double quotient of each @xmath10 ( @xmath11 ) ; also , there exists a graph @xmath12 such that each @xmath10 is a double quotient of @xmath12 .
[ main2 ] using notation of the previous theorem , let @xmath13 be the set of non - isolated points of the closure of @xmath8 .
then * @xmath13 is a minimal invariant set for the action of the grigorchuk group @xmath0 on @xmath14 ; * the action of @xmath0 on @xmath13 is a continuous extension of the action on the boundary of the binary rooted tree ; the extension is one - to - one everywhere except for a countable set , where it is three - to - one ; * there exists a unique borel probability measure @xmath15 on @xmath14 invariant under the action of @xmath0 and supported on the set @xmath13 ; * the action of @xmath0 on @xmath13 with the invariant measure @xmath15 is isomorphic to the action of @xmath0 on @xmath16 with the uniform measure . the paper is organized as follows .
section [ graph ] contains a detailed construction of the space of marked graphs .
the construction is more general than that in @xcite .
section [ act ] contains notation and definitions concerning group actions . in section [ sch ]
we introduce the schreier graphs of a finitely generated group , the space of marked schreier graphs , and the action of the group on that space . in section [ sub ]
we study the space of subgroups of a countable group and establish a relation of this space with the space of marked schreier graphs .
section [ tree ] is devoted to general considerations concerning groups of automorphisms of a regular rooted tree and their actions on the boundary of the tree . in section [ grig ]
we apply the results of the previous sections to the study of the grigorchuk group and prove theorems [ main1 ] and [ main2 ] .
the exposition in sections [ graph][tree ] is as general as possible , to make their results applicable to the actions of groups other than the grigorchuk group .
the author is grateful to rostislav grigorchuk , anatoly vershik , and oleg ageev for useful discussions . ]
a _ graph _ @xmath17 is a combinatorial object that consists of _ vertices _ and _ edges _ related so that every edge joins two vertices or a vertex to itself ( in the latter case the edge is called a _ loop _ ) .
the vertices joined by an edge are its _
endpoints_. let @xmath18 be the _ vertex set _ of the graph @xmath17 and @xmath19 be the set of its edges .
traditionally @xmath19 is regarded as a subset of @xmath20 , i.e. , any edge is identified with the pair of its endpoints . in this paper , however , we are going to consider graphs with multiple edges joining the same vertices .
also , our graphs will carry additional structure . to accomodate this , we regard @xmath19 merely as a reference set whereas the actual information about the edges is contained in their _ attributes _ , which are functions on @xmath19 . in a plain graph
any edge has only one attribute : its endpoints , which are an unordered pair of vertices .
other types of graphs involve more attributes .
a _ directed graph _ has directed edges .
the endpoints of a _ directed edge _
@xmath21 are ordered , namely , there is a _ beginning _ @xmath22 and an _ end _ @xmath23 .
clearly , an undirected loop is no different from a directed one .
an undirected edge joining two distinct vertices may be regarded as two directed edges @xmath24 and @xmath25 with the same endpoints and opposite directions , i.e. , @xmath26 and @xmath27 . this way we can represent any graph with undirected edges as a directed graph .
conversely , some directed graphs can be regarded as graphs with undirected edges ( we shall use this in section [ grig ] ) . a _ graph with labeled edges _
is a graph in which each edge @xmath21 is assigned a _ label _
the labels are elements of a prescribed finite set .
a _ marked graph _ is a graph with a distinguished vertex called the _ marked vertex_. the vertices of a graph are pictured as dots or small circles .
an undirected edge is pictured as an arc joining its endpoints .
a directed edge is pictured as an arrow going from its beginning to its end .
the label of an edge is written next to the edge .
alternatively , one might think of labels as colors and picture a graph with labeled edges as a colored graph .
let @xmath17 be a graph and @xmath18 be its vertex set . to any subset @xmath29 of @xmath18
we associate a graph @xmath30 called a _ subgraph _ of @xmath17 . by definition ,
the vertex set of the graph @xmath30 is @xmath29 and the edges are those edges of @xmath17 that have both endpoints in @xmath29 ( all attributes are retained ) .
if @xmath17 is a marked graph and the marked vertex is in @xmath29 , it will also be the marked vertex of the subgraph @xmath30 . otherwise the subgraph is not marked .
suppose @xmath31 and @xmath32 are graphs of the same type .
for any @xmath33 let @xmath34 be the vertex set of @xmath35 and @xmath36 be the set of its edges .
the graph @xmath31 is said to be _ isomorphic _ to @xmath32 if there exist bijections @xmath37 and @xmath38 that respect the structure of the graphs .
first of all , this means that @xmath39 sends the endpoints of any edge @xmath40 to the endpoints of @xmath41 .
if @xmath31 and @xmath32 are directed graphs , we additionally require that @xmath42 and @xmath43 for all @xmath40 .
if @xmath31 and @xmath32 have labeled edges , we also require that @xmath44 preserve labels .
if @xmath31 and @xmath32 are marked graphs , we also require that @xmath39 map the marked vertex of @xmath31 to the marked vertex of @xmath32 . assuming the above requirements are met , the mapping @xmath39 of the vertex set is called an _ isomorphism of the graphs _ @xmath31 and @xmath32 . if @xmath45 then @xmath39 is also called an _ automorphism _ of the graph @xmath31 .
we call the mapping @xmath44 a _ companion mapping _ of @xmath39 .
if neither of the graphs @xmath31 and @xmath32 admits multiple edges with identical attributes , the companion mapping is uniquely determined by the isomorphism @xmath39 .
further , we say that the graph @xmath32 is a _ quotient _ of @xmath31 if all of the above requirements are met except the mappings @xmath39 and @xmath44 need not be injective .
moreover , @xmath32 is a _
@xmath46-fold quotient _ of @xmath31 if @xmath39 is @xmath46-to-@xmath47 .
finally , we say that the graphs @xmath31 and @xmath32 coincide _ up to renaming edges _ if they have the same vertices and there is a one - to - one correspondence between their edges that preserves all attributes .
an equivalent condition is that the identity map on the common vertex set is an isomorphism of these graphs . a _ path _ in a graph
@xmath17 is a sequence of vertices @xmath48 together with a sequence of edges @xmath49 such that for any @xmath50 the endpoints of the edge @xmath51 are @xmath52 and @xmath53 .
we say that the vertex @xmath54 is the beginning of the path and @xmath55 is the end .
the path is _ closed _ if @xmath56 .
the _ length _ of the path is the number of edges in the sequence ( counted with repetitions ) , which is a nonnegative integer .
the path is a _ directed path _ if the edges are directed and , moreover , @xmath57 and @xmath58 for @xmath50 . if the graph @xmath17 has labeled edges then the path is assigned a _ code word _
@xmath59 , which is a string of labels read off the edges while traversing the path .
we say that a vertex @xmath60 of a graph @xmath17 is _ connected to _
a vertex @xmath61 if there is a path in @xmath17 such that the beginning of the path is @xmath60 and the end is @xmath61 .
the length of the shortest path with this property is the _ distance _ from @xmath60 to @xmath61 .
the connectivity is an equivalence relation on the vertex set of @xmath17 .
the subgraphs of @xmath17 corresponding to the equivalence classes are _ connected components _ of the graph @xmath17 .
a graph is _ connected _ if all vertices are connected to each other . clearly , the connected components of any graph are its maximal connected subgraphs .
let @xmath60 be a vertex of a graph @xmath17 .
for any integer @xmath62 , the _ closed ball _ of radius @xmath63 centered at @xmath60 , denoted @xmath64 , is the subgraph of @xmath17 whose vertex set consists of all vertices in @xmath17 at distance at most @xmath63 from the vertex @xmath60 .
a graph is _ locally finite _ if every vertex is the endpoint for only finitely many edges .
if the graph @xmath17 is locally finite then any closed ball of @xmath17 is a _ finite graph _ , i.e. , it has a finite number of vertices and a finite number of edges .
let @xmath65 denote the set of isomorphism classes of all marked directed graphs with labeled edges . for convenience , we regard elements of @xmath65 as graphs ( i.e. , we choose representatives of isomorphism classes ) .
it is easy to observe that connectedness and local finiteness of graphs are preserved under isomorphisms .
let @xmath66 denote the subset of @xmath65 consisting of connected , locally finite graphs .
we endow the set @xmath66 with a topology as follows .
the topology is generated by sets @xmath67 , where @xmath68 runs over all finite graphs in @xmath66 and @xmath69 can be any subset of the vertex set of @xmath68 . by definition
, @xmath70 is the set of all isomorphism classes in @xmath66 containing any graph @xmath17 such that @xmath68 is a subgraph of @xmath17 and every edge of @xmath17 with at least one endpoint in the set @xmath69 is actually an edge of @xmath68 . in other words , there is no edge in @xmath17 that joins a vertex from @xmath69 to a vertex outside the vertex set of @xmath68 .
for example , @xmath71 is the set of all graphs in @xmath66 that have a subgraph isomorphic to @xmath68 . on the other hand
, if @xmath69 is the entire vertex set of @xmath68 then @xmath70 contains only the graph @xmath68 . as a consequence ,
every finite graph in @xmath66 is an isolated point .
the following lemma implies that sets of the form @xmath70 constitute a base of the topology .
[ graph1 ] any nonempty intersection of two sets of the form @xmath70 can be represented as the union of some sets of the same form .
let @xmath72 be finite graphs and @xmath73 be subsets of their vertex sets .
consider an arbitrary graph @xmath74 .
for any @xmath33 let @xmath75 be an isomorphism of the graph @xmath35 with a subgraph of @xmath17 such that no edge of @xmath17 joins a vertex from the set @xmath76 to a vertex outside @xmath77 .
denote by @xmath68 the finite subgraph of @xmath17 with the vertex set @xmath78 . since the subgraphs of @xmath17 with vertex sets @xmath79 and @xmath80 are both connected and both contain the marked vertex of @xmath17 , the subgraph @xmath68 is also marked and connected .
besides , no edge of @xmath17 joins a vertex from the set @xmath81 to a vertex outside @xmath82 .
hence @xmath83 .
it is easy to observe that the entire set @xmath70 is contained in the intersection @xmath84 .
the lemma follows .
next we introduce a distance function on @xmath66 .
consider arbitrary graphs @xmath72 .
let @xmath85 be the marked vertex of @xmath31 and @xmath86 be the marked vertex of @xmath32 .
we let @xmath87 if the graphs @xmath31 and @xmath32 are isomorphic ( i.e. , they represent the same element of @xmath66 ) .
otherwise we let @xmath88 , where @xmath63 is the smallest nonnegative integer such that the closed balls @xmath89 and @xmath90 are not isomorphic .
[ graph2 ] the graphs @xmath31 and @xmath32 are isomorphic if and only if the closed balls @xmath89 and @xmath90 are isomorphic for any integer @xmath62 . for any @xmath33 let @xmath34 denote the vertex set of the graph @xmath35 and @xmath36 denote its set of edges .
further , for any integer @xmath62 let @xmath91 and @xmath92 denote the vertex set and the set of edges of the closed ball @xmath93 .
first assume that the graph @xmath31 is isomorphic to @xmath32 .
let @xmath37 be an isomorphism of these graphs and @xmath38 be its companion mapping .
clearly , @xmath94 .
it is easy to see that any isomorphism of graphs preserves distances between vertices .
it follows that @xmath39 maps @xmath95 onto @xmath96 for any @xmath62 .
consequently , @xmath44 maps @xmath97 onto @xmath98 .
hence the restriction of @xmath39 to the set @xmath95 is an isomorphism of the graphs @xmath89 and @xmath90 .
now assume that for every integer @xmath62 the closed balls @xmath89 and @xmath90 are isomorphic .
let @xmath99 be an isomorphism of these graphs and @xmath100 be its companion mapping
. clearly , @xmath101 .
note that the closed ball @xmath93 is also the closed ball with the same center and radius in any of the graphs @xmath102 , @xmath103 .
it follows that the restriction of the mapping @xmath104 to the set @xmath95 is an isomorphism of the graphs @xmath89 and @xmath90 while the restriction of @xmath105 to @xmath97 is its companion mapping .
since the graphs @xmath31 and @xmath32 are locally finite , the sets @xmath106 are finite .
hence there are only finitely many distinct restrictions @xmath107 or @xmath108 for any fixed @xmath63 .
therefore one can find nested infinite sets of indices @xmath109 such that the restriction @xmath107 is the same for all @xmath110 and the restriction @xmath108 is the same for all @xmath110 .
for any integer @xmath62 let @xmath111 and @xmath112 , where @xmath110 . by construction ,
@xmath113 is a restriction of @xmath114 and @xmath115 is a restriction of @xmath116 whenever @xmath117 .
hence there exist maps @xmath37 and @xmath38 such that all @xmath113 are restrictions of @xmath39 and all @xmath115 are restrictions of @xmath44 .
since the graphs @xmath31 and @xmath32 are connected , any finite collection of vertices and edges in either graph is contained in a closed ball centered at the marked vertex . as for any @xmath62 the mapping @xmath113 is an isomorphism of @xmath89 and @xmath90 and @xmath115 is its companion mapping , it follows that @xmath39 is an isomorphism of @xmath31 and @xmath32 and @xmath44 is its companion mapping .
lemma [ graph2 ] implies that @xmath118 is a well - defined function on @xmath119 .
this is a distance function , which makes @xmath66 into an ultrametric space .
[ graph3 ] the distance function @xmath118 is compatible with the topology on @xmath66 .
the base of the topology on @xmath66 consists of the sets @xmath70 .
the base of the topology defined by the distance function @xmath118 is formed by open balls @xmath120 , where @xmath31 can be any graph in @xmath66 and @xmath121 .
we have to show that any element of either base is the union of some elements of the other base .
first consider an open ball @xmath122 .
if @xmath123 then @xmath124 , which is the union of all sets @xmath70 . otherwise let @xmath63 be the largest integer such that @xmath125 . clearly , @xmath126
let @xmath127 , where @xmath85 is the marked vertex of the graph @xmath31 , and let @xmath69 be the set of all vertices of @xmath31 at distance at most @xmath128 from @xmath85 .
consider an arbitrary graph @xmath129 such that @xmath68 is a subgraph of @xmath17 .
clearly , @xmath68 is also a subgraph of the closed ball @xmath130 .
if @xmath60 is a vertex of @xmath17 at distance @xmath46 from the marked vertex @xmath85 , then any vertex joined to @xmath60 by an edge is at distance at most @xmath131 and at least @xmath132 from @xmath85 .
moreover , if @xmath133 then @xmath60 is joined to a vertex at distance exactly @xmath132 from @xmath85 .
it follows that @xmath134 if and only if no vertex from the set @xmath69 is joined in @xmath17 to a vertex that is not a vertex of @xmath68 .
thus @xmath135 .
now consider the set @xmath70 , where @xmath68 is a finite graph in @xmath66 and @xmath69 is a subset of its vertex set .
denote by @xmath54 the marked vertex of @xmath68 .
let @xmath63 be the smallest integer such that every vertex of @xmath68 is at distance at most @xmath63 from @xmath54 and every vertex from @xmath69 is at distance at most @xmath128 from @xmath54 .
take any graph @xmath129 such that @xmath68 is a subgraph of @xmath17 and there is no edge in @xmath17 joining a vertex from @xmath69 to a vertex outside the vertex set of @xmath68 .
let @xmath136 and @xmath137 be the set of all vertices of @xmath17 at distance at most @xmath128 from @xmath54 . by the above , @xmath138 . at the same time , @xmath139 since @xmath68 is a subgraph of @xmath31 and @xmath69 is a subset of @xmath137 . thus for any graph @xmath83 the entire open ball @xmath140
is contained in @xmath70 . in particular , @xmath70 is the union of those open balls . given a positive integer @xmath141 and a finite set @xmath142 , let @xmath143 denote the subset of @xmath65 consisting of all graphs in which every vertex is the endpoint for at most @xmath141 edges and every label belongs to @xmath142 .
further , let @xmath144 .
[ graph4 ] @xmath145 is a compact subset of the metric space @xmath66 .
we have to show that any sequence of graphs @xmath146 in @xmath145 has a subsequence converging to some graph in @xmath145 . for any positive integer @xmath63 let @xmath147 denote the vertex set of the graph @xmath148 , @xmath149 denote its sets of edges , and @xmath55 denote the marked vertex of @xmath148 . first consider the special case when each @xmath148 is a subgraph of @xmath150 .
let @xmath17 be the graph with the vertex set @xmath151 and the set of edges @xmath152 .
we assume that any edge @xmath153 retains its attributes ( beginning , end , and label ) in the graph @xmath17 .
the common marked vertex of the graphs @xmath148 is set as the marked vertex of @xmath17 .
note that any finite collection of vertices and edges of the graph @xmath17 is already contained in some @xmath148 . as the graphs @xmath146 belong to @xmath145 , it follows that @xmath154 as well .
in particular , for any integer @xmath155 the closed ball @xmath156 is a finite graph .
then it is a subgraph of some @xmath148 .
clearly , @xmath156 is also a subgraph of the graphs @xmath157 .
moreover , it remains the closed ball of radius @xmath46 centered at the marked vertex in all these graphs .
it follows that @xmath158 for @xmath159 .
since @xmath46 can be arbitrarily large , the sequence @xmath146 converges to @xmath17 in the metric space @xmath66 .
next consider a more general case when each @xmath148 is isomorphic to a subgraph of @xmath150 .
this case is reduced to the previous one by repeatedly using the following observation : if a graph @xmath160 is isomorphic to a subgraph of a graph @xmath161 then there exists a graph @xmath162 isomorphic to @xmath161 such that @xmath160 is a subgraph of @xmath162 . finally consider the general case .
for any graph in @xmath145 , the closed ball of radius @xmath46 with any center contains at most @xmath163 vertices while the number of edges is at most @xmath141 times the number of vertices .
hence for any fixed @xmath46 the number of vertices and edges in the balls @xmath164 is uniformly bounded , which implies that there are only finitely many non - isomorphic graphs among them .
therefore one can find nested infinite sets of indices @xmath165 such that the closed balls @xmath164 are isomorphic for all @xmath166 . choose an increasing sequence of indices @xmath167 such that @xmath168 for all @xmath46 , and
let @xmath169 be the closed ball of radius @xmath46 in the graph @xmath170 centered at the marked point @xmath171 . clearly , @xmath172 and @xmath173 . by construction ,
@xmath169 is isomorphic to a subgraph of @xmath174 whenever @xmath175 . by the above the sequence @xmath176
converges to a graph @xmath154 .
since @xmath173 for all @xmath155 , the subsequence @xmath177 converges to the graph @xmath17 as well .
let @xmath178 be an arbitrary nonempty set .
invertible transformations @xmath179 form a transformation group .
an _ action _ @xmath180 of an abstract group @xmath1 on the set @xmath178 is a homomorphism of @xmath1 into that transformation group .
the action can be regarded as a collection of invertible transformations @xmath181 , @xmath182 , where @xmath183 is the image of @xmath184 under the homomorphism .
the transformations are to satisfy @xmath185 for all @xmath186 .
we say that @xmath183 is the action of an element @xmath184 within the action @xmath180 .
alternatively , the action of the group @xmath1 can be given as a mapping @xmath187 such that @xmath188 for all @xmath182 and @xmath189 .
such a mapping defines an action of @xmath1 if and only if the following two conditions hold : * @xmath190 for all @xmath186 and @xmath189 ; * @xmath191 for all @xmath189 , where @xmath192 is the unity of the group @xmath1 . a nonempty set @xmath193 is called a _ generating set _ for the group @xmath1 if any element @xmath182 can be represented as a product @xmath194 where each factor @xmath195 is an element of @xmath2 or the inverse of an element of @xmath2 .
the elements of the generating set are called _ generators _ of the group @xmath1 .
the generating set @xmath2 is _ symmetric _ if it is closed under taking inverses , i.e. , @xmath196 whenever @xmath197 . if @xmath2 is a generating set for @xmath1 then any action @xmath180 of the group @xmath1 is uniquely determined by transformations @xmath198 , @xmath197 .
suppose @xmath1 is a topological group .
an action of @xmath1 on a topological space @xmath178 is a _ continuous action _ if it is continuous as a mapping of @xmath199 to @xmath178 .
similarly , an action of @xmath1 on a measured space @xmath178 is a _ measurable action _ if it is measurable as a mapping of @xmath199 to @xmath178 .
a measurable action @xmath180 of the group @xmath1 on a measured space @xmath178 with a measure @xmath200 is _ measure - preserving _ if the action of every element of @xmath1 is measure - preserving , i.e. , @xmath201 for all @xmath182 and measurable sets @xmath202 . in what follows
, the group @xmath1 will be a discrete countable group . in that case ,
an action @xmath180 of @xmath1 is continuous if and only if all transformations @xmath183 , @xmath182 are continuous .
likewise , the action @xmath180 is measurable if and only if every @xmath183 is measurable . given an action @xmath180 of a group @xmath1 on a set @xmath178 , the _ orbit _ @xmath203 of a point @xmath189 under the action @xmath180 is the set of all points @xmath204 , @xmath182 .
a subset @xmath205 is _ invariant _ under the action @xmath180 if @xmath206 for all @xmath182 .
clearly , the orbit @xmath203 is invariant under the action .
moreover , this is the smallest invariant set containing @xmath207 .
the _ restriction _ of the action @xmath180 to a nonempty invariant set @xmath208 is an action of @xmath1 obtained by restricting every transformation @xmath183 to @xmath208 .
equivalently , one might restrict the mapping @xmath187 to the set @xmath209 . the action @xmath180 is _ transitive _ if the only invariant subsets of @xmath178 are the empty set and @xmath178 itself
equivalently , the orbit of any point is the entire set @xmath178 . assuming the action @xmath180 is continuous , it is _ topologically transitive _ if there is an orbit dense in @xmath178 , and _ minimal _ if every orbit of @xmath180 is dense .
the action is minimal if and only if the empty set and @xmath178 are the only closed invariant subsets of @xmath178 . assuming the action @xmath180 is measure - preserving , it is _ ergodic _ if any measurable invariant subset of @xmath178 has zero or full measure .
a continuous action on a compact space @xmath178 is _ uniquely ergodic _ if there exists a unique borel probability measure on @xmath178 invariant under the action ( the action is going to be ergodic with respect to that measure ) . given an action @xmath180 of a group @xmath1 on a set @xmath178 , the _ stabilizer _ @xmath210 of a point @xmath189 under the action @xmath180 is the set of all elements @xmath182 whose action fixes @xmath207 , i.e. , @xmath211 .
the stabilizer @xmath210 is a subgroup of @xmath1 .
the action is _ free _ if all stabilizers are trivial . in the case
when the action @xmath180 is continuous , we define the _ neighborhood stabilizer _
@xmath212 of a point @xmath189 to be the set of all @xmath182 whose action fixes the point @xmath207 along with its neighborhood ( the neighborhood may depend on @xmath184 ) .
the neighborhood stabilizer @xmath212 is a normal subgroup of @xmath210 .
let @xmath213 and @xmath214 be actions of a group @xmath1 on sets @xmath215 and @xmath216 , respectively .
the actions @xmath180 and @xmath217 are _ conjugated _ if there exists a bijection @xmath218 such that @xmath219 for all @xmath182 .
an equivalent condition is that @xmath220 for all @xmath182 and @xmath221 .
the bijection @xmath39 is called a _
conjugacy _ of the action @xmath180 with @xmath217 .
two continuous actions of the same group are _ continuously conjugated _ if they are conjugated and , moreover , the conjugacy can be chosen to be a homeomorphism .
similarly , two measurable actions are _ measurably conjugated _ if they are conjugated and , moreover , the conjugacy @xmath39 can be chosen so that both @xmath39 and the inverse @xmath222 are measurable . also
, two measure - preserving actions are _ isomorphic _ if they are conjugated and , moreover , the conjugacy can be chosen to be an isomorphism of spaces with measure .
the measure - preserving actions are _ isomorphic modulo zero measure _ if each action admits an invariant set of full measure such that the corresponding restrictions are isomorphic . given two actions @xmath213 and @xmath214 of a group @xmath1 , the action @xmath180 is an _ extension _ of @xmath217 if there exists a mapping @xmath39 of @xmath215 onto @xmath216 such that @xmath223 for all @xmath182 .
the extension is _
@xmath46-to-@xmath47 _ if @xmath39 is @xmath46-to-@xmath47 .
the extension is _ continuous _ if the actions @xmath180 and @xmath217 are continuous and @xmath39 can be chosen continuous .
let @xmath1 be a finitely generated group .
let us fix a finite symmetric generating set @xmath2 for @xmath1 . given an action @xmath180 of the group @xmath1 on a set @xmath178 , the _ schreier graph
_ @xmath224 of the action relative to the generating set @xmath2 is a directed graph with labeled edges .
the vertex set of the graph @xmath224 is @xmath178 , the set of edges is @xmath225 , and the set of labels is @xmath2 .
for any @xmath189 and @xmath197 the edge @xmath226 has beginning @xmath207 , end @xmath227 , and carries label @xmath228 .
clearly , the action @xmath180 can be uniquely recovered from its schreier graph . given another action @xmath229 of @xmath1 on some set @xmath230 ,
the schreier graph @xmath231 is isomorphic to @xmath224 if and only if the actions @xmath180 and @xmath229 are conjugated .
indeed , a bijection @xmath232 is an isomorphism of the schreier graphs if and only if @xmath233 for all @xmath197 , which is equivalent to @xmath39 being a conjugacy of the action @xmath180 with @xmath229 .
any graph of the form @xmath224 is called a _ schreier graph _ of the group @xmath1 ( relative to the generating set @xmath2 ) .
notice that any graph isomorphic to a schreier graph is also a schreier graph up to renaming edges .
this follows from the next proposition , which explains how to recognize a schreier graph of @xmath1 .
[ sch1 ] a directed graph @xmath17 with labeled edges is , up to renaming edges , a schreier graph of the group @xmath1 relative to the generating set @xmath2 if and only if the following conditions are satisfied : * all labels are in @xmath2 ; * for any vertex @xmath60 and any generator
@xmath197 there exists a unique edge with beginning @xmath60 and label @xmath228 ; * given a directed path with code word @xmath234 , the path is closed whenever the reversed code word @xmath235 equals @xmath192 when regarded as a product in @xmath1 .
first suppose @xmath17 is a schreier graph @xmath224 .
consider an arbitrary directed path in the graph @xmath17 .
let @xmath60 be the beginning of the path and @xmath234 be its code word .
then the consecutive vertices of the path are @xmath236 , where @xmath237 for @xmath238 .
hence the end of the path is @xmath239 , where @xmath184 denotes @xmath235 regarded as a product in @xmath1 .
clearly , the path is closed whenever @xmath240 .
thus any schreier graph of the group @xmath1 satisfies the condition ( iii ) .
the conditions ( i ) and ( ii ) are trivially satisfied as well .
it is easy to see that the conditions ( i ) , ( ii ) , and ( iii ) are preserved under isomorphisms of graphs . in particular , they hold for any graph that coincides with a schreier graph up to renaming edges .
now suppose @xmath17 is a directed graph with labeled edges that satisfies the conditions ( i ) , ( ii ) , and ( iii ) .
let @xmath178 denote the vertex set of @xmath17 . given a word @xmath241 over the alphabet @xmath2
, we define a transformation @xmath242 as follows .
the condition ( ii ) implies that for any vertex @xmath243 there is a unique directed path in @xmath17 with beginning @xmath60 and code word @xmath235 ( the word @xmath244 reversed ) .
we set @xmath245 to be the end of that path . for any words @xmath241 and @xmath246 over the alphabet @xmath2
let @xmath247 denote the concatenated word @xmath248 .
then @xmath249 for all @xmath243 .
any word over the alphabet @xmath2 can be regarded as a product in the group @xmath1 thus representing an element @xmath250 .
clearly , the concatenation of words corresponds to the multiplication in the group .
the condition ( iii ) means that @xmath251 is the identity transformation whenever the word @xmath244 represents the unity @xmath192 .
this implies that transformations @xmath251 and @xmath252 are the same if the words @xmath244 and @xmath253 represent the same element @xmath182 . indeed , let @xmath241 , @xmath246 and consider the third word @xmath254 .
the word @xmath255 represents the inverse @xmath256 .
therefore the words @xmath257 and @xmath258 both represent the unity . then @xmath259 .
now for any @xmath182 we let @xmath260 , where @xmath244 is an arbitrary word over the alphabet @xmath2 representing @xmath184 . by the above
@xmath183 is a well - defined transformation of @xmath178 , @xmath261 is the identity transformation , and @xmath262 for all @xmath263 .
hence the transformations @xmath183 , @xmath182 constitute an action @xmath180 of the group @xmath1 on the vertex set @xmath178 . by construction , for any @xmath243 and @xmath197 the vertex @xmath264 is the end of the edge with beginning @xmath60 and label @xmath228 . in view of the conditions ( i ) and
( ii ) , this means that the graph @xmath17 coincides with the schreier graph @xmath224 up to renaming edges . for any @xmath189 let @xmath265 denote the schreier graph of the restriction of the action @xmath180 to the orbit of @xmath207 .
we refer to @xmath265 as the schreier graph of the orbit of @xmath207 .
it is easy to observe that @xmath265 is the connected component of the graph @xmath224 containing the vertex @xmath207 . in particular , the schreier graph of the action @xmath180 is connected if and only if the action is transitive , in which case @xmath266 for all @xmath189 .
let @xmath267 denote a marked graph obtained from @xmath265 by marking the vertex @xmath207 .
we refer to it as the _ marked schreier graph _ of the point @xmath207 ( under the action @xmath180 ) .
notice that the point @xmath207 and the restriction of the action @xmath180 to its orbit are uniquely recovered from the graph @xmath267 .
any graph of the form @xmath267 is called a marked schreier graph of the group @xmath1 ( relative to the generating set @xmath2 ) .
let @xmath3 denote the set of isomorphism classes of all marked schreier graphs of the group @xmath1 relative to the generating set @xmath2 .
a graph @xmath268 belongs to @xmath3 if it is a marked directed graph that is connected and satisfies conditions ( i ) , ( ii ) , ( iii ) of proposition [ sch1 ] .
the group @xmath1 acts naturally on the set of the marked schreier graphs of @xmath1 by changing the marked vertex .
the action @xmath269 is given by @xmath270 , @xmath182 .
it turns out that @xmath269 is well defined as an action on @xmath3 .
indeed , let @xmath271 be a marked schreier graph isomorphic to @xmath267 .
then any isomorphism @xmath39 of the latter graph with the former one is simultaneously a conjugacy of the restriction of the action @xmath180 to the orbit of @xmath207 with the restriction of the action @xmath217 to the orbit of @xmath272 . since @xmath273
, it follows that @xmath274 for all @xmath182 .
hence for any @xmath182 the map @xmath39 is also an isomorphism of the graph @xmath275 with @xmath276 .
[ sch2 ] @xmath3 is a compact subset of the metric space @xmath66 .
the action of the group @xmath1 ( regarded as a discrete group ) on @xmath3 is continuous .
let @xmath141 be the number of elements in the generating set @xmath2 .
then every vertex @xmath60 of a graph @xmath277 is the beginning of exactly @xmath141 edges .
furthermore , @xmath60 is the end of an edge with beginning @xmath61 and label @xmath228 if and only if @xmath61 is the end of the edge with beginning @xmath60 and label @xmath278 .
it follows that @xmath60 is also the end of exactly @xmath141 edges .
hence any vertex of @xmath17 is an endpoint for at most @xmath279 edges .
therefore @xmath280 .
since all marked schreier graphs are connected , we have @xmath281 .
now let us show that the set @xmath3 is closed in the topological space @xmath66 . take any graph @xmath129 not in that set
. then @xmath17 does not satisfy at least one of the conditions ( i ) , ( ii ) , and ( iii ) in proposition [ sch1 ] .
first consider the case when the condition ( i ) or ( iii ) does not hold . since the graph @xmath17 is locally finite
, it has a finite subgraph @xmath68 for which the same condition does not hold .
since @xmath17 is connected , we can choose the subgraph @xmath68 to be marked and connected so that @xmath282 .
clearly , the same condition does not hold for any graph @xmath30 such that @xmath68 is a subgraph of @xmath30 .
it follows that the neighborhood @xmath71 of the graph @xmath17 is disjoint from @xmath3 .
next consider the case when @xmath17 does not satisfy the condition ( ii ) .
let @xmath60 be the vertex of @xmath17 such that for some generator @xmath197 there are either several edges with beginning @xmath60 and label @xmath228 or no such edges at all . since @xmath129 , there exists a finite connected subgraph @xmath68 of @xmath17 that contains the marked vertex , the vertex @xmath60 , and all edges for which @xmath60 is an endpoint
. then @xmath282 and the open set @xmath283 is a neighborhood of @xmath17 . by construction , the condition ( ii ) fails in the entire neighborhood so that @xmath283 is disjoint from @xmath3 .
thus the set @xmath284 is open in @xmath66 . therefore the set @xmath3 is closed .
since the closed set @xmath3 is contained in @xmath285 , which is a compact set due to proposition [ graph4 ] , the set @xmath3 is compact as well .
an action of the group @xmath1 is continuous whenever the generators act continuously . to prove that the transformations @xmath286 , @xmath197 are continuous , we are going to show that @xmath287 for any graphs @xmath288 and any generator @xmath197 . if the graphs @xmath17 and @xmath30 are isomorphic , then the graphs @xmath289 and @xmath290 are also isomorphic so that @xmath291
. otherwise @xmath292 for some nonnegative integer @xmath63 .
since the distance between any graphs in @xmath66 never exceeds @xmath47 , it is enough to consider the case @xmath293 .
let @xmath60 denote the marked vertex of @xmath17 and @xmath61 denote the marked vertex of @xmath30 .
by definition of the distance function , the closed balls @xmath294 and @xmath295 are isomorphic .
consider an isomorphism @xmath39 of these graphs .
clearly , @xmath296 .
let @xmath85 denote the marked vertex of the graph @xmath289 and @xmath297 denote the marked vertex of @xmath290 .
then @xmath85 is the end of the edge with beginning @xmath60 and label @xmath228 in the graph @xmath17 .
similarly , @xmath297 is the end of the edge with beginning @xmath61 and label @xmath228 in @xmath30 .
it follows that @xmath298 .
since the vertex @xmath85 is joined to @xmath60 by an edge , the closed ball @xmath299 is a subgraph of @xmath294 .
note that @xmath299 remains the closed ball with the same center and radius in the graph @xmath294 .
similarly , @xmath300 is a subgraph of @xmath301 and it is also the closed ball of radius @xmath302 centered at @xmath297 in the graph @xmath301 . since @xmath298 and any isomorphism of graphs preserves distance between vertices , the restriction @xmath303 of @xmath39 to the vertex set of @xmath299 is an isomorphisms of the graphs @xmath299 and @xmath300 .
it remains to notice that the closed ball @xmath304 differs from @xmath299 in that the marked vertex is @xmath85 and , similarly , @xmath305 differs from @xmath300 in that the marked vertex is @xmath297
. therefore @xmath303 is also an isomorphism of @xmath304 and @xmath305 . by definition of the distance function , @xmath306 .
let @xmath180 be an action of the group @xmath1 on a set @xmath178 . to any point @xmath189
we associate three subgroups of @xmath1 : the stabilizer @xmath210 of @xmath207 , the stabilizer @xmath307 of the marked schreier graph @xmath308 , and the neighborhood stabilizer @xmath309 ( if the action @xmath180 is continuous then there is the fourth subgroup , the neighborhood stabilizer of @xmath207 ) . clearly , the graph @xmath310 coincides with @xmath311 if and only if @xmath211 .
however this does not imply that the stabilizer of the graph is the same as the stabilizer of @xmath207 .
since @xmath269 is an action on isomorphism classes of graphs , we have @xmath312 if and only if the graph @xmath310 is isomorphic to @xmath311 .
[ sch3 ] ( i ) @xmath210 is a normal subgroup of @xmath313 .
\(ii ) the quotient of @xmath313 by @xmath210 is isomorphic to the group of all automorphisms of the unmarked graph @xmath265 .
\(iii ) @xmath210 is a subgroup of @xmath314 . without loss of generality
we can assume that the action @xmath180 is transitive .
for brevity , let @xmath315 denote the marked graph @xmath267 , @xmath17 denote the unmarked graph @xmath265 , and @xmath316 denote the group of all automorphisms of @xmath17 .
consider an arbitrary @xmath317 .
for any vertex @xmath318 and any label @xmath197 the unique edge of @xmath17 with beginning @xmath272 and label @xmath228 has end @xmath319 .
it follows that @xmath320 .
since the action @xmath180 is transitive , the automorphism @xmath39 commutes with transformations @xmath198 , @xmath197
. then @xmath39 commutes with @xmath183 for all @xmath182 .
notice that the automorphism @xmath39 is uniquely determined by the vertex @xmath321 .
indeed , any vertex @xmath272 of @xmath17 is represented as @xmath204 for some @xmath182 , then @xmath322 . in particular
, @xmath39 is the identity if @xmath323 . to prove the statements ( i ) and ( ii )
, we are going to construct a homomorphism @xmath324 of the stabilizer @xmath325 onto the group @xmath316 with kernel @xmath210 .
an element @xmath182 belongs to @xmath325 if the graph @xmath315 is isomorphic to @xmath275 .
an isomorphism of these marked graphs is an automorphism of the unmarked graph @xmath17 that sends @xmath207 to @xmath204 .
hence @xmath326 if and only if @xmath327 for some @xmath328 . by the above the automorphism @xmath329
is uniquely determined by @xmath204 .
now we define a mapping @xmath330 by @xmath331 .
it is easy to observe that @xmath324 maps @xmath325 onto @xmath316 and the preimage of the identity under @xmath324 is @xmath210 .
further , for any @xmath332 we have @xmath333 .
recall that the automorphism @xmath334 commutes with the action @xmath180 , in particular , @xmath335 .
then @xmath336 , which implies that @xmath337 .
thus @xmath324 is a homomorphism .
we proceed to the statement ( iii ) .
take any element @xmath338 .
it can be represented as a product @xmath234 , where each @xmath339 is in @xmath2 .
let @xmath340 denote the unique directed path in @xmath315 with beginning @xmath207 and code word @xmath235 . by construction ,
the end of the path @xmath340 is @xmath204 so that the path is closed . let @xmath341 denote the subgraph of @xmath315 whose vertex set consists of all vertices of the path @xmath340 . clearly , @xmath341 is a marked graph , finite and connected . hence @xmath342 .
any graph @xmath343 admits a closed directed path with beginning at the marked point and code word @xmath235 . if @xmath344 , this implies that @xmath345 .
hence @xmath346 .
thus the transformation @xmath347 fixes the set @xmath348 , which is an open neighborhood of the graph @xmath315 in @xmath3 .
any group @xmath1 acts naturally on itself by left multiplication .
the action @xmath349 , called _ adjoint _ , is given by @xmath350 .
the schreier graph of this action relative to any generating set @xmath2 is the _ cayley graph _ of the group @xmath1 relative to @xmath2 .
given a subgroup @xmath351 of @xmath1 , the adjoint action of the group @xmath1 descends to an action on @xmath352 .
the action @xmath353 is given by @xmath354 .
the schreier graph of the latter action relative to a generating set @xmath2 is denoted @xmath355 .
it is called a _
schreier coset graph_. the _ marked schreier coset graph _
@xmath356 is the marked schreier graph of the coset @xmath351 under the action @xmath357 .
it is obtained from @xmath355 by marking the vertex @xmath351 .
[ sch4 ] a marked schreier graph @xmath267 is isomorphic to a marked schreier coset graph @xmath356 if and only if @xmath358 .
let @xmath359 denote the stabilizer @xmath210 .
suppose @xmath360 for some @xmath361 .
then @xmath362 so that @xmath363
. hence @xmath364 and @xmath365 .
conversely , if @xmath365 then @xmath366 for some @xmath367 .
it follows that @xmath368 .
let us define a mapping @xmath369 by @xmath370 . by the above @xmath39
is well defined and one - to - one .
clearly , it maps @xmath371 onto the entire orbit @xmath203 . for any @xmath372 we have @xmath373 .
therefore @xmath39 is a conjugacy of the action @xmath374 with the restriction of the action @xmath180 to the orbit @xmath203 .
it follows that @xmath39 is also an isomorphism of the unmarked graphs @xmath375 and @xmath265 . as @xmath376 ,
the mapping @xmath39 is an isomorphism of the marked graphs @xmath377 and @xmath267 as well . since any isomorphism of schreier graphs of the group @xmath1 is also a conjugacy of the corresponding actions , it preserves stabilizers of vertices .
in particular , marked schreier graphs can not be isomorphic if the stabilizers of their marked vertices do not coincide .
for any subgroup @xmath351 of @xmath1 the stabilizer of the coset @xmath351 under the action @xmath357 is @xmath351 itself .
therefore the graph @xmath356 is not isomorphic to @xmath267 if @xmath378 .
let @xmath1 be a discrete countable group .
denote by @xmath379 the set of all subgroups of @xmath1 .
we endow the set @xmath379 with a topology as follows .
first we consider the product topology on @xmath380 .
the set @xmath380 is in a one - to - one correspondence with the set of all functions @xmath381 .
also , any subset @xmath382 ( in particular , any subgroup ) is assigned the indicator function @xmath383 defined by @xmath384 0 \mbox { if } g\notin h. \end{array } \right.\ ] ] this gives rise to a mapping @xmath385 , which is an embedding .
now the topology on @xmath379 is the smallest topology such that the embedding @xmath386 is continuous . by definition ,
the base of this topology consists of sets of the form @xmath387 where @xmath388 and @xmath389 run independently over all finite subsets of @xmath1 .
notice that @xmath390 .
the topological space @xmath379 is ultrametric and compact ( since @xmath380 is ultrametric and compact , and @xmath391 is closed in @xmath380 ) .
suppose @xmath392 is a complete list of elements of the group @xmath1 .
for any subgroups @xmath393 let @xmath394 if @xmath395 ; otherwise let @xmath396 , where @xmath63 is the smallest index such that @xmath397 belongs to the symmetric difference of @xmath398 and @xmath399 . then @xmath400 is a distance function on @xmath379 compatible with the topology .
note that the above construction also applies to a finite group @xmath1 , in which case @xmath379 is a finite set with the discrete topology .
the following three lemmas explore properties of the topological space @xmath379 .
[ sub1 ] the intersection of subgroups is a continuous operation on the space @xmath379 .
we have to show that the mapping @xmath401 defined by @xmath402 is continuous .
take any finite sets @xmath403 .
given subgroups @xmath393 , the intersection @xmath404 is an element of the set @xmath405 if and only if @xmath406 and @xmath407 for some sets @xmath408 and @xmath409 such that @xmath410 . clearly , the sets @xmath408 and @xmath409 are finite .
it follows that @xmath411 it remains to notice that any open subset of @xmath379 is a union of sets of the form @xmath405 while any set of the form @xmath412 is open in @xmath413 .
[ sub2 ] for any subgroups @xmath398 and @xmath399 of the group @xmath1 , let @xmath414 denote the subgroup generated by all elements of @xmath398 and @xmath399 . then @xmath415 is a borel measurable operation on @xmath379 .
we have to show that the mapping @xmath416 defined by @xmath417 is borel measurable .
take any @xmath182 and consider arbitrary subgroups @xmath418 such that @xmath419 , i.e. , @xmath414 contains @xmath184 .
the element @xmath184 can be represented as a product @xmath420 , where each @xmath421 belongs to @xmath398 or @xmath399 .
let @xmath408 denote the set of all elements of @xmath398 in the sequence @xmath422 and @xmath409 denote the set of all elements of @xmath399 in the same sequence .
then the element @xmath184 belongs to @xmath423 for any subgroups @xmath424 and @xmath425 .
hence the pair @xmath426 is contained in the preimage of @xmath427 under the mapping @xmath428 along with its open neighborhood @xmath429 .
thus the preimage @xmath430 is an open set .
since the set @xmath431 is the complement of @xmath427 , its preimage under @xmath428 is closed .
given finite sets @xmath403 , the set @xmath405 is the intersection of sets @xmath427 , @xmath432 and @xmath433 , @xmath434 . by the above
@xmath435 is a borel set , the intersection of an open set with a closed one .
finally , any open subset of @xmath379 is the union of some sets @xmath405 .
moreover , it is a finite or countable union since there are only countably many sets of the form @xmath405 .
it follows that the preimage under @xmath428 of any open set is a borel set .
[ sub3 ] suppose @xmath351 is a subgroup of @xmath1 .
then @xmath436 is a closed subset of @xmath379 .
moreover , the intrinsic topology on @xmath436 coincides with the topology induced by @xmath379 .
the intrinsic topology on @xmath436 is generated by all sets of the form @xmath437 , where @xmath438 and @xmath439 are finite subsets of @xmath351 .
the topology induced by @xmath379 is generated by all sets of the form @xmath440 , where @xmath388 and @xmath389 are finite subsets of @xmath1 . clearly , @xmath441
if @xmath442 and @xmath443 otherwise . it follows that the two topologies coincide . for any @xmath182 the open set @xmath431
is also closed in @xmath379 as it is the complement of another open set @xmath427 .
then the set @xmath436 is closed in @xmath379 since it is the intersection of closed sets @xmath431 over all @xmath444 .
let @xmath180 be an action of the group @xmath1 on a set @xmath178 .
let us consider the stabilizer @xmath210 of a point @xmath189 under the action ( see section [ act ] ) as the value of a mapping @xmath445 .
[ sub4 ] suppose @xmath180 is a continuous action of the group @xmath1 on a hausdorff topological space @xmath178 .
then * the mapping @xmath446 is borel measurable ; * @xmath446 is continuous at a point @xmath189 if and only if the stabilizer of @xmath207 under the action coincides with its neighborhood stabilizer : @xmath447 ; * if a sequence of points in @xmath178 converges to the point @xmath207 and the sequence of their stabilizers converges to a subgroup @xmath351 , then @xmath448 . for any @xmath182 let @xmath449 denote the set of all points in @xmath178 fixed by the transformation @xmath183
let us show that @xmath449 is a closed set .
take any point @xmath189 not in @xmath449 .
since the points @xmath207 and @xmath204 are distinct , they have disjoint open neighborhoods @xmath450 and @xmath451 , respectively . since @xmath183 is continuous , there exists an open neighborhood @xmath452 of @xmath207 such that @xmath453
. then @xmath454 is an open neighborhood of @xmath207 and @xmath455 is disjoint from @xmath454 . in particular
, @xmath454 is disjoint from @xmath449 . for any finite sets
@xmath403 the preimage of the open set @xmath405 under the mapping @xmath446 is @xmath456 this is a borel set as @xmath449 is closed for any @xmath182 .
since sets of the form @xmath457 constitute a base of the topology on @xmath379 , the mapping @xmath446 is borel measurable .
the mapping @xmath446 is continuous at a point @xmath189 if and only if @xmath207 is an interior point in the preimage under @xmath446 of any set @xmath405 containing @xmath210 .
the latter holds true if and only if @xmath207 is an interior point in any set @xmath449 containing this point .
clearly , @xmath207 is an interior point of @xmath449 if and only if @xmath184 belongs to the neighborhood stabilizer @xmath458 .
thus @xmath446 is continuous at @xmath207 if and only if any element of @xmath210 belongs to @xmath458 as well .
now suppose that a sequence @xmath459 of points in @xmath178 converges to the point @xmath207 and , moreover , the stabilizers @xmath460 converge to a subgroup @xmath351 .
consider an arbitrary @xmath182 . in the case
@xmath461 , the subgroup @xmath351 belongs to the open set @xmath427 .
since @xmath462 as @xmath463 , we have @xmath464 for large @xmath63 .
in other words , @xmath465 for large @xmath63 . since the set @xmath449 is closed , it contains the limit point @xmath207 as well .
that is , @xmath338 . in the case
@xmath466 , the subgroup @xmath351 belongs to the open set @xmath431 .
then @xmath467 for large @xmath63 .
in other words , @xmath468 for large @xmath63 . since @xmath469 as @xmath463 , the action of @xmath184 fixes no neighborhood of @xmath207 .
that is , @xmath470 .
the group @xmath1 acts naturally on the set @xmath379 by conjugation .
the action @xmath471 is given by @xmath472 .
this action is continuous .
indeed , one easily observes that @xmath473 for all @xmath182 and finite sets @xmath474 .
[ sub5 ] the action @xmath475 of the group @xmath1 on @xmath379 is continuously conjugated to the action @xmath269 on the space @xmath3 of the marked schreier graphs of @xmath1 relative to a generating set @xmath2 .
moreover , the mapping @xmath476 given by @xmath477 is a continuous conjugacy .
proposition [ sch4 ] implies that the mapping @xmath39 is bijective .
consider arbitrary element @xmath184 and subgroup @xmath351 of the group @xmath1 .
the stabilizer of the coset @xmath478 under the action @xmath357 consists of those @xmath479 for which @xmath480 .
the latter condition is equivalent to @xmath481 .
therefore the stabilizer is @xmath482 .
as @xmath483 , it follows from proposition [ sch4 ] that @xmath484 .
thus @xmath39 conjugates the action @xmath475 with @xmath269 .
now we are going to show that for any finite sets @xmath403 the image of the open set @xmath405 under the mapping @xmath39 is open in @xmath3 .
let @xmath267 be an arbitrary graph in that image .
any element @xmath182 can be represented as a product @xmath485 , where @xmath486 .
let us fix such a representation and denote by @xmath487 the unique directed path in @xmath267 with beginning @xmath207 and code word @xmath235 .
then the end of the path @xmath487 is @xmath204 . in particular ,
the path @xmath487 is closed if and only if @xmath338 . by proposition [ sch4 ] , the preimage of the graph @xmath267 under @xmath39 is @xmath210 . since @xmath488 ,
the path @xmath487 is closed for @xmath432 and not closed for @xmath489 .
let @xmath68 denote the smallest subgraph of @xmath267 containing all paths @xmath487 , @xmath490 .
clearly , @xmath68 is a marked graph , finite and connected .
hence @xmath282 .
for any marked schreier graph @xmath271 in @xmath71 , the directed path with beginning @xmath272 and the same code word as in @xmath487 is closed for all @xmath432 and not closed for all @xmath489 .
it follows that @xmath491 .
therefore the graph @xmath267 is contained in @xmath492 along with its neighborhood @xmath493 .
any open set in @xmath379 is the union of some sets @xmath405 .
hence it follows from the above that the mapping @xmath39 maps open sets onto open sets . in other words ,
the inverse mapping @xmath222 is continuous . since the topological spaces @xmath379 and @xmath3 are compact , @xmath39 is continuous as well .
proposition [ sub5 ] allows for a short ( although not constructive ) proof of the following statement .
[ sub6 ] any subgroup of finite index of a finitely generated group is also finitely generated .
suppose @xmath1 is a finitely generated group and @xmath351 is a subgroup of @xmath1 of finite index .
let @xmath2 be a finite symmetric generating set for @xmath1 .
by proposition [ sub5 ] , the space @xmath379 of subgroups of @xmath1 is homeomorphic to the space @xmath3 of marked schreier graphs of @xmath1 relative to the generating set @xmath2 .
moreover , there is a homeomorphism that maps the subgroup @xmath351 to the marked schreier coset graph @xmath356 .
the vertices of the graph are cosets of @xmath351 in @xmath1 .
since @xmath351 has finite index in @xmath1 , the graph @xmath356 is finite .
notice that any finite graph in the topological space @xmath66 , which contains @xmath3 , is an isolated point .
it follows that @xmath351 is an isolated point in @xmath379 .
then there exist finite sets @xmath403 such that @xmath351 is the only element of the open set @xmath405 .
let @xmath359 be the subgroup of @xmath1 generated by the finite set @xmath388 .
since @xmath442 and @xmath494 , the subgroup @xmath359 is disjoint from @xmath389 .
thus @xmath495 so that @xmath496 .
consider an arbitrary graph @xmath17 .
let @xmath340 be a path in this graph , @xmath497 be consecutive vertices of @xmath340 , and @xmath498 be consecutive edges .
backtracking _ in the path @xmath340 occurs if @xmath499 for some @xmath500 ( then @xmath501 ) .
the graph @xmath17 is called a _ tree _ if it is connected and admits no closed path of positive length without backtracking . in particular , this means no loops and no multiple edges . a _ rooted tree _ is a tree with a distinguished vertex called the _
root_. clearly , the root is a synonym for the marked vertex . for any integer @xmath62
the _ level _ @xmath63 ( or the @xmath63th level ) of the tree is defined as the set of vertices at distance @xmath63 from the root . if @xmath502 then any vertex on the @xmath63th level is joined to exactly one vertex on the level @xmath128 and , optionally , to some vertices on the level @xmath503 .
the rooted tree is called _ @xmath46-regular _ if every vertex on any level @xmath63 is joined to exactly @xmath46 vertices on level @xmath503 .
the @xmath504-regular rooted tree is also called _
binary_. all @xmath46-regular rooted trees are isomorphic to each other .
a standard model of such a tree is built as follows .
let @xmath450 be a set of cardinality @xmath46 referred to as the _ alphabet _ ( usually @xmath505 ) .
a _ word _ ( or _ finite word _ ) in the alphabet @xmath450 is a finite string of elements from @xmath450 ( referred to as _ letters _ ) .
the set of all words in the alphabet @xmath450 is denoted @xmath506 .
@xmath506 is a monoid with respect to the concatenation ( the unit element is the empty word , denoted @xmath507 ) .
moreover , it is the free monoid generated by elements of @xmath450 .
now we define a plain graph @xmath508 with the vertex set @xmath506 in which two vertices @xmath509 and @xmath510 are joined by an edge if @xmath511 or @xmath512 for some @xmath513 .
then @xmath508 is a @xmath46-regular rooted tree with the root @xmath507 .
the @xmath63th level of the tree @xmath508 consists of all words of length @xmath63 .
a bijection @xmath514 is an automorphism of the rooted tree @xmath508 if and only if it preserves the length of any word and the length of the common beginning of any two words .
given an automorphism @xmath39 and a word @xmath515 , there exists a unique transformation @xmath516 such that @xmath517 for all @xmath518 .
it is easy to see that @xmath519 is also an automorphism of the tree @xmath508 .
this automorphism is called the _ section _ of @xmath39 at the word @xmath520 and denoted @xmath521 .
a set of automorphisms of the tree @xmath508 is called _ self - similar _ if it is closed under taking sections . for any automorphisms @xmath39 and @xmath519 and
any word @xmath515 one has @xmath522 and @xmath523 .
it follows that any group of automorphisms generated by a self - similar set is itself self - similar .
suppose @xmath1 is a group of automorphisms of the tree @xmath508 .
let @xmath524 denote the natural action of @xmath1 on the vertex set @xmath506 .
given a word @xmath515 , the section mapping @xmath525 is a homomorphism when restricted to the stabilizer @xmath526 .
if @xmath1 is self - similar then this is a homomorphism to @xmath1 .
the self - similar group @xmath1 is called _ self - replicating _ if for any @xmath515 the mapping @xmath525 maps the subgroup @xmath526 onto the entire group @xmath1 .
suppose that letters of the alphabet @xmath450 are canonically ordered : @xmath527 .
for any permutation @xmath528 on @xmath450 and automorphisms @xmath422 of the tree @xmath508 we denote by @xmath529 a transformation @xmath514 given by @xmath530 for all @xmath518 and @xmath238 .
it is easy to observe that @xmath39 is also an automorphism of @xmath508 and @xmath531 for @xmath238 .
the expression @xmath529 is called the _ wreath recursion _ for @xmath39 .
any self - similar set of automorphisms @xmath532 , @xmath533 satisfies a system of `` self - similar '' wreath recursions @xmath534 where @xmath535 , @xmath533 are permutations on @xmath450 and @xmath536 maps @xmath537 to @xmath428 .
[ tree0 ] any system of self - similar wreath recursions over the alphabet @xmath450 is satisfied by a unique self - similar set of automorphisms of the regular rooted tree @xmath508 .
consider a system of wreath recursions @xmath538 , @xmath533 , where @xmath535 , @xmath533 are permutations on @xmath450 and @xmath536 is a mapping of @xmath537 to @xmath428 .
we define transformations @xmath539 , @xmath533 of the set @xmath506 inductively as follows .
first @xmath540 for all @xmath533 .
then , once the transformations are defined on words of a particular length @xmath62 , we let @xmath541 for all @xmath533 , @xmath238 , and words @xmath244 of length @xmath63 . by definition , each @xmath539 preserves the length of words . besides
, it follows by induction on @xmath63 that @xmath539 is bijective when restricted to words of length @xmath63 and that @xmath539 preserves having a common beginning of length @xmath63 for any two words .
therefore each @xmath539 is an automorphism of the tree @xmath508 . by construction ,
the automorphisms @xmath539 , @xmath533 form a self - similar set satisfying the above system of wreath recursions .
moreover , they provide the only solution to that system . an _ infinite path _ in the tree @xmath508 is an infinite sequence of vertices @xmath542 together with a sequence of edges @xmath543 such that the endpoints of any @xmath51 are @xmath52 and @xmath53 .
the vertex @xmath54 is the beginning of the path .
clearly , the path is uniquely determined by the sequence of vertices alone .
the _ boundary _ of the rooted tree @xmath508 , denoted @xmath16 , is the set of all infinite paths without backtracking that begin at the root .
there is a natural one - to - one correspondence between @xmath16 and the set @xmath544 of infinite words over the alphabet @xmath450 .
namely , an infinite word @xmath545 corresponds to the path going through the vertices @xmath546 .
the set @xmath544 is equipped with the product topology and the uniform bernoulli measure .
this allows us to regard the tree boundary @xmath16 as a compact topological space with a borel probability measure ( called _ uniform _ ) .
suppose @xmath1 is a group of automorphisms of the regular rooted tree @xmath508 .
the natural action of @xmath1 on the vertex set @xmath506 gives rise to an action on the boundary @xmath16 .
the latter is continuous and preserves the uniform measure on @xmath16 .
[ tree1 ] let @xmath1 be a countable group of automorphisms of a regular rooted tree @xmath508 .
then the following conditions are equivalent : * the group @xmath1 acts transitively on each level of the tree ; * the action of @xmath1 on the boundary @xmath16 of the tree is topologically transitive ; * the action of @xmath1 on @xmath16 is minimal ; * the action of @xmath1 on @xmath16 is ergodic with respect to the uniform measure ; * the action of @xmath1 on @xmath16 is uniquely ergodic . let @xmath1 be a countable group of automorphisms of a regular rooted tree @xmath508 .
let @xmath524 denote the natural action of @xmath1 on the vertex set of the tree @xmath508 and @xmath547 denote the induced action of @xmath1 on the boundary @xmath16 of the tree .
[ tree2 ] the mapping @xmath548 is continuous on a residual ( dense @xmath549 ) set . for any @xmath182 let @xmath550 denote the set of all points in @xmath16 fixed by the transformation @xmath551 .
if @xmath552 but @xmath553 , then @xmath554 is a boundary point of the set @xmath550 , and vice versa . since @xmath550 is a closed set , its boundary is a closed , nowhere dense set .
it follows that the set of points @xmath555 such that @xmath556 is the intersection of countably many dense open sets ( it is dense since @xmath16 is a complete metric space ) . by lemma [ sub4 ] ,
the latter set consists of points at which the mapping @xmath548 is continuous .
the mapping @xmath548 is borel due to lemma [ sub4 ] .
if @xmath548 is injective then , according to the descriptive set theory , it also maps borel sets onto borel sets ( see , e.g. , @xcite ) .
the following two lemmas show the same can hold under a little weaker condition .
[ tree3 ] assume that for any points @xmath557 either @xmath558 or @xmath559 is not contained in @xmath560 .
then the mapping @xmath548 maps any open set , any closed set , and any intersection of an open set with a closed one onto borel sets .
first let us show that @xmath548 maps any closed subset @xmath561 of the boundary @xmath16 onto a borel subset of @xmath379 .
for any positive integer @xmath63 let @xmath562 denote the set of all words of length @xmath63 in the alphabet @xmath450 that are beginnings of infinite words in @xmath561 .
further , let @xmath563 be the union of sets @xmath564 over all words @xmath565 .
finally , let @xmath566 be the intersection of the sets @xmath563 over all @xmath502 .
by lemma [ sub3 ] , the set @xmath436 is closed in @xmath379 for any subgroup @xmath567 .
hence each @xmath563 is closed as the union of finitely many closed sets .
then the intersection @xmath566 is closed as well . the stabilizer @xmath559 of an infinite word @xmath555 is a subgroup of the stabilizer @xmath568 of a finite word @xmath518 whenever @xmath244 is a beginning of @xmath554 .
it follows that @xmath569 for all @xmath570 . by construction of the set @xmath566 , any subgroup of an element of @xmath566 is also an element of @xmath566 .
hence @xmath566 contains all subgroups of the groups @xmath559 , @xmath571 .
conversely , for any subgroup @xmath572 there is a sequence of words @xmath573 such that @xmath574 and @xmath575 for @xmath576 . since the number of words of a fixed length is finite , one can find nested infinite sets of indices @xmath577 such that the beginning of length @xmath46 of the word @xmath578 is the same for all @xmath166 .
choose an increasing sequence of indices @xmath579 such that @xmath168 for all @xmath46 , and let @xmath580 be the beginning of length @xmath46 of the word @xmath581 .
then @xmath582 as @xmath583 .
besides , @xmath584 , in particular , the group @xmath351 is a subgroup of @xmath585 . by construction ,
the word @xmath580 is a beginning of @xmath586 whenever @xmath175
. therefore all @xmath580 are beginnings of the same infinite word @xmath587 .
since every beginning of @xmath588 coincides with a beginning of some infinite word in @xmath561 and the set @xmath561 is closed , it follows that @xmath589 .
the stabilizer @xmath590 is the intersection of stabilizers @xmath585 over all @xmath591 .
hence @xmath351 is a subgroup of @xmath590
. by the above a subgroup @xmath351 of @xmath1 belongs to the set @xmath566 if and only if it is a subgroup of the stabilizer @xmath559 for some @xmath571 .
the assumption of the lemma implies that stabilizers @xmath559 , @xmath571 can be distinguished as the maximal subgroups in the set @xmath566 .
that is , such a stabilizer is an element of @xmath566 which is not a proper subgroup of another element of @xmath566 . for any @xmath182
we define a transformation @xmath329 of @xmath379 by @xmath592 , where @xmath593 is a cyclic subgroup of @xmath1 generated by @xmath184 .
the group @xmath594 is generated by @xmath184 and all elements of the group @xmath351 .
clearly , a subgroup @xmath572 is not maximal in @xmath566 if and only if @xmath595 for some @xmath466 .
an equivalent condition is that @xmath351 belongs to the set @xmath596 .
it follows from lemma [ sub2 ] that the mapping @xmath329 is borel measurable .
therefore @xmath597 is a borel set . now
the image of the set @xmath561 under the mapping @xmath548 is the difference of the closed set @xmath566 and the union of borel sets @xmath597 , @xmath182 .
hence this image is a borel set .
any open set @xmath598 is the union of a finite or countable collection of cylinders @xmath599 , which are both open and closed sets . by the above each cylinder
is mapped by @xmath548 onto a borel set in @xmath379 .
then the union @xmath600 is mapped onto the union of images of the cylinders , which is a borel set as well .
further , for any closed set
@xmath601 the intersection @xmath602 is the union of closed sets @xmath603 .
hence it is also mapped by @xmath548 onto a borel set .
[ tree4 ] under the assumption of lemma [ tree3 ] , if the mapping @xmath548 is finite - to - one , i.e. , the preimage of any subgroup in @xmath379 is finite , then it maps borel sets onto borel sets .
recall that the class @xmath604 of the borel sets in @xmath16 is the smallest collection of subsets of @xmath16 that contains all closed sets and is closed under taking countable intersections , countable unions , and complements .
let @xmath605 denote the smallest collection of subsets of @xmath16 that contains all closed sets and is closed under taking countable intersections of nested sets and countable unions of any sets .
note that @xmath605 is well defined ; it is the intersection of all collections satisfying these conditions .
in particular , @xmath606 .
further , let @xmath607 denote the collection of all borel sets in @xmath16 mapped onto borel sets in @xmath379 by the mapping @xmath548 . for any mapping @xmath608 and any sequence @xmath609 of subsets of @xmath16
the image of the union @xmath610 under @xmath39 is the union of images @xmath611 . on the other hand ,
the image of the intersection @xmath612 under @xmath39 is contained in @xmath613 but need not coincide with the latter when the mapping @xmath39 is not one - to - one .
the two sets do coincide if @xmath39 is finite - to - one and @xmath614 .
since the mapping @xmath548 is assumed to be finite - to - one , it follows that the collection @xmath607 is closed under taking countable intersections of nested sets and countable unions of any sets . by lemma [ tree3 ] , @xmath607 contains all closed sets
. therefore @xmath615 . to complete the proof
, we are going to show that @xmath616 , which will imply that @xmath617 .
given a set @xmath618 , let @xmath619 denote the collection of all sets @xmath620 such that the intersection @xmath621 also belongs to @xmath605 . for any sequence @xmath609 of elements of @xmath619
we have @xmath622 besides , the sets @xmath623 are nested whenever the sets @xmath609 are nested .
it follows that the class @xmath619 is closed under taking countable intersections of nested sets and countable unions of any sets .
consequently , @xmath624 whenever @xmath619 contains all closed sets .
the latter condition obviously holds if the set @xmath451 is itself closed .
notice that for any sets @xmath625 we have @xmath626 if and only if @xmath627 .
since @xmath624 for any closed set @xmath451 , it follows that @xmath628 contains all closed sets for any @xmath629 .
then @xmath630 for any @xmath629 .
in other words , the class @xmath605 is closed under taking finite intersections . combining finite intersections with countable intersections of nested sets
, we can obtain any countable intersection of sets from @xmath605 .
namely , if @xmath609 are arbitrary elements of @xmath605 , then their intersection coincides with the intersection of sets @xmath631 , @xmath576 , which are nested : @xmath632 .
therefore @xmath605 is closed under taking any countable intersections .
let @xmath633 be the collection of complements in @xmath16 of all sets from @xmath605 .
for any subsets @xmath609 of @xmath16 the complement of their union is the intersection of their complements @xmath634 while the complement of their intersection is the union of their complements . since the class @xmath605 is closed under taking countable intersections and countable unions ,
so is @xmath633 .
further , any open subset of @xmath16 is the union of at most countably many cylinders , which are closed ( as well as open ) sets .
therefore @xmath605 contains all open sets .
then @xmath633 contains all closed sets .
now it follows that @xmath635 . in other words ,
the class @xmath605 is closed under taking complements .
thus the collection @xmath605 is closed under taking any countable intersections and complements .
this implies that @xmath616 .
let @xmath180 be a continuous action of a countable group @xmath1 on a compact metric space @xmath178 .
let @xmath13 denote the image of @xmath178 under the mapping @xmath446 .
[ tree5 ] assume that for any distinct points @xmath636 the neighborhood stabilizer @xmath212 is not contained in the stabilizer @xmath637 .
then the inverse of @xmath446 , defined on the set @xmath13 , can be extended to a continuous mapping of the closure of @xmath13 onto @xmath178 .
since @xmath212 is a subgroup of @xmath210 for any @xmath189 , the assumption of the lemma implies that the mapping @xmath446 is one - to - one so that the inverse is well defined on @xmath13 .
to prove that the inverse can be extended to a continuous mapping of the closure of @xmath13 onto @xmath178 , it is enough to show that any sequence @xmath459 of points in @xmath178 is convergent whenever the sequence of stabilizers @xmath638 converges in @xmath379 .
suppose that @xmath462 as @xmath463
. since @xmath178 is a compact metric space , the sequence @xmath459 has at least one limit point . by lemma [ sub4](iii )
, any limit point @xmath207 satisfies @xmath639 .
in particular , @xmath640 for any limit points @xmath207 and @xmath272 . then @xmath641 due to the assumption of the lemma .
it follows that the sequence @xmath459 is convergent .
let @xmath642 be the binary alphabet , @xmath506 be the set of finite words over @xmath450 regarded as the vertex set of a binary rooted tree @xmath508 , and @xmath544 be the set of infinite words over @xmath450 regarded as the boundary @xmath16 of the tree @xmath508 .
we define the grigorchuk group @xmath0 as a self - similar group of automorphisms of the tree @xmath508 ( for alternative definitions , see @xcite ) .
the group is generated by four automorphisms @xmath643 that , together with the trivial automorphism , form a self - similar set .
consider the following system of wreath recursions : @xmath644 by lemma [ tree0 ] , this system uniquely defines a self - similar set of automorphisms of the tree @xmath508 .
the automorphism @xmath21 is clearly the identity ( e.g. , by lemma [ tree0 ] ) .
it is the unity of the group @xmath0 .
we shall denote the unity by @xmath645 to avoid confusion with a letter of the alphabet @xmath450 .
the set @xmath646 shall be considered the standard set of generators for the group @xmath0 .
all @xmath647 generators of the grigorchuk group are involutions . indeed , the transformations @xmath648 form a self - similar set satisfying wreath recursions @xmath649 , @xmath650 , @xmath651 , @xmath652 , and @xmath653
. then lemma [ tree0 ] implies that @xmath654 .
this fact allows us to regard the schreier graphs of the group @xmath0 relative to the generating set @xmath2 as graphs with undirected edges ( as explained in section [ graph ] ) . since @xmath655 , the automorphisms @xmath656 , @xmath657 , @xmath658 , and @xmath645 form a self - similar set satisfying wreath recursions @xmath659 , @xmath660 , @xmath661 , and @xmath653 .
lemma [ tree0 ] implies that @xmath662 .
then @xmath663 .
it follows that @xmath664 is a subgroup of @xmath0 isomorphic to the klein @xmath647-group .
we denote by @xmath524 the generic action of the group @xmath0 on vertices of the binary rooted tree @xmath508 .
the induced action on the boundary @xmath16 of the tree is denoted @xmath547 . for brevity ,
we write @xmath665 instead of @xmath666 .
the action of the generator @xmath667 is very simple : it changes the first letter in every finite or infinite word while keeping the other letters intact .
in particular , the empty word is the only word fixed by @xmath667 . to describe the action of the other generators , we need three observations .
first of all , @xmath668 , @xmath669 , and @xmath400 fix one - letter words . secondly , any word beginning with @xmath670 is fixed by @xmath400 while @xmath668 and @xmath669 change only the second letter in such a word .
thirdly , the section mapping @xmath671 induces a cyclic permutation on the set @xmath672 .
it follows that a finite or infinite word @xmath244 is simultaneouly fixed by @xmath668 , @xmath669 , and @xmath400 if it contains no zeros or the only zero is the last letter .
otherwise two of the three generators change the letter following the first zero in @xmath244 ( keeping the other letters intact ) while the third generator fixes @xmath244 . in the latter case , it is the position @xmath46 of the first zero in @xmath244 that determines the generator fixing @xmath244 .
namely , @xmath673 if @xmath674 , @xmath675 if @xmath676 , and @xmath677 if @xmath678 .
we have to show that for any word @xmath518 the section mapping @xmath679 maps the stabilizer @xmath568 onto the entire group @xmath0 .
let @xmath566 be the set of all words with this property . clearly , @xmath680 as @xmath681 and @xmath682 for all @xmath683
suppose @xmath684 .
given an arbitrary @xmath683 , there exists @xmath685 such that @xmath686 and @xmath687 .
further , there exists @xmath688 such that @xmath689 and @xmath690 . then @xmath691 and @xmath692 . since @xmath184 is arbitrary , @xmath693 .
that is , the set @xmath566 is closed under concatenation .
any automorphism of the tree @xmath508 either interchanges the vertices @xmath670 and @xmath47 or fixes them both .
hence the stabilizer @xmath694 coincides with @xmath695 .
this stabilizer contains the elements @xmath696 .
the wreath recursions for these elements are @xmath697 , @xmath698 , @xmath699 , @xmath700 , @xmath701 , @xmath702 .
it follows that the images of the group @xmath694 under the section mappings @xmath703 and @xmath671 contain the generating set @xmath2 .
as the restrictions of these mappings to @xmath694 are homomorphisms , both images coincide with @xmath0 . therefore the words @xmath670 and @xmath47 are in the set @xmath566 . by the above @xmath566
is closed under concatenation and contains the empty word .
this implies @xmath704 .
[ grig2 ] the group @xmath0 acts transitively on each level of the binary rooted tree @xmath508 .
any two infinite words in @xmath16 are in the same orbit of the action @xmath547 if and only if they differ in only finitely many letters . for any infinite word @xmath555 and any generator @xmath705 the infinite word @xmath706 differs from @xmath554 in at most one letter .
any @xmath683 can be represented as a product @xmath420 , where each @xmath421 is in @xmath707 .
it follows that for any @xmath555 the infinite words @xmath665 and @xmath554 differ in at most @xmath46 letters .
thus any two infinite words in the same orbit of the action @xmath547 differ in only finitely many letters .
now we are going to show that for any finite words @xmath708 of the same length there exists @xmath683 such that @xmath709 and @xmath710 .
equivalently , @xmath711 for all @xmath555 .
this will complete the proof of the lemma .
indeed , the claim contains the statement that the group @xmath0 acts transitively on each level of the tree @xmath508 .
moreover , it implies that two infinite words in @xmath16 are in the same orbit of the action @xmath547 whenever they differ in a finite number of letters .
we prove the claim by induction on the length @xmath63 of the words @xmath509 and @xmath510 .
the case @xmath712 is trivial . here
@xmath509 and @xmath510 are the empty words so that we take @xmath713 .
now assume that the claim is true for all pairs of words of specific length @xmath62 and consider words @xmath509 and @xmath510 of length @xmath503 .
let @xmath714 be the first letter of @xmath509 and @xmath715 be the first letter of @xmath510
. then @xmath716 and @xmath717 , where @xmath718 and @xmath719 are words of length @xmath63 .
by the inductive assumption , there exists @xmath720 such that @xmath721 for all @xmath555 .
since the group @xmath0 is self - replicating , there exists @xmath722 such that @xmath723 for all @xmath724 .
in particular , @xmath725 for all @xmath555 .
it remains to take @xmath726 if @xmath727 and @xmath728 otherwise .
then @xmath729 for all @xmath555 .
[ grig3 ] suppose @xmath509 and @xmath510 are words in the alphabet @xmath730 such that @xmath509 is not a beginning of @xmath510 while @xmath510 , even with the last two letters deleted , is not a beginning of @xmath509 .
then there exists @xmath683 that does not fix @xmath510 while fixing all words with beginning @xmath509 .
first we consider a special case when @xmath731 . to satisfy the assumption of the lemma
, the word @xmath509 has to begin with @xmath670
. then we can take @xmath732 .
indeed , the transformation @xmath400 fixes all words that begin with @xmath670 , which includes all words with beginning @xmath509 . at the same time , @xmath733 .
next we consider a slightly more general case when @xmath510 is an arbitrary word of length @xmath734 . by lemma [ grig2 ] ,
the group @xmath0 acts transitively on the third level of the tree @xmath508 .
therefore @xmath735 for some @xmath720 .
the words @xmath736 and @xmath737 satisfy the assumption of the lemma since the words @xmath509 and @xmath510 do . by the above , @xmath738 while @xmath739 for all @xmath515 .
let @xmath740 .
then @xmath741 while @xmath742 for all @xmath518 . finally , consider the general case .
let @xmath743 be the longest common beginning of the words @xmath509 and @xmath510 . then @xmath744 and @xmath745 , where the words @xmath718 and @xmath719 also satisfy the assumption of the lemma .
in particular , @xmath718 is nonempty and the length of @xmath719 is at least @xmath734 .
we have @xmath746 , where @xmath747 and the length of @xmath748 is @xmath734 .
since the first letters of the words @xmath718 and @xmath748 are distinct , these words satisfy the assumption of the lemma . by the above there
exists @xmath722 such that @xmath749 and @xmath750 for all @xmath515 .
since the group @xmath0 is self - replicating , there exists @xmath683 such that @xmath751 for all @xmath518 .
then @xmath184 does not fix the word @xmath752 while fixing all words with beginning @xmath509 .
since @xmath752 is a beginning of @xmath510 , the transformation @xmath184 does not fix @xmath510 as well .
let @xmath63 denote the length of the longest common beginning of the distinct infinite words @xmath554 and @xmath754 .
let @xmath509 be the beginning of @xmath554 of length @xmath503 and @xmath510 be the beginning of @xmath754 of length @xmath755 .
it is easy to see that the words @xmath509 and @xmath510 satisfy the assumption of lemma [ grig3 ] .
therefore there exists a transformation @xmath683 that does not fix @xmath510 while fixing all finite words with beginning @xmath509 . clearly , the action of @xmath184 on @xmath16 fixes all infinite words with beginning @xmath509 . as such infinite words
form an open neighborhood of the point @xmath554 , we have @xmath756 . at the same time
, @xmath184 does not fix the infinite word @xmath754 since it does not fix its beginning @xmath510 .
hence @xmath757 so that @xmath758 .
we are going to show that , given an automorphism @xmath683 and an infinite word @xmath555 with infinitely many zeros , one has @xmath759 for a sufficiently long beginning @xmath244 of @xmath554 .
this claim implies the lemma . indeed , in the case @xmath760 the action of @xmath184 fixes all infinite words with beginning @xmath244 , which form an open neighborhood of @xmath554 .
let @xmath316 be the set of all @xmath683 such that the claim holds true for @xmath184 and any @xmath555 with infinitely many zeros .
the set @xmath316 contains the generating set @xmath2 .
indeed , @xmath761 for any nonempty word @xmath518 and @xmath762 for any word @xmath244 that contains a zero which is not the last letter of @xmath244 .
now suppose @xmath763 and consider an arbitrary @xmath555 with infinitely many zeros .
then @xmath764 for a sufficiently long beginning @xmath244 of @xmath554 .
lemma [ grig2 ] implies that the infinite word @xmath706 also has infinitely many zeros .
since @xmath765 is a beginning of @xmath706 and @xmath766 , we have @xmath767 provided @xmath244 is long enough . since @xmath768 ,
we have @xmath769 provided @xmath244 is long enough .
thus @xmath770 .
that is , the set @xmath316 is closed under multiplication . since @xmath771 and all generators are involutions , it follows that @xmath772 .
for any positive integer @xmath63 let @xmath781 denote the infinite word over the alphabet @xmath450 that has a single zero in the position @xmath63 .
the sequence @xmath782 converges to @xmath783 .
one observes that any of the generators @xmath668 , @xmath669 , and @xmath400 fixes @xmath781 only if @xmath63 leaves a specific remainder under division by @xmath734 ( @xmath670 for @xmath668 , @xmath504 for @xmath669 , and @xmath47 for @xmath400 ) .
it follows that @xmath779 .
now let us show that any @xmath784 is contained in the set @xmath785 .
the proof is by strong induction on the length @xmath63 of @xmath184 , which is the smallest possible number of factors in an expansion @xmath786 such that each @xmath486 .
the case @xmath712 is trivial as @xmath645 is the only element of length @xmath670 .
assume that the claim is true for all elements of length less than some @xmath787 and consider an arbitrary element @xmath784 of length @xmath63 .
we have @xmath788 , where each @xmath339 is a generator from @xmath2 .
let @xmath789 , @xmath790 .
if @xmath791 for some @xmath792 , then @xmath793 and @xmath794 both fix @xmath783 . since the length of @xmath795 and @xmath796 is less than @xmath63 , they belong to @xmath785 by the inductive assumption . as @xmath785 is a group , so does @xmath797 . if @xmath798 for all @xmath792 , then @xmath799 for any @xmath800 and sufficiently long beginning @xmath801 of the infinite word @xmath802
it follows that @xmath759 for a sufficiently long beginning @xmath244 of @xmath783 .
thus @xmath803 .
recall that we consider the schreier graphs of the group @xmath0 relative to the generating set @xmath646 as graphs with undirected edges .
the schreier graphs of all orbits of the action @xmath547 except @xmath804 are similar .
any vertex is joined to two other vertices .
moreover , it is joined to one of the neighbors by a single edge labeled @xmath667 and to the other neighbor by two edges . also , there is one loop at each vertex .
hence the schreier graph has a linear structure ( see figure [ fig1 ] ) and all such graphs are isomorphic as graphs with unlabeled edges . the schreier graph of the orbit of @xmath805 is different in that there are three loops labeled @xmath668 , @xmath669 , and @xmath400 at the vertex @xmath783 ( see figure [ fig2 ] ) .
let @xmath806 be the mapping that assigns to any point on the boundary of the binary rooted tree @xmath508 its marked schreier graph under the action @xmath547 .
using notation of section [ sch ] , @xmath807 for all @xmath555 .
let @xmath68 denote the marked graph with a single vertex and three loops labeled @xmath668 , @xmath669 , and @xmath400 . recall that @xmath71 is an open subset of @xmath66 consisting of all graphs in @xmath66 that have a subgraph isomorphic to @xmath68 .
hence @xmath808 is an open subset of @xmath809 .
given @xmath555 , the graph @xmath810 belongs to that open subset if and only if @xmath811 and @xmath812
. the latter conditions are satisfied only for @xmath813 .
the lemma follows .
it turns out that the image @xmath8 is not closed in @xmath809 .
the following construction will help to describe the closure of @xmath8 .
let us take two copies of the schreier graph @xmath814 .
we remove two out of three loops at the vertex @xmath783 ( loops with the same labels in both copies ) and replace them with two edges joining the two copies .
let @xmath815 and @xmath816 denote labels of the removed loops and @xmath817 denote the label of the retained loop .
then @xmath818 is a permutation of @xmath819 . to be rigorous
, the new graph has the vertex set @xmath820 , the set of edges @xmath821 , and the set of labels @xmath2 .
an arbitrary edge @xmath822 has beginning @xmath823 and label @xmath228 .
the end of this edge is @xmath824 unless @xmath813 and @xmath825 or @xmath826 , in which case the end is @xmath827 .
there are three ways to perform the above construction depending on the choice of @xmath817 .
we denote by @xmath828 , @xmath829 , and @xmath830 the graphs obtained when @xmath831 , @xmath832 , and @xmath833 , respectively .
further , for any @xmath834 we denote by @xmath835 a marked graph obtained from @xmath10 by marking the vertex @xmath836 ( see figure [ fig3 ] ) . for any @xmath502
we define a map @xmath843 as follows . given @xmath844 and @xmath845 , let @xmath846 be an infinite word obtained from @xmath781 after replacing the first @xmath847 letters by the first @xmath847 letters of @xmath554 and adding @xmath848 to the @xmath849-th letter .
clearly , @xmath850 .
let @xmath851 be the remainder of @xmath839 under division by @xmath734 .
one can check that the restriction of @xmath852 to the vertex set of the closed ball @xmath853 is an isomorphism of this ball with the closed ball @xmath854 whenever @xmath855 .
therefore @xmath856 as @xmath463 .
one consequence of lemma [ grig8 ] is that the graphs @xmath828 , @xmath829 , and @xmath830 are schreier graphs of the group @xmath0 . by construction ,
each of these graphs admits a nontrivial automorphism , which interchanges vertices corresponding to the same vertex of @xmath814 .
this property distinguishes @xmath828 , @xmath829 , and @xmath830 from the schreier graphs of orbits of the action @xmath547 .
it follows from proposition [ sch4 ] and lemma [ grig4 ] that marked schreier graphs @xmath810 and @xmath858 are isomorphic only if @xmath859 .
therefore the schreier graphs @xmath857 , @xmath555 admit no nontrivial automorphisms .
the graphs @xmath828 , @xmath829 , and @xmath830 have linear structure .
namely , one can label their vertices by @xmath860 , @xmath861 so that each @xmath860 is adjacent only to @xmath862 and @xmath863 .
if @xmath39 is an automorphism of such a graph , then either @xmath864 for some @xmath865 and all @xmath861 or @xmath866 for some @xmath865 and all @xmath861 .
assume that some @xmath10 has more than one nontrivial automorphism .
then we can choose @xmath39 above so that the latter option holds with @xmath867 .
take any path in @xmath10 that begins at @xmath54 and ends at @xmath55 and let @xmath244 be the code word of that path . since @xmath868 and @xmath869 for any integer @xmath536 , the path in @xmath10 with beginning @xmath870 and code word @xmath244 ends at @xmath871 .
it follows that for any integer @xmath872 the path with beginning @xmath54 and code word @xmath873 ends at @xmath870 .
in particular , this path is not closed .
however every element of the grigorchuk group @xmath0 is of finite order ( see @xcite ) so that for some @xmath872 the reversed word @xmath873 equals @xmath645 when regarded as a product in @xmath0 .
this conradicts with proposition [ sch1 ] .
thus the graph @xmath10 admits only one nontrivial automorphism .
[ grig10 ] the schreier graph @xmath814 is a double quotient of each of the graphs @xmath828 , @xmath829 , and @xmath830 . on the other hand ,
each of the graphs @xmath828 , @xmath829 , and @xmath830 is a double quotient of the schreier coset graph @xmath874 .
the schreier coset graph of the subgroup @xmath775 is shown in figure [ fig4 ] . in view of lemmas [ grig6 ] and [ grig9 ] ,
the automorphism group of this graph is the klein @xmath647-group .
the quotient of the graph by the entire automorphism group is the schreier graph of the orbit of @xmath783 .
the quotients by subgroups of order @xmath504 are the graphs @xmath828 , @xmath829 , and @xmath830 .
theorem [ main1 ] we are concerned with the mapping @xmath806 given by @xmath875 .
let us also consider a mapping @xmath876 given by @xmath877 and a mapping @xmath878 given by @xmath879 . by proposition [ sch4 ] , @xmath880 for all @xmath555 . by proposition [ sub5 ]
, @xmath39 is a homeomorphism .
lemma [ grig4 ] implies that the mapping @xmath881 is injective .
it is borel measurable due to lemma [ sub4 ] . also , @xmath881 is continuous at a point @xmath555 if and only if @xmath556 .
lemmas [ grig5 ] and [ grig6 ] imply that the latter condition fails only if the infinite word @xmath554 contains only finitely many zeros .
according to lemma [ grig2 ] , an equivalent condition is that @xmath554 is in the orbit of @xmath805 under the action @xmath547 . since the mapping @xmath5 is @xmath39 postcomposed with a homeomorphism , it is also injective , borel measurable , and continuous everywhere except the orbit of @xmath783 . by lemma [ grig7 ] ,
the graph @xmath7 is an isolated point of the image @xmath8 .
since @xmath882 for any @xmath555 and @xmath683 and since the action @xmath269 is continuous ( see proposition [ sch2 ] ) , the graph @xmath883 is an isolated point of @xmath8 for all @xmath683 . on the other hand ,
if @xmath555 is not in the orbit of @xmath783 , then the graph @xmath810 is not an isolated point of @xmath8 as the mapping @xmath5 is injective and continuous at @xmath554 .
it follows from lemma [ grig9 ] that the image @xmath8 and the orbits @xmath884 , @xmath834 are disjoint sets .
note that the orbit @xmath884 consists of marked graphs obtained from the graph @xmath10 by marking an arbitrary vertex .
lemma [ grig8 ] implies the union of those @xmath647 sets is the closure of @xmath8 .
theorem [ main2 ] lemma [ tree5 ] combined with lemma [ grig4 ] implies that the action of @xmath0 on the closure of @xmath8 is a continuous extension of the action @xmath547 .
the extension is one - to - one everywhere except for the orbit @xmath804 where it is four - to - one .
namely , for any @xmath683 the point @xmath885 is covered by @xmath647 graphs @xmath883 , @xmath886 , @xmath887 , and @xmath886 . according to theorem [ main1 ] ,
the graph @xmath883 is an isolated point of the closure of @xmath8 .
when we restrict our attention to the set @xmath13 of non - isolated points of the closure , we still have a continuous extension of the action @xmath547 , but it is three - to - one on the orbit @xmath804 .
by lemma [ grig2 ] , the group @xmath0 acts transitively on each level of the binary rooted tree @xmath508 .
then proposition [ tree1 ] implies that the action @xmath547 is minimal and uniquely ergodic , the only invariant borel probability measure being the uniform measure on @xmath16 .
since the action of @xmath0 on the set @xmath13 is a continuous extension of the action @xmath547 that is one - to - one except for a countable set and since this action has no finite orbits , it follows that the action is minimal , uniquely ergodic , and isomorphic to @xmath547 as the action with an invariant measure .
r. grigorchuk , _ solved and unsolved problems around one group_. l. bartholdi ( ed . ) et al . ,
infinite groups : geometric , combinatorial and dynamical aspects .
basel , birkhuser . _
progress in mathematics _
* 248 * , 117218 ( 2005 ) . | the paper is concerned with the space of the marked schreier graphs of the grigorchuk group and the action of the group on this space . in particular , we describe the invariant set of the schreier graphs corresponding to the action on the boundary of the binary rooted tree and dynamics of the group action restricted to this invariant set . |
layered cobaltites are investigated eagerly because of their structural and compositional variety and also their good thermoelectric properties . at present , the following three groups of cobaltites are known to be good thermoelectrics , because they display metallic conductivities as well as high thermoelectric powers @xmath18 , for reasons which are currently not fully understood .
a sodium cobaltite , na@xmath19coo@xmath0 , was the first compound reported as a good thermoelectric material @xcite .
then , the finding of [ ca@xmath0coo@xmath13@xmath14[coo@xmath0 ] @xcite , and then [ sr@xmath0bi@xmath20pb@xmath21o@xmath3@xmath4[coo@xmath0 ] @xcite , followed @xcite , where rs denotes a rocksalt - type block .
all share a common structural component , the coo@xmath0 planes , in which a two - dimensional - triangular lattice of co ions is formed by a network of edge - sharing coo@xmath22 octahedra .
charge carrier transport in these cobaltites is thought to be restricted mainly to the coo@xmath0 planes , as in the case of the cuo@xmath0 planes for the high-@xmath23 cuprates .
recent positive muon spin rotation and relaxation ( @xmath7sr ) experiments on [ ca@xmath0coo@xmath13@xmath14[coo@xmath0 ] @xcite indicated the existence of an incommensurate ( ic ) spin density wave ( sdw ) state below 100 k , which was not detected previously by other magnetic measurements @xcite .
the latter two experiments suggested that a long - range ic - sdw order completed below @xmath24 30 k , while a short - range order appeared below 100 k@xcite . since the @xmath25 curve exhibits a broad minimum around 80 k @xcite ,
the behavior of conduction electrons is found to be strongly affected even by their short - range magnetic order .
a new thermoelectric layered cobaltite [ ca@xmath0co@xmath1cu@xmath2o@xmath3@xmath14[coo@xmath0 ] was found recently @xcite ; the crystal structure consists of alternating layers of the quadruple rocksalt - type [ ca@xmath0co@xmath1cu@xmath2o@xmath3 subsystem and the single cdi@xmath0-type [ coo@xmath0 ] subsystem stacked along the @xmath12 axis .
there is a misfit between these subsystems along the @xmath11-axis , similar to the case of [ ca@xmath0coo@xmath13@xmath14[coo@xmath0 ] @xcite .
polycrystalline [ ca@xmath0co@xmath1cu@xmath2o@xmath3@xmath14[coo@xmath0 ] samples have values of thermopower @xmath18 = 150 @xmath26vk@xmath27 and resistivity @xmath28 = 15 m@xmath29 cm at 300 k @xcite . as a result ,
their thermoelectric power factor (= @xmath30 ) is @xmath24 20% larger than that of polycrystalline [ ca@xmath0coo@xmath13@xmath14[coo@xmath0 ] samples .
since the @xmath25 curve exhibited a broad minimum and the @xmath31 curve a broad maximum around 130 k , there seems to exist a transition from a high - temperature metallic to a low - temperature insulator state around 130 k. on the other hand , susceptibility ( @xmath32 ) measurements indicate no anomalies around 130 k , although @xmath32 showed a small change at @xmath24 80 k. therefore , @xmath7sr experiments on [ ca@xmath0co@xmath1cu@xmath2o@xmath3@xmath14[coo@xmath0 ] are also expected to provide crucial information on the correlation between magnetism and transport properties in the layered cobaltites . furthermore , such experiment is significant to clarify the universal behavior of magnetism in thermoelectric layered cobaltites . here ,
we report both weak ( @xmath24 100 oe ) transverse - field ( wtf- ) @xmath7sr and zero field ( zf- ) @xmath7sr measurements for a @xmath12 axis aligned [ ca@xmath0co@xmath1cu@xmath2o@xmath3@xmath14[coo@xmath0 ] sample at temperatures below 300 k. the former method is sensitive to local magnetic order _ via _ the shift of the @xmath7 spin precession frequency and the enhanced @xmath7 spin relaxation , while zf-@xmath7sr is sensitive to weak local magnetic [ dis]order in samples exhibiting quasi - static paramagnetic moments .
a @xmath12-axis aligned polycrystalline [ ca@xmath0co@xmath1cu@xmath2o@xmath3@xmath14[coo@xmath0 ] plate ( @xmath33 mm@xmath34 ) was synthesized by a reactive templated grain growth technique @xcite . in addition , @xmath12-axis aligned cobaltites with the triple rocksalt - type subsystem , _
i.e. _ , [ ca@xmath0coo@xmath13@xmath14[coo@xmath0 ] and [ ca@xmath35coo@xmath13@xmath4[coo@xmath0 ] ( @xmath36 = sr , y and bi ) plates were prepared for comparison .
then , the @xmath12-aligned plates were annealed at 450 @xmath37c for 12 hours in an oxygen flow .
powder x - ray diffraction ( _ xrd _ ) studies indicated that all the samples were single phase with a monoclinic structure and almost 100% aligned along the @xmath12 axis .
the preparation and characterization of these samples were described in detail elsewhere @xcite . magnetic susceptibility ( @xmath32 )
was measured using a superconducting quantum interference device ( _ squid _ ) magnetometer ( mpms , _ quantum design _ ) in a magnetic field of less than 55 koe .
the @xmath7sr experiments were performed on the * m15 * or * m20 * surface muon beam lines at triumf . the experimental setup and techniques
are described elsewhere @xcite .
[ [ ca_2co_43cu_23o_4_0.62rm - rscoo_2 ] ] temperature dependences of ( a ) susceptibility @xmath32 and ( b ) @xmath38 for @xmath12-aligned [ ca@xmath0co@xmath1cu@xmath2o@xmath3@xmath14[coo@xmath0 ] .
, width=302 ] figures [ fig : chi](a ) and [ fig : chi](b ) show the temperature dependences of @xmath32 and @xmath38 for the @xmath12-aligned [ ca@xmath0co@xmath1cu@xmath2o@xmath3@xmath14[coo@xmath0 ] sample . in order to determine anisotropy ,
the magnetic field @xmath39 was applied parallel or perpendicular to the _ ab _ plane .
hereby , we abbreviate @xmath32 obtained with @xmath40 as @xmath41 and @xmath42 as @xmath43 , respectively .
the @xmath44 curve in a zero - field cooling ( zfc ) mode exhibits a cusp at @xmath24 85 k ; also , a clear thermal hysteresis is seen between the data obtained in a zfc mode and a field cooling ( fc ) mode . on the other hand ,
as @xmath45 decreases , @xmath46 decreases monotonically with increasing slope ( d@xmath46/d@xmath45 ) , although there is a small anomaly at @xmath24 85 k , probably due to a misalignment between the sample axis and @xmath39 . nevertheless , the magnetization ( @xmath47 curve did not show a clear loop even at 5 k. these results suggest that [ ca@xmath0co@xmath1cu@xmath2o@xmath3@xmath14[coo@xmath0 ] undergoes a magnetic transition to either a ferrimagnetic or a spin glass state at 80 k with the easiest magnetization direction parallel to the @xmath12-axis .
it is worth noting that there are no marked anomalies in figure [ fig : chi ] except for the cusp in the @xmath44 at @xmath24 85 k. temperature dependences of ( a ) @xmath48 , ( b ) @xmath49 , ( c ) @xmath50 and ( d ) @xmath51 ( @xmath52 = fast and slow ) in @xmath12-aligned [ ca@xmath0co@xmath1cu@xmath2o@xmath3@xmath14[coo@xmath0 ] .
the data were obtained by fitting the wtf-@xmath7sr spectra using equation ( [ eq : tffit ] ) . , width=302 ] the wtf-@xmath7sr spectra in a magnetic field of @xmath53 oe in the @xmath12-aligned [ ca@xmath0co@xmath1cu@xmath2o@xmath3@xmath14[coo@xmath0 ] sample exhibit a clear reduction of the @xmath7 precession amplitude below 200 k. the wtf-@xmath7sr spectrum below 200 k
was well fitted in the time domain with a combination of a slowly relaxing precessing signal and two non - oscillatory signals , one fast and the other slow relaxing : @xmath54 where @xmath55 is the initial asymmetry , @xmath56 is the muon spin polarization function , @xmath57 is the muon larmor frequency , @xmath58 is the initial phase of the precession and @xmath59 and @xmath60 ( @xmath52 = para , fast and slow ) are the asymmetries and exponential relaxation rates of the three signals .
the latter two signals ( @xmath52 = fast and slow ) have finite amplitudes below @xmath61 k. figures [ fig : wtf - musr](a ) - [ fig : wtf - musr](d ) show the temperature dependences of @xmath48 , @xmath49 , @xmath50 ( @xmath52 = fast and slow ) and @xmath51 in the @xmath12-aligned [ ca@xmath0co@xmath1cu@xmath2o@xmath3@xmath14[coo@xmath0 ] sample . the large decrease in @xmath48 below 180
k ( and the accompanying increase in @xmath49 ) indicate the existence of a magnetic transition with an onset temperature @xmath62 k , a transition width @xmath63 k and an endpoint @xmath64 k , respectively .
since @xmath48 is roughly proportional to the volume of a paramagnetic phase , this result ( @xmath65 0 below @xmath66 ) suggests that almost all the sample changes into a magnetically ordered state below 140 k. zf-@xmath7sr time spectra of the @xmath12-aligned [ ca@xmath0co@xmath1cu@xmath2o@xmath3@xmath14[coo@xmath0 ] plate at 2.1 k. the configurations of the sample and the muon beam are ( top ) @xmath67 and ( bottom ) @xmath68 , where @xmath69 indicates the initial muon spin direction .
, width=302 ] figure [ fig : zf - musr ] shows zf-@xmath7sr time spectra at 2.1 k in the @xmath12-aligned sample ; the top spectrum was obtained with the initial @xmath7 spin direction @xmath69 perpendicular to the @xmath12-axis and the bottom one with @xmath68 .
a clear oscillation due to quasi - static internal fields is observed only when @xmath67 .
this oscillating spectrum is reasonably well fitted with a combination of two zeroth - order bessel functions of the first kind @xmath70 ( for the ic - sdw)@xcite and an exponential relaxation function : @xmath71 @xmath72 where @xmath55 is the empirical maximum muon decay asymmetry , @xmath73 , @xmath74 and @xmath75 are the asymmetries and @xmath76 , @xmath77 and @xmath78 are the exponential relaxation rates associated with the three signals . also , @xmath79 is the muon precession frequency in the characteristic local magnetic field @xmath80 due to an ic - sdw and @xmath81 is muon gyromagnetic ratio . the two bessel functions in equation ( [ eq : zffit ] ) indicate that there are two inequivalent muon sites in the [ ca@xmath0co@xmath1cu@xmath2o@xmath3@xmath14[coo@xmath0 ] lattice , because the three signals have a finite value below 140 k. due to some broadening of the ic - sdw field distribution , the two bessel functions exhibit an exponential damping .
we therefore conclude that [ ca@xmath0co@xmath1cu@xmath2o@xmath3@xmath14[coo@xmath0 ] undergoes a magnetic transition from a paramagnetic state to an ic - sdw state , similar to the case of [ ca@xmath0coo@xmath13@xmath14[coo@xmath0 ] .
@xcite the absence of a clear oscillation in the bottom spectrum of figure [ fig : zf - musr ] indicates that the internal magnetic field @xmath82 is roughly parallel to the @xmath12 axis @xcite , since the muon spins do not precess in a parallel magnetic field .
the ic - sdw is unlikely to propagate along the @xmath12 axis due both to the two - dimensionality and to the misfit between the two subsystems .
the ic - sdw is therefore considered to propagate in the @xmath10-@xmath11 plane , with oscillating moments directed along the @xmath12 axis . in the [ ca@xmath0co@xmath1cu@xmath2o@xmath3 subsystem , 1/3 of the co sites
are randomly substituted by cu @xcite .
this means that 2 or 3 of the eight of the first nearest neighboring co ions are replaced by cu for each co ion in the [ ca@xmath0co@xmath1cu@xmath2o@xmath3 subsystem .
nevertheless , a clear precession was observed in the zf-@xmath7sr spectrum below 140 k. in addition , the precession frequency ( @xmath83 mhz ) at zero temperature is almost the same as to [ ca@xmath0coo@xmath13@xmath14[coo@xmath0 ] ( @xmath84 mhz ) .
since the long - range order of the co moments in the [ ca@xmath0co@xmath1cu@xmath2o@xmath3 subsystem should be strongly hindered by cu , it is concluded that the ic - sdw exists not in the [ ca@xmath0co@xmath1cu@xmath2o@xmath3 subsystem but in the [ coo@xmath0 ] plane .
temperature dependences of ( a ) @xmath50 ( @xmath52 = sdw1 , sdw2 and f ) , ( b ) @xmath85 ( @xmath52 = 1 and 2 ) and ( c ) @xmath51 ( @xmath52 = sdw1 , sdw2 and f ) for the @xmath12-aligned [ ca@xmath0co@xmath1cu@xmath2o@xmath3@xmath14[coo@xmath0 ] .
the data was obtained by fitting of the zf-@xmath7sr time spectra using equation ( [ eq : zffit ] ) .
, width=302 ] temperature dependences of ( a ) @xmath73 , ( b ) @xmath86 , ( c ) @xmath76 , and ( d ) in - plane resistivity ( @xmath87 ) for the @xmath12-aligned [ ca@xmath0co@xmath1cu@xmath2o@xmath3@xmath14[coo@xmath0 ] . in figure
[ fig : sdw1](a ) , @xmath48 obtained by the wtf-@xmath7sr experiment is also shown for comparison .
the solid line in figure [ fig : sdw1](b ) represents the temperature dependence of the bcs gap energy .
, width=302 ] figures [ fig : sdw1&2](a ) - [ fig :
sdw1&2](c ) show the temperature dependences of @xmath50 ( @xmath52 = sdw1 , sdw2 and f ) , @xmath85 ( @xmath52 = 1 and 2 ) and @xmath51 ( @xmath52 = sdw1 , sdw2 and f ) .
it is clearly seen that the signal associated with the sdw1 is the predominant one among the three signals .
the volume fraction of the signal from the sdw1 is estimated as @xmath24 100% at 2.1 k. this suggests that almost all the @xmath7 are bound to the oxygen ions in the [ coo@xmath0 ] plane , with only at the very small portion in the [ ca@xmath0co@xmath1cu@xmath2o@xmath3 subsystem .
therefore , we consider the signal of the sdw1 and ignore the contribution from the other signals . figures [ fig : sdw1](a ) - [ fig : sdw1](d ) show the temperature dependences of @xmath73 , @xmath86 , @xmath76 , and in - plane resistivity ( @xmath87 ) . in particular , @xmath73 increases monotonically with decreasing @xmath45 from 140 k , although @xmath48 obtained by the wtf-@xmath7sr measurement exhibits a rapid decrease below 200 k and levels off to almost 0 below 140 k ( see figure [ fig : sdw1](a ) ) .
this suggests that a long - range ic - sdw order completes below @xmath24 140 k (= @xmath88 ) , while a short - range order appears below 200 k (= @xmath89 ) , as in the case of [ ca@xmath0coo@xmath13@xmath14[coo@xmath0 ] @xcite .
actually , this is in good agreement with the temperature dependence of @xmath87 ; that is , @xmath90 is metallic above 140 k and semiconducting below 140 k. on the other hand , there was no clear anomalies around 140 k in the @xmath91 curve ( see fig .
[ fig : chi ] ) probably due to effects of the grain boundaries , a possible inhomogeneous distribution of the cations and oxygen ion , and magnetic anisotropy .
indeed , the @xmath91 curve for the [ ca@xmath0coo@xmath13@xmath14[coo@xmath0 ] single crystals exhibited a clear but small maximum at 27 k indicating the formation of the ic - sdw order only when @xmath92@xcite , whereas those for the @xmath12-aligned and random polycrystalline samples did not .
therefore , if a large single crystal of [ ca@xmath0co@xmath1cu@xmath2o@xmath3@xmath14[coo@xmath0 ] is available , its @xmath93 curve would also have an anomaly around 140 k. the @xmath94 curve is well explained by the energy gap function in the bcs weak - coupling theory ( see figure [ fig : sdw1](b ) ) , as expected for the ic - sdw state @xcite .
as seen in figure [ fig : sdw1](c ) , @xmath76 seems to increase monotonically with decreasing @xmath45 .
the exponential damping of the ic - sdw oscillation is most likely to be caused by the misfit between the [ ca@xmath0co@xmath1cu@xmath2o@xmath3 and the [ coo@xmath0 ] subsystem along the @xmath11 axis . since we used the @xmath12-aligned sample for the current zf-@xmath7sr measurements ,
the result obtained with @xmath67 is the average information along the @xmath10 and @xmath11 axis .
thus , the anisotropic ic modulation in the @xmath95 plane is considered to be the origin of the broadening of the ic - sdw field distribution at the @xmath7 sites .
temperature dependences of @xmath96 for the @xmath12-aligned [ ca@xmath0co@xmath1cu@xmath2o@xmath3@xmath14[coo@xmath0 ] and @xmath12-aligned pure and doped [ ca@xmath0coo@xmath13@xmath14[coo@xmath0 ] .
the dotted lines represent the temperature dependence of the bcs gap energy . , width=302 ] similar anisotropic zf-@xmath7sr spectra were also observed for the @xmath12-aligned [ ca@xmath0coo@xmath13@xmath14[coo@xmath0 ] below @xmath24 30 k @xcite , and the zf-@xmath7sr time spectra with @xmath67 were well fitted using the bessel function , @xmath97exp(@xmath98 ) .
figure [ fig : sdw_others ] shows the @xmath99 curve for the @xmath12-aligned pure and doped [ ca@xmath0coo@xmath13@xmath14[coo@xmath0 ] and [ ca@xmath0co@xmath1cu@xmath2o@xmath3@xmath14[coo@xmath0 ] samples .
the data were obtained by fitting the zf-@xmath7sr time spectra with @xmath67 using equation ( [ eq : zffit ] ) .
all the temperature dependences of @xmath86 for the @xmath12-aligned layered cobaltites are well fitted by the bcs weak - coupling theory .
it should be noted that all the samples show approximately the same precession frequency at zero temperature , although the transition temperatures are different from 27 to @xmath24140 k ; _ i.e. _ , @xmath88 = 27 k for [ ca@xmath15coo@xmath13@xmath14[coo@xmath0 ] , @xmath24 45 k for [ ca@xmath100sr@xmath101coo@xmath13@xmath4[coo@xmath0 ] , @xmath24 80 k for [ ca@xmath100y@xmath101coo@xmath13@xmath4[coo@xmath0 ] and [ ca@xmath100bi@xmath101coo@xmath13@xmath4[coo@xmath0 ] and @xmath24 140 k for [ ca@xmath0co@xmath1cu@xmath2o@xmath3@xmath14[coo@xmath0 ] .
this suggests that the local magnetic field @xmath102(0 k ) is independent of both dopant and the number of layers in the rocksalt - type subsystem , if we assume that the muon sites are essentially the same in all these compounds .
this supports the conclusion that the ic - sdw exists not in the rocksalt - type subsystem but in the [ coo@xmath0 ] plane and , as a result , the @xmath7 sites are considered to be the vicinity of the o ions in the [ coo@xmath0 ] plane .
although the possible @xmath7 sites are bound to the oxygen ions in the [ coo@xmath0 ] plane and the two inequivalent oxygen ions in the [ ca@xmath0co@xmath1cu@xmath2o@xmath3 subsystem , the @xmath7 sites are most likely to the vicinity of the o ions in the [ coo@xmath0 ] plane .
we , therefore , calculate the dipole field at the o ions in the [ coo@xmath0 ] plane at first . the bond length @xmath103 of co - o in the [ coo@xmath0 ] plane is 0.184 nm in [ ca@xmath0co@xmath1cu@xmath2o@xmath3@xmath14[coo@xmath0 ] @xcite , whereas it is 0.197 nm in [ ca@xmath0coo@xmath13@xmath14[coo@xmath0 ] @xcite . a simple estimate of dipole field from one ion is given by ; @xmath104 where @xmath105 is the dipole moment of the co ions in the ic - sdw state and @xmath106 is the permeability of free space . using the above @xmath103 values and the observed @xmath102
, we estimate @xmath107 for [ ca@xmath0co@xmath1cu@xmath2o@xmath3@xmath14[coo@xmath0 ] and 3.3 @xmath108 for [ ca@xmath0coo@xmath13@xmath14[coo@xmath0 ] ( see table 1 ) .
@ccc cobaltite & [ ca@xmath0co@xmath1cu@xmath2o@xmath3@xmath14[coo@xmath0 ] & [ ca@xmath0coo@xmath13@xmath14[coo@xmath0 ] + @xmath109 ( mhz ) & 63.9 & 55.5 + @xmath102 ( koe ) & 4.71 & 4.08 + @xmath110 in [ coo@xmath0 ] ( nm ) &
0.184 & 0.197 + @xmath111 & 3.2 & 3.3 + if we ignore the effect of the distortion of the coo@xmath22 octahedra in the [ coo@xmath0 ] plane , the number of the nearest neighboring co ions for the o ion is three @xcite ; hence , the ordered co moment in the ic - sdw state is therefore roughly estimated as @xmath112/co ion for both compounds .
this is in good agreement with the amplitude of the ic - sdw estimated by the mean field theory ( @xmath113/co ion ) @xcite . here
, it should be noted that the muon locates probably @xmath240.1 nm away from the oxygen ions , and that there is no space for it in the coo@xmath22 octahedra in the [ coo@xmath0 ] plane as in the case for the high-@xmath114 cuprates@xcite .
thus , the accuracy of the above estimation is very limited ; _
i.e. _ , @xmath105=0.3 - 11.7 @xmath108 for [ ca@xmath0co@xmath1cu@xmath2o@xmath3@xmath14[coo@xmath0 ] , if @xmath115 nm .
in order to determine the @xmath7 sites , both further experiments on the layered cobaltites and a theoretical research are necessary .
there are two co sites in the [ ca@xmath0co@xmath1cu@xmath2o@xmath3@xmath14[coo@xmath0 ] lattice .
thus , it is difficult to determine the co valence in the [ coo@xmath0 ] plane by a @xmath32 measurement or a chemical titration technique alone . here
, we assume that the magnitude of @xmath18 depends on the concentration @xmath116 of co@xmath117 ions in the [ coo@xmath0 ] plane as @xcite ; @xmath118 where @xmath119 is the boltzmann constant , @xmath120 is the elementary charge and @xmath121 and @xmath122 are the multiplying numbers of the spin configurations of co@xmath123 and co@xmath117 , respectively .
since the electron configurations of both co@xmath123 and co@xmath117 are in the low - spin state ( @xmath124 and @xmath125 ) @xcite , @xmath121=1 and @xmath122=6 .
then , using the value of @xmath18(300 k ) @xmath126vk@xmath27 , we obtain @xmath127 0.51 , _ i.e. _ the average valence of the co ions in the [ coo@xmath0 ] plane is + 3.51 .
this indicates that almost the same amounts of co@xmath123 and co@xmath117 coexist in the [ coo@xmath0 ] plane . in other words ,
the co spin ( * s*=1/2 ) occupies about half of the sites of the two - dimensional - triangular lattice of co ions .
this is significant to achieve the ic - sdw long - range order in the triangular lattice .
in addition , the average co moment is calculated as @xmath113/co ion , which is consistent with the value estimated above .
it is worth noting that @xmath7 sites are bound to the o ions in the [ coo@xmath0 ] plane .
this means that the @xmath7 feel mainly the magnetic field in the [ coo@xmath0 ] plane .
thus , the ic - sdw is most unlikely to be caused by the misfit between the two subsystems , but to be an intrinsic behavior of the [ coo@xmath0 ] plane .
indeed , the recent @xmath7sr experiments on [ na]@xmath19[coo@xmath0 ] , which consist of the alternating stack of na and [ coo@xmath0 ] planes , also indicate the existence of a commensurate sdw or a ferrimagnetic state below 22 k for [ na]@xmath128[coo@xmath0 ] @xcite and an ic - sdw state below 19 k for [ na]@xmath129[coo@xmath0 ] @xcite .
therefore , it is concluded that the ic - sdw state is a common behavior for the layered cobaltites , although the magnitude of @xmath130 depends on the co valence in the [ coo@xmath0 ] plane and the structure of the subsystem sandwiched by the two [ coo@xmath0 ] planes ( see figure [ fig : sdw_others ] ) .
the ic - sdw order in the two layered cobaltites is assigned to be a spin ( * s*=1/2 ) order on the two - dimensional triangular lattice ( _ i.e. _ the coo@xmath0 plane ) at non - half filling .
such a problem was investigated by several workers using the hubbard model within a mean field approximation @xcite ; @xmath131 where @xmath132 creates ( destroys ) an electron with spin @xmath133 on site @xmath134 , @xmath135 is the number operator , @xmath136 is the nearest - neighbor hopping amplitude and @xmath137 is the hubbard on - site repulsion .
the electron filling @xmath52 is defined as @xmath52 = ( 1/2@xmath138)@xmath139 , where @xmath138 is the total number of sites . at @xmath45=0 and @xmath52=0.5 ( _
i.e. _ , the average valence of the co ions in the [ coo@xmath0 ] plane is + 4 ) , as @xmath137 increased from 0 , the system is paramagnetic metal up to @xmath140 , and changes into a metal with a spiral ic - sdw , and then at @xmath141 , a first - order metal - insulator transition occurs @xcite .
also the calculations predict that @xcite , as @xmath52 increases from 0 , the magnitude of @xmath142 at the boundary between the paramagnetic and sdw phases decreases with increasing slope ( @xmath143 ) up to @xmath52=0.75 . even for @xmath142=0 ,
the sdw phase is stable at @xmath52=0.75 .
@xmath142 increases with further increasing @xmath52 , with decreasing slope .
in other words , at @xmath52=0.75 ( _ i.e. _ the average valence of the co ions in the [ coo@xmath0 ] plane is + 3.5 ) , a spiral ic - sdw state is expected to appear at the highest temperature in the @xmath52 range between 0.5 and 1 @xcite .
the value of @xmath52 is estimated as 0.74 for [ ca@xmath0co@xmath1cu@xmath2o@xmath3@xmath14[coo@xmath0 ] and 0.715 , 0.70 , 0.73 and 0.73 for pure , sr- , y- , and bi - doped [ ca@xmath0coo@xmath13@xmath14[coo@xmath0 ] respectively , using eq .
( [ eq : tep ] ) and @xmath18(300 k ) . therefore ,
the fact that @xmath88 for [ ca@xmath0co@xmath1cu@xmath2o@xmath3@xmath14[coo@xmath0 ] is higher than those for pure and doped [ ca@xmath0coo@xmath13@xmath14[coo@xmath0 ] ( see figure [ fig : sdw_others ] ) is roughly explained by the model calculations , if we ignore the data for sr - doped [ ca@xmath0coo@xmath13@xmath14[coo@xmath0 ] .
also , it should be pointed out that the increased two - dimensionality induced by the increase in the interlayer distance between coo@xmath0 planes plays a significant role to increase @xmath88 .
this is because @xmath88 for [ ca@xmath0co@xmath1cu@xmath2o@xmath3@xmath14[coo@xmath0 ] is considerably higher than those for pure and doped [ ca@xmath0coo@xmath13@xmath14[coo@xmath0 ] .
here , we repeat the prediction of the model calculations .
that is , the sdw phase is stable at @xmath52=0.75 even for @xmath142=0 ; this means that , on the ideal two - dimensional triangular lattice at @xmath52=0.75 , the sdw spin structure appears at temperatures below its melting point .
indeed , the large observed transition width ( 60 k ) are consistent with enhanced two - dimensionality and resulting spin fluctuations .
therefore , not only the shift of @xmath52 towards the optimal value ( 0.75 ) but also the enhanced two - dimensionality are considered to increase @xmath88 for [ ca@xmath0co@xmath1cu@xmath2o@xmath3@xmath14[coo@xmath0 ] .
since the magnitude of @xmath86 indicates the amplitude of the ic - sdw directly , there are two predominant factors for @xmath86 .
that is , the magnitude and direction of the co spin in the [ coo@xmath0 ] plane .
the former is a constant at low temperatures , because the spin configurations for co@xmath123 and co@xmath117 ions ( * s*=0 and 1/2 ) are not affected by both dopant and the number of layers in the rocksalt - type subsystem .
the latter is a function of magnetic anisotropy in the lattice . here
the magnetic anisotropy in [ ca@xmath0co@xmath1cu@xmath2o@xmath3@xmath14[coo@xmath0 ] is most likely to be equivalent to that in [ ca@xmath0coo@xmath13@xmath14[coo@xmath0 ] due to similar environment around the [ coo@xmath0 ] plane in the two cobaltites .
that is , the first and second nearest adjacent plane for the [ coo@xmath0 ] plane are the ca - o and ( co@xmath2cu@xmath144)-o or co - o plane in the rocksalt - type subsystem , although the third nearest plane is different for the two cobaltites .
therefore , @xmath86(0 k ) is considered to be approximately same for all the cobaltites investigated here ( see fig .
[ fig : sdw_others ] ) .
the @xmath91 curve exhibits a clear anomaly at @xmath24 85 k , while the results of both wtf-@xmath7sr and zf-@xmath7sr do not ( see figures [ fig : wtf - musr ] , [ fig : sdw1&2 ] and [ fig : sdw1 ] ) . on the other hand
, there is no marked anomaly at 140 - 200 k in the @xmath91 curve ( see figure [ fig : chi ] ) , although @xmath7sr detects the formation of the ic - sdw state .
this is quite similar to the case for [ ca@xmath0coo@xmath13@xmath14[coo@xmath0 ] ; where according to the @xmath7sr experiments , it was found that the short - range ic - sdw order appeared below @xmath24 100 k and the long - range order completed below 27 k , whereas the @xmath91 curve below 300 k only exhibited a ferrimagnetic transition at 19 k @xcite .
the structures of the ic - sdw of both compounds are considered to be essentially the same , as discussed above . for [ ca@xmath0coo@xmath13@xmath14[coo@xmath0 ] ,
the ic - sdw is induced by ordering of the co moments in the [ coo@xmath0 ] plane , whereas the ferrimagnetic ordering arises because of the interlayer coupling between the co spins in the [ coo@xmath0 ] and [ ca@xmath0coo@xmath13 subsystems@xcite .
as seen in figure [ fig : sdw1](b ) , the microscopic internal ic - sdw field does not change around 85 k. furthermore , the extremely small value of the asymmetry of the exponential relaxation function @xmath75 , even at 2.1 k , ( see figure [ fig : sdw1](a ) ) suggests that the anomaly at @xmath24 85 k is not a transition to a spin - glass state @xcite .
therefore , the anomaly at @xmath24 85 k is most likely caused by some order between the two subsystems , _
i.e. _ , the [ ca@xmath0co@xmath1cu@xmath2o@xmath3 and the [ coo@xmath0 ] .
the interlayer coupling between both subsystems is expected to be basically antiferromagnetic ( af ) , because there was no clear @xmath145 loop even at 5 k. in other words , a two - dimensional af ( ic - sdw ) order of the co spins completes at 140 k , whereas a three - dimensional af order occurs below @xmath24 85 k. the [ ca@xmath0co@xmath1cu@xmath2o@xmath3 subsystem consists of two ca - o planes and two ( co@xmath2cu@xmath144)-o planes ; so that , the two ( co@xmath2cu@xmath144)-o planes are sandwiched by the two ca - o planes . considering the fact that 1/3 of the co ions are replaced by cu ions and that the thickness of the rocksalt - type subsystem increases due to the extra ( co@xmath2cu@xmath144)-o plane , the coupling along the @xmath12-axis for [ ca@xmath0co@xmath1cu@xmath2o@xmath3@xmath14[coo@xmath0 ] is rather weak compared with that for [ ca@xmath0coo@xmath13@xmath14[coo@xmath0 ] .
this weak coupling is likely to be a significant factor for the thermal hysteresis of the @xmath91 curve .
that is , the ic - sdw in each [ coo@xmath0 ] plane is considered to be moved by the external magnetic field ( @xmath24 10 koe ) at 85 - 140 k , but fixed below @xmath24 85 k.
in order to elucidate the magnetism in good thermoelectric layered cobaltites , @xmath7sr spectroscopy has been used on a @xmath12-aligned polycrystalline [ ca@xmath0co@xmath1cu@xmath2o@xmath3@xmath14[coo@xmath0 ] sample at temperatures below 300 k. it was found that [ ca@xmath0co@xmath1cu@xmath2o@xmath3@xmath14[coo@xmath0 ] exhibits a transition at around 140 k from a paramagnetic to an incommensurate spin density wave ic - sdw state , although a short - range order appears below @xmath24 180 k. by comparison with the @xmath7sr results on pure and doped [ ca@xmath0coo@xmath13@xmath14[coo@xmath0 ] , the ic - sdw appears to be common behavior for these cobaltites .
the characteristic features of the ic - sdw are as follows ; 1 .
a long - range ic - sdw order completes below @xmath88 , while a short - range order is observed at 40 - 60 k higher than @xmath88 with a transition width @xmath146 = 40 - 60 k. 2 .
the ic - sdw propagates in the @xmath10-@xmath11 plane , with oscillating moments directed along the @xmath12 axis .
the ic - sdw exists not in the rocksalt - type subsystem but in the [ coo@xmath0 ] plane .
the magnitude of @xmath130 is found to be sensitive both to the co valence in the [ coo@xmath0 ] plane , _
i.e. _ , the occupancy of the co spin ( @xmath18=1/2 ) in the triangular lattice , and to the structure of the subsystem sandwiched by the two [ coo@xmath0 ] planes .
therefore , physical properties of the layered cobaltites should be investigated systematically as functions of the co valence in the [ coo@xmath0 ] plane and the distance between the two adjacent [ coo@xmath0 ] planes .
we thank dr .
s.r . kreitzman and dr .
arseneau of triumf for help with the @xmath7sr experiments .
also , we thank mr .
a. izadi - najafabadi and mr .
laroy of university of british columbia for help with the experiments .
we appreciate useful discussions with dr .
r. asahi of toyota crdl , professor u. mizutani , professor h. ikuta and professor t. takeuchi of nagoya university .
this work was supported at toyota crdl by joint research and development with international center for environmental technology transfer in 2002 - 2004 , commissioned by the ministry of economy trade and industry of japan , at ubc by the canadian institute for advanced research , the natural sciences and engineering research council of canada , and at triumf by the national research council of canada .
30 molenda j , delmas c , dordor p and stoklosa a 1989 _ solid state ionics _ * 12 * 473 yakabe h , kikuchi k , terasaki i , sasago y and uchinokura k 1997 _ proc .
16th int . conf . on thermoelectrics ( dresden ) _
( piscataway : institute of electrical and electronics engineers , inc . )
pp 523 - 527 terasaki i , sasago y and uchinokura k 1997 _ phys .
b _ * 56 * r12685 funahashi r , matsubara i , ikuta h , takeuchi t , mizutani u and sodeoka s 2000 _ jpn .
* 39 * l1127 masset a c , michel c , maignan a , hervieu m , toulemonde o , studer f , raveau b and hejtmanek j 2000 _ phys . rev .
b _ * 62 * 166 miyazaki y , kudo k , akoshima m , ono y , koike y and kajitani t 2000 _ jpn . j. appl .
phys . _ * 39 * l531 tsukada i , yamamoto t , takagi m , tsubone t , konno s and uchinokura k 2001 _ j. pys . soc .
_ * 70 * 834 yamamoto t , uchinokura k and tsukada i 2002 _ pys .
b _ * 65 * 184434 fujii t , terasaki i , watanabe t and matsuda a 2002 _ jpn .
_ * 41 * l783 sugiyama j , itahara h , tani t , brewer j h and ansaldo e j 2002 _ phys . rev . b _ * 66 * 134413 sugiyama j , brewer j h , ansaldo e j , bayer m , itahara h and tani t 2003 _ physica b _ * 326 * 518 sugiyama j , brewer j h , ansaldo e j , itahara h , seno y , dohmae k , xia c and tani t 2003 _ phys
b _ in press sugiyama j , xia c and tani t 2003 _ phys . rev .
b _ * 67 * 104410 miyazaki y , miura t , ono y and kajitani t 2002 _ jpn .
phys . _ * 41 * l849 tani t , itahara h , xia c and sugiyama j 2003 _ j. mater .
* 13 * 1865 itahara h _ et al .
_ 2003 unpublished le l p , keren a , luke g m , sternlieb b j , wu w d , uemura y j , brewer j h , riseman t m , upasani r v , chiang l y , kang w , chaikin p m , csiba t and gr " uner g 1993 _ phys . rev .
b _ * 48 * 7284 uemura y j 1999 _ muon science _
( bristol : institute of physics publishing ) pp 85 - 114 and references cited therein kalvius g m , noakes d r and hartmann o 2001 _ handbook on the physics and chemistry of rare earths _
vol 32 ( amsterdam : north - holland ) pp 55 - 451 and references cited therein gr " uner g 1994 _ density waves in solids _
( reading : addison - wesley - longmans ) pp 71 - 85 and references cited therein miyazaki y _ et al .
_ 2003 unpublished miyazaki y , onoda m , oku t , kikuchi m , ishii y , ono y , morii y and kajitani t 2002 _ j. phys .
jpn . _ * 71 * 491 koshibae w , tsutsui k and maekawa s 2000 _ phys . rev .
b _ * 62 * 6869 h ' ebert s , lambert s , pelloquin d and maignan a 2001 _ phys . rev .
b _ * 64 * 172101 sugiyama j , itahara h , brewer j h , ansaldo e j , motohashi t , karppinen m and yamauchi h 2003 _ phys .
b _ * 67 * 214420 sugiyama j
_ et al . _
2003 unpublished campbell i a 1999 _ muon science _
( bristol : institute of physics publishing ) pp 137 - 148 and references cited therein krishnamurthy h r , jayaprakash c , sarker s and wenzel w 1990 _ phys .
lett . _ * 64 * 950 fujita m , ichimura m and nakao k 1991 _ j. phys .
jpn . _ * 60 * 2831 fujita m , nakanishi t and machida k 1992 _ phys .
b _ * 45 * 2190 | magnetism of a misfit layered cobaltite [ ca@xmath0co@xmath1cu@xmath2o@xmath3@xmath4[coo@xmath0 ] ( @xmath6 0.62 , rs denotes a rocksalt - type block ) was investigated by a positive muon spin rotation and relaxation ( @xmath7sr ) experiment . a transition to an incommensurate ( ic )
spin density wave ( sdw ) state was found below 180 k (= @xmath8 ) ; and a clear oscillation due to a static internal magnetic field was observed below 140 k (= @xmath9 ) .
furthermore , an anisotropic behavior of the zero - field @xmath7sr experiment indicated that the ic - sdw propagates in the @xmath10-@xmath11 plane , with oscillating moments directed along the @xmath12 axis .
these results were quite similar to those for the related compound [ ca@xmath0coo@xmath13@xmath14[coo@xmath0 ] , _ i.e. _ , ca@xmath15co@xmath16o@xmath17 .
since the ic - sdw field in [ ca@xmath0co@xmath1cu@xmath2o@xmath3@xmath14[coo@xmath0 ] was approximately same to those in pure and doped [ ca@xmath0coo@xmath13@xmath14[coo@xmath0 ] , it was concluded that the ic - sdw exist in the [ coo@xmath0 ] planes . |
in the standard model of particle physics ( sm ) , neutrinos are regarded as massless particles .
however , existence of nonzero masses of neutrinos has been established by a sequence of success of neutrino oscillation measurements @xcite . if neutrino masses are generated similarly to the other fermion masses via a yukawa interaction ( introducing right - handed neutrinos @xmath0 ) with the @xmath1-doublet scalar field @xmath2 in the sm
, the yukawa coupling constants must be extremely small ( @xmath3 ) .
since the coupling constant seems too different from the other yukawa coupling constants to regard it as natural , we might expect that neutrino masses are generated in a different mechanism .
the most famous example will be the seesaw mechanism @xcite where extremely heavy gauge - singlet fermions are introduced . in this short review ,
we consider extensions of the sm with new scalars ( in contrast with the seesaw models ) along the viewpoint `` higgs as a probe of new physics '' of this workshop . along my viewpoint `` neutrino as a guide to new physics '' , we investigate only leptophilic scalars which couple only with leptons among fermions because such scalars would contribute to mechanism of generating neutrino masses . in order to give predictions
, it will not be preferred that two new particles appear in an interaction ( lepton``new fermion''``new scalar '' ) .
therefore , let us concentrate on the following types of yukawa interactions : lepton
lepton``leptophilic scalar '' .
the leptophilic scalars here are assumed to be light enough ( tev - scale ) to be produced at collider experiments . if the yukawa interaction relates to the neutrino mass matrix , the flavor structure of decays of the new scalar into leptons
would be predicted by using current knowledge on the neutrino oscillation parameters .
if the prediction is experimentally confirmed in the future , we would obtain information on the mechanism of the neutrino mass generation and on parameters ( e.g. the lightest neutrino mass ) which can not be measured in neutrino oscillation experiments .
we deal with @xmath1-singlet , doublet , and triplet scalar fields .
their yukawa interactions are listed in table [ tab : table ] . for simplicity ,
mixings between scalars are ignored throughout this article .
leptophilic neutral scalars in doublet and triplet fields are not discussed in this article because their decays via yukawa interactions are into neutrinos which do not provide information on the flavor structure . .
yukawa interactions of leptophilic scalar fields and leptonic decays of their singly - charged and doubly - charged components . [ cols= " < ,
< , < , < " , ]
neutrinos @xmath4 in the flavor basis are superpositions of mass eigenstates @xmath5 : @xmath6 , where the unitary matrix @xmath7 is the so - called maki - nakagawa - sakata matrix @xcite . when the neutrino mass term is @xmath8 , neutrinos are referred to as the dirac neutrinos .
the mass matrix for dirac neutrinos is diagonalized with @xmath7 as @xmath9 , where mass eigenvalues @xmath10 are taken to be real and positive . on the other hand ,
if the neutrino mass term is @xmath11 , we call the neutrinos as the majorana neutrinos which break the lepton number conservation .
the majorana neutrino mass matrix is diagonalized as @xmath12 , where @xmath13 and @xmath14 are majorana phases @xcite which are physical parameters only for majorana neutrinos .
the matrix @xmath7 can be parameterized as @xmath15 where @xmath16 and @xmath17 stand for @xmath18 and @xmath19 , respectively .
current data of neutrino oscillation measurements @xcite constrains mixing angles and squared - mass differences ( @xmath20 ) as @xmath21 singly charged scalar --------------------- a singly charged scalar @xmath22 of an @xmath1-singlet with the hypercharge @xmath23 couples to the lepton doublet @xmath24 as @xmath25 = - 2 f_{\ell\ell^\prime}^ { } \bigl [ \overline{(\ell_l)^c}\ , \nu_{\ell^\prime l}^ { } s^+ \bigr ] , \label{eq : s - yukawa}\end{aligned}\ ] ] where the matrix of yukawa coupling constants is antisymmetric ( @xmath26 ) and @xmath27 are the pauli matrices . the scalar @xmath22 is introduced in e.g.the so - called zee model @xcite where light majorana neutrino masses are generated at the one - loop level . the simplest version of the zee model @xcite where there is no flavor - changing - neutral - current ( fcnc ) was excluded by neutrino oscillation measurements ( see e.g. ref .
@xcite ) . motivated by the original version of the zee model where the fcnc exists , we consider a majorana neutrino mass matrix @xmath28 of the following structure in the flavor basis : @xmath29 where @xmath30 is an arbitrary matrix and @xmath31 .
let us assume that contributions of @xmath32 and @xmath33 to @xmath34 are negligible . then eq .
is simplified as @xmath35 note that @xmath34 is a rank-2 matrix under this assumption although each term in the right - hand side of eq .
is a rank-1 matrix .
thus , the lightest neutrino becomes massless ( @xmath36 or @xmath37 ) while the other two have non - zero masses as required .
the @xmath34 includes four parameters : @xmath38 , @xmath39 , @xmath40 , and @xmath41 .
the latter three combinations can be expressed with neutrino mixing parameters by using conditions that @xmath7 diagonalizes the @xmath42 ( three off - diagonal parts must be zero ) .
since @xmath38 is an overall factor for neutrino mass eigenstates , the ratio of nonzero mass eigenstates does not depend on @xmath38 .
we see that @xmath43 ( the so - called inverted hierarchy where @xmath44 ) is allowed while @xmath36 ( the so - called normal hierarchy where @xmath45 ) can not be obtained .
when we use simple values @xmath46 and @xmath47 which are almost consistent with neutrino oscillation measurements , @xmath44 results in @xmath48 the `` predicted '' value @xmath49 seems reasonably agree with observations @xcite . see ref .
@xcite for more detailed analysis of the zee model .
partial decay widths @xmath50 for @xmath51 ( where neutrino species are summed ) are proportional to @xmath52 . in the discussion above
, @xmath39 is constrained by neutrino mixing parameters .
thus , we obtain the following `` prediction '' for a combination of branching ratios @xmath53 : @xmath54 if this relation is confirmed experimentally , neutrino mass matrix might be of eq . with parameters in eq . .
the singly - charged scalar @xmath22 is utilized also in the so - called zee - babu model @xcite where majorana neutrino masses are generated at the two - loop level .
another example is a model of loop - induced dirac neutrino masses @xcite .
neutrino mass matrix might be given by @xmath55 for majorana neutrinos ( similarly to the zee - babu model ) or @xmath56 = ( x f)_{i\ell } \bigl [ \overline{\nu_{ir}^{}}\ , \nu_{\ell l}^ { } \bigr ] \label{eq : s - dirac}\end{aligned}\ ] ] for dirac neutrinos ( similarly to the model in refs .
the matrix @xmath57 is symmetric while @xmath30 is arbitrary
. ratios of three elements of @xmath58 can be easily obtained as functions of neutrino mixing parameters @xcite ( see also refs . @xcite ) .
results are the same for eqs . and .
for @xmath46 and @xmath47 , we obtain @xmath59 if experiments confirm these ratios , the structure of the neutrino mass matrix in eq . or might be true .
neutrinos might obtain their dirac masses via a vacuum expectation value ( vev ) of an additional @xmath1-doublet scalar field @xmath60 @xcite .
we refer to the model as the neutrinophilic two - higgs - doublet model .
the yukawa interaction of neutrinos with @xmath60 is written as @xmath61 = ( y_\nu^{})_{i\ell}^ { } \bigl [ \overline{\nu_{ir}^{}}\ , \nu_{\ell l}^ { } \phi_\nu^0 \bigr ] - ( y_\nu^{})_{i\ell}^ { } \bigl [ \overline{\nu_{ir}^{}}\ , \ell_l \phi_\nu^+ \bigr ] , \end{aligned}\ ] ] where @xmath62 are right - handed components of mass eigenstates @xmath63 ( so , we do not regard @xmath62 as new particles here ) .
the mass matrix of the dirac neutrinos is simply given by @xmath64 .
the branching ratios @xmath65 , where neutrino species are summed , are proportional to @xmath66 .
figure [ fig : h+ ] ( taken from ref .
@xcite ) shows behaviors of these branching ratios with respect to the lightest neutrino mass ( @xmath67 for the left panel , and @xmath68 for the right one ) for the case @xmath69 decays only into leptons . by measuring @xmath70 mode ,
it would be possible to extract information on the value of the lightest neutrino mass and on whether @xmath71 ( left panel in fig . [ fig : h+ ] ) or not .
if experiments show that @xmath72 is very different from @xmath73 , the charged scalar might not contribute to the mechanism of generating neutrino masses in a simple way .
a singly charged scalar exists in an @xmath1-triplet field @xmath74 with @xmath23 , which can be expressed as @xmath75 the triplet scalar field interacts with the lepton doublet as @xmath76 = - h_{\ell\ell^\prime}^ { } \bigl [ \overline{(\ell_l)^c}\ , \ell_l^\prime \delta^{++ } \bigr ] - \sqrt{2}\ , h_{\ell\ell^\prime}^ { } \bigl [ \overline{(\nu_{\ell l}^{})^c}\ , \ell_l^\prime \delta^+ \bigr ] + h_{\ell\ell^\prime}^ { } \bigl [ \overline{(\nu_{\ell l}^{})^c}\ , \nu_{\ell^\prime l}^ { } \delta^0 \bigr ] , \end{aligned}\ ] ] where the yukawa coupling constants satisfy @xmath77 .
the vev of @xmath78 can generate neutrino masses @xcite as @xmath79 hereafter , we refer to this solo mechanism of generating neutrino masses as the higgs triplet model . branching ratios @xmath80 in the higgs triplet model
are proportional to @xmath81 identically to those in the neutrinophilic two - higgs - doublet model . therefore , the discussion in the previous subsection is applicable also for @xmath82 .
see e.g. fig . 16 in ref .
@xcite to compare with fig .
[ fig : h+ ] in this article .
if non - leptonic decays ( e.g. @xmath83 ) are not negligible , a ratio of branching ratios of @xmath70 and @xmath84 modes would be reliable .
an @xmath1-singlet scalar @xmath85 with @xmath86 has the following yukawa interaction : @xmath87 , \end{aligned}\ ] ] where the yukawa coupling constants satisfy @xmath88 .
the scalar @xmath85 is introduced in e.g.the zee - babu model @xcite where an @xmath1-singlet scalar @xmath22 ( see also sections [ subsec : singly - singlet - i ] and [ subsec : singly - singlet - ii ] ) is also introduced . the yukawa interaction with @xmath22
is shown in eq . . motivated by the zee - babu model , let us take a case in which the structure of neutrino mass matrix @xmath89 is given by @xmath90_{\ell\ell^\prime } , \label{eq : mnu - zbm}\end{aligned}\ ] ] where @xmath91 . discussion in this subsection is based on refs .
@xcite where the zee - babu model was studied .
the lightest neutrino becomes massless ( @xmath36 or @xmath37 ) because of @xmath92 .
we simply assume that @xmath93 in eq .
can be ignored so that we can have some `` prediction '' on the flavor structure of branching ratios @xmath94 which are proportional to @xmath95 .
then , because of a large mixing angle @xmath96 , we expect @xmath97 which results in @xmath98 .
the branching ratios become @xmath99 if eq .
turns out to be consistent with measurements , the neutrino mass matrix might be the structure in eq . where @xmath36 or @xmath37 is predicted . a doubly charged scalar exists also in the higgs triplet model ( see also sec . [ subsec : singly - triplet ] )
let us take a scenario that @xmath100 dominantly decays into a pair of same - signed leptons : @xmath101 . then branching ratios of leptonic decays are determined by @xmath102 , and the flavor structure of the branching ratios
can provide direct information on the neutrino mass matrix @xcite .
for example , if branching ratios @xmath103 and @xmath104 are observed in the shaded region in fig .
[ fig : htm ] , it would be excluded that the lightest neutrino mass is zero in this model @xcite .
note that information on the lightest neutrino mass can not be obtained by neutrino oscillation measurements .
we discussed relations between the neutrino mass matrix and the flavor structure of decays of leptophilic charged scalars . by assuming how a matrix of yukawa coupling constants for a leptophilic scalar appears in the neutrino mass matrix
, we obtained predictions on the leptonic decays of the scalar . if the antisymmetric matrix @xmath105 of yukawa coupling constants for an @xmath1-singlet singly - charged scalar @xmath22 appears in the neutrino mass matrix as @xmath106 , a combination of branching ratios for @xmath107 satisfies @xmath108 . for @xmath46 and @xmath47
, we obtained @xmath49 , @xmath109 , and @xmath110 . if the matrix @xmath105 appears as @xmath111 or @xmath112 , we predicted @xmath113 for @xmath36 and @xmath114 for @xmath43 .
on the other hand , if a leptophilic singly - charged scalar is a member of @xmath1-doublet or triplet , a branching ratio for a decay into @xmath115 ( and @xmath116 ) would provide information on the value of the lightest neutrino mass and on whether @xmath71 or not .
the symmetric matrix @xmath117 of yukawa coupling constants for an @xmath1-singlet doubly - charged scalar @xmath85 might contribute to the neutrino mass matrix as @xmath118 .
when we assume that a contribution of @xmath93 to @xmath42 is negligible and @xmath97 , decays of @xmath85 into @xmath119 and @xmath120 become negligible in comparison with the @xmath121 mode . for the case of @xmath100 in an @xmath1-triplet field , the flavor structure of @xmath101 directly relates to the neutrino mass matrix .
we showed that information on the lightest neutrino mass ( and majorana phases etc . )
could be obtained by observing the structure of the @xmath100 decays .
we hope that the mechanism of the neutrino mass generation is uncovered by discovery of such leptophilic scalars at collider experiments .
peaks of their signals may remind us of the tateyama peaks in toyama !
i thank a.g .
akeroyd , m. aoki , s. kanemura , and t. nabeshima for collaboration on which this article is partially based .
i also thank members of the higgs working group and participants in this workshop for valuable discussions .
this work was supported in part by the grant - in - aid for young scientists ( b ) no .
23740210 .
99 b. t. cleveland _ et al . _ ,
astrophys .
j. * 496 * , 505 ( 1998 ) ; w. hampel _ et al .
_ [ gallex collaboration ] , phys .
b * 447 * , 127 ( 1999 ) ; j. n. abdurashitov _ et al . _ [ sage collaboration ] , phys .
c * 80 * , 015807 ( 2009 ) ; k. abe _ et al .
_ [ super - kamiokande collaboration ] , phys . rev .
d * 83 * , 052010 ( 2011 ) ; g. bellini
_ et al . _
[ borexino collaboration ] , phys .
lett . * 107 * , 141302 ( 2011 ) ; b. aharmim _ et al .
_ [ sno collaboration ] , arxiv:1109.0763 [ nucl - ex ] .
m. h. ahn _
[ k2k collaboration ] , phys .
d * 74 * , 072003 ( 2006 ) ; p. adamson _ et al .
_ [ minos collaboration ] , phys . rev .
lett . * 106 * , 181801 ( 2011 ) ; k. abe _ et al . _
[ t2k collaboration ] , phys .
d * 85 * , 031103 ( 2012 ) .
m. apollonio _ et al . _
[ chooz collaboration ] , eur .
j. c * 27 * , 331 ( 2003 ) ; y. abe _ et al . _
[ double - chooz collaboration ] , phys .
lett . * 108 * , 131801 ( 2012 ) ; f. p.
an _ et al . _
[ daya - bay collaboration ] , phys .
lett . *
108 * , 171803 ( 2012 ) ; j. k. ahn _
[ reno collaboration ] , phys .
lett . * 108 * , 191802 ( 2012 ) .
p. minkowski , phys .
b * 67 * , 421 ( 1977 ) ; t. yanagida , conf .
c * 7902131 * , 95 ( 1979 ) ; prog .
* 64 * , 1103 ( 1980 ) ; m. gell - mann , p. ramond and r. slansky , conf .
c * 790927 * , 315 ( 1979 ) ; r. n. mohapatra and g. senjanovic , phys . rev .
* 44 * , 912 ( 1980 ) .
w. konetschny and w. kummer , phys .
b * 70 * , 433 ( 1977 ) ; m. magg and c. wetterich , phys .
b * 94 * , 61 ( 1980 ) ; t. p. cheng and l. f. li , phys .
d * 22 * , 2860 ( 1980 ) ; g. lazarides , q. shafi and c. wetterich , nucl . phys .
b * 181 * , 287 ( 1981 ) . | this is a short review about relations between new scalars and mechanisms to generate neutrino masses .
we investigate leptohilic scalars whose yukawa interactions are only with leptons .
we discuss possibilities that measurements of their leptonic decays provide information on how neutrino masses are generated and on parameters in the neutrino mass matrix ( e.g. the lightest neutrino mass ) . |
living at the intersection of spirals and ellipticals on the hubble tuning fork , lenticular ( or s0 ) galaxies are often thought to be remnants of spirals after star formation ( sf ) has ceased ( or has slowed down substantially ) .
this transition in the sf rate ( sfr ) must be directly related to the change of cold gas content and/or density in the disk of such a galaxy .
the change may be caused by such processes as ( e.g. , martig et al .
2009 and references therein ) : ram - pressure stripping ( of cold gas ) , termination or strangulation ( removing surrounding hot gas , hence a source of new cold gas ) , morphological quenching ( heating of a stellar disk or even transforming to a spheroid ) , and exhaustion ( consuming gas via sf ) .
while the ram - pressure stripping of the cold or hot gas is expected to work primarily in a dense environment ( in galaxy clusters and possibly in rich groups ) , the morphological quenching via galaxy - galaxy mergers should be most effective in groups .
the exhaustion , an internal process , can occur in both rich and poor environments .
for example , this process may occur for a relatively isolated galaxy when its halo mass @xmath9 grows to be @xmath10 and its gas accretion mode turns from cold to hot ( e.g. , kere et al . 2005 ; kere & hernquist 2009 and references therein ) . in general , more than one of these processes may be involved in the transition of a galaxy .
their relative importance can not be determined easily ; great uncertainty remains in our understanding of the processes , especially in terms of their interplay with various other galactic activities such as the feedback from stars and agns . while observational studies of galaxy transformation have been concentrated on stellar properties , theoretical efforts ( or computational simulations )
are primarily on gaseous components on galaxy scales .
this gap must be bridged to make significant progress in the field . here
we present a case study of the nearby edge - on s0 galaxy ngc 5866 ( also referred as m 102 or `` spindle '' galaxy ; see table [ table : basicpara ] for the salient parameters ) .
our goal is to understand the potential role of the stellar energy and mass feedback in regulating the evolution of various gaseous components in this relatively isolated galaxy .
ngc 5866 belongs to the poor group lgg 396 with a velocity dispersion of @xmath11 ( e.g. , mulchaey et al .
two gas - rich companions , ngc 5907 ( sc ) and ngc 5879 ( sbc ) , are known to be located several hundred kpc away ; their interactions with ngc 5866 should be minimal .
so we can focus on internal activities of the galaxy .
as will be discussed later , this isolated environment in the `` field '' may explain the relatively richness of molecular gas in the galaxy ( table [ table : basicpara ] ; welch & sage 2003 ; sage & welch 2006 ) and its slow evolution with ongoing low - level sf many gigayears after a major merger .
there is also no evidence for strong nuclear activity in the galaxy ( terashima & wilson 2003 ) ; the nucleus is classified in optical to be an intermediate object between a weak region and a liner ( ho , filippenko & sargent 1997 ) .
therefore , we can examine the effect of the stellar feedback with little confusion from other potential sources of mechanical energy inputs .
furthermore , the nearly complete edge - on inclination of the galaxy is optimal for the study of the galactic disk / halo interaction . indeed ,
archival _ hst _ optical images ( first studied by cantiello , blakeslee & raimondo 2007 for globular clusters ) show numerous off - disk cold gas spurs , in addition to a well - defined dust lane . to our knowledge
, no characterization of such spurs has ever been made for any s0 galaxies , let alone their nature and role in the galactic gaseous evolution .
a previous x - ray study based on _ rosat _ observation revealed that ngc 5866 is an x - ray faint galaxy and that the x - ray emission is dominated by stellar sources ( pellegrini 1994 ) .
there was also a study based on a snap - shot ( 2 ks ) observation made with _
chandra _ to see if an agn is present in the galaxy ( terashima & wilson 2003 ) ; only an upper limit of @xmath12 was obtained , assuming the distance of the galaxy to be @xmath13 ( pellegrini 2005 ) .
our study is based on a later 34 ks _ chandra _ observation ( previously studied by david et al .
2006 in a sample of early - type galaxies ) , which allows us to resolve out much of the discrete source contribution and to examine the potentially diffuse x - ray emission from the galaxy . in short , with various available tracers ( optical absorption as well as h@xmath3/p@xmath3 , infrared ( ir ) and soft x - ray emissions )
, we can now characterize the cold molecular / atomic , warm ionized , dusty and hot phases of the interstellar medium ( ism ) and explore their relationships with the stellar feedback in the galaxy .
the organization of the paper is as follows . in [ sec : data ] , we briefly describe various data sets used for our examinations ( mainly , , and ) .
spatial and spectral analysis of these multi - wavelength data is presented in [ sec : analysis ] .
we discuss our results and their implications for understanding the energetics and dynamics of the gas phases in [ sec : discussion ] .
we summarize the results and conclusions in [ sec : summary ] .
the _ chandra _ observation of ngc 5866 was taken on nov .
14 , 2002 .
we reprocess the archived data for our study . after the removal of background flares ,
a total effective exposure of 24.3 ks remains . a `` stowed background '' , normalized by the count rate in the 10 - 12 kev range ,
is used to account for the non - x - ray event contribution .
our study is based on the data from the acis - s3 chip , although part of the s2 chip data is also used for our source detection , which is performed in the broad ( b , @xmath14 ) , soft ( s , @xmath15 ) and hard ( h , @xmath16 ) bands ( fig .
[ fig : pointsrc ] ) , following the procedure detailed in wang ( 2004 ) . to study the diffuse x - ray emission , we remove detected discrete sources from the data .
circular regions are excluded within twice the 90 per cent energy enclosed radius ( eer ) around each source of a count rate @xmath17 . for brighter sources ,
the source removal radius is further multiplied by a factor of @xmath18 .
generally about 96 per cent of the source counts are excluded in such a removal . for our spectral analysis of the diffuse x - ray emission ,
we further remove a local sky background .
this background spectrum is first extracted from the source - removed data in a region surrounding the galaxy ( fig .
[ fig : pointsrc ] ) .
the spectrum , with the stowed background subtracted , is modeled with a thermal plasma model ( mekal in xspec ) and a power - law ( with a photon index of 1.4 , lumb et al .
2002 ) , representing both the galactic and extragalactic background contributions .
this local sky background model , normalized for the relative area coverage , is to be subtracted from on - source spectral data .
we use _ hst _ acs ( pixel size@xmath19 ) and nicmos ( pixel size@xmath20 ) observations in both broad and narrow bands . an h@xmath3 image is constructed from the acs on - line narrow band ( f658n ) and off - line broad band ( f625w ) observations .
the off - line image , used for the continuum subtraction , is normalized by a factor of 0.09 to account for the bandwidth difference , which is determined by the average flux ratio of several bright foreground stars in the field ( thought to have no h@xmath3 line emission ) .
we further construct a p@xmath3 image from nicmos data .
the f190n continuum observation is directly subtracted from the on - line f187n observation because the two bands have approximately the same bandwidth ; there is no bright foreground stars found in the small field of view in the nicmos observations .
we further use archival _ spitzer _ irac 3.6 @xmath21 , 8 @xmath22 and mips 24 @xmath21 images as well as 2mass near - ir data .
the 24 @xmath21 image is used to trace the dust emission , the stellar - light - subtracted 8 @xmath21 image is used to trace the pah emission , while the 3.6 @xmath21 image and the 2mass images are used to trace the emission from old stars .
we make an aperture correction for the _ spitzer _ images , following dale et al .
( 2007 ) . to subtract the old stellar contribution from the 8 @xmath23 emission ,
we adopt a ratio of 4.23 between the stellar light in 3.6 @xmath21 and 8 @xmath21 , since 3.6 @xmath21 emission in galaxies are dominated by old stellar light ( hunter et al .
the pixel size of irac ( @xmath24 ) is smaller than that of mips 24 @xmath21 image ( @xmath25 ) , so we rebin the 24 @xmath21 image to have the same pixel scale as the 3.6 @xmath21 image ( but note that the resolution of mips at 24 @xmath21 is only @xmath26 ) , and plot them in fig .
[ fig : tricolorimgs]c . the 2mass near - ir image ( with a spatial resolution of @xmath27 ) is also dominated by old stellar light in galaxies , and the j - k color , as detailed in [ subsec : sf ] ,
is adopted as a tracer of the intrinsic extinction near the disk .
we present the acis - s intensity images of in fig . [
fig : pointsrc ] ( a and b ) . in particular
, the morphology of the source - removed emission appears substantially rounder than that of the stellar bulge ( fig .
[ fig : pointsrc]b ) , clearly indicating the presence of non - stellar diffuse x - ray sources around the galaxy , probably hot gas .
the emission is further compared to the optical and ir images in fig .
[ fig : tricolorimgs ] .
the dusty disk is prominent in the optical images as absorption feature , with a diameter of @xmath28 ( @xmath29 ) , less extended than the stellar disk which is also presented in the optical images ( fig .
[ fig : tricolorimgs]a ) .
the dusty disk is slightly tilted from the stellar disk at the northwestern edge .
in addition to the prominent dust lane , there are also lots of dusty spurs evidenced as extinction features in the optical images ( fig .
[ fig : tricolorimgs]a ) .
diffuse soft x - ray emission is found primarily in the bulge , while the h@xmath3 emission is strongest just off the dust lane ( fig .
[ fig : tricolorimgs]b ) .
pah and 24 @xmath21 emission is concentrated in the disk ( fig .
[ fig : tricolorimgs]c ) , in similar fashion as the extinction materials .
but there is also a bright core on the 24 @xmath21 image ( fig .
[ fig : tricolorimgs]c ) , probably a nuclear sf region or a weak agn .
[ fig : multiprof ] gives a more quantitative comparison of these various intensity distributions along the minor axis of the galaxy .
it is clear that optical and soft x - ray intensities drop significantly toward the galactic plane within an off - plane distance of @xmath30 ( @xmath31 , assume a distance of 11.21 mpc as listed in table [ table : basicpara ] throughout the paper ) , due to the extinction and absorption by dusty gas in the galactic disk .
we define distinct features outside this characteristic distance ( @xmath32 ) as `` extraplanar '' .
assume the diameter of the dusty disk to be @xmath28 ( @xmath29 ) and adopt the inclination angle in table [ table : basicpara ] , the projected vertical extension of the dusty disk is @xmath33 ( @xmath34 ) , so the region with @xmath35 ( @xmath36 ) is out of the dusty disk , comparable to our definition of the extraplanar region . in the following
we detail the analysis and results from the data sets .
we use the optical and infrared images to characterize the interstellar extinction as well as the cold and warm gas components in ngc 5866 ( figs .
[ fig : pointsrc]b , [ fig : tricolorimgs ] ) . fig .
[ fig : multiimg ] presents multi - wavelength images of the central region of the galaxy .
an unsharp - masked ( howk & savage 1997 ) b - band image is constructed from the _ hst _ acs f435w image .
it is first smoothed with a median filter of a @xmath37 box and is then subtracted from the original image .
this unsharp - masked image could emphasize structures on scales smaller than the smoothing box .
as shown in figs .
[ fig : multiimg]a and [ fig : dustfeature ] , it highlights the filamentary structure of the extinction .
in addition to the prominent central dust lane along the galactic disk , various prominent dusty spurs are located typically within @xmath38 ( @xmath39 ) from the galactic plane .
there is a shell - like dusty feature near the nucleus , with a diameter of @xmath40 ( @xmath0 ) ( fig .
[ fig : multiimg]a ; i.e. , feature `` 04 '' in fig .
[ fig : dustfeature ] ) .
the unsharp - masked image also shows more extended , but weaker extinction features in regions farther away from the mid - plane ( up to @xmath2 ( @xmath41 ) scale ; e.g. , feature `` 16 '' in fig .
[ fig : dustfeature ] ) . for a distinct small - scale feature , we may estimate its extinction locally .
the extinction at a particular wavelength @xmath42 can be expressed as @xmath43 where @xmath44 and @xmath45 are the intensities on and off the feature ( see also howk & savage 1997 ) .
this estimation assumes that all the stellar light arises behind the feature , which is reasonable for the stellar bulge with a centrally concentrated light distribution .
but in general , @xmath46 should be considered as a lower limit to the true extinction @xmath47 .
we adopt the relation between the total neutral hydrogen column density and the color excess : @xmath48 ( bohlin , savage & drake 1978 ) , assuming @xmath49 and an extinction curve to convert @xmath47 to @xmath50 ( cox 1999 ) .
we further assume that the depth of a dust feature is similar to its width and that the hydrogen to total mass conversion factor is 1.37 , appropriate for gas of solar abundance .
the resultant estimates of the hydrogen column density , number density and total mass are presented in table [ table : dustfeature ] .
a typical spur has mass of @xmath51 .
we estimate that the mass of the extraplanar cold gas is @xmath52 in total , much smaller than the total cold gas mass estimated from co observations ( @xmath6 , table [ table : basicpara ] ) .
thus most of the cold gas must have a smoother distribution , concentrated in the disk .
to characterize the global extinction as a function of the vertical distance @xmath53 , we use the b - v color calculated with the _ hst _ f435w and f555w images .
the optical color can not be an exact measure of the absorption column density when the intrinsic color is unknown or in an optical thick case .
here we just use the color to show the general distribution of the dusty gas and to give a crude estimate of the vertical scale height of the cold gas .
as described above , @xmath54 , so it can be written as @xmath55 , where @xmath56 and @xmath57 are the v- and b - band surface brightness , @xmath58 , @xmath59 is a parameter determined by the intrinsic color of the stellar population . assuming a constant intrinsic color , the cold gas distribution is determined by the term @xmath60 , which is plotted in fig .
[ fig : multiprof]b and is fitted with an exponential model for the extraplanar region with @xmath61 ( @xmath31 ) .
the scale height of the cold gas so obtained is @xmath0 .
the flatten of the @xmath62 profile in the central region with @xmath63 ( @xmath31 ) is due to the tilt of the disk , that is why we have defined regions with @xmath61 ( @xmath31 ) as `` extraplanar '' .
the innermost trough of the @xmath62 profile is caused by the extinction saturation in b - band .
warm ionized gas , as traced by the h@xmath3 or p@xmath3 emission , shows a relatively smooth morphology and extends to @xmath64 ( @xmath38 ) from the mid - plane ( e.g. , fig .
[ fig : multiprof]b , fig .
[ fig : multiimg]b , c ) .
although its overall distribution is similar to that of the dusty cold gas , there is little obvious feature - to - feature correlation .
indeed , some of the dusty features appear in regions with weak h@xmath3 emission ( e.g. , feature `` 02 '' in fig . [
fig : dustfeature ] ) .
we fit the vertical distribution of the h@xmath3 emission , instead of p@xmath3 ( because of the small field coverage of the nicmos ) , in the region with @xmath61 ( @xmath31 ) ( fig .
[ fig : multiprof]b ) .
the fitted exponential scale height is @xmath65 .
since the h@xmath3 intensity is proportional to @xmath66 , the scale height of the warm gas density should be @xmath0 , comparable to the scale height of the optical absorption features ( or cold gas ) .
as the h@xmath3 emission suffers stronger extinction than the p@xmath3 emission , we use the p@xmath3 luminosity to estimate the recombination rate of the warm ionized gas .
the total observed p@xmath3 luminosity in the region represented by the second x - ray data point at @xmath67 ( @xmath68 ) ( fig .
[ fig : multiprof]b ) is @xmath69 , accounting for the uncertainty in the normalization factor of the continuum subtraction ( [ sec : data ] ) and corrected for the extinction .
this luminosity corresponds to a recombination rate of @xmath70 .
we detect 15 point sources within the bulge region ( fig .
[ fig : pointsrc ] ) .
in particular , source 30 coincides with the optical center of the galaxy ( fig .
[ fig : pointsrc]c ) .
this source , detected only in the 1.5 - 7 kev band , has a count rate of @xmath71 ( table [ table : acis_source_list ] ) .
assuming a typical power law spectrum with a photon index of 1.7 , we estimate the corresponding 1.5 - 7 kev luminosity as @xmath72 . the apparent strong soft x - ray absorption is consistent with the nucleus being behind the dusty disk . in any case
, the low luminosity of the nucleus indicates that it is in a relatively quiescent state and probably has little impact mechanically on the global ism of the galaxy .
the spectrum of other sources outside the disk ( fig .
[ fig : pointsrc]b ) , the bulk of which should be lmxbs in the galactic bulge , can be represented by a power law of a photon index @xmath73 and a very low absorption , consistent with the galactic foreground absorption value .
their integrated intrinsic luminosity is @xmath74 in the 0.5 - 2 kev band . by removing the detected sources
, we study the apparent diffuse emission . to isolate the truly diffuse hot gas in the galaxy , however , we need to quantify remaining stellar contributions : the emission from cataclysmic variables ( cvs ) and coronal active binaries ( abs ) as well as the emission from unresolved lmxbs and the residual photons out of the discrete source removal circles ( [ sec : data ] ) .
we estimate the cv and ab contribution , @xmath75 , from the calibrated ratio between the 0.5 - 2 kev luminosity and the bulge stellar mass ( @xmath76 ; revnivtsev et al . 2008 ; refer to [ subsec : sf ] for the estimation of the stellar mass ) .
the contributions in other bands are calculated with a spectral model consisting of a 0.5 kev _ mekal _ optically - thin thermal plasma and a @xmath77 power law ( revnivtsev et al .
the residual emission from the source removal is @xmath78 of the detected source luminosity .
the emission from unresolved lmxbs below our detection limit ( @xmath79 in the 0.5 - 8 kev band ) is estimated from the luminosity function of gilfanov ( 2004 ) , assuming the same spectrum as that for the detected sources .
we then find that the total residual lmxb contribution is @xmath80 in 0.5 - 2 kev band .
this contribution is confirmed independently from the spectral analysis of the diffuse emission , in which the normalization of the residual lmxb contribution is fitted ( the luminosity of this component is fitted to be @xmath81 in 0.5 - 2 kev band , see below ) .
these stellar contributions are added together and are assumed to have the same spatial distribution as the irac @xmath82 intensity of the galaxy .
they are then subtracted from the unresolved x - ray emission in spatial ( fig .
[ fig : multiprof]a ) and spectral ( see below ) analysis .
[ fig : multiprof]a compares the intensity distributions of the stellar contributions with the actually observed diffuse x - ray emission in the 0.5 - 1.5 kev and 1.5 - 7 kev bands .
the two data points at the galactic plane in fig .
[ fig : multiprof]a shows large deviations from the global distribution of the 0.5 - 1.5 kev intensity profile .
these deviations are clearly due to the absorption of soft x - ray by the cold gas disk .
the stellar contribution as traced by near - ir emission matches well the observed 1.5 - 7 kev emission , indicating a stellar origin of the hard x - ray emission . using the stellar contribution in 1.5 - 7 kev band as fitted and the spectral model of stellar sources adopted in the x - ray spectral analysis , we calculate the stellar contribution in 0.5 - 1.5 kev band ( the red solid line in fig .
[ fig : multiprof]a ) .
the excess of the observed 0.5 - 1.5 kev emission above the stellar contribution clearly shows the presence of the truly diffuse hot gas widespread in the galactic bulge , with a luminosity of @xmath83 in a region with @xmath84 ( roughly the bulge region shown in fig .
[ fig : pointsrc ] ) . fig .
[ fig : bulgespec ] presents the spectrum of the diffuse x - ray emission from the bulge , together with our best - fit model .
in addition to the inclusion of the stellar contributions as described above , we model the hot gas emission as an optically - thin thermal plasma ( _ vmekal _ ) .
the fitted temperature of the plasma is @xmath85 .
the fe and o abundances are @xmath86 and @xmath87 solar .
the inferred intrinsic luminosity of the hot gas @xmath88 in the 0.5 - 2 kev band is consistent with the above estimation based on the spatial distribution .
the corresponding emission measure of the hot gas is @xmath89 .
note that the exact value of the metal abundance is somewhat uncertain and is model - dependent due to the limited counting statistics and spectral resolution of the spectrum , but the assumption of isothermality typically leads to an under - estimate of the metal abundance .
therefore , the x - ray spectrum strongly indicates supersolar iron abundance , consistent with significant enrichment by type ia sne .
based on the above multi - wavelength characterizations of the stellar and interstellar components of , we now discuss their interplays , focusing on the physical state , dynamics , and evolution of the extraplanar gas . but first let us characterize the sf history of the galaxy , important for estimating the stellar mass and energy feedback . for an s0 galaxy , we may assume that the bulk of stars is formed long ago during or before the last major merger that presumably forms the galactic bulge .
one way to estimate the age of this merger is to use the globular cluster ( gc ) specific frequency ( @xmath90 ) , the number of gcs per unit luminosity in a given galaxy .
gcs are believed to form in starbursts triggered by major mergers of gas - rich galaxies .
assuming that the evolution after mergers preserves the gc number , while the overall stellar luminosity of the galaxy may decrease with time .
therefore , @xmath90 can be used as an age estimate for an s0 galaxy after its last merger and accompanied starburst ( barr et al .
2007 ) . with @xmath91 ( cantiello , blakeslee & raimondo 2007 )
, we estimate the age of the ngc 5866 bulge to be @xmath92 , which is typical for s0 galaxies ( barr et al .
2007 ) .
we now estimate the stellar mass of ngc 5866 . using the _ 2mass _ j- and k - band images , we get the j- and k - band intensity distributions .
we further assume an optically thin case for near - ir emission , and an intrinsic color j - k=0.90 , suitable for old stars in early type galaxies ( jarrett et al .
we then map out the extinction value in different parts of the galaxy . using the extinction - corrected k - band luminosity and adopting a color - dependent stellar mass - to - light ratio as detailed in bell & de jong ( 2001 )
, we estimate the masses of the galactic bulge and disk regions ( fig .
[ fig : pointsrc]b ) as @xmath93 and @xmath94 .
there are lines of evidence for ongoing sf in the disk of .
the most direct evidence comes from the observed 24 @xmath95 emission , which is usually thought to be mainly produced by hot dust heated by ultraviolet radiation , and is observed to concentrate in the cold gas disk in ( fig .
[ fig : tricolorimgs]c ) .
such hot dust is found in both sf regions and circum - stellar material around evolved stars .
the latter contribution , tightly correlated with the k - band luminosity in elliptical and s0 galaxies ( e.g. , temi , brighenti & mathews 2007 , 2009 ) , is estimated to be @xmath96 for , significantly smaller than the total observed 24 @xmath95 luminosity of @xmath97 ( excluding the nucleus which may be related to the agn ) .
this excess of @xmath98 predicts an ongoing sf rate of @xmath99 in the galactic disk .
here we have assumed that the relationship , sfr @xmath100^{0.8850}$ ] , obtained for active star forming galaxies ( calzetti et al .
2007 ) , applies to ngc 5866 .
the low level of sf inferred here is consistent with the constraint ( @xmath101 ) obtained from the model fit to the observed sed of ngc 5866 over the @xmath102 range ( draine et al .
this means the sfr estimated using the mid - ir emission is reliable within the errors for ngc 5866 , and further indicates that the mips 24 @xmath95 luminosity is mainly produced by the weak sf process in the disk .
the ongoing sf also explains the ionizing photon flux needed for producing the p@xmath3 luminosity .
the relationship @xmath103 ( rieke et al . 2009 ) gives a p@xmath3 luminosity of @xmath104 , fully consistent with the observed value , @xmath105 , which corresponds to a recombination rate of @xmath70 ( [ sec : coolgas ] ) .
this may be a lower limit , since we may have still missed part of the p@xmath3 emission from the disk , due to its strong extinction .
alternatively , could the ionizing radiation be supplied by evolved stars ? we use the ionizing radiation with a rate of @xmath106 for post - agb stars ( binette et al .
1994 ) and @xmath107 for hot horizontal branch stars ( han et al .
2007 ) . adopting the total stellar mass of the galaxy , @xmath108 as inferred in
[ sec : coolgas ] , we estimate that the total expected hydrogen recombination rate for the galaxy is @xmath109 , where @xmath110 is the fraction of the evolved ionizing photons intercepted by the gas .
while most of the stars are in the galactic bulge , @xmath110 should be considerably smaller than one .
therefore , the evolved stars alone seem to be insufficient to provide the ionizing photons . furthermore , observed various emission lines such as [ ] @xmath111 and [ ]
@xmath112 appear to be considerably stronger than what are expected from ionizing photons purely from the evolved stars ( malhotra et al .
2000 , 2001 ) .
therefore , we conclude that the presence of low - level sf provides a unified explanation for the radiative heating and ionization of the dust and warm - ionized gas in .
hot gas apparently dominates the extraplanar space . adopting the spectral fit parameters in [ sec : hotgas ] and approximating the x - ray - emitting region as a cylindric volume with the radius @xmath113 ( @xmath114 ) and the hight @xmath115 ( @xmath116 ) ( fig .
[ fig : pointsrc]b ) , we estimate the mean density , thermal pressure , and radiative cooling time scale of the hot gas as @xmath117 , @xmath118 , and @xmath119 . the total hot gas mass is @xmath120 .
the estimation of these parameters are all based on the most reliable spectral fit parameters , i.e. , the temperature , the luminosity and the emission measure of the thermal component ( [ sec : hotgas ] ) .
the most natural source of the gas that is subsequently heated is the mass loss of evolved stars .
assuming the mass loss by each planetary nebula ( pn ) is @xmath121 ( faber & gallagher 1976 ) , and using the pn birth rate in our galaxy @xmath122 ( peimbert 1993 ) , the mass loss rate ( @xmath123 ) is @xmath124 . adopting the stellar mass estimated in
[ subsec : sf ] and the age of @xmath92 ( [ subsec : sf ] ) , the total mass return in ngc 5866 is then @xmath125 , which is substantially greater than what is observed in the hot gas .
furthermore , the mass - averaged fe abundance as estimated from the mass loss of the evolved stars and metal enrichment of the type ia sn ejecta is about 6 times the solar value ( grevesse & suval 1998 ) , much larger than our spectral fit result ( [ sec : hotgas ] ) .
one possibility of this relatively low observed fe abundance is that the bulk of the fe - rich sn ejecta is being blown out of the bulge before being fully mixed with the materials from the stellar mass loss ( tang et al .
2009 ) . to see
if the heating provided by type ia sne is sufficient to explain the x - ray cooling luminosity , we compare the x - ray luminosity with the sn heating rate . the supersolar iron abundance , as inferred from the spectral analysis of the hot gas , is consistent with heating primarily by type ia sne .
assuming that each sn releases @xmath126 and using the empirical type ia sn rate of mannucci et al .
( 2005 ) , which can be approximately expressed as @xmath127 , and the stellar mass @xmath128 ( [ subsec : sf ] ) , we estimate the total type ia sn rate of @xmath129 and the mechanical energy injection rate as @xmath130 .
this energy injection rate is nearly two orders of magnitude greater than the x - ray luminosity !
while a fraction of the energy injected into the cool gas disk may be radiated in other energy bands ( e.g. , extreme- and far - uv ) , avoiding a detection so far , the general lack of cool gas in the galactic bulge makes it hard to hide and convert the mechanical energy .
therefore , there is a missing stellar feedback problem in ngc 5866 , similar to situations in other low - to - intermediate mass early - type galaxies ( e.g. li , wang & hameed 2007 ; li , & wang 2008 ; david et al . 2006 ) .
this missing stellar feedback indicates the presence of an outflow from the galactic bulge .
the dynamics of this outflow depends critically on the mass - loading from the cool gas disk , which is indicated by the observed low temperature of the hot gas in .
the specific enthalpy of the stellar feedback , @xmath131 per particle , would give an average gas temperature of @xmath132 at the bulge center , which is substantially greater than @xmath133 inferred from the x - ray spectrum of the bulge ( [ sec : hotgas ] ) .
in the inner bulge , the kinetic energy is often negligible compared to the thermal energy of the hot gas ( tang et al .
2009 ) , so it can not account for this large temperature difference .
part of the difference can be attributed to the sn energy stored in other forms ( possibly magnetic field and cosmic rays as well as turbulent motion ) , the equal - partition among which may account for a factor up to 3 - 4 .
the emission measure - weighted temperature , as in the x - ray spectral analysis , also tends to under - estimate the real gas temperature by a factor @xmath134 ( tang et al .
at least an additional factor of 2 - 3 is needed to explain our measured low temperature .
this factor can be naturally explained by the mass - loading from the cool gas disk , as evidenced by the presence of extraplanar warm - ionized gas and dusty spurs , which represent the interface between the hot gas in the bulge and the cool gas in the disk .
assuming a temperature of @xmath135 and an approximate pressure balance with the hot gas , we estimate the parameters of the warm ionized gas based on the observed p@xmath3 emission measure . the volume filling factor , number density and total mass of the warm gas estimated this way
are @xmath136 , @xmath137 and @xmath138 .
the total mass is roughly consistent with the value of @xmath7 as listed in table [ table : basicpara ] .
however , the number density is too high compared to that in other nearby galaxies ( e.g. collins et al . 2000 ) , indicating that the warm and hot gas may be not in a pressure balance state , or the warm ionized gas may present as skins of some unresolved cool gas filaments .
similar dusty spurs and their correlation with the extraplanar ionized gas have been observed previously in nearby edge - on _ active star - forming _ galaxies ( e.g. , howk & savage 1999 ; li et al .
2008 ; rossa et al .
in particular , howk & savage ( 1997 ) have presented several mechanisms for the formation of extraplanar dusty features observed in ngc 891 .
these mechanisms include hydrodynamical processes like galactic fountain , radiation pressure driven outflow , and magnetic instability presumably induced by the differential rotation of galactic gaseous disks .
the galactic fountain , the radiation pressure , and some magnetic instabilities like the parker instability are closely related to recent star forming activities .
this is consistent with an apparent correlation between the amount of extraplanar gas and the sf rate of galaxies ( lehnert & heckman 1996 ) .
the heights and masses of the extraplanar cold gas features in ngc 5866 are significantly smaller than that in late - type galaxies ( howk & savage 1999 ) .
thus the processes responsible for producing the features are probably less energetic and would be overwhelmed in an active star forming galaxy .
the very detection of the extraplanar extinction features in ngc 5866 is largely due to its nearly perfect edge - on perspective and its prominent stellar bulge background light .
to check out the mechanical energy requirement for the extraplanar extinction features , we estimate their gravitational potential energies .
following howk & savage ( 1997 ) , we estimate the energy @xmath139 of a feature at height @xmath53 from the mid - plane as : @xmath140\rm~ergs\ ] ] where @xmath141 is the mass of the cold gas feature , @xmath142 is the scale height of the mass density and is taken to be @xmath143 ( @xmath38 ) , and @xmath144 is the mass density at the mid - plane derived from the mass of the stellar disk ( [ subsec : sf ] ) and its size .
the estimated gravitational potential energies for individual spurs are included in table [ table : dustfeature ] .
the mechanical energy needed to overcome the gravitational potential energy for an individual spur is mostly comparable to that expected for a single sn .
some of the spurs may represent super - impositions of multiple features .
in such cases , the required energy could be considerably less .
so in principle such spurs can be produced by individual sne ( e.g. , raley , shelton & plewa 2007 ) , although some may be due to concerted energy inputs from massive stars , which may be formed in small groups or associations in . using the empirical sfr - dependent relationship of cappellaro , evans & turatto ( 1999 ) , we estimate the core collapsed sn rate as @xmath145 , which is smaller than the type ia sn rate of @xmath129 as estimated in [ sec : ism ] .
thus in ngc 5866 , type ia sne may be the primary energy source of the extraplanar dusty spurs , although core collapsed sne can also be important .
in contrast , the required energy for a high-@xmath53 cold gas feature in late - type galaxies typically equals tens or hundreds of the sn energy ( howk & savage 1997 , 1999 ; thompson , howk & savage 2004 ) .
this large energy requirement means most of the high-@xmath53 cold gas features in these galaxies should be produced by the concerted energy injection from massive star forming regions .
similarly , large scale cold gas filaments ( e.g. , feature `` 16 '' in fig .
[ fig : dustfeature ] ) detected in ngc 5866 also require substantially more energy to form .
they are probably produced by mechanical energy inputs provided collectively by core collapsed sne in the disk and/or type ia sne in the bulge of the galaxy .
we further estimate the disk mass loss rate via outflows as indicated by the extraplanar dusty spurs in . from the simulation of extraplanar sne driven disk outflow ( raley , shelton & plewa 2007 ) , the typical timescale to form an extraplanar cold gas spur through sne driven hydrodynamical processes is @xmath146 . from the mass of the observed extraplanar cool gas ( @xmath147 ; [ sec : coolgas ] )
, we infer the disk mass loss rate as @xmath148 . a mass - loading to the hot phase at a comparable rate
can naturally explain the low temperature of the hot gas ( [ sec : ism ] ) .
but , assume a constant mass - loss rate of @xmath148 , the disk would have lost @xmath149 over 8 gyr , which is an order of magnitude more than what remains in the disk . the current mass loss from the disk
is driven by individual sne ( [ sec : coolgasenergetics ] ) , a scenario different from that assumed for more energetic features detected in more active star forming galaxies . however , during the evolution of the galaxy , different processes ( e.g. type ia sne and sf ) may dominate in different stages , so the mass loss rate may not be constant .
in addition , accretion of cold gas from the intergalactic space could to some degree replenish the mass loss from the cold gas disk , which will also affect the observed amount of cold gas in the disk . some of the outflowing hot gas will eventually cool and return to the disk .
indeed , the mass - loaded hot gas with a temperature of @xmath150 kev is not expected to escape the gravitational potential of the galaxy .
the corresponding sound speed of the hot gas is only about @xmath151 , which is considerably smaller than the escape velocity , which should roughly be about three times of the circular velocity ( @xmath152 ; neistein et al .
1999 ) for a typical dark halo potential . in this case
, the hot gas will be confined by the gravitational potential of the galactic halo .
the outflow is likely to be subsonic , explaining the moderate x - ray luminosity of the hot gas emission from ( tang et al .
the upper limit to the outflow speed , @xmath153 , may be estimated from the accumulation time scale ( @xmath154 yrs ) of the hot gas in the bulge via the mass loss of evolved stars and from the cool gas mass loading from the disk .
the lower limit can be set to have the cooling time scale ( @xmath155 yrs ) of the hot gas longer than its escape time scale , assuming a characteristic bulge x - ray emission radius to be @xmath156 kpc .
we thus have @xmath157 .
the gas will cool in the large - scale galactic halo and will eventually return to the disk / bulge of the galaxy , which may be partly responsible for the present dusty galactic disk .
therefore , we expect a gas circulation , or a halo fountain ( i.e. , an outflow of heated gas to the halo and an inflow of cool gas to the disk / bulge ) , some of the extraplanar dusty structures may be such cool gas inflow ( fig . [
fig : dustfeature ] ) .
now let us see how may fit into the existing picture of galaxy formation and evolution .
s0 galaxies in general are believed to be evolved from spirals when they have exhausted gas for massive sf .
much of the existing work has been focused on s0 galaxies in cluster environments , where the exhaustion of the gas is typically attributed to the ram - pressure stripping by the ambient medium ( e.g. van dokkum et al . 1998 ; bekki , couch & shioya 2002 ; goto et al .
2003 ; moran et al .
outside clusters , the ram - pressure is unlikely to be strong enough to remove cool gas from galactic disks , although it may still be important in stripping gas from extended galactic halos and/or regulating the accretion from the intergalactic medium ( igm ; e.g. , kawata & mulchaey 2008 ) .
the evolution of such a field galaxy at present is typically dominated by internal processes , with galactic mass as the key parameter , according to recent simulations of galaxy formation ( birnboim & dekel 2003 ; birnboim et al .
2007 ; kere et al . 2005 , 2009 ) .
a galaxy with @xmath158 is expected to obtain its gas primarily in a so - called hot mode , in which accreted gas gets shock - heated near the virial radius to @xmath159 k. this virial shock presumably does not develop in lower mass galaxies , and the accretion is in a cold mode , in which the gas flows along narrow filaments that extend well inside the virial radius .
this simple picture may provide a natural interpretation for the observed galaxy bi - modality of stellar color and morphology , although complications arise when galaxy - galaxy mergers and galactic feedback are considered . in this evolution picture , s0 galaxies may be a direct result of their mass growth , passing through the mass threshold , through mergers and/or accretion .
however , a field s0 galaxy formed this way ( shutting off the cool gas supply ) is expected to contain certain amount of cool gas in the galactic disk .
in addition to the accumulation of the stellar mass loss , as discussed earlier , we also expect residual gas in the galactic disk from past sf .
the consumption via sf becomes inefficient as the amount of cool gas slowly decreases ( e.g. , considering the schmidt - kennicutt law and the toomre column density criterion of sf ; kennicutt 1998 ; toomre 1964 ) .
so substantial amounts of the disk gas should remain when the sf has more or less ceased , which may be the case for ngc 5866 and other field s0 galaxies ( e.g. , sage & welch 2006 ) .
of course , as demonstrated in the present work , the feedback from old stars in the galactic disk and bulge , though relatively gentle , is long lasting , which can help to evaporate the cool gas disk .
the effectiveness of this process on the long run depends on whether or not the evaporated gas can cool and then fall back into the disk , which is in turn sensitive to the intergalactic ram - pressure that the gaseous halo is subject to .
ngc 5866 is relatively rich in cool gas as an s0 galaxy ( welch & sage 2003 ; sage & welch 2006 ) , which may be due to a lower than average ram - pressure a result of the galaxy s relative isolation in a low - density environment . how the environment and s0 galaxy evolution affect the gas content and other galactic properties will be discussed in a following paper
we have presented a multi - wavelength study of the edge - on s0 galaxy ngc 5866 in a very low density environment .
we characterize various stellar and gaseous components and discuss the galactic disk / halo interaction and its role in the evolution of s0 galaxies .
the main results are as follows : * based on a acis observation , we detect 15 point - like sources within the galactic bulge of the galaxy .
one of them , with a 1.5 - 7 kev luminosity of @xmath160 , probably represents a very weak agn of the galaxy . * after removing the detected sources and subtracting remaining stellar contributions , we reveal the x - ray emission from the diffuse hot gas in and around the galactic bulge .
the emission extends as far as 3.5 kpc away from the galactic plane and has a total luminosity of @xmath161 in 0.5 - 2 kev . with a characteristic temperature of @xmath162 and a total mass of @xmath120 , the hot gas most likely represents a combination of the expected stellar mass loss and the mass - loading from the cool gas disk of the galaxy .
the hot gas is over - abundant in iron and under - abundant in oxygen , indicating a strong chemical enrichment by type ia sne .
* we detect warm ionized gas based on observations of h@xmath3/p@xmath3 emission .
this gas has an exponential scale height of @xmath0 and extends as far as @xmath163 from the mid - plane , indicating the presence of an extraplanar component .
massive stars in the galactic disk are primarily responsible for the ionization of this warm ionized gas .
* we characterize numerous cold gas features off the galactic disk , based on their extinction against the bulge stellar light .
dusty spurs typically have extensions similar to the warm ionized gas and have individual masses of @xmath164 .
such spurs may be generated mostly by individual sne in the galactic disk .
but there are also more energetic kpc - scale filaments , which may be produced by sne collectively in the disk and/or bulge of the galaxy . *
we show strong evidence for ongoing massive sf in the disk of the galaxy at a rate of @xmath99 .
the presence of the massive stars is required for the heating of the dust and warm ionized gas .
the stars , which tend to be formed in groups , may be responsible for some of the energetic extraplanar cool gas features . * the relative richness of cool gas in the galactic disk
is probably related to the isolation of the galaxy in a low density environment .
this gas , while being slowly consumed by the sf , may be replenished by the mass loss of evolved star , mostly in the galactic bulge .
the gas appears to be undergoing a circulation between the cool and hot phases via the galactic disk / halo interaction .
this circulation becomes possible for the galaxy because of both the low temperature of the mass - loaded hot gas and the inefficiency of ram - pressure stripping in the environment . * a field s0 galaxy may form mainly from the development of a hot gaseous halo , which stops or slows down the supply of cool gas from the igm .
this development may be a natural consequence of the mass growth of a galaxy , according to recent structure formation simulations .
but the inclusion of the galactic feedback as well as the environmental effects , as indicated in the present work , may also be important .
this work is supported by nasa through the cxc / sao grants ar7 - 8016a and g08 - 9088b , by nsfc through the grants 10725312 and 10673003 and the china 973 program grant 2009cb824800 , and by china scholarship council .
lrrrrrrrr morphology@xmath167 & @xmath168 , /liner + center position@xmath167 & r.a . 15h06m29.5s + ( j2000 ) & dec .
+ 55d45m47.6s + distance ( mpc)@xmath167 & @xmath169 ( 1@xmath170 54 pc ) + inclination angle@xmath171 & @xmath172 + @xmath173 ( arcmin)@xmath171 & @xmath174 + total b magnitude@xmath171 & @xmath175 + total k magnitude@xmath167 & @xmath176 + total b - v color@xmath171 & 0.85 + foreground galactic extinction ( e(b - v))@xmath167 & 0.013 + @xmath177@xmath171 & 11.044 + galactic foreground @xmath178 & 1.46 + @xmath179@xmath180 & @xmath181 + @xmath182@xmath183 & @xmath184 + @xmath185@xmath186 & @xmath187 + @xmath188@xmath189 & @xmath190 + @xmath191@xmath189 & @xmath192 + @xmath193@xmath194 & 0.29 + sfr(@xmath195)@xmath194 & @xmath196 + @xmath197@xmath189 & 2% + @xmath198@xmath189 & 0.005 + [ table : basicpara ] lrrrrrrrr 1 & j150552.44 + 554415.8 & 1.4 & @xmath199 & & & b , s + 2 & j150555.82 + 554556.9 & 1.2 & @xmath200 & & & b , s + 3 & j150559.07 + 554455.2 & 1.1 & @xmath201 & & & b , s + 4 & j150600.69 + 554921.6 & 2.0 & @xmath202 & @xmath203 & & h + 5 & j150601.04 + 554513.0 & 0.9 & @xmath204 & @xmath205 & @xmath206 & b , s , h + 6 & j150603.95 + 554623.3 & 0.9 & @xmath207 & & & b , s + 7 & j150605.82 + 554649.4 & 0.8 & @xmath208 & & @xmath209 & b , s , h + 8 & j150608.65 + 554919.2 & 1.4 & @xmath210 & @xmath211 & & h + 9 & j150613.31 + 554359.4 & 0.7 & @xmath212 & & & b , s , h + 10 & j150616.84 + 554715.2 & 0.5 & @xmath213 & @xmath214 & @xmath215 & b , s , h + 11 & j150618.38 + 554547.7 & 0.7 & @xmath216 & & & b + 12 & j150620.41 + 554530.3 & 0.6 & @xmath217 & & & b , h , s + 13 & j150620.46 + 554622.7 & 0.4 & @xmath218 & & & b , s , h + 14 & j150620.48 + 554542.2 & 0.4 & @xmath219 & @xmath220 & @xmath221 & b , s , h + 15 & j150620.66 + 554221.4 & 1.0 & @xmath222 & & & b , h + 16 & j150620.76 + 554442.7 & 0.6 & @xmath223 & @xmath224 & & s + 17 & j150621.30 + 554906.7 & 0.6 & @xmath225 & @xmath226 & @xmath227 & b , s , h + 18 & j150621.36 + 554344.1 & 1.1 & @xmath228 & & & b + 19 & j150622.50 + 554721.8 & 0.6 & @xmath229 & & & s , b + 20 & j150623.05 + 554542.1 & 0.7 & @xmath230 & & & b , s + 21 & j150624.53 + 554707.5 & 0.8 & @xmath231 & & & b , s , h + 22 & j150625.18 + 554556.6 & 0.5 & @xmath232 & & & b , s + 23 & j150627.77 + 554537.8 & 0.5 & @xmath233 & & & b , s , h + 24 & j150628.11 + 554631.6 & 0.9 & @xmath234 & & & b , s + 25 & j150628.43 + 554556.4 & 0.7 & @xmath235 & & & b + 26 & j150628.73 + 554555.1 & 0.5 & @xmath236 & & & b , h , s + 27 & j150628.75 + 554538.7 & 0.5 & @xmath237 & & & b , s , h + 28 & j150629.14 + 554551.4 & 0.8 & @xmath238 & & & b + 29 & j150629.15 + 554515.6 & 0.7 & @xmath239 & & & b , s + 30 & j150629.51 + 554547.5 & 0.6 & @xmath240 & @xmath241 & & h , b + 31 & j150629.65 + 554552.7 & 1.2 & @xmath242 & & & b + 32 & j150629.73 + 554309.0 & 0.4 & @xmath243 & @xmath244 & @xmath245 & s , b , h + 33 & j150629.84 + 554533.2 & 0.6 & @xmath246 & & & b , s + 34 & j150630.33 + 554648.8 & 0.6 & @xmath247 & @xmath248 & & h , b + 35 & j150630.43 + 554139.5 & 0.6 & @xmath249 & @xmath250 & @xmath251 & b , s , h + 36 & j150630.45 + 554515.4 & 0.7 & @xmath252 & & & b , h + 37 & j150631.26 + 554458.1 & 0.5 & @xmath253 & & & b , s , h + 38 & j150631.58 + 554251.3 & 0.6 & @xmath235 & & & b , s , h + 39 & j150631.83 + 554232.3 & 0.7 & @xmath254 & & & b , s , h + 40 & j150632.17 + 554220.7 & 0.6 & @xmath255 & @xmath256 & & b , h + 41 & j150632.70 + 554658.9 & 0.5 & @xmath257 & & @xmath258 & b , s , h + 42 & j150632.73 + 554525.6 & 0.6 & @xmath222 & & & b , h , s + 43 & j150634.07 + 554536.6 & 0.7 & @xmath259 & & & b , h , s + 44 & j150635.60 + 554445.6 & 1.7 & @xmath260 & & & b + 45 & j150635.78 + 554531.4 & 0.6 & @xmath261 & @xmath248 & & b , h + 46 & j150637.04 + 554527.3 & 1.7 & @xmath262 & @xmath211 & & b + 47 & j150638.02 + 554842.2 & 0.5 & @xmath263 & & & b , s , h + 48 & j150638.74 + 554505.6 & 0.6 & @xmath264 & & & b , s , h + 49 & j150640.17 + 554612.5 & 1.4 & @xmath265 & & & h , b + 50 & j150640.26 + 554514.0 & 0.9 & @xmath266 & & & b , s + 51 & j150640.48 + 554806.6 & 0.4 & @xmath267 & & & s , b , h + 52 & j150643.97 + 554522.9 & 0.7 & @xmath222 & & & b , h , s + 53 & j150645.29 + 554845.3 & 0.6 & @xmath268 & @xmath248 & & b + 54 & j150652.51 + 554639.5 & 1.0 & @xmath269 & & & b , h + 55 & j150653.12 + 554800.1 & 0.6 & @xmath270 & & & b , h + 56 & j150653.79 + 554432.9 & 0.5 & @xmath271 & & & b , s , h + 57 & j150655.64 + 554437.7 & 0.5 & @xmath272 & & & b , h + 58 & j150656.64 + 554550.5 & 0.4 & @xmath273 & & & b , h , s + 59 & j150701.15 + 554715.6 & 0.9 & @xmath274 & & & b , s + 60 & j150703.66 + 554833.4 & 1.7 & @xmath275 & & & b + 61 & j150703.70 + 554900.6 & 1.2 & @xmath276 & & & b , h + 62 & j150704.60 + 554448.9 & 0.7 & @xmath277 & & & s , b + 63 & j150707.34 + 554435.4 & 1.0 & @xmath278 & @xmath248 & & h , b + 64 & j150707.50 + 554527.1 & 1.0 & @xmath279 & & & b + 65 & j150708.00 + 554613.9 & 0.6 & @xmath280 & @xmath281 & @xmath282 & b , s , h + 66 & j150708.48 + 554545.8 & 0.7 & @xmath283 & & & b , h + 67 & j150708.78 + 554144.7 & 1.3 & @xmath284 & @xmath285 & & b , h + 68 & j150715.72 + 554955.2 & 2.1 & @xmath286 & & & b , s + 69 & j150719.26 + 554449.3 & 1.0 & @xmath287 & @xmath288 & & b , s , h + 70 & j150721.08 + 554559.0 & 1.0 & @xmath289 & @xmath290 & @xmath291 & b , s , h + 71 & j150736.05 + 554605.6 & 2.3 & @xmath292 & & & b , s + [ table : acis_source_list ] lrrrrrrrr 01 & @xmath293 & 110 & 0.55 & 0.45 & 8.0 & 3.3 & 5 & 3 + 02 & @xmath294 & 150 & 0.41 & 0.32 & 5.7 & 2.3 & 10 & 10 + 03 & @xmath295 & 100 & 0.78 & 0.56 & 10.1 & 6.2 & 5 & 2 + 04 & @xmath296 & 140 & 0.31 & 0.22 & 3.9 & 4.9 & 2 & 2 + 05 & @xmath297 & 270 & 0.10 & 0.07 & 1.2 & 0.9 & 1 & 3 + 06 & @xmath298 & 120 & 0.47 & 0.37 & 6.7 & 8.3 & 2 & 1 + 07 & @xmath299 & 190 & 0.17 & 0.11 & 2.0 & 1.0 & 2 & 3 + 08 & @xmath300 & 70 & 0.39 & 0.29 & 5.1 & 3.2 & 1 & 0.3 + 09 & @xmath301 & 80 & 0.16 & 0.11 & 2.0 & 2.5 & 0.5 & 0.2 + 10 & @xmath302 & 70 & 0.43 & 0.31 & 5.5 & 4.6 & 1 & 0.3 + 11 & @xmath303 & 90 & 0.36 & 0.22 & 4.0 & 3.8 & 2 & 0.6 + 12 & @xmath304 & 80 & 0.36 & 0.29 & 5.3 & 2.8 & 3 & 0.8 + 13 & @xmath305 & 70 & 0.40 & 0.32 & 5.7 & 2.8 & 4 & 0.9 + 14 & @xmath306 & 60 & 0.42 & 0.31 & 5.6 & 7.0 & 2 & 0.3 + 15 & @xmath307 & 90 & 0.41 & 0.32 & 5.8 & 3.6 & 4 & 2 + 16 & @xmath308 & 780 & 0.14 & 0.11 & 1.9 & 1.2 & 10 & 17 + [ table : dustfeature ] | s0 galaxies are often thought to be passively evolved from spirals after star formation is quenched . to explore
what is actually occurring in such galaxies , we present a multi - wavelength case study of ngc 5866 a nearby edge - on s0 galaxy in a relatively isolated environment .
this study shows strong evidence for dynamic activities in the interstellar medium , which are most likely driven by supernova explosions in the galactic disk and bulge . understanding these activities can have strong implications for studying the evolution of such galaxies .
we utilize _ chandra _ , _ hst _ , and _ spitzer _ data as well as ground - based observations to characterize the content , structure , and physical state of the medium and its interplay with the stellar component in ngc 5866 .
a cold gas disk is detected with an exponential scale height of @xmath0 .
numerous distinct off - disk dusty spurs are also clearly present : prominent ones can extend as far as @xmath1 from the galactic plane and are probably produced by _
individual _
sne , whereas faint filaments can have @xmath2 scale and are likely produced by sne collectively in the disk / bulge .
we also detect substantial amounts of diffuse h@xmath3- and p@xmath3-emitting gas with a comparable scale height as the cold gas .
we find that the heating of the dust and warm ionized gas can not be explained by the radiation from evolved stars alone , strongly indicating the presence of young stars in the galactic disk , though at a slow formation rate of @xmath4 .
we further reveal the presence of diffuse x - ray - emitting hot gas , which extends as far as 3.5 kpc away from the galactic plane and can be heated easily by type ia sne in the bulge .
however , the mean temperature of this gas is @xmath5 , substantially lower than what might be expected from the mass - loss of evolved stars and type ia sne heating alone in the galaxy , indicating that the mass loading from the cool gas is important .
the total masses of the cold , warm ionized , and hot gases are @xmath6 , @xmath7 and @xmath8 , respectively .
the relative richness of the gases , apparently undergoing circulations between the disk and halo of the galaxy , is perhaps a result of its relative isolation . |
the motion of atoms in a liquid can be divided into two constituent : ( a ) oscillation in a valley of the liquid s many - body potential and ( b ) transits between many - body valleys .
the latter process is responsible for self - diffusion . as we showed in a previous paper @xcite , which drew upon earlier results of clements and wallace @xcite
, the former motion can be modeled very precisely under the assumption that the valleys are nearly harmonic , with the majority of valleys ( the random valleys ) sharing a common spectrum of frequencies .
specifically , a purely harmonic model provides an extremely accurate formula for @xmath0 , the normalized velocity autocorrelation function , in the nondiffusing regime . here
we will use this picture and our previous work to justify a `` mean atom trajectory '' model , a single - atom model that approximates the behavior of an average atom in the liquid and correctly reproduces its nondiffusing behavior .
then we will introduce a simple intuitive account of the transit process that allows us to extend the model to the self - diffusing regime . in sec.[model ] we develop this model and explain how it is used to calculate @xmath0 for a diffusing liquid .
then we fit the model to md simulations at various temperatures in sec .
[ md ] , and we comment on the quality of the results . in sec .
[ concl ] we compare the present mean atom trajectory model with previous work based on instantaneous normal modes ( inm ) , and with an earlier independent atom model @xcite , and we summarize our conclusions .
to form an appropriate basis for our model , we must begin with some initial reasonable approximations about the nature of the valleys and transits in the real liquid . as mentioned in the introduction
, available evidence suggests that the valleys are nearly harmonic , and we will continue to assume that here .
further , we will assume that transits between valleys occur instantaneously and are local in character ; that is , each transit involves only a few neighboring atoms .
now for the nondiffusing supercooled liquid , there are no transits , and as we ve seen the system s motion is accurately expressed in terms of the harmonic normal modes .
but it is also legitimate to consider this motion from the point of view of a single atom , and when we do so we see that each atom moves along a complicated trajectory within a single - particle potential well which fluctuates because of the motion of neighboring atoms , but whose center is fixed in space .
an important timescale for this motion is the single - atom mean vibrational period @xmath1 , which we take as @xmath2 , where @xmath3 is the rms frequency of the set of normal modes .
let us now follow the single - atom description as the temperature is increased .
once the glass transition is passed , the atom will begin to make transits from one single - particle well to another @xcite , and even before the melting temperature is reached , the transit rate will be on the order of one per mean vibrational period .
since each atom has approximately ten neighbors , roughly ten transits will occur in its immediate vicinity every period , changing the set of normal mode eigenvectors each time ; thus we conclude that a decomposition of the motion into normal modes will not be useful when the liquid is diffusing , so we have no choice but to follow a single - atom description for the liquid state @xcite .
that being the case , we shall start from the beginning with a single - atom description , in order to construct a unified model for diffusing and nondiffusing motion alike .
our starting point will be in the nondiffusing regime , where a normal mode analysis is still valid . as shown in @xcite , the @xmath4th coordinate of the @xmath5th atom in an @xmath6-body harmonic valley
can be written @xmath7 where the @xmath8 form a @xmath9 orthogonal matrix , the @xmath10 are the frequencies of the normal modes , and the @xmath11 are the amplitudes of the modes .
three of the modes have zero frequency and correspond to center of mass motion ; here we demand that the center of mass is stationary , so the system has only @xmath12 independent degrees of freedom , the zero frequency modes are absent , and the sum over @xmath13 runs from 1 to @xmath12 . to make this an equation for
a `` mean '' atom , we first drop the index @xmath5 : @xmath14 now we must reinterpret the @xmath15 since they no longer form an orthogonal ( or even square ) matrix .
let @xmath16 so @xmath17 we will ultimately consider situations in which the well center is allowed to move , so let @xmath18 be the atom s position , be the location of the center of the well , and @xmath18 be the atom s displacement from the well center ; then @xmath19 with velocity @xmath20 we now have the basic formula , but we must decide how to assign values to the @xmath21@xmath22 , @xmath11 , and @xmath23 .
let s do this by calculating @xmath24 , the velocity autocorrelation function , in this model and comparing it to the harmonic result derived in @xcite .
@xmath25 let s assign the @xmath23 randomly and average over each @xmath23 _ separately _ ; then if @xmath26 , @xmath27 but if @xmath28 , then @xmath29 thus the @xmath26 terms in eq .
( [ 1stz ] ) are eliminated and we have @xmath30 note that the formula for @xmath24 in the harmonic theory also lacks off - diagonal terms @xcite , but for a different reason : the orthogonality of the matrix @xmath8 removes them in that case .
now let us make the assignment @xmath31 where @xmath32 is a randomly chosen unit vector ; then @xmath33 this is the same expression for @xmath24 that one derives in the harmonic model ( see eq .
( 10 ) of @xcite ) for _ any _ distribution of normal mode amplitudes @xmath11 , not just the thermal equilibrium distribution .
thus our model with these choices of @xmath23 and @xmath22 correctly reproduces @xmath24 for any equilibrium or nonequilibrium ensemble in the harmonic theory . to recover the equilibrium result
, we make the final substitution @xmath34 with the result @xmath35 which is eq .
( 13 ) from @xcite .
thus our model assumes that the motion of a mean nondiffusing atom in thermal equilibrium at temperature @xmath36 is given by @xmath37 where the phases @xmath23 and unit vectors @xmath32 are randomly chosen . by construction this model
gets the same result for @xmath24 , and thus @xmath38 , as the harmonic model ; does it correctly reproduce any other functions ?
we can check by calculating the analog of @xmath24 for positions , @xmath39 .
using eq .
( [ uharmoft ] ) for @xmath40 and calculating for the harmonic model as in @xcite , and using eq .
( [ u1oft ] ) for @xmath18 and calculating as above , one finds in both cases that @xmath41 again the substitution @xmath42 recovers the correct thermal equilibrium result .
so this model correctly reproduces @xmath43 and @xmath44 as well .
notice that because we have a closed form for @xmath18 , all of the correlation functions calculated so far also have a closed form .
once we introduce transits , this will no longer be the case . while this model compares well with the harmonic model
, one might wonder how well it compares with molecular dynamics ( md ) results . in fig .
[ vsqvsmd ] a graph of @xmath45 as a function of @xmath46 for the model in equilibrium at 6.69 k is compared to a randomly chosen particle from an md run for a randomly chosen atom in an md simulation of na at 6.69 k ( top ) is compared with @xmath45 for the mean atom trajectory model in the nondiffusing regime ( bottom ) over 5000 timesteps of duration @xmath47 s. notice that both fluctuations have roughly the same magnitude and oscillate at approximately the same frequencies.,title="fig : " ] for a randomly chosen atom in an md simulation of na at 6.69 k ( top ) is compared with @xmath45 for the mean atom trajectory model in the nondiffusing regime ( bottom ) over 5000 timesteps of duration @xmath47 s. notice that both fluctuations have roughly the same magnitude and oscillate at approximately the same frequencies.,title="fig : " ] of liquid na also at 6.69 k , a temperature at which it is known the sample is nondiffusing ( see @xcite for details ) .
note that in both graphs @xmath45 has approximately the same amplitude , and the gaps between peaks are roughly the same size , indicating oscillations at frequencies in the same ranges .
thus not only functions of the motion but the motion itself shows strong qualitative agreement with md in the nondiffusing regime .
the md system with which we will be comparing our model has @xmath48 particles , so it has 1497 normal mode frequencies .
( remember that the zero frequency modes are removed at the start . )
when we introduce transits , we will have to evaluate @xmath24 numerically , and for speed of computation we would like to use only a representative subset of the normal mode frequencies , say 75 instead of all 1497 .
the original set of frequencies is determined as described in @xcite .
we decided which subset to use by calculating three moments of the full frequency distribution defined below : @xmath49^{1/2 } \nonumber \\ \omega_{0 } & = & \exp \langle \ln(\omega_{\lambda } ) \rangle \nonumber \\ \omega_{-2 } & = & \left [ \frac{1}{3 } \langle \omega_{\lambda}^{-2 } \rangle \right]^{-1/2}.\end{aligned}\ ] ] note that by this definition @xmath50 .
we then calculated the same three moments for several sets of 75 frequencies evenly spaced throughout the full set , and we chose the set that best fit the moments to use in further calculations .
this process is somewhat subjective , because skewing the sample in favor of lower frequencies improves the accuracy of @xmath51 but reduces the accuracy of @xmath52 , and the opposite is true if one skews in favor of high frequencies .
the table below shows the values of the moments for the full set and the reduced set of 75 that we ultimately chose .
the frequencies are in units of @xmath53 , where @xmath54 s is the timestep of our md simulations @xcite . [ cols="^,^,^",options="header " , ] at @xmath55 compared with the md result for supercooled liquid na at @xmath56 k. ] at @xmath57 compared with the md result for supercooled liquid na at @xmath58 k. ] at @xmath59 compared with the md result for liquid na at @xmath60 k. ] at @xmath61 compared with the md result for liquid na at @xmath62 k. ] at @xmath63 compared with the md result for liquid na at @xmath64 k. ] notice that in all cases @xmath65 is of the same order of magnitude as @xmath66 , indicating roughly one transit per mean vibrational period , as mentioned in subsection [ general ] , and as predicted in @xcite . the most obvious trend exhibited by @xmath0 from
the five md runs is that its first minimum is rising with increasing @xmath36 ; as we mention below , this is the primary reason for the increasing diffusion coefficient @xmath67 . note that the model is able to reproduce this most important feature quite satisfactorily .
in fact , all five fits of the model to the md results capture their essential features , but we do see systematic trends in the discrepancies . first , note that the location of the first minimum barely changes at all in the model as @xmath65 is raised , but in md the first minimum moves steadily to earlier times as the temperature rises .
the first minimum occurs at a time roughly equal to half of the mean vibrational period ( @xmath68 in this system ) , so the steady drift backward suggests that the md system is sampling a higher range of frequencies at higher @xmath36 .
also , for the three lowest temperatures the model tends to overshoot the md result in the vicinity of the first two maxima after the origin , and at the highest two temperatures this overshoot is accompanied by a positive tail that is slightly higher than the ( still somewhat long ) tail predicted by md .
these overshoots should clearly affect the diffusion coefficient @xmath67 , which is the integral of @xmath24 . to check this , we calculated the reduced diffusion coefficient @xmath69 , the integral of @xmath0 , which is related to @xmath67 by @xmath70 the results are compared to the values of @xmath69 calculated from the md runs in fig .
the results from the two nondiffusing runs discussed in @xcite are also included . as a function of @xmath36 for both the model and md .
] in all of the diffusing cases , the model overestimates @xmath69 by roughly the same amount , which we take to be the effect of the overshoots at the first two maxima . at the higher temperatures the discrepancy is also higher , presumably due to the model s long tail .
it is interesting to note that this md system produces results that agree very closely with experiment : for example , the md predicts that at @xmath60 k and @xmath71 g/@xmath72 , @xmath73 @xmath74/s , while experiments by larsson , roxbergh , and lodding @xcite find that at @xmath60 k and @xmath75 g/@xmath72 , @xmath76 @xmath74/s .
hence our agreement with md results genuinely reflects agreement with properties of real liquid na .
we have presented a single - atom model of a monatomic liquid that provides a unified account of diffusing and nondiffusing behavior .
the nondiffusing motion is modeled as a sum of oscillations at the normal mode frequencies ( eq . ( [ uoft ] ) ) , simulating the trajectory of an average atom in a complicated single - body potential well that fluctuates due to the motion of its neighbors .
self - diffusion is accounted for in terms of instantaneous transits between wells , which occur at a temperature - dependent rate @xmath65 .
since this model gives a simple and straightforward account of the motion itself , it can easily be used to calculate any single - atom correlation function one wishes ; here we have focussed on the velocity autocorrelation function .
it is interesting to note that in this model the velocity correlations persist through a transit , instead of being washed out entirely by the transit process ; we will return to this point below .
the relaxation of correlations seen by the decay of @xmath0 arises here from two distinct processes : dephasing as a result of the large number of frequencies in the single - well motion , and transits between wells .
the dephasing effect produces relaxation but not diffusion : it causes @xmath0 to decay @xcite but its integral remains zero . on the other hand , transits certainly contribute to relaxation ( see @xcite , where they provide the only relaxation mechanism ) , but in addition they raise the first minimum of @xmath0 substantially , increasing its integral and providing a nonzero @xmath67 .
the comparison of this model to md results is generally quite positive , particularly for a one - parameter model ; the two calculations of @xmath0 agree strongly over 1000 timesteps .
in addition , the match is encouraging over a very large range of temperatures , from essentially 0 k to @xmath77 .
the most noticeable discrepancies are the backward drift in the location of the first minimum of @xmath0 , which is present in md but not the model , and the tendency of the model to exaggerate certain characteristics of the md results ( the maxima at intermediate times and the high-@xmath36 positive tail ) .
this latter effect is responsible for the model s overestimate of the diffusion coefficient , though we hasten to add that the model @xmath69 is still in satisfactory agreement with md results , especially for the liquid at @xmath78 ( see fig .
[ dvst ] ) . as in @xcite ,
it is useful to compare this model and the accompanying results to the work others have done using the formalism of instantaneous normal modes ( inm ) and similar methods .
previously , we discussed the advantages of our general approach and the superior quality of its results when applied to nondiffusing states ; here we will consider matters relating explicitly to mechanisms of diffusion .
as is noted explicitly by vallauri and bermejo @xcite , the account of inm by stratt ( see , for example , @xcite ) does not consider at all ; their @xmath24 is essentially a sum over cosines , and as such it integrates to zero .
this is understandable , because as stratt et al .
repeatedly emphasize , their approximation is valid only for very short times , so they are not attempting to model effects with longer timescales .
they compare their results with md calculations of @xmath0 for states of an lj system ranging from a moderately supercooled liquid to well above the melting temperature , and our fits are of roughly the same quality or better in all cases .
authors who do attempt to model diffusion usually follow the path suggested initially by zwanzig @xcite , who thought of the liquid s phase space as divided into in which each atom spends its time before finding a saddle point in the potential and jumping from one cell to the next .
he imagined as a first approximation that the jumps destroy all correlations between the cells ; since atoms are jumping all the time , he suggested that the net result was to multiply the nondiffusing form of @xmath24 by a damping factor @xmath79 for some timescale @xmath1 representing the lifetime of a stay in a single cell .
notice that zwanzig provided no dynamical model of the jumping process itself .
this suggestion has been developed and transformed extensively by madan , keyes , and seeley @xcite , who take a general zwanzig - like functional form for @xmath24 and use a combination of heuristic arguments and constraints on its moments to specify its dependence on a `` hopping rate '' @xmath80 , the analog of @xmath66 for zwanzig , which they then extract from the unstable lobe of the inm spectrum .
although we can not be entirely sure , as indicated in @xcite , we think it most likely that their simulations of lj ar at 80 k , 120 k , and 150 k consist of states comparable to our figs .
[ 216.3kvsmodel ] and [ 309.7kvsmodel ] , and again we would argue that our fits are somewhat better .
finally , cao and voth @xcite also approach diffusion by means of a damping factor , and they consider factors of two different types , each of which contains parameters that can be determined from other calculated or experimental quantities .
their matches with md are actually quite good , but again ours are of at least comparable quality .
having claimed that our matches to md simulations are as good as or better than all of the others we have surveyed , let us emphasize a fundamental difference between our approaches to diffusion : we provide an account of the _ process _ of transiting from well to well , so we have a model of the actual _ motion _ of a mean atom in space that is valid to arbitrary times , and given this model we can _ calculate _ the effect of transits on correlations .
all other approaches we know of begin with an account of the motion valid only for very short times , calculate @xmath24 from this motion , and then try to model the effects of diffusion on @xmath24 directly , using parameters that are thought to be characteristic of jumps between wells . in the diffusing regime ,
no one else we have seen actually calculates @xmath81 to find @xmath24 , as we do . in the process
, we find that some of the assumptions made by others , in particular zwanzig s hypothesis that jumps between wells simply erase correlations , are not true .
this approach is already yielding insights into the actual motion of an atom undergoing a transit .
finally , we would like to compare the present mean atom trajectory model with an earlier independent atom model by wallace @xcite . in developing the independent atom model ,
two arguments were made : ( a ) the high rate of transits in the liquid state shows the need to abandon the normal mode description of motion , and instead picture the motion of a single atom among a set of fluctuating wells , and ( b ) the leading approximation to a fluctuating well is its smooth time average well .
accordingly , the independent atom oscillates with frequency @xmath82 in a smooth isotropic well , and it transits with probability @xmath83 to an adjacent identical well at each turning point @xcite .
what we have now learned by considering low temperatures , and especially the nondiffusing states of @xcite , is that the md system exhibits a velocity decorrelation process which results from the presence of many frequencies in the single - particle motion , making assumption ( b ) less reasonable . in the current model these many frequencies
are retained , and they are intended to represent the strong fluctuations in each single - particle well ; in this way the current model makes an important improvement over the independent atom model , where the well fluctuations were averaged out . beyond this difference
, the two models contain similar but not identical treatments of transits , whose rate increases with increasing temperature , and which produce self - diffusion .
the less detailed but simpler independent atom model has proven useful in a description of the glass transition @xcite , and the corresponding transit parameter has been used to relate shear viscosity and self - diffusion in liquid metals @xcite . | a recent description of the motion of atoms in a classical monatomic system in liquid and supercooled liquid states divides the motion into two parts : oscillations within a given many - particle potential valley , and transit motion which carries the system from one many - particle valley to another . building on this picture ,
we construct a model for the trajectory of an average atom in the system .
the trajectory consists of oscillations at the normal - mode distribution of frequencies , representing motion within a fluctuating single - particle well , interspersed with position- and velocity - conserving transits to similar adjacent wells . for the supercooled
liquid in nondiffusing states , the model gives velocity and displacement autocorrelation functions which exactly match those found in the many - particle harmonic approximation , and which are known to agree almost precisely with molecular dynamics ( md ) simulations of liquid na . at higher temperatures , by allowing transits to proceed at a temperature - dependent rate ,
the model gives velocity autocorrelation functions which are also in remarkably good agreement with md simulations of na at up to three times its melting temperature .
two independent processes in the model relax velocity autocorrelations : ( a ) dephasing due to the presence of many frequency components , which operates at all temperatures but which produces zero diffusion , and ( b ) the transit process , which increases with increasing temperature and which produces diffusion .
compared to several treatments of velocity autocorrelations based on instantaneous normal modes , the present model offers an advantage : it provides a single - atom trajectory in real space and time , including transits and valid for arbitrary times , from which all single - atom correlation functions can be calculated , and they are also valid at all times . |
recent detections of distant dusty galaxies with the scuba camera ( the submillimeter common user bolometer array ; holland et al .
1999 ) on the 15 m james clerk maxwell telescope constitute a substantial fraction of the cosmic fir background detected by the _ firas _ and _ dirbe _ experiments on the _ cobe _ satellite ( puget et al . 1996 ; guiderdoni et al .
1997 ; schlegel , finkbeiner , & davis 1998 ; fixsen et al .
1998 ; hauser et al . 1998 ; lagache et al . 1999 ) .
since the observed fir background is comparable to the total unobscured emission at ultraviolet / optical wavelengths , a full determination of the global star formation history of the universe requires a comprehensive understanding of this dust - enshrouded galaxy population .
the 850 @xmath0 m scuba surveys to date have reported galaxy number counts that are in general agreement ( smail , ivison & blain 1997 ; barger et al .
1998 ; hughes et al . 1998 ; blain et al . 1999a ; eales et al . 1999 ; barger , cowie , & sanders 1999 ) ; the cumulative surface density above 2 mjy is about @xmath9 .
the discrete sources have bolometric luminosities that are characteristically @xmath10 if they lie at @xmath11 .
moreover , the sources for which measurements exist at multiple wavelengths ( e.g. ivison et al .
1998 ) show thermal spectral energy distributions ( seds ) .
thus , the scuba sources are inferred to be the distant analogs of the local ultraluminous infrared galaxy ( ulig ; sanders & mirabel 1996 ) population .
an essential observational goal is to determine the redshift distribution of the submillimeter population in order to trace the extent and evolution of obscured emission in the distant universe .
however , identifying the optical / nir counterparts to the submillimeter sources is difficult due to the uncertainty in the scuba positions .
barger et al . ( 1999b ) presented a spectroscopic survey of possible optical counterparts to a flux - limited sample of galaxies selected from the 850 @xmath0 m survey of massive lensing clusters by smail et al .
candidate optical counterparts in the scuba error - boxes were identified using moderately deep ground - based and _ hst _ exposures ( @xmath12 and @xmath13 , respectively ) .
one - quarter of the sources could be reliably identified , and those had redshifts in the range @xmath8 . a lower limit of 20 per cent of the full sample showed signs of agn activity . however , for the majority of the submillimeter sources there were either no optical counterparts or the optical associations were not secure .
such sources could either be at very high redshift or be so highly obscured that they emit their energy almost entirely in the submillimeter .
high resolution radio continuum maps with subarcsecond positional accuracy and resolution offer new opportunities for locating submillimeter sources and determining their physical properties .
the unique advantage of centimeter and fir observations is that galaxies and the intergalactic medium are transparent at these wavelengths , so observed flux densities are proportional to intrinsic luminosities . in galaxies without a powerful agn , the radio luminosity is dominated by diffuse synchrotron emission from relativistic electrons accelerated in supernovae remnants from stars more massive than @xmath14 .
these massive stars live @xmath15 yr ; the relativistic electrons probably live @xmath16 yr ( condon 1992 ) . thus , radio observations probe very recent star formation .
fir observations of starburst galaxies are also a direct measure of massive star formation . as summarized by condon ( 1992 ) , radio continuum emission and thermal dust emission
are empirically observed to be tightly correlated due to both being linearly related to the massive star formation rate . if the fir - radio correlation applies to high redshift , as is plausibly the case ( though at the very highest redshifts compton cooling of the relativistic electrons by the microwave background may suppress the radio emission ) , then very sensitive radio observations can be used to pinpoint distant submillimeter sources . in this paper
we investigate the feasibility of using radio data to identify and characterize the bright submillimeter source population .
richards ( 1999b ) recently obtained an extremely deep very large array ( vla ) 1.4 ghz image centered on the hubble deep field ( hdf ) .
richards et al .
( 1999 ) matched ground - based optical data from barger et al .
( 1999a ) to the radio image and found that @xmath17 per cent of the galaxies in the sample could not be identified to optical magnitude limits of @xmath18 .
in other respects , such as radio size and spectral index , the optically - faint objects were not any different from the remaining population .
richards et al .
( 1999 ) proposed four possible scenarios to explain this population , including @xmath19 obscured starbursts ( beyond @xmath20 the sensitivity to star forming galaxies cuts off due to the flux density limits of the radio data ) , extreme redshift ( @xmath21 ) agn , @xmath22 obscured agn , or one - sided radio jets .
in the first phase of our program we observed with lris on the keck ii 10 m telescope a complete subsample of the radio sources . our primary objective was to determine the redshifts of the optical / nir - faint ( @xmath23 ) radio sources .
although we were able to spectroscopically identify nearly all the objects in our subsample to @xmath24 ( all had @xmath25 ) , we were unable to obtain redshifts for the fainter objects .
either these sources are distant ( @xmath26 ) with spectral features lying outside the optical wavelength range , or the visibility of remarkable features is strongly affected by dust in the galaxies .
if the optical / nir - faint radio sources are highly dust obscured systems , then it is possible that they will be detectable in the submillimeter . in the second phase of our program we observed with scuba 15 of the 22 optical / nir - faint radio sources in the central @xmath27 square arcminute region of the radio map ;
another 4 were observed by hughes et al .
( 1998 ; hereafter h98 ) in the hdf - proper .
the jiggle map mode enabled simultaneous observations of a large fraction ( 31/48 ) of the optical / nir - bright radio sources .
even with relatively shallow scuba observations ( a @xmath2 detection limit of 6 mjy at 850 @xmath0 m ) , we detected 5 of the optical / nir - faint radio sources ; a sixth with submillimeter flux @xmath28 mjy was detected in the deep hdf - proper submillimeter map of h98 .
in contrast , none of the optical / nir - bright radio sources were detected .
we additionally detected two @xmath5 mjy sources that did not have radio counterparts .
thus , our targeted scuba survey of optical / nir - faint radio sources turned up @xmath29 per cent of the bright submillimeter sources in our surveyed areas . in the final phase of our program
, we explored the feasibility of obtaining redshift estimates from the submillimeter - to - radio flux ratios , as recently suggested by carilli & yun ( 1999 ) .
we find that the redshifted arp 220 sed reasonably describes both redshifted local ulig data and known high redshift submillimeter sources and hence can be used as a rough redshift estimator .
we estimate that all of our bright submillimeter sources fall in the redshift range @xmath6 , consistent with the redshifts for the lensed submillimeter sources of barger et al .
( 1999b ) .
once we have redshifts for the distant submillimeter source population , we can determine the global evolution of star formation in dust - obscured galaxies .
previous studies of the star formation rate density ( sfrd ) have primarily used rest - frame ultraviolet data ( e.g. madau et al .
however , the ultraviolet emission from a galaxy is heavily affected by the presence of even small amounts of dust , and the extinction corrections are highly uncertain ; for example , corrections for dust obscuration at @xmath30 range from factors of @xmath31 ( pettini et al . 1997 ) to factors of @xmath32 ( meurer et al .
1997 ) , though the more recent estimate of meurer , heckman , & calzetti ( 1999 ) lowers the latter to a factor of @xmath33
. only direct measurements of the reradiated light at submillimeter wavelengths can securely address the sfrd at high redshifts .
the true sfrd will receive contributions from both the ultraviolet / optical and the submillimeter .
we find that the submillimeter contribution to the sfrd in the @xmath6 range from our @xmath5 mjy sources is comparable to the observed ultraviolet / optical sfrd contribution .
the ultraviolet lyman - break galaxies are on the average undetected in the submillimeter at a @xmath34 level of @xmath35 mjy , and thus the lyman - break galaxy population is largely distinct from the bright submillimeter population ( chapman et al . 1999 ) .
if we assume that fainter submillimeter sources have the same redshift distribution and properties as the @xmath5 mjy sample , then the sfrd of the entire population contributing to the submillimeter background is about an order of magnitude higher than the observed ultraviolet / optical sfrd . the contribution from uligs to the sfrd increases by more than two orders of magnitude from @xmath7 to @xmath6 , which supports a scenario in which the distant submillimeter sources are the progenitors of massive spheroidal systems .
this rapid evolution in the ulig population sampled by scuba can be reproduced in models ( e.g. , blain et al . 1999b , c ) .
2 we present our radio sample and optical / nir imaging , along with our new scuba and lris observations . in
3 we determine our radio - selected submillimeter source counts . in
4 we introduce the predicted high redshift submillimeter - radio flux correlation and obtain millimetric redshift estimates for our sources . in
5 we use the complementary information from radio and submillimeter fluxes to gain insights into the characteristics of our radio - selected submillimeter source population . in
6 we compare the rest - frame colors of the radio sources that have spectroscopic redshifts with the rest - frame colors of our submillimeter sources and find that the latter are likely to fall in the extremely red object category . in 7 we calculate the luminosities , number densities , and sfrds of our submillimeter sources .
we compare the submillimeter contribution to the sfrd with contributions from radio and ultraviolet / optical wavebands over a range of redshifts . in 8 we summarize our main conclusions .
we take @xmath36 km s@xmath37 mpc@xmath37 and consider both @xmath38 , @xmath39 and @xmath40 , @xmath41 , which should cover the full range of possible cosmologies .
the present study is based on deep radio maps centered on the hdf that were observed with the vla at 1.4 ghz ( richards 1999b ) and 8.5 ghz ( richards et al .
the primary 1.4 ghz image covers a @xmath42 diameter region with an effective resolution of @xmath43 and a @xmath44 completeness limit of 40 @xmath0jy .
the 8.5 ghz images have an effective resolution of @xmath45 and a @xmath44 completeness limit of 8 @xmath0jy over a radius of @xmath46 from the hdf center , rising to 40 @xmath0jy at @xmath47 .
the 1.4 ghz hdf map was trimmed to match the 79.4 arcmin@xmath48 nir and optical imaging of the field obtained by barger et al .
( 1999a ) .
the absolute radio positions are known to @xmath49 rms in the hdf ; the alignment of the optical data to the radio data left residual astrometric uncertainties of @xmath50 ( richards et al .
there are 70 sources in our final radio sample .
[ tab1 ] table 1 gives the 1.4 ghz radio catalog for the central 79.4 arcmin@xmath48 region discussed above .
the first five columns are catalog number , ra(2000 ) , dec(2000 ) , 1.4 ghz radio flux , and radio flux uncertainty ( richards 1999b ) .
the remaining columns , to be discussed in the following subsections , are @xmath51 , @xmath52 , @xmath53 , @xmath54 , @xmath55 , and @xmath56 magnitudes , redshift , submillimeter flux , submillimeter flux uncertainty , and @xmath57 and 15 @xmath0 m isocam fluxes ( aussel et al .
the last column of table 1 gives the radio spectral index @xmath58 ( @xmath59 ) or limit for the sources where 8.4 ghz data are available and can be used to determine the origin of the radio emission ( richards 1999b ) .
inverted spectrum sources ( @xmath60 ) invariably have self - absorbed synchrotron emission associated with an agn .
flat spectrum sources ( @xmath61 ) can be produced by agn activity or by optically thin bremsstrahlung radiation from star formation at higher radio frequencies ( @xmath62 ghz ) .
steep spectrum sources ( @xmath63 ) consist of diffuse synchrotron emission often associated with either radio jets or star formation in galaxies .
in this paper we use the radio spectral index as a crude discriminator between agn and star formation activity .
we arbitrarily classify any source with an available spectral index that has @xmath64 ( there are six such sources in our sample ) to be primarily powered by agn activity .
wide - field and deep @xmath51 observations of the hdf and flanking fields were obtained using the university of hawaii quick infrared camera ( quirc ; hodapp et al .
1996 ) on the 2.2 m university of hawaii ( uh ) telescope and the 3.6 m canada - france - hawaii telescope ( cfht ) .
the @xmath51 ( @xmath65 @xmath0 m ) filter is described in wainscoat & cowie ( in preparation ) ; the empirical relation between @xmath51 and @xmath66 is @xmath67 , which simplifies to @xmath68 , assuming the median @xmath69 galaxy color ( barger et al .
in table 1 we use a @xmath70 limit of @xmath71 for our wide - field image and a @xmath70 limit of @xmath72 for our deep image in the area around the hdf proper .
deep johnson @xmath53 and kron - cousins @xmath52-band observations were made with the cfht over a much larger area using the uh8k ccd mosaic camera built by metzger , luppino , and miyazaki .
details of the above @xmath51 and optical observations can be found in barger et al .
( 1999a ) .
in table 1 we use @xmath70 limits of @xmath73 and @xmath74 . in february 1998 we observed four of the sources in our radio sample with the near - infrared camera ( nirc ; matthews & soifer 1994 ) on the keck i 10 m telescope .
these sources previously had only @xmath51 limits but were detected in the nirc observations .
conditions were photometric with seeing @xmath75 fwhm .
nirc has a @xmath76 insb array with @xmath77 pixels , giving a @xmath78 field of view .
we imaged at 2.1 @xmath0 m ( @xmath79 ) with total exposure times for each object of @xmath80 s. the data were obtained in sets of 120 s exposures , and the center of the field was moved in a @xmath81 grid pattern with @xmath82 on a side .
the centers of successive grids were moved by @xmath83 between each set .
a fifth source was subsequently detected with nirc with a longer exposure under non - photometric conditions .
the data were processed using median sky flats generated from the dithered images and calibrated onto the @xmath51 images using other galaxies in the field .
we used the low - resolution imaging spectrometer ( lris ; oke et al .
1995 ) on the keck ii 10 m telescope in march 1997 and february 1998 to obtain @xmath55-band and kron - cousins @xmath54-band images , respectively , of a strip region @xmath84 in size that crosses the hdf .
the total exposure times were 1680 s and 1600 s for @xmath55 and @xmath54 , respectively , and the seeing was @xmath85 in @xmath54 and @xmath86 in @xmath55 .
the @xmath70 limits are @xmath87 and @xmath88 .
details of the observations can be found in cowie & hu ( 1998 ) and cowie , songaila , & barger ( 1999 ) .
finally , a deep @xmath89 ) image of an area @xmath90
arcmin@xmath48 centered on the hdf - proper was obtained using the orbit ccd on the uh telescope .
the @xmath70 limit for the @xmath56 image is @xmath91 .
details of the @xmath56 observations can be found in wilson et al .
( in preparation ) .
all photometric magnitudes were measured in @xmath82 diameter apertures and then corrected to @xmath92 diameter ( near total ) magnitudes following the procedures of cowie et al .
( 1994 ) .
serjeant et al . (
1997 ) obtained deep isocam observations of the hdf and flanking fields at 6.75 and 15 @xmath0 m .
the aussel et al .
( 1999 ) reductions of the data produced a main source list of 49 objects ( @xmath93 ) and a supplementary list of an additional 51 objects ( @xmath2 for 15 @xmath0 m and @xmath44 for 6.75 @xmath0 m ) . despite the large point spread function ( @xmath94 at 15 @xmath0 m ) , in most cases
the optical or nir identification of the isocam source was straightforward .
the redshifts for sources in the sample are in the range @xmath95 to @xmath96 ( median @xmath97 ) .
we used lris during a two night run on keck ii in march 1999 to obtain spectroscopic observations of a sub - sample of the 1.4 ghz sources in a strip region centered on the hdf .
although the hdf and flanking field region has been intensely studied by a number of groups ( see cohen et al .
1999 for a summary and references ) , most of the sources in our sample had not been previously observed due to their faint optical fluxes .
we used @xmath98 wide slits and the 400 lines mm@xmath37 grating blazed at 8500 , which gives a wavelength resolution of @xmath99 and a wavelength coverage of @xmath100 .
the wavelength range for each object depends on the exact location of the slit in the mask but is generally between @xmath101 and 10000 .
three of the slit masks were constructed at a position angle of @xmath102 , and the remaining three were nearly identical versions constructed at a position angle of @xmath103 .
this procedure enabled us to obtain sufficient wavelength coverage for all the objects in our sample , including those that fell close to the edges of the masks .
the observations were 1.5 hr per slit mask , broken into three sets of 30 minute exposures .
three hdf slit masks were observed per night . some of the objects were in all of the slit masks and
hence were observed for 9 hrs ( see table [ tab2 ] ) .
conditions were photometric with seeing @xmath104 fwhm both nights .
the objects were stepped along the slit by @xmath105 in each direction , and the sky backgrounds were removed using the median of the images to avoid the difficult and time - consuming problems of flat - fielding lris data .
details of the spectroscopic reduction procedures can be found in cowie et al .
( 1996 ) .
ccccc 16 & 23.39 & 25.30 & & 6 + 17 & 18.75 & 22.49 & 1.013 & 9 + 23 & 16.36 & 19.09 & 0.456 & + 24 & 19.10 & 22.29 & 1.219 & + 25 & 19.25 & 22.71 & & 3 + 26 & 22.29 & 25.30 & & 6 + 28 & 21.22 & 25.30 & & 6 + 30 & 21.23 & 25.30 & & 3 + 31 & 18.06 & 20.86 & 0.857 & + 32 & 17.85 & 21.04 & 1.013 & + 33 & 21.65 & 24.72 & & 3 + 35 & 18.36 & 20.76 & 0.961 & + 36 & 22.60 & 25.30 & & 3 + 37 & 21.00 & 23.20 & & 3 + 38 & 18.74 & 21.09 & 0.475 & + 40 & 17.36 & 19.88 & 0.410 & + 41 & 21.50 & 25.30 & & + 42 & 16.52 & 18.70 & 0.321 & + 43 & 19.31 & 21.88 & 1.275 & + 44 & 21.50 & 25.30 & & 6 + 46 & 22.60 & 25.30 & & 9 + 47 & 19.55 & 22.64 & 0.474 & + 48 & 18.52 & 20.92 & 0.761 & + 50 & 20.14 & 24.81 & 0.884 & + 51 & 19.21 & 21.89 & 1.243 & + 52 & 18.18 & 20.55 & 0.902 & + 53 & 20.29 & 24.70 & & 3 + 54 & 20.76 & 25.13 & & 3 + 56 & 17.98 & 19.87 & 0.422 & + 59 & 19.51 & 22.66 & & 3 + 60 & 20.08 & 22.90 & & 6 + 61 & 20.38 & 22.90 & & + 62 & 17.36 & 19.80 & 0.558 & + 64 & 17.59 & 19.90 & 0.411 & 7.5 + 65 & 20.90 & 25.30 & & 3 + 66 & 19.39 & 22.14 & 1.019 & 3 + 69 & 21.50 & 25.30 & & 1.5 + in table [ tab2 ] we list all the objects in the radio sample that fall in the lris strip region .
the columns are radio catalog number from table 1 , @xmath51 mag , @xmath52 mag , redshift , and exposure time for any object targeted in our march 1999 spectroscopic survey . over an area @xmath106
arcmin@xmath48 , 19 of the 37 radio sources now have secure redshift identifications ; all are at @xmath25 .
figure [ figzhkandzi]a , b shows @xmath51 versus redshift and @xmath52 versus redshift .
spectroscopic redshifts are in general relatively straightforward to obtain for radio objects with @xmath24 .
we note that waddington et al .
( 1999 ) claim a redshift identification of @xmath107 for object 30 based on a ly@xmath108 detection .
however , the position of their detection is @xmath109 away from the radio source position and counterpart optical galaxy , and so an association with the radio source is not secure .
the radio source is detected in the mid - ir , which would suggest @xmath110 , consistent with the redshift estimated using the millimetric redshift technique described in [ secze ] .
however , we did not see any strong [ oii ] 3727 feature in our spectrum from @xmath111 , which would suggest @xmath26 .
thus , the redshift of this object remains uncertain . in all of our slit - masks ( equivalent to a 9 hr integration ) we also included the position of the brightest scuba source , hdf850.1 , from the deep 850 @xmath0 m map of the hdf - proper by h98 .
we centered our slit across the position of the 1.3 mm detection of hdf850.1 reported by downes et al .
( 1999 ) , which also coincides with the 8.5 ghz supplemental radio source 3651 + 1226 ( richards et al .
we oriented the slit such that it fell across both the arc - like feature 3 - 593.0 that is favored by downes et al .
( 1999 ) as the optical counterpart to hdf850.1 and the nearby red galaxy 3 - 586.0 . however , we were unable to determine a secure redshift for either source from our spectroscopic data .
our scuba jiggle map observations were flexibly scheduled in mostly excellent observing conditions during two runs in april and june 1999 for a total of five observing shifts .
the maps were dithered to prevent any regions of the sky from repeatedly falling on bad bolometers .
the chop throw was fixed at a position angle of @xmath102 so that the negative beams would appear 45 arcsec on either side east - west of the positive beam .
the data were reduced using beam weighted extraction routines that included both the positive and negative portions of the chopped images , thereby increasing the effective exposure times .
regular `` skydips '' ( lightfoot et al . 1998 ) were obtained to measure the zenith atmospheric opacities at 450 and 850 @xmath0 m , and the 225 ghz sky opacity was monitored at all times to check for sky stability .
the median 850 @xmath0 m optical depth for all nights together was 0.265 .
pointing checks were performed every hour during the observations on the blazars 0954 + 685 , 1418 + 546 , 0923 + 392 , or 1308 + 326 .
the data were calibrated using 30 arcsec diameter aperture measurements of the positive beam in beam maps of the primary calibration source , mars , and one of three secondary calibration sources , crl618 , irc+10216 , or oh231.8 .
the data were reduced in a standard and consistent way using the dedicated scuba user reduction facility ( surf ; jenness & lightfoot 1998 ) . due to the variation in the density of bolometer samples across the maps , there is a rapid increase in the noise levels at the very edges .
we have clipped the low exposure edges from our images .
we present our scuba maps in fig .
[ figmaps ] .
we also re - reduced the archival hdf - proper scuba data of h98 in order to make a consistent analysis with the present data .
we could only make use of the maps that were taken with a fixed ra chop ( 90 per cent of the data sample ) , which included 34 hours with a 30 arcsec chop throw and 28 hours with a 45 arcsec chop throw .
we combined these data separately to form two independent maps . in table 1
we quote submillimeter fluxes determined from the weighted average of measurements made in each map .
the surf reduction routines arbitrarily normalize all the data maps in a reduction sequence to the central pixel of the first map ; thus , the noise levels in a combined image are determined relative to the quality of the central pixel in the first map . in order to determine the absolute noise levels of our maps , we first eliminated the @xmath112 real sources in each field by subtracting an appropriately normalized version of the beam profile .
we then iteratively adjusted the noise normalization until the dispersion of the signal - to - noise ratio measured at random positions became @xmath113 .
our noise estimate includes both fainter sources and correlated noise .
we centered on the positions of the radio sources and measured the submillimeter fluxes in both the positive and negative beams .
the extracted fluxes were calibrated to the 30 arcsec diameter aperture fluxes of the brightest sources .
the brighter submillimeter sources were subtracted from the maps before the fainter sources were extracted . in cases where the positions of two radio sources are very close , we may be slightly overestimating the submillimeter flux allocated to the bright source with this procedure and underestimating that allocated to the faint source , but since the total submillimeter flux should be reasonable , we will not be making any gross errors in our later estimates of the total star formation rate . from the total sample of 70 radio - selected galaxies in the @xmath27 square arcminute central region of the radio map , we take the 22 with @xmath23 to be our optical / nir - faint radio sample , for which there are now submillimeter observations of 19 . even though our scuba observations were relatively shallow ( the 850 @xmath0 m @xmath2 limit was 6 mjy ) , we detected 5 of the 15 optical / nir - faint radio sources that we observed ; a sixth significant source ( @xmath114 mjy ) was detected in the hdf - proper scuba map of h98 ( hdf850.2 in their notation ) .
the submillimeter flux we measure for the brightest source ( hdf850.1 ) in the hdf - proper map is @xmath115 mjy , which is slightly lower than the value of @xmath116 found by h98 but is consistent within the statistical and systematic errors .
the jiggle map observing mode enabled simultaneous observations of a large fraction ( 31/48 ) of the optical / nir - bright radio sources , none of which were detected .
there are two @xmath5 mjy submillimeter sources in our jiggle maps that were not in the radio sample . in fig .
[ figcolor]a , b we show the 1.4 ghz fluxes of the radio sources in our sample versus their @xmath51 and @xmath52 magnitudes , indicating those with measured redshifts and those with submillimeter detections .
we find that the radio selection technique is effective in locating the majority of bright submillimeter sources .
we document this in fig .
[ figcounts1 ] where we compare the combined 850 @xmath0 m source counts from blank field submillimeter surveys ( h98 ; eales et al . 1999 ; barger , cowie , & sanders 1999 ) with our new radio - selected 850 @xmath0 m source counts .
a source of flux strength @xmath117 contributes @xmath118 to the counts per unit area , where @xmath119 is the area over which there is @xmath120 sensitivity to @xmath117 .
the cumulative counts , @xmath121 , are given by the sum of the inverse areas of all sources brighter than @xmath122 .
however , for our radio - targeted search , the effective area ( 65 square arcminutes ) is greater than if we had done an untargeted search with the same number of pointings and is given by the fraction ( number @xmath23 sources observed)/(total number @xmath23 sources in sample ) times the radio survey area ( 79.4 square arcminutes ) . because of the correspondence between the optical / nir - faint radio population and the bright submillimeter population , we do not need to survey the entire radio field in order to observe the bright submillimeter population .
it is interesting to speculate whether the correspondence between optical / nir - faint radio sources and bright submillimeter sources also holds at fainter submillimeter flux levels .
a necessary requirement for this to be the case is that the surface density of faint radio sources be comparable to or exceed the surface density of submillimeter sources .
extrapolating a power - law parameterization of the 1.4 ghz counts over the range @xmath123 @xmath0jy ( richards 1999b ) down to a 1 @xmath0jy cut - off gives @xmath124jy ) @xmath125 arcmin@xmath126 .
the empirically estimated surface density of submillimeter sources ( barger , cowie , & sanders 1999 ) is @xmath127 arcmin@xmath126 .
a possible counter - argument to a close correspondence between radio and submillimeter sources at faint submillimeter flux levels comes from the study by smail et al .
( 1999b ) who compared their lensed submillimeter survey data with radio data .
they found that the carilli & yun ( 1999 ) radio - submillimeter spectral indices differed for their bright and faint submillimeter subsamples .
they suggested that the difference could arise if the faint submillimeter sources lie at higher redshifts or have different fractions of radio - loud agn or have different dust temperatures .
however , with the low statistics , strong conclusions can not be drawn at this time .
optical spectroscopic surveys of submillimeter galaxies are difficult because of the very different behaviors of the @xmath66-corrections in the optical and submillimeter and the poor submillimeter resolution .
spectroscopic surveys to date ( barger et al .
1999b ; lilly et al .
1999 ) give potentially conflicting results and are very limited in size .
photometric redshift estimates of submillimeter sources have been made from candidate optical counterparts ( h98 ; lilly et al .
1999 ) , but these are questionable both because the counterpart identification is not secure and because the seds of the submillimeter sources may be unlike those of the optical sources used in making the redshift estimates due to dust extinction and possible agn contributions . consequently , a new approach is required if the positions of the submillimeter sources are to be reliably determined and their nature and redshift distribution understood .
the remarkably tight local correlation between the global fir and nonthermal radio luminosities provides a promising alternative method for identifying and studying individual submillimeter sources , provided that the correlation holds to high redshift . in [ secsmmcounts ] we found that targeting with scuba optical / nir - faint 1.4 ghz sources is an efficient technique for identifying the majority of the bright submillimeter source population .
our results indicate that a large fraction of bright sources in submillimeter surveys have extremely faint optical / nir counterparts and hence are inaccessible to optical spectroscopy ( see also smail et al .
this conclusion is consistent with results from the barger et al .
( 1999b ) spectroscopic survey of lensed submillimeter sources discussed in the introduction .
although we are unable to obtain spectroscopic redshifts for the optical / nir - faint radio - selected submillimeter sources , we can use the submillimeter - to - radio flux ratios to obtain millimetric redshift estimates .
figure [ figarp ] illustrates how the slope of a dusty galaxy s sed changes abruptly at frequencies higher than 100 ghz .
below 30 ghz synchrotron radio emission is dominant , free - free emission is largest in the range @xmath128 ghz , and thermal dust emission dominates above 200 ghz . because of the opposing spectral slopes of the blackbody spectrum in the submillimeter and the synchrotron spectrum in the radio , the submillimeter - to - radio flux ratio rises extremely rapidly with increasing redshift ( see fig .
[ figulig ] ) .
carilli & yun ( 1999 ; hereafter cy99 ) have therefore suggested using the submillimeter - to - radio flux ratio as a redshift estimator . the dust emission from local luminous infrared galaxies is well described by optically thin single - temperature modified blackbodies with extinction coefficient @xmath129 , where @xmath130 to 2 .
the dust temperatures derived lie between 30 and 60 k. arp 220 , with its high bolometric luminosity produced almost entirely from starburst activity ( downes & solomon 1998 ) , is an appropriate prototype for high redshift submillimeter sources , as we will justify subsequently . over the interval @xmath131 ghz ( @xmath132 m )
arp 220 s sed is well represented by a modified blackbody with @xmath133 and a dust temperature @xmath134 k ; the luminosity of this dominant cooler component is @xmath135 ( klaas et al .
1997 ) . in terms of the blackbody distribution @xmath136 the submillimeter flux for an arp 220-like galaxy at redshift @xmath137
is given by @xmath138 ^ 2 \label{smmexact}\ ] ] where @xmath139 is the observed submillimeter frequency , @xmath140 is the dust temperature , and @xmath141 is the luminosity distance . we allow for an overall strength factor , @xmath142 , relative to arp 220 , for which @xmath143 .
the above normalization constant is fixed by a @xmath144 fit to the arp 220 flux measurements . in the rayleigh - jeans long - wavelength limit , the submillimeter flux is approximately given by @xmath145 ^ 2 \label{smmlong}\ ] ] which is normalized to the measured 375 ghz ( 800@xmath0 m ) flux value for arp 220 from rigopoulou , lawrence , & rowan - robinson ( 1996 ) ; here @xmath146 .
the approximation in eq .
[ smmlong ] is reasonably accurate for @xmath147 ; at higher @xmath137 it extrapolates above the result of eq .
[ smmexact ] .
the radio flux is given by @xmath148 ^ 2 \label{radio}\ ] ] which is normalized to the measured 8.4 ghz flux for arp 220 from condon et al .
( 1991 ) . since both @xmath149 and @xmath150 depend linearly on the star formation rate , the scale factor @xmath142 is the same . in local star forming galaxies
there is an observed spectral flattening at 1.4 ghz due to free - free absorption ( condon 1992 ) . since 1.4 ghz measurements at higher redshifts sample higher frequencies that are not sensitive to these absorption effects , we adopt the standard value @xmath151 from condon ( 1992 ) to extrapolate local 8.4 ghz fluxes to 1.4 ghz fluxes , in order to be consistent with high redshift observations . the ratio of eqs .
[ smmlong ] and [ radio ] with @xmath152k gives @xmath153 a relation of this type was obtained by cy99 based on the sed of m82 . in their relation , the factor of 1.1 in eq
. [ ratio ] is replaced by 0.25 and the power - law index @xmath154 is replaced by 4.3 .
however , m82 , with more than an order of magnitude lower luminosity than arp 220 ( @xmath155 ; hughes , dunlop , & rawlings 1997 ) , is not a ulig , and so we would argue that our result is more appropriate for the distant submillimeter sources .
in addition to their analytic result , cy99 give empirical results for the flux ratio versus redshift for both arp 220 and m82 .
other authors have used the span of the four cy99 curves in their fig . 1 to estimate redshift ranges
; however , the use of m82 again introduces an inappropriate uncertainty in obtaining redshifts .
inverting eq .
[ ratio ] , the value of @xmath137 can be related to the flux ratio by @xmath156 which holds for @xmath147 . for higher redshifts
the ratio of eqs .
[ smmexact ] and [ radio ] must be used . the flux ratio method for estimating redshifts has the advantage that the luminosity distance drops out , and
thus there is no dependence on @xmath157 or on the cosmology .
we make the assumption that unevolved local uligs are representative of the distant ulig population .
we show in the following that the arp 220 sed provides a good representation of the submillimeter - to - radio flux ratios of the ensemble of local uligs placed at appropriate redshifts .
we use a compilation of submillimeter measurements at 3000 , 857 , 667 , and 375 ghz from rigopoulou et al .
( 1996 ) to infer 353 ghz fluxes at redshifts of 7.5 , 1.44 , 0.9 , and 0.06 .
fluxes at 1.4 ghz were extrapolated from the 8.4 ghz measured values of condon et al .
( 1991 ) using a standard synchrotron emission spectrum with @xmath158 , as justified earlier .
we plot the mean 353 ghz to 1.4 ghz flux ratios versus redshift as solid diamonds in fig .
[ figulig]a with ranges from the minimum to maximum flux ratios .
the number of objects in the bins are , in order of decreasing redshift , 25 , 2 , 4 , and 7 .
the data are to be compared with the redshifted arp 220 flux ratio calculated from eqs .
[ smmexact ] and [ radio ] and plotted as a solid line in the figure .
the spread in measured ratios is about a multiplicative factor of two relative to arp 220 , as indicated by the dashed curves .
thus , the redshifted arp 220 flux ratio is likely to approximate the flux ratios of uligs at high redshifts to this accuracy .
the above conclusion is strengthened by fig .
[ figulig]b where we have plotted as solid ( star forming galaxies ) and open ( agn ) squares available submillimeter observations with corresponding radio and redshift information .
the star forming galaxies fall neatly on the curves .
the average submillimeter - to - radio flux ratio of the galaxies in our survey which have spectroscopic redshifts and have been observed in the submillimeter are shown in two redshift bins as the large filled diamonds with @xmath34 uncertainties .
the lowest point is consistent with a @xmath2 null detection , but in the higher redshift bin there is a strong positive detection consistent with the arp 220 ratio , showing that these objects have cool dust emission and obey the fir - radio relation at these redshifts .
the short horizontal lines show the redshift ranges for our six significant radio - selected submillimeter detections ; all are in the @xmath6 redshift range .
the top portion of table [ tab3 ] gives the fluxes and the millimetric redshifts and ranges for these six sources , as well as their bolometric luminosities for both @xmath39 and @xmath41 ( see [ secfluxz ] ) .
crrrrrrr 33 & 11.9 & 0.124 & 1.0 & 1.5 & 2.1 & 5.0 & 9.1 + 49 & 10.7 & 0.324 & 1.6 & 2.3 & 3.1 & 4.5 & 9.0 + 7 & 7.8 & 0.151 & 1.3 & 1.8 & 2.5 & 3.3 & 6.3 + 11 & 7.5 & 0.148 & 1.2 & 1.8 & 2.4 & 3.2 & 6.0 + 13 & 7.1 & 0.071 & 1.8 & 2.4 & 3.3 & 3.0 & 6.0 + 46 ( hdf850.2 ) & 2.4 & 0.046 & 1.3 & 1.8 & 2.5 & 1.0 & 1.9 + + & 19.0 & @xmath159 & @xmath160 & @xmath161 & @xmath162 & @xmath163 & @xmath164 + & 8.8 & @xmath159 & @xmath165 & @xmath166 & @xmath167 & @xmath168 & @xmath169 + ( hdf850.1 ) & 5.4 & @xmath159 & @xmath170 & @xmath171 & @xmath172 & @xmath173 & @xmath174 + the ulig sources plotted in fig .
[ figulig ] have a range of emissivity indices , @xmath175 , and dust temperatures in the vicinity of 50 k. it is therefore quite remarkable that all the sources follow a constrained envelope around the arp 220 submillimeter - to - radio flux ratio .
the implication of this empirical correlation is that for practical purposes a temperature - redshift degeneracy ( blain 1999 ) does not present a serious problem to millimetric redshift estimation .
a recent empirical analysis by carilli & yun ( 2000 ) , which appeared subsequent to the submission of our paper , finds conclusions that are consistent with our analysis above .
in our analysis we assume a flat universe , @xmath176 , where @xmath177 is the matter density and @xmath178 the vacuum density . the absolute flux values in eqs .
[ smmexact ] and [ radio ] depend on the cosmology through the luminosity distance ( carroll , press , & turner 1992 ) @xmath179^{1/2}\ ] ] the basic physics underlying the joint use of radio and submillimeter observations is contained in figs .
[ figfalloff ] and [ figsvss ] .
figure [ figfalloff ] shows predicted arp 220 radio and submillimeter fluxes versus redshift for the @xmath38 , @xmath39 cosmology ( @xmath180 km s@xmath37 mpc@xmath37 ) with selected overall strengths @xmath181 , 1 , 3 , and 6 relative to arp 220 .
@xmath182 is flat with redshift for @xmath11 due to the negative @xmath66-correction , whereas @xmath183 falls sharply with increasing redshift due to the @xmath184 dependence .
these behaviors are characteristic of other cosmologies as well .
thus , radio surveys detect a high proportion of low redshift sources whereas submillimeter surveys offer comparable sensitivity to both moderate ( @xmath185 ) and high redshifts . in fig .
[ figsvss ] we illustrate the dependence on cosmology of the absolute fluxes in the @xmath183 versus @xmath182 plane .
figure [ figsvss]a shows predicted arp 220 fluxes for the @xmath39 cosmology , and fig .
[ figsvss]b shows the same for the @xmath41 cosmology favored by recent distant supernovae type ia observations ( perlmutter et al .
1999 ; riess et al .
redshift labels are given on the @xmath143 , 3 , and 6 curves . in fig .
[ figsvss]a the bright submillimeter sources fall near the @xmath186 contour , indicating that these sources have luminosities several times the luminosity of the arp 220 prototype . for the @xmath41 cosmology we would infer somewhat higher luminosities from the curves , although the redshift estimates remain the same . from fig .
[ figsvss ] , we see that any submillimeter sources above our detection threshold of 6 mjy would not be detectable in the radio if @xmath187 . of the significant submillimeter sources detected in our survey , two had no radio counterparts and one ,
hdf850.1 , had only a supplemental 8.5 ghz radio detection .
based on the millimetric redshift estimator , these sources are potentially at very high redshift .
the bottom three lines of table [ tab3 ] give the fluxes and the millimetric redshifts and ranges , along with the bolometric luminosities for @xmath39 and @xmath41 , for the three significant scuba detections without corresponding 1.4 ghz detections ; here the numerical entries are based on the 0.040 mjy ( @xmath44 ) radio flux limit , which corresponds to the lowest possible redshift and @xmath188 .
since we have covered only one - third of our @xmath27 square arcminute area with our targeted scuba observations , there could in principle be more high redshift submillimeter sources with no radio detections .
however , our present radio counts already saturate the bright submillimeter counts distribution from the combined blank field submillimeter surveys ( see fig .
[ figcounts1 ] ) and hence argue against the probability of finding many such additional sources .
in contrast , sources of arp 220 or sub - arp 220 strength could be seen in the radio but not in the submillimeter with our present 6 mjy threshold if @xmath189 .
figure [ figfalloff]b shows that we have many such candidates . our targeted scuba observations of optical / nir - faint radio sources are therefore selecting the high redshift end of the faint radio source population .
in fig . [ figrest ] we plot rest - frame ab(2800)ab(8140 ) color versus redshift for the radio sample with spectroscopic identifications .
the colors of field galaxies in the hdf and ssa22 fields are indicated with tiny solid circles to illustrate how the radio galaxies ( solid triangles for sources with @xmath64 but otherwise solid circles ) generally have very red ab colors .
these colors range from @xmath190 to @xmath191 .
there is no evidence for a color trend with redshift .
the rest - frame ab colors of the submillimeter galaxies with @xmath51 detections ( solid pentagons ) , although mostly lower limits , are not inconsistent with the colors of the lower - redshift radio sources . at @xmath192 , ab(2800)ab(8140 ) roughly corresponds to ab(@xmath52)ab(@xmath51 ) , and thus submillimeter sources with similar colors to the radio sources would have observed vega - based @xmath193 colors in the range @xmath194 , which would mostly place them in the extremely red object ( ero ) category ( @xmath195 or @xmath196 ; e.g. , mccracken et al .
1999 and references therein ) .
this result is consistent with the recent detections of bright ( @xmath197 ) eros as submillimeter sources ( cimatti et al .
1998 ; dey et al .
1999 ; smail et al .
. in general , however , the submillimeter sources are so optically faint that identifying them as eros is difficult .
the luminosities of our submillimeter sources were obtained by scaling the arp 220 luminosity at redshift @xmath137 by the relative source strengths , @xmath142 ; these are given in table [ tab3 ] .
the fir luminosity provides a direct measure of the current star formation rate ( sfr ) ; with present detection capabilities , the radio provides greater sensitivity to sfrs at low redshifts ( @xmath198 ) while the submillimeter is superior at high redshifts . in this
sense radio and submillimeter data complement each other in getting the sfr over the full range of redshifts .
this relative capability is illustrated in fig .
[ figradiolum ] where the strength factors , @xmath142 , as determined from the radio luminosities relative to redshifted arp 220 luminosities , are plotted versus redshift in the @xmath39 cosmology . here
the dotted curve represents the radio threshold of 40 @xmath0jy , and the solid line at a value of two represents our 6 mjy submillimeter threshold of roughly two times the luminosity of arp 220 for the @xmath39 cosmology ( see fig .
[ figsvss ] ) .
these two threshold curves cross at @xmath199 .
since the bulk of the radio detections are at @xmath25 and have strength factors @xmath200 , their non - detection in the submillimeter is as expected ; our @xmath201 sources are detected in both the radio and submillimeter . at redshifts
@xmath202 the submillimeter may provide the only access to distant uligs .
figure [ figradiolum ] also provides a means to establish the completeness of the ulig number distribution versus redshift .
our distant ulig sample is complete over the redshift range @xmath6 ( except possibly for the 4 out of 22 optically - faint radio sources that were not observed ) because these @xmath6 sources are above both radio and submillimeter detection thresholds . at @xmath203
there is the risk that we may lose sources that fall below the rising 1.4 ghz threshold relative to the arp 220 luminosity at redshift @xmath137 .
the space density of our five significant @xmath204 mjy submillimeter sources in the redshift range @xmath6 is @xmath205 in units @xmath206 . in comparison ,
the space density of local uligs ( @xmath207 ; kim & sanders 1998 ) with bolometric luminosities above @xmath208 is only @xmath209 for @xmath39 and @xmath210 for @xmath41 , about a factor of 200 and 80 lower than our high redshift results .
the discrepancy in space density is still larger if we compare with local uligs that are comparably luminous to the @xmath5 mjy population .
thus , enormous evolution in the ulig population must take place between @xmath201 and the present .
it is interesting to note a possible empirical similarity between the redshift distribution of quasars and that of submillimeter sources , which might be expected if quasars are a successor stage to merger processes forming submillimeter sources ( sanders & mirabel 1996 ) .
the relative space density of a complete radio - selected quasar sample ( shaver et al . 1998 ) , which should be unaffected by dust obscuration , is approximately described by a gaussian as a function of redshift , given by @xmath211 the peaking of this distribution at @xmath199 and the rapid fall - off at low redshift seems consistent with our measured submillimeter redshift distribution .
recent rosat x - ray studies have found a very high redshift tail ( @xmath212 ) to the quasar distribution that may be above the gaussian expectation ( hasinger 1998 ) .
since the rosat sources were not seen in the optical , it is plausible that they are dust obscured .
similarly , we have found a few submillimeter sources that plausibly lie at @xmath203 .
the observed ultraviolet light can most simply be translated to a metal production rate where , as first pointed out in cowie ( 1988 ) , there is an extremely tight relation since both metals and ultraviolet light are produced by the same massive stars .
however , it is usual ( particularly in the fir ) to translate the metal production rate to a total stellar mass production rate , which may be more intuitively interpreted .
unfortunately , this extrapolation requires knowledge of the shape of the initial mass function ( imf ) , which is quite uncertain . in the following
we shall assume a salpeter imf ( @xmath213 , @xmath214 ) .
the sfr is related to the rest - frame 2800 luminosity per unit frequency by @xmath215 with @xmath216 in ergs s@xmath37 hz@xmath37 and @xmath217 in m@xmath218 .
the cowie ( 1988 ) estimate of @xmath219 from nucleosynthesis arguments and galaxy sed modelling is @xmath220 for the above imf limits .
later songaila , cowie , & lilly ( 1989 ) suggested that @xmath219 should be about a factor 1.7 lower based on stellar synthesis .
madau et al .
( 1996 ) obtained a value of @xmath221 based on the evolutionary models of bruzual & charlot ( 1993 ) .
the relatively invariant estimates of @xmath219 are a direct consequence of the nucleosynthesis arguments underlying the modelling ; the primary uncertainty in @xmath217 lies in the stellar imf .
we adopt the madau et al .
value for consistency with other recent papers that follow this normalization . in terms of the solar luminosity , the sfr relation is then @xmath222 where @xmath223 .
the conversion of the bolometric luminosity to a total stellar mass production rate is an even more invariant prediction of the nucleosynthesis arguments since knowledge of the exact shape of the ultraviolet sed of the galaxies is not needed .
the bolometric luminosity in units of the solar luminosity is @xmath224 where @xmath225 is the bolometric correction . for a flat @xmath226 over the wavelength range 912 to 22000 , @xmath227 .
rowan - robinson et al . ( 1997 ) computed a bolometric correction of 3.5 from the bruzual & charlot ( 1993 ) models , which we adopt .
then the sfr is related to @xmath188 by @xmath228 @xmath217 would be increased by a factor of 3.3 if we had instead assumed a miller - scalo imf over the same mass range .
equations [ mdot2 ] and [ mdot3 ] therefore provide a self - consistent description of the mass production rates seen in the optical and submillimeter .
light which directly escapes the galaxy can be mapped with the ultraviolet light , while reprocessed light , which escapes the galaxy in the submillimeter , can be calibrated with eq .
[ mdot3 ] .
we now use the arp 220 bolometric luminosity @xmath229 ( klaas et al . 1997 ) and the 8.4 ghz radio flux of 0.148 jy ( condon et al . 1991 ) to calibrate the sfrs for our radio sources .
we find @xmath230 ; thus @xmath231 where the normalization factor is in units of ergs s@xmath37 hz@xmath37 . from the equations of condon ( 1992 ) translated to the same assumptions about the imf , we would obtain a normalization of approximately @xmath232 or about a factor of 3.5 smaller than our result .
our normalization translated with the @xmath233 spectral shape to 1.4 ghz is @xmath234 using eq .
[ radiosfr ] , we can convert our radio luminosities into sfrds .
we know that 95 per cent of the micro - jansky radio sources in the hdf region have been resolved at @xmath235 resolution with the multi - element radio linked interferometer ( merlin ) at 1.4 ghz .
the median angular size for the radio emission is @xmath236 ( richards 1999a ) , which suggests radio emission on galactic or sub - galactic size scales .
thus , the radio emission in most of these systems may originate primarily from star formation , although we can not exclude contributions to the radio flux densities from embedded agn .
we divide our radio sources with spectroscopic redshifts into two bins , @xmath237 and @xmath238 , excluding sources with radio spectral indices @xmath239 that might be agn . of the 38 presumed star forming objects with @xmath240 in our sample , 23 have spectroscopic redshifts .
most of the remaining 15 sources have not been observed and thus may be expected to follow the same redshift distribution as the identified sources .
we therefore calculate the differential comoving volume over a survey area @xmath241
the volume is integrated from @xmath242 , the lower redshift limit of the bin , to @xmath243 , either the radio luminosity limit of the survey or the upper redshift limit of the bin , whichever is smaller .
the comoving volume element in mpc@xmath244 is @xmath245\ ] ] where @xmath246 is the area in arcmin@xmath48 and @xmath141 is in mpc .
we find that for the @xmath247 bin , the sfrd in units @xmath248 is @xmath249 and for the @xmath238 bin , @xmath250 where the uncertainties are poissonian based on the number of sources .
these represent lower limits to the sfrd since we have not attempted to correct for the contributions of faint sources below our flux limits .
the sfrd from our submillimeter sources can be estimated under the assumption that star formation dominates agn contributions .
we can calculate the contribution of the @xmath204 mjy submillimeter sources at @xmath6 to the sfrd using eq .
[ mdot3 ] ; we find @xmath251 unlike the number distribution versus redshift , the sfrd needs large corrections for completeness due to the fact that we are detecting only relatively bright sources here ( at about six times the luminosity at which the submillimeter background is primarily resolved ) , and the distribution @xmath252 increases rapidly as @xmath122 decreases . to estimate the completeness correction
, we assume that the dependences of @xmath253 on @xmath122 and @xmath137 factorize , @xmath254 .
this is a plausible assumption in the submillimeter where the fluxes are nearly independent of redshift ; nonetheless , in view of the smail et al .
( 1999b ) study , this assumption remains to be confirmed .
we can determine the completeness correction for the sfrd using the empirical number distribution at 850 @xmath0 m versus @xmath122 @xmath255 that describes the measured submillimeter counts above 2 mjy ( barger , cowie , & sanders 1999 ) .
we are making the assumption that the flux to @xmath256 conversion based on arp 220 applies in the low submillimeter flux region ; the justification is that even the dominant @xmath113 mjy population are near - ulig sources .
the completeness correction over all submillimeter fluxes is therefore the measured @xmath257 m extragalactic background light ( ebl ) divided by the @xmath257 m light above 6 mjy .
the @xmath257 m ebl measurement of @xmath258 mjy deg@xmath126 from puget et al .
( 1996 ) and @xmath259 mjy deg@xmath126 from fixsen et al .
( 1998 ) imply correction factors in the submillimeter of 11 and 15 , respectively .
thus , the estimated total submillimeter contribution to the sfrd in units of @xmath248 is @xmath260 where we have used the factor of 11 completeness correction .
more speculatively , we can determine the sfrd from higher redshift sources using our two @xmath5 mjy submillimeter sources without radio counterparts .
in this case we use the volume from @xmath261 and the actual area surveyed in the submillimeter .
we find @xmath262 after including the factor of 11 completeness correction , this becomes @xmath263 h98 used photometric redshift estimates to infer that four of their five @xmath264 mjy sources were in the redshift range @xmath265 .
while these source identifications were problematic , it appears likely from the present work that the redshifts do lie in this rough redshift range . using our parameters and a scaled arp 220 sed
, we find that the sfrd for their sources with @xmath39 is @xmath266 .
if we make a completeness correction to include the contribution below 2 mjy , we obtain sfrd=@xmath267 , in good agreement with our result in eq .
[ eqtruesfrdsmma ] .
the presence of a substantial fraction of agn - dominated ulig sources would reduce the above sfrds . in a recent near - infrared spectroscopic study of 64 local uligs , veilleux , sanders , & kim ( 1999 ) found agn characteristics in @xmath268 per cent of the sample , which increased to @xmath269 per cent for the sample with @xmath270 .
thus , our @xmath5 mjy contributions to the sfrd may need to be reduced by a factor @xmath271 .
however , the lower agn fraction in fainter uligs seen locally suggests that agn contamination may be less of an issue for the extrapolated sfrd of the whole submillimeter population .
the determination of the sfrd from optical observations has been a subject of intense investigation .
observations first indicated a rather rapid rise in the sfrd from @xmath272 followed by a sharp decline at higher redshifts with the peak sfrd being @xmath273 ( madau et al .
a recent modification in the inferred optical sfrd at low redshifts was made by cowie , songaila , & barger ( 1999 ) , whose data indicated a more gradual rise in the sfrd than had previously been found by lilly et al .
( 1996 ) .
it was realized that dust obscuration effects could result in factors of 3 to 5 ( pettini et al . 1997 ; meurer , heckman , & calzetti 1999 ) increases in the sfrd at high redshift . with these rather uncertain dust corrections taken into account
, it has been argued that the sfrd flattens at a constant @xmath274 in the @xmath39 cosmology for @xmath275 ( steidel et al .
1999 ) . in fig .
[ figsfrdvsz ] we compare the star formation history in the optical ( without extinction corrections ) with that which we obtain in the submillimeter both before ( eqs .
[ eqsfrdsmma ] and [ eqhisfrdsmma ] ; solid triangles ) and after ( eqs .
[ eqtruesfrdsmma ] and [ eqhitruesfrdsmma ] ; solid circles ) correcting for incompleteness .
we also include our newly determined radio sfrd limits ( eqs . [ eqlowradioa ] and [ eqhiradioa ] ) on the figure as solid squares .
the submillimeter contribution to the sfrd inferred from our @xmath5 mjy observations is comparable to the ultraviolet / optical contribution to the sfrd .
the two wavelength regimes are likely sampling different stages in galaxy formation .
the submillimeter detects the formation of massive spheroids while the ultraviolet / optical detects the formation of smaller disk or bulge systems .
the approximate equality of the optical and submillimeter backgrounds supports this hypothesis ; the metal density in present - day disks is roughly comparable to that in the spheroidal components of galaxies , so comparable amounts of light are expected to be produced in their formation ( cowie 1988 ) .
the completeness - corrected submillimeter sfrd , shown by the solid circles in fig .
[ figsfrdvsz ] , is based on the assumption that fainter submillimeter sources have the same redshift distribution and properties as the @xmath5 mjy sample .
hence the sfrd from the entire population contributing to the submillimeter background is about an order of magnitude higher than the extinction - uncorrected ultraviolet / optical sfrd .
since the submillimeter measures reradiated optical light , the observed ultraviolet / optical sfrd contribution should be added to the submillimeter contribution .
we also plot in fig .
[ figsfrdvsz ] the sfrd determined from the sum of the sfrd of the local ulig data ( @xmath207 ; @xmath276 ) from kim & sanders ( 1998 ) and the local near - ulig data ( @xmath277 ; @xmath278 ) from sanders et al .
( in preparation ) .
the latter @xmath188 selection is imposed in order to be able to compare with our completeness - corrected submillimeter data points , which should include all sources with luminosities @xmath279 .
we note that the radio data in fig .
[ figsfrdvsz ] can not be straightforwardly compared to the ulig data because the radio sources may not satisfy the same luminosity criteria .
the completeness - corrected sfrd in fig .
[ figsfrdvsz ] shows a very rapid evolution , @xmath280 , in the sfrd of uligs from @xmath7 to @xmath8 .
fast evolution is not surprising if the distant submillimeter sources are associated with major merger events giving rise to the formation of massive spheroidal systems ( smail et al .
1998 ; eales et al . 1999 ; lilly et al . 1999 ; trentham et al . 1999 ) .
barger , cowie , & sanders ( 1999 ) showed that the volume density of submillimeter sources at high redshift ( @xmath281 mpc@xmath282 for @xmath283 ) is comparable to the volume density of present - day elliptical galaxies ( @xmath284 mpc@xmath282 ) .
we have carried out an observational program designed to establish the overlap between the optical / nir - faint radio population and the submillimeter population and to exploit the complementary information that the radio and submillimeter wavelengths provide to gain insights into the evolution of the galaxy populations .
our major conclusions are as follows @xmath285 we have found that submillimeter sources at bright flux levels ( @xmath286 mjy ) are associated with optical / nir - faint radio sources ( @xmath287jy ) .
this association provides a powerful means to conduct submillimeter surveys by preselecting potential bright submillimeter sources through high - resolution deep radio maps .
@xmath285 we have shown that the redshifted submillimeter - to - radio flux ratio of an unevolved arp 220 sed reproduces the ensemble of local ulig data placed at appropriate redshifts , and thus arp 220 provides a prototype for ulig sources at high redshift . from the arp 220 model
we have derived a _ millimetric _
redshift estimator to determine the redshifts of the submillimeter sources using the ratio of the submillimeter to radio fluxes .
our estimator is consistent with another recent study by carilli & yun ( 2000 ) .
the flux strengths relative to the redshifted arp 220 sed indicate that the @xmath6 bright ( @xmath5 mjy ) submillimeter sources are @xmath288 the luminosity of arp 220 for @xmath283 and @xmath289 the luminosity of arp 220 for @xmath290-dominated models .
@xmath285 through radio versus submillimeter flux plots of the arp 220 predictions with redshift , we have shown that our present survey is sensitive only to sources with strengths comparable to or greater than arp 220 and that our @xmath291jy radio threshold precludes the detection of very high redshift sources ( @xmath292 ) if they are radio sources ; however , we serendipitously observed two bright submillimeter sources without radio detections that may lie at extreme redshifts .
an alternate possibility is that they could be colder galaxies than arp 220 located at @xmath293 .
it remains to be established whether the very faint radio population down to @xmath294jy have submillimeter counterparts .
@xmath285 at @xmath295 the ultraviolet / optical contribution to the sfrd dominates . at @xmath296
the submillimeter contribution from @xmath5 mjy sources is comparable to the observed ultraviolet / optical contribution .
the ultraviolet / optical and bright submillimeter are likely sampling two different stages in galaxy formation .
the bright submillimeter detects the formation of massive spheroids while the ultraviolet / optical detects the formation of smaller disk or bulge systems .
joint radio / submillimeter surveys are a powerful way to explore dust - obscured galaxies in the distant universe since the radio and submillimeter approaches are complementary and allow us to infer redshifts and sfrds for galaxy sources that we now conclusively see are inaccessible in shorter wavelength observations . 0.5 cm we thank dave sanders for valuable discussions , nicolas biver for expert observing assistance , and an anonymous referee for helpful comments about the manuscript .
support for this work was provided by nasa through hubble fellowship grants hf-01117.01 - 99a and hf-01123.01 - 99a awarded by the space telescope science institute , which is operated by the association of universities for research in astronomy , inc .
, for nasa under contract nas 5 - 26555 . shaver , p.a . , hook , i.m . , jackson , c.a . ,
wall , j.v . ,
kellermann , k.i .
1998 , in highly redshifted radio lines , asp conf .
series vol .
156 , eds .
carilli , s.j.e .
radford , k.m .
menten , g.i .
langston , p. 163 | direct submillimeter imaging has recently revealed the 850 @xmath0 m background to be mostly composed of a population of distant ultraluminous infrared galaxies , but identifying the optical / near - infrared ( nir ) counterparts to these sources has proved difficult due to the poor submillimeter spatial resolution .
however , the proportionality of both centimeter and submillimeter data to the star formation rate suggests that high resolution radio continuum maps with subarcsecond positional accuracy can be exploited to locate submillimeter sources . in this paper we present results from a targeted scuba survey of micro - jansky radio sources in the flanking fields of the hubble deep field .
the sources were selected from the uniform ( 8 @xmath0jy at 1@xmath1 ) 1.4 ghz vla image of richards ( 1999b ) .
even with relatively shallow scuba observations ( a @xmath2 detection limit of 6 mjy at 850 @xmath0 m ) , we were successful at making submillimeter detections of optical / nir - faint ( @xmath3 and @xmath4 ) radio sources , and our counts closely match the bright counts from submillimeter surveys .
an important corollary is that a large fraction of the bright ( @xmath5 mjy ) submillimeter sources in untargeted submillimeter surveys have extremely faint optical / nir counterparts and hence are inaccessible to optical imaging and spectroscopy .
however , redshift estimates can be made from the ratio of the submillimeter flux to the radio flux across the 100 ghz break in the spectral energy distribution .
this procedure , which we refer to as _ millimetric _
redshift estimation , places the bright submillimeter population at @xmath6 where it forms the high redshift tail of the faint radio population .
the star formation rate density ( sfrd ) due to ultraluminous infrared galaxies increases by more than two orders of magnitude from @xmath7 to @xmath8 .
the sfrd at high redshift inferred from our @xmath5 mjy submillimeter observations is comparable to that observed in the ultraviolet / optical . |
this work was supported by the national science council of taiwan under grants no .
100 - 2628-m-007 - 001 and no . 101 - 2923-m-007 - 002 and by national tsing hua university under grant no .
s. d. was supported by the hong kong research grants council ( project no .
the hsinchu team is the partner in eu fp7 irses project colima ( contract no .
pirses - ga-2009 - 247475 ) .
the authors thank professor p. k. lam for valuable comments on the manuscript .
99 t. chanelire , d. n. matsukevich , s. d. jenkins , s. y. lan , t. a. b. kennedy , and a. kuzmich , nature * 438 * , 833 ( 2005 ) . m. d. eisaman , a. andr , f. massou , m. fleischhauer , a. s. zibrov , and m. d. lukin , nature * 438 * , 837 ( 2005 ) . k. s. choi , h. deng , j. laurat , and h. j. kimble , `` mapping photonic entanglement into and out of a quantum memory , '' nature * 452 * , 67 ( 2008 ) . c. clausen , i. usmani , f. bussires , n. sangouard , m. afzelius , h. de riedmatten , and n. gisin , nature , * 469 * , 508 ( 2011 ) .
e. saglamyurek , n. sinclair , j. jin , j. a. slater , d. oblak , f. bussires , m. george , r. ricken , w. sohler , and w. tittel , nature * 469 * , 512 ( 2011 ) .
m. hosseini , b. m. sparkes , g. campbell , p. k. lam , and b. c. buchler , nature commun .
* 2 * , 174 ( 2011 ) .
m. hosseini , g. campbell , b. m. sparkes , p. k. lam , and b. c. buchler , nature phys .
* 7 * , 794 ( 2011 ) .
k. f. reim , j. nunn , x. m. jin , p. s. michelberger , t. f. m. champion , d. g. england , k. c. lee , w. s. kolthammer , n. k. langford , and i. a. walmsley , phys .
lett . * 108 * , 263602 ( 2012 ) .
n. gisin , g. ribordy , w. tittel , and h. zbinden , rev .
74 * , 145 ( 2002 ) .
a. kuzmich , w. p. bowen , a. d. boozer , a. boca , c. w. chou , l. m. duan , and h. j. kimble , nature * 423 * , 731 ( 2003 ) .
j. f. sherson , h. krauter , r. k. olsson , b. julsgaard , k. hammerer , i. cirac , and e. s. polzik , nature * 443 * , 557 ( 2006 ) .
e. knill , r. laflamme , and g. j. milburn , nature * 409 * , 46 ( 2001 ) . l. m. duan , m. d. lukin , j. i. cirac , and p. zoller , nature * 414 * , 413 ( 2001 ) .
j. briegel , w. dur , j. i. cirac , and p. zoller , phys . rev
. lett . * 81 * , 5932 ( 1998 ) .
n. sangouard , c. simon , h. de riedmatten , and n. gisin , rev .
83 * , 33 ( 2011 ) .
g. htet , a. peng , m. t. johnsson , j. j. hope , and p. k. lam , phys .
a * 77 * , 012323 ( 2008 ) .
m. p. hedges , jevon j. longdell , y. li , and m. j. sellars , nature * 465 * , 1052 ( 2010 ) .
m. varnava , d. e. browne , and t. rudolph , phys .
lett . * 97 * , 120501 ( 2006 ) .
s. e. harris , phys . today * 50 * , no . 7 , 36 ( 1997 ) .
m. fleischhauer and m. d. lukin , phys .
a * 65 * , 022314 ( 2002 ) .
a. v. gorshkov , a. andr , m. fleischhauer , and a. s. srensen , and m. d. lukin , phys .
lett . * 98 * , 123601 ( 2007 ) .
i. novikova , a. v. gorshkov , d. f. phillips , a. s. srensen , m. d. lukin , and r. l. walsworth , phys .
lett . * 98 * , 243602 ( 2007 ) .
n. b. phillips , a. v. gorshkov , and i. novikova , phys .
a * 78 * , 023801 ( 2008 ) .
n. b. phillips , a. v. gorshkov , and i. novikova , phys .
a * 83 * , 063823 ( 2011 ) .
s. zhang , s. zhou , m. m. t. loy , g. k. l. wong , and s. du , opt . lett . * 36 * , 4530 ( 2011 ) .
s. zhou , s. zhang , c. liu , j. f. chen , m. m. t. loy , g. k. l. wong , and s. du , opt .
express * 20 * , 24124 ( 2012 ) .
y. f. chen , y. c. liu , z. h. tsai , s. h. wang , and i. a. yu , phys .
a * 72 * , 033812 ( 2005 ) .
y. w. lin , h. c. chou , p. p. dwivedi , y. c. chen , and i. a. yu , opt . express * 16 * , 3753 ( 2008 ) .
s. w. su , y. h. chen , s. c. gou , t. l. horng , and i. a. yu , phys .
a * 83 * , 013827 ( 2011 ) .
y. h. chen , m. j. lee , w. hung , y. c. chen , y. f. chen , and i. a. yu , phys . rev . lett . *
108 * , 173603 ( 2012 ) .
y. w. lin , w. t. liao , t. peters , h. c. chou , j. s. wang , h. w. cho , p. c. kuan , and i. a. yu , phys . rev .
102 * , 213601 ( 2009 ) .
see details of the theoretical calculation in appendix .
s. e. harris and y. yamamoto , phys .
lett . * 81 * , 3611 ( 1998 ) .
d. a. braje , v. bali , s. goda , g. y. yin , and s. e. harris , phys .
lett . * 93 * , 183601 ( 2004 ) .
we made the theoretical predictions by numerically solving the maxwell - schrdinger equation of the light pulse and the optical bloch equation of the atomic density - matrix operator .
the inset in fig .
1 of the main text shows the three - level eit system formed by the two ground states of @xmath4 and @xmath9 and the excited state of @xmath8 .
the probe and coupling fields drive the @xmath37 and @xmath38 transitions with the rabi frequencies of @xmath5 and @xmath39 , respectively . considering the very weak probe pulse as the perturbation , we used the following equations to calculate the theoretical predictions of slow light and storage and retrieval of light [ 31 ] : @xmath40 where @xmath41 s are the elements of the density - matrix operator ; @xmath11 is the spontaneous decay rate of the excited state and equal to 2@xmath42@xmath356 mhz in our case ; @xmath43 = @xmath44 ; @xmath45 and @xmath19 are the optical density ( od ) and length of the medium ; @xmath24 is the relaxation rate of the ground - state coherence . in the four - wave mixing ( fwm ) process
, the same coupling field in the eit scheme also drives the far - detuned @xmath37 transition and generates the sum - frequency field ( @xmath46 ) of the @xmath47 transition with the help of @xmath48 .
we follow the formulas given by ref . [ 24 ] , which experimentally and theoretically studied this fwm effect , and modify eqs .
( s1)-(s3 ) to @xmath49 where @xmath50 , @xmath51 , and @xmath52 are the rabi frequency , detuning , and spectral linewidth of the coupling s @xmath37 transition , respectively . because @xmath50 and @xmath39 are resulted from the same coupling field , they only differ by the ratio of the clebsch - gordan coefficients of the @xmath37 and @xmath38 transitions . in our case , @xmath51 is about -1140@xmath11 and @xmath52 of cold atoms is approximately equal to the natural linewidth of the excited state @xmath11 .
hence , such large @xmath53 makes the fwm effect negligible . based on our experimental condition , the numerical predictions from eqs .
( s1)-(s3 ) and from eqs .
( s4)-(s7 ) are nearly the same . on the other hand ,
@xmath52 of hot or room - temperature atoms is about 100-fold larger due to the doppler broadening , reducing @xmath53 by two orders of magnitude .
therefore , the fwm effect becomes rather significant in the hot medium of a large od [ 24 ] . in the study of time - space - reversing method , the probe pulse was sent to the medium in the forward direction , stored by turning off @xmath13 , and retrieved in the backward direction by turning on @xmath14 .
the equations are given by [ 31 ] @xmath54 where the @xmath55 and @xmath56 signs indicate the fields and optical coherences in the forward and backward propagation directions , e.g. @xmath57 is the probe pulse sent to the medium in the forward direction and @xmath58 is that retrieved from the medium in the backward direction and , @xmath59 and @xmath60 are their corresponding optical coherences .
in the paper , the theoretical predictions shown in fig .
2(a ) and the two blue lines in fig .
2(b ) were calculated with eqs .
( s1)-(s3 ) ; those in fig .
4(a ) and the green line in fig .
2(b ) were with eqs .
( s8)-(s12 ) . | a high - storage efficiency and long - lived quantum memory for photons is an essential component in long - distance quantum communication and optical quantum computation . here , we report a 78% storage efficiency of light pulses in a cold atomic medium based on the effect of electromagnetically induced transparency ( eit ) . at 50% storage efficiency
, we obtain a fractional delay of 74 , which is the best up - to - date record .
the classical fidelity of the recalled pulse is better than 90% and nearly independent of the storage time , as confirmed by the direct measurement of phase evolution of the output light pulse with a beat - note interferometer .
such excellent phase coherence between the stored and recalled light pulses suggests that the current result may be readily applied to single photon wave packets .
our work significantly advances the technology of eit - based optical memory and may find practical applications in long - distance quantum communication and optical quantum computation .
quantum memory @xcite is essential for quantum information processing , including quantum communication @xcite and quantum computation @xcite .
using quantum repeaters will be a practical protocol for implementing long - distance quantum communication without suffering from transmission loss @xcite .
the scheme proposed in ref .
@xcite divides a long distance into shorter elementary channels , stores and retrieves entanglement pairs , and extends the transmission distance via entanglement swapping .
quantum memory , a storage device mapping quantum state between light and matter , is a crucial component of the quantum repeater . processing a particular task or waiting for the completion of others
requires a quantum state to be stored in memory for a long enough time .
therefore , quantum memory with high storage efficiency ( se ) , which is defined as the ratio of recalled to input photon energies , and long coherence time are the keys to successful operation of long - distance quantum communication and quantum information processing .
the fractional delay ( fd ) or delay - bandwidth product at 50% se is a possible figure of merit for a memory in the no - cloning limit @xcite or in the one - way quantum computation @xcite , where fd is defined as the ratio of storage time to the full - width - half - maximum ( fwhm ) pulse duration .
several memory devices based on different mechanisms , such as gradient photon echo , raman interaction , and electromagnetically induced transparency ( eit ) , have been proposed and experimentally demonstrated . with the gradient echo memory , a maximum se of 87% as well as the fd of 11 at 50% se for classical light was demonstrated @xcite , and the recall fidelity can be as high as 98% for coherent pulses containing around one photon @xcite . with a far - off - resonant raman transition , ref .
@xcite showed raman memory with an se of 43% and a coherence time of approximately 1 @xmath0s for 300-ps coherent light pulses .
slowing and storing light using the eit effect @xcite has been intensively explored in the past two decades owing to its potential applications in optical information processing and optical communication , as well as its compatibility with quantum state operations .
previous theory predicted that the se of the eit - based memory could approach unity in an atomic ensemble at a high optical density ( od ) @xcite .
however , experiments in hot vapor cells have revealed that a maximum se is only about 43% at the optimum od and further increasing od even reduces the se @xcite .
recent demonstrations with cold atoms at high od also showed a saturation value of se at 50% for both coherent pulses
@xcite and single - photon wave packets @xcite .
these previous experiments seemingly excluded the eit medium as an efficient quantum memory . in this paper , we report a high - performance eit - based memory with an se of 78% and a coherence time of 98@xmath16 @xmath0s in a cold atomic medium .
this work is the first experimental demonstration that the time - space - reversing method proposed in ref .
@xcite can enhance the se of the eit - based memory .
the fd at 50% se is 74@xmath15 , which , to our knowledge , is the best record to date .
we also demonstrated that the classical fidelity of the recall pulse is better than 90% as measured by a beat - note interferometer @xcite .
our results suggest that eit - based quantum memory can be used for practical quantum information applications .
the experiment was carried out in a cigar - shaped cloud of cold @xmath2rb atoms produced by a magneto - optical trap ( mot ) @xcite .
typically , we trapped @xmath3 atoms with a temperature of about 300 @xmath0k @xcite .
the experimental setup is shown in fig . 1 .
we pumped the whole population to a single zeeman ground state @xmath4 as shown in the inset of fig . 1 .
the probe ( rabi frequency of which is denoted as @xmath5 ) and coupling fields all had @xmath6 polarization .
they drove the @xmath4@xmath7@xmath8 and @xmath9@xmath7@xmath8 transitions , respectively , forming the @xmath10 configuration of eit .
the spontaneous decay rate @xmath11 of the excited state @xmath8 is @xmath12 mhz . in this study
, the stored probe pulse can be retrieved by the coupling field @xmath13 ( or @xmath14 ) in a forward ( or backward ) direction .
the probe pulse was focused to a beam size with an @xmath15 full width of 150 @xmath0 m and propagated along the major axis of the atom cloud , denoted as @xmath16 .
the forward and backward coupling fields were collimated with beam diameters of 5 and 3 mm , propagating along the directions of @xmath17 and @xmath18 .
we set the peak power of the input probe pulse to about 100 nw . the collection efficiencies of the photo detectors for the forward and backward probe pulses were 52% and 68% , respectively . before performing each measurement , we employed the dark and compressed mot for about 7 ms to increase the atomic density and the od of the system .
other details of the experimental setup can be found in ref .
@xcite . : zero - order quarter - wave plate ; @xmath191-@xmath193 : lenses with the focal length @xmath20 specified in units of mm .
@xmath21 : mirror ; aom : 80-mhz acoustic optical modulator .
inset : relevant energy levels and laser excitations in the experiment.,width=264 ] we first worked with the forward - retrieval configuration .
the experimental data of the input , slow - light output , and stored and then recalled pulses are shown by the open circles in fig .
2(a ) . the temporal gaussian profile of the input probe pulse had an fwhm of 1.5 @xmath0s .
the solid lines in the figure are the theoretical predictions . in the calculation
, we numerically solved the maxwell - schrdinger equation of the light pulse and the optical bloch equation of the atomic density - matrix operator @xcite .
the good agreement between the data and the predictions indicates that the measurements were taken under od = 156 and @xmath13 = 1.1@xmath11 .
we directly integrated the areas below the input and output probe pulses and used their ratio to determine the se .
as shown in fig .
2(a ) , the se of a 4-@xmath0s storage time was about 69% which already exceeded previous results for eit - based storage .
some studies suggest that an unwanted four - wave mixing ( fwm ) process degrades the se and limits the maximum achievable se to about 43% with hot atomic vapors @xcite . here , the experimental data with the cold atoms agree with the theoretical predictions calculated without the fwm process .
we also found theoretically that the ses with and without the fwm process differ by less than 0.2% .
thus , the fwm effect can be negligible in our system and cold atomic samples .
all the equations used for the theoretical predictions and their details are described in appendix .
the measured se as a function of od is shown by the black squares in fig .
2(b ) .
for each od , the width of the gaussian probe pulse was the same and we optimized the coupling rabi frequency @xmath13 to maximize the se . because the eit bandwidth is proportional to @xmath22 ,
a higher od resulted in a larger optimum @xmath13 as expected .
the se data initially increased with the od but eventually became saturated .
this saturation phenomenon made us realize the coupling field also drove the @xmath9@xmath7@xmath23 transition with a detuning of 44@xmath11 , as shown in fig .
1 .
the undesired transition induced the switching effect , destroying the probe pulse @xcite .
a larger rabi frequency of the @xmath9@xmath7@xmath23 transition or a longer probe pulse made the switching more noticeable .
the switching effect is equivalent to increasing the relaxation rate of the ground - state coherence , @xmath24 . in fig .
2(b ) , the blue lines are the theoretical predictions of two different @xmath24 s . at lower ( higher ) ods , the se data points are closer to the predictions of the smaller ( larger ) @xmath24 as expected , because the @xmath13 used in the measurement was smaller ( larger ) .
this explains the observed saturation phenomenon of se .
please note that during the storage the coupling field was turned off so that the switching effect did not exist .
hence , the coherence time during the storage , @xmath25 , could not be directly derived from @xmath24 .
the measurement of @xmath25 will be shown later .
memory must possess high fidelity for practical applications .
the classical fidelity is defined by @xmath26 \left [ \int |e_{\rm out}(t)|^2 dt \right]},\ ] ] where @xmath27 and @xmath28 are the electric fields of input and output pulses and @xmath29 is the delay time .
equation ( 1 ) is similar to @xmath30 , but we replaced the wave function @xmath31 with the electric field here . to obtain the fidelity , we measured the phase evolution of pulses with a beat - note interferometer @xcite .
the pulsed probe field spatially overlapped the continuous - wave reference field , as depicted in the bottom - right corner of fig . 1 .
the intensity of the reference field was about one tenth of that of the probe pulse peak .
the combined fields produced the beat note at 80 mhz and entered a polarization - maintained fiber ( pmf ) .
coming out of the pmf , the combined fields split into two beams .
one beam was directly detected by pd3 , the beat note signal of which served as the phase reference .
the other propagated through the atoms and was detected by pd1 producing the data shown in fig .
3 .
other details of the beat - note interferometer can be found in ref .
@xcite .
the high - frequency signal in fig .
3(a ) is the beat note that was averaged over 1436 traces .
we can zoom in on different time intervals and clearly see the phase evolution of the beat note as demonstrated in figs .
3(b)-3(e ) .
the beat note provided the phase as well as amplitude information on the electric field in the output probe pulse .
using eq .
( 1 ) ( the integration range covers the @xmath15 pulse width ) and the beat note signals , we determined that the classical fidelity of the recalled pulse after a 7-@xmath0s ( or 55-@xmath0s ) storage time was 94% ( or 90% ) .
the theoretical fidelity was 97% due to the output pulse width being broadened to 1.7 times the input one .
the measured fidelity was limited by the signal - to - noise ratio in the beat - note signal .
the beat - note amplitude in the data yielding a fidelity of 94% was about twice that yielding a fidelity of 90% .
we increased the reference field intensity to double the beat note amplitude and the above fidelity was improved to 97% ( or 93% ) .
thus , the classical fidelity was influenced very little by storage time .
the excellent phase coherence between the stored and recalled light pulses suggests that the current results may be readily applied to single - photon wave packets .
s ( cyan ) and 55 @xmath0s ( blue ) storage times .
green dashed lines are the coupling field employed in the light - storage measurement .
( b)-(e ) zooming in on the time intervals around the pulse peaks with the same horizontal span of 200 ns .
the ratio of the vertical spans in ( b)-(e ) is 4:2:2:1 .
the frequency of beat notes is 80 mhz.,width=264 ] reference @xcite proposed a time - space - reversing method for improving the se with a backward - retrieval configuration .
a representative example is shown in fig .
4(a ) .
the probe pulse was sent to the medium in the forward direction , stored by turning off @xmath13 , and retrieved in the backward direction by turning on @xmath14 .
the retrieved probe pulse provided the reversed temporal profile for the input probe pulse in the next run . the forward write - in and backward read - back process
was run iteratively until the input and output pulse shapes were nearly in the time - reversal symmetry and the se approached the maximum value asymptotically .
the optimized input probe shape had a slowly varying front , with the frequency components well within the eit bandwidth , and a sharp - edged rear , corresponding high - frequency components .
the slowly varying front propagated through the entire medium with minimum loss .
the sharp - edged rear was forward stored near the medium entrance and backward retrieved without propagating through the medium , leading to a low energy loss .
representative data at storage times of 4 and 54 @xmath0s are shown by the open circles in fig .
4(a ) .
the theoretcial predictions shown by the solid lines and the experimental data had nearly the same shape but with slight differences in the amplitudes @xcite .
the measured se was enhanced to 78% with the backward retrieval plus optimum pulse shape .
this is also the first experimental demonstration confirming the time - space - reversing theory in the eit system @xcite . the red circles in fig .
2(b ) show the measured optimal se as a function of od in the backward - retrieval configuration .
because the coupling rabi frequency does not affect the se and only determines the optimum pulse shape in this method @xcite , we kept @xmath32 approximately the same for all ods . each measured datum was smaller than its theoretical prediction .
the discrepancy may be due to phase mismatch during the backward retrieval @xcite .
a phase mismatch caused by a beam misalignment of 0.1@xmath33 can reduce the retrieved energy by about 4% .
both the measured and predicted ses behaved similarly , increasing with the od and approaching an asymptotic value . figure 4(b ) shows the storage efficiency as a function of storage time .
the experimental data behaved like a gaussian - decay function , i.e. , @xmath34 where @xmath25 is the coherence time , suggesting that the atomic motion was the dominant decoherence mechanism .
best fit of the data gives @xmath25 .
the average of day - to - day @xmath25 is 98@xmath16 @xmath0s . at the 50% se
, we obtain an excellent fraction delay of 74@xmath15 . in the backward - retrieval configuration , the input and recall pulses propagated in opposite directions and were detected by two different detectors ( pd1 and pd2 in fig . 1 ) .
this makes it extremely difficult to measure their phase correlation . because the beat - note data of fig .
3 showed excellent phase correlation between the input and forward - recalled pulses , the phase of the input pulse must have been encoded to the atomic spin wave during the storage . the retrieval process in either the forward or backward direction maps both amplitude and phase of the spin wave to the recall pulse
therefore , the input and backward - recalled pulses should also be phase coherent . under an assumption of perfect phase correlation , we took the square root of the input and retrieved pulses ( after time reversal ) as the fields in eq .
( 1 ) and estimate the classical fidelities of the two recall data in fig .
4(a ) all to be greater than 99% , which is better than the theoretical fidelity in the forward retrieval .
the superior result is due to nearly perfect similarity between the input and time - reversed output pulse shapes . in conclusion
, we have demonstrated an efficient eit - based optical memory with cold atoms .
our results in the forward retrieval show a se of 69% . using the space - time - reversing method plus optimum pulse shape , we further enhanced the se to 78% .
this efficiency is normalized with respect to the collection efficiencies of the photodetectors in the forward and backward directions .
the saturation of se at a high od is mainly caused by the dephasing rate of about 5.5@xmath35@xmath36@xmath11 . in our system ,
the saturation resulted from the far - detuned switching effect and the laser frequency instability , rather than from the fwm process which limits the se with hot atomic vapors .
the measured decay time of the memory is about 98 @xmath0s . allowing the se to be 50%
, we obtained a large fd of 74 .
we also employed a beat - note interferometer to verify that the classical fidelity of the recalled pulse was better than 90% and nearly independent of the storage time .
there is excellent phase coherence between input and output pulses as well as good agreement between theoretical and experimental values of fidelity .
although the experiments were done with coherent optical pulses , the eit light - matter interface may possibly be applied to single - photon quantum states due to its quantum nature .
our work may bring the eit - based quantum memory toward practical quantum information applications . |
Subsets and Splits
No saved queries yet
Save your SQL queries to embed, download, and access them later. Queries will appear here once saved.